Vertebral Artery Injuries Associated with Cervical Spine Trauma

Published on 17/03/2015 by admin

Last modified 22/04/2025

Print this page

This article have been viewed 2456 times

CHAPTER 82 Vertebral Artery Injuries Associated with Cervical Spine Trauma

Adam L. Shimer, MD, Alexander R. Vaccaro, MD, PhD

Vertebral artery injury (VAI) due to blunt cervical trauma is often initially asymptomatic; however, potentially devastating sequelae including embolism, arterial dissection, and stroke make this condition a diagnostic challenge and treatment dilemma. Recent improvements in diagnostic imaging have led to a higher rate of VAI detection, but the optimal treatment of VAI in the setting of cervical spinal trauma remains controversial. Although treatment of symptomatic VAI is prudent in nearly all settings, the treating physician must balance the treatment of asymptomatic patients with the complications of anticoagulation therapy and possibly the delay of surgical intervention when a diagnosis of vertebral arterial disruption is encountered.

Anatomy

The right vertebral artery (VA) arises from the right subclavian artery distal to the origin of the common carotid artery. On the left side, the VA usually arises directly from the subclavian artery, but in 3% to 6% of cases it may arise from the arch of the aorta as a separate vessel. From the vessel’s origin at the subclavian to its intracranial termination, the VA can be divided into four anatomic segments. The first portion (V1) extends from the subclavian takeoff to the foramen transversarium of C6 as it travels between the longus colli and anterior scalene muscles. The second portion (V2) of the artery then courses cephalad through successive transverse foramina until the level of C2, where the arteries diverge and then course superiorly to enter the more laterally situated foramen transversarium of the C1 vertebral body. The third segment (V3) begins on exiting the C1 foramen transversarium, where it lies in a groove along the superior margin of the posterior arch of the atlas; it then turns anteromedially and passes through the foramen magnum, lying anterior to the medulla oblongata. The fourth and final segment (V4) is intracranial after it pierces the dura mater. The VA then joins with the contralateral artery to form the basilar artery, which feeds the circle of Willis and contributes to the posterior circulation of the brain. This confluence is the anatomic basis of unilateral VA occlusion often being asymptomatic.

Epidemiology

As improved, higher-resolution imaging protocols have been developed and instituted, the incidence of detectable VAIs has increased. Although VAI has been reported as a result of low-energy movements such as chiropractic manipulation,^1,² yoga, and sudden head turning,³^–⁵ discussion of incidence is most appropriate for consideration in high-energy cervical spine trauma such as motor vehicle collisions and falls.

Investigation into incidence of VAI began as reports on small cohort groups and then later matured into larger, prospective studies. Louw and colleagues ⁶ were the first to investigate VAI in bilateral facet dislocation in a prospective fashion using digital subtraction angiography (DSA) and detected arterial occlusion in 9 of 12 (75%) patients. Willis and colleagues⁷ defined “high-risk” groups as those with criteria of either cervical spine fracture involving the transverse foramen or bilateral facet dislocations and, using DSA, detected VAI in 12 of 26 (46%) patients. Miller and colleagues⁸ described a broader definition of “high-risk” criteria including (1) cervical spine fracture/dislocation, (2) neurologic findings not explained by brain imaging, (3) Horner syndrome, (4) Le Fort II or III fracture, (5) skull base fracture, or (6) expanding neck soft tissue injury. Of 216 patients who met the criteria, they found 49 (22.7%) VAIs. Biffl and colleagues’⁹ selection criteria was even more inclusive and found 92 VAI among 605 patients (15.2%). Using magnetic resonance angiography (MRA), Vaccaro and colleagues¹⁰ noted a 37% rate of VAI with unilateral and bilateral cervical facet dislocations. In a follow-up study Vaccaro found a 20% (12/61) incidence of VA occlusion in patients with subaxial cervical spine fractures.¹¹

Later epidemiologic studies have used large, level I trauma registries to more accurately define overall incidence of VAI in blunt trauma admissions. The incidence of VAI among total blunt trauma admissions ranged from 0.075% to 0.77%.^8,¹²^–¹³ The studies had inherent variations due to location and referral pattern of the center and different imaging modalities used including digital subtraction angiography, computed tomography angiogram, and magnetic resonance arteriography. Despite the relatively low overall incidence of blunt VAI in the overall trauma population, this small population of patients has a substantial risk of stroke and mortality,¹⁴ thereby demanding screening criteria and structured protocols for detection of clinically silent lesions.

Mechanism and Types of Arterial Injury

Although VAI has been described at all levels, the V2 segment is the most commonly cited level of injury due to its confinement and tethering within the foramen transversarium, making it susceptible to shear forces during forceful cervical rotation or direct injury from fracture fragments.^15,¹⁶ Although less common, the tortuous V3 segment is vulnerable to injury due to its tethering between the foramen transversarium of the atlas and the atlantal-occipital membrane.¹⁷

Cervical flexion-distraction fracture patterns most commonly result in VAIs, followed by flexion-compression injuries (Figs. 82-1 and 82-2).¹¹ A review of the literature suggests that a rotational component to the cervical spine may also be a risk factor for VA occlusion.

FIGURE 82–1 Flexion-distraction stage II injury (lordotic position). A unilateral facet dislocation has been created at C4-5. The anteroposterior (A) and lateral (B) projections show an hourglass-like stenosis of the left vertebral artery (black arrows). The contralateral vessel (right) has a normal course in the anteroposterior (C) and lateral (D) projections. E, Contrast extravasation by rupture of a local radicular artery (black arrow) and loosening of a muscular arterial ligature (white arrow) is demonstrated at dye injection into the left vertebral artery.

(From Sim E, Vaccaro AR, Berzianovich A, Simon P: The effects of staged static cervical flexion-distraction deformities on the patency of the vertebral arterial vasculature. Spine 25:2180-2186, 2002.)

FIGURE 82–2 Flexion-distraction stage III injury. A bilateral facet dislocation has been created at C4-5. The anteroposterior (A) and lateral projections (B, right, and C, left vertebral artery) reveal excessive thinning of dye flow without complete obstruction (white and black arrows) in both the vertebral vessels.

In an attempt to better refine patient screening criteria for VAI, Cothren and colleagues ¹³ reviewed 17,007 blunt trauma patients of whom 125 had cerebral vascular injuries associated with cervical spine injuries. The injuries were characterized as subluxations in 56 (48%) patients, C1 to C3 cervical spine fractures in 42 (36%), and extension of the fracture through the foramen transversarium in 19 (16%). Cervical spine fractures were the sole indication for screening in 90% of the study population. Screening yield of all patients admitted with one of these three fracture patterns was 37%. Therefore they recommended focused screening for patients with cervical subluxations, fractures through the transverse foramen, or fractures of C1 through C3.

Injury to the VA in the setting of nonpenetrating cervical trauma is most likely due to a stretching mechanism.¹⁸ Sim and colleagues¹⁹ demonstrated in a cadaveric cervical spine model that the static deformity of a flexion-distraction stage II to IV subaxial cervical injury resulted in significant compression/stretch of the vertebral arterial vasculature (Figs. 82-3 to 82-5). Initially the vessel, or a portion of its intimal lining, is injured through excessive distraction in a flexion-distraction injury owing to the soft tissue attachments of the vessel in the foramina transversaria. Secondary events such as thrombus formation may then lead to occlusion of the vessel lumen. Recanalization over time after blunt injury to the VA was not a feature observed in a long-term follow-up study reported by Vaccaro and colleagues¹⁰ of image-confirmed vertebral vessel disruption. The lack of recanalization suggests that significant vessel intimal disruption occurs and that thrombus formation is not the only factor involved in vessel occlusion.

FIGURE 82–3 A 73-year-old woman incurred a flexion-distraction injury at C5-6 after a fall. A magnetic resonance angiography coronal projection image from a two-dimensional time-of-flight dataset shows evidence of normal antegrade flow in the common carotid arteries bilaterally and the right vertebral artery. There is absence of flow in the expected course of the left vertebral artery (arrows).

(From Vaccaro AR, Gregg KR, Flanders AE, et al: Long-term evaluation of vertebral artery injuries following cervical spine trauma using magnetic resonance angiography. Spine 23:789-794, 1998.)

FIGURE 82–4 A 51-year-old woman sustained a C4-5 flexion-distraction injury after a motor vehicle accident. This patient underwent a posterior cervical fusion. Magnetic resonance imaging revealed a C4-5 central herniated disc resulting in spinal cord compression. Cord edema was seen from C3 to C5. An axial magnetic resonance angiography gradient-echo image obtained at the mid-C4 level shows a hypointense focus (compared with the normal [hyperintense] left foramen transversarium) contained within the right foramen transversarium (arrow) that is consistent with acute thrombosis of the right vertebral artery.

(From Vaccaro AR, Gregg KR, Flanders AE, et al: Long-term evaluation of vertebral artery injuries following cervical spine trauma using magnetic resonance angiography. Spine 23:789-794, 1998.)

FIGURE 82–5 A 28-year-old man sustained a flexion-distraction injury at the C5-6 level following a motor vehicle accident. A, Coronal projection image from a two-dimensional time-of-flight magnetic resonance angiography dataset shows the junction of the right vertebral artery with the basilar artery at the expected location of the vertebrobasilar junction. The distal left vertebral artery is absent (arrow). B, Four transaxial contiguous images from a two-dimensional time-of-flight dataset show absence of flow-related enhancement in the left foramen transversarium, which is consistent with occlusion of the left vertebral artery.

(From Vaccaro AR, Gregg KR, Flanders AE, et al: Long-term evaluation of vertebral artery injuries following cervical spine trauma using magnetic resonance angiography. Spine 23:789-794, 1998.)

The intima of the VA is highly sensitive to shear forces experienced during high-energy blunt trauma. Because the vessel adventitia is more resilient, isolated intimal disruption can lead to a spectrum of clinical entities including an intimal flap with resultant thrombus formation or a dissection as the intima and adventitia separates cranially. Thrombus and dissection can lead to occlusion. If the energy transmitted is high enough, the adventitia can fail, leading to a pseudoaneurysm formation ²⁰ or transection if complete.

DSA provides the most anatomic detail of arterial injuries. A dissection of the VA has been shown to be the most common pattern of VAI, followed by occlusion. Multiple studies have demonstrated that dissections tend to resolve, whereas occlusions, as a rule, do not recanalize regardless of therapeutic attempts.^10,^21,²²

Clinical Diagnoses

Abundant collateral circulation feeding the vertebrobasilar system and posterior circulation of the brain often allows unilateral disruption of the VA to remain asymptomatic. When this collateral circulation is diminished in situations of atherosclerosis or anatomic variations, patients may initially present with symptoms of vertebrobasilar insufficiency. Symptoms of VA insufficiency may manifest as dizziness, dysarthria, dysphagia, diplopia, blurred vision, and tinnitus.¹⁷

Initially asymptomatic patients can have progression of symptoms either acutely or delayed due to embolism, thrombus extension, or dissection. Heros and colleagues ²³ described a patient who sustained a VAI with a resultant cerebellar infarction despite having a normal contralateral artery. To explain the infarction, the authors postulated that the thrombosis had extended distally to the intracranial portion of the VA. Interestingly, Six and colleagues²⁴ reported a case of an asymptomatic post-traumatic bilateral VA occlusion. Angiography demonstrated occlusion of both vertebral arteries but with the presence of blood flow through the intramuscular collateral vessels of the thyrocervical trunk and by collaterals from the superficial occipital artery. In fact, Blam and colleagues²⁵ conducted a retrospective review of 1283 patients with cervical spine trauma and determined that a normal neurologic examination is not predictive of VA patency after significant cervical spine trauma. These authors also noted that VAI was observed in similar frequency in both neurologically intact and motor incomplete patients (ASIA C and D).²⁵

The interval between spinal injury and the development of vertebrobasilar symptoms varies greatly from immediately after trauma to up to 3 months later.^17,²⁶ The late onset of symptoms suggests a process of vertebral vessel thrombus formation at the injury site with clot propagation or embolism and subsequent infarction. Therefore clinicians should be aware that late symptoms and signs of vertebrobasilar insufficiency after cervical spine trauma may be a manifestation of VAI that occurred at the time of trauma.¹¹ Vaccaro and colleagues¹⁰ demonstrated that 83% of patients with a known traumatic vertebral vessel injury demonstrated no flow reconstitution after an average follow-up of 25.8 months. The one patient who did show reconstitution of flow was thought to have a resolution of vertebral arterial spasm that initially compromised flow.

Imaging Modalities

Digital subtraction angiography (DSA) is a type of fluoroscopy technique used in interventional radiology to clearly visualize blood vessels in a bony or dense soft tissue environment. Images are produced using contrast medium by subtracting a “precontrast image” or the mask from later images, once the contrast medium has been introduced into a structure. DSA has routinely been described as the “gold standard” for detection of VA abnormalities. It is the only imaging technique that can reproducibly detect subtle intimal abnormalities. DSA is also the only technique that allows an interventional neuroradiologist to perform interventions if deemed necessary and appropriate. DSA, however, is time demanding, technique dependent, and invasive, with inherent risks including iatrogenic arterial perforation, puncture-site hematoma/false aneurysm, and dye toxicity. Because of this, DSA may not be a suitable screening tool, leading to development of alternate tests including computed tomography angiography (CTA) and MRA. However, neither of these techniques has proven to match the sensitivity or specificity of DSA.

CTA is a technique that combines precisely timed high-speed, multidetector computed tomography (CT) imaging shortly after a bolus of contrast dye through a peripheral intravenous line. Using bolus tracking, a specific anatomic site, in this example the vertebral arteries, can be imaged without the need for cannulating a catheter directly to the site, thereby eliminating many possible complications associated with DSA. In addition, the amount of dye used in CTA is substantially less then that used in DSA. The majority of high-energy trauma patients require head and cervical spine scans; therefore this imaging modality can be done in the same clinical setting as other required trauma-specific imaging studies. The previous generation, four-detector CT scanner was widely found to be inadequate for VAI injury. Newer high-speed, 16-detector CT scanner imaging is far superior in resolution and in its ability to bolus track. Therefore this technique has experienced near universal acceptance within the radiology community as a replacement for DSA. Eastman and colleagues ²⁷ found that CTA was nearly identical to DSA for detection of VAI in 162 patients determined to be at risk of injury.

Although these less-invasive imaging techniques are attractive for the reasons stated earlier, and despite wide acceptance in the radiology literature, some trauma surgeons remain skeptical of CTA accuracy. Biffl and colleagues ²¹ have questioned their ability to detect all vascular lesions when compared with the “gold-standard” DSA. The investigators developed a grading scale to evaluate the performance of MRA and CTA: grade I, dissection with less than 25% luminal narrowing; grade II, dissection with more than 25% luminal narrowing; grade III, pseudoaneurysm; grade IV, occlusion; and grade V, transection. Forty-six consecutive patients underwent CTA or MRA in conjunction with DSA. Although adequate for dramatic lesions, the more subtle grade I, II, and III lesions were routinely missed by CTA or MRA.

Malhotra and colleagues ²⁸ also noted that increasingly CTA has been replacing DSA in suspected VAI injuries and selected 119 patients who met screening criteria (facial/cervical-spinal fractures; unexplained neurologic deficit; anisocoria; lateral neck soft tissue injury; clinical suspicion) out of 7000 level I blunt trauma admissions and performed DSA and CTA. In agreement with Biffl’s study they found an alarming rate of missed VAIs using CTA that were visualized using DSA. The authors went as far as to say using CTA was dangerous due to the risk of missed injuries. Miller and colleagues’ study also found CTA and MRA to be inadequate screening tools. Although many investigators have shown failures of CTA to detect subtle lesions, Berne and colleagues²⁹ have asserted CTA is sufficient for detection of clinically relevant injuries (i.e., high-grade injuries). Although screening criteria may be employed to direct a dedicated CTA of the cerebrovascular system, another consideration is to integrate a CTA of the VA into the standard CT trauma series protocol. Langner and colleagues³⁰ developed such a protocol and found that an optimized craniocervical CTA can be easily integrated into a whole-body CT protocol for polytrauma patients. They assert that no additional screening technique is necessary to identify clinically relevant vascular injuries. This type of protocol could lead to a more liberalized screening of all trauma patients as recommended by some investigators.³¹

MRA uses MRI with specifically designed sequences or gadolinium dye to image arterial blood flow. Similar to CTA, MRA offers a minimally invasive alternative to DSA. In addition, MRA is nonionizing. MRA can be performed in the same clinical setting as a spine or brain MRI. Although many studies have compared CTA to the “gold-standard” DSA, there are relatively fewer studies comparing the sensitivity and specificity of MRA in detecting VAI.

The most common noncontrast, flow-based method is inflow angiography or “time-of-flight” (TOF), which is simply ordinary MRI, using settings that make flowing blood much brighter than stationary tissue.

MRA based on an injected contrast medium (usually containing gadolinium) is currently the most common form of MRA. The contrast medium is injected into a vein, and images are acquired during the first pass of the agent through the arteries. Provided that the timing is correct, this may result in images of high quality. An alternative is to use a contrast agent that does not, as most agents do, leave the vascular system within a few minutes but remains in the circulation up to an hour (a “blood-pool agent”). Because longer time is available for image acquisition, higher-resolution imaging is possible. A problem, however, is the fact that both arteries and veins are enhanced at the same time.

Veras and colleagues ³² performed a retrospective review of patients with VAI to determine the ideal MRA imaging sequence. They noted that two-dimensional time-of-flight sequencing was a more effective imaging sequence than the three-dimensional time-of-flow because of a reduced false-positive signal dropout. Axial T1-weighted MRI was useful because of its ability to delineate the absence of flow from an intimal occlusion including an intraluminal clot. In a dog model Ren and colleagues³³ found that MRA was comparable with DSA for evaluation of VAI.

Treatment

Treatment considerations should be divided into (1) symptomatic with evidence of stroke, (2) symptomatic without evidence of stroke, and (3) asymptomatic VAI.

Several different methods of therapy have been used to treat symptomatically occluded VAs. The American Association of Neurological Surgeons/Congress of Neurological Surgeons Guidelines consensus concluded that patients with posterior circulation stroke and VAI have a better outcome when treated with intravenous heparin than patients who do not receive this treatment.³⁴ Fibrinolysis with streptokinase has also been described to successfully restore flow with varied results.³⁵

Although the exact mechanism is ill defined, be it ischemia or embolic, the outcome of patients who develop symptoms of posterior circulation ischemia without stroke and are treated with intravenous heparin ^1,³⁶^–³⁸ is similar to that of patients receiving no treatment.^6,³⁹ In the symptomatic patient, Schellinger and colleagues³⁷ advocated the initial use of heparin with a target activated partial thromboplastin time of at least twice the baseline value followed by oral anticoagulation (warfarin) for at least 3 to 6 months as a secondary prophylaxis.

Complications associated with anticoagulation in the trauma population range from 25% to 54%.⁴⁰ Bleeding complications include intracranial hemorrhage, gastrointestinal bleeding, retroperitoneal hemorrhage, and excessive surgical site bleeding. In fact, in a study by Eachempati and colleagues,⁴¹ only 14% of trauma patients were deemed candidates for full anticoagulation. These facts have led physicians to focus more on antiplatelet therapy as opposed to heparinization.

In appropriately selected candidates with a symptomatic dissecting VA, surgical intervention may include vertebral vessel resection followed by grafting. Another method of arresting vessel flow is through the use of catheter embolization with Guglielmi detachable coils.⁴²^–⁴⁴ Today the gold standard surgical procedure for a symptomatic vertebral dissection remains ligation of the injured VA proximal and distal to the site of the lesion rather than primary repair.⁴³ As mentioned previously, ligation is well tolerated if there is adequate collateral flow.

The treatment of asymptomatic VAIs is controversial. Today, in the majority of level I trauma centers, formal treatment is not recommended for a patient with an asymptomatic VAI after trauma.^24,⁴⁴ Biffl and colleagues² found, in a retrospective review, that systemic heparin improved the long-term neurologic outcome of both initially symptomatic and asymptomatic patients sustaining blunt VAI. The authors postulated that this treatment approach may prevent the arteriographic progression of lesions, protect against postinjury stroke, and prevent neurologic deterioration from treatment to discharge. However, the authors also noted that systemic anticoagulation introduces risks of hemorrhagic complications, especially in multisystem trauma victims with a reported 10% complication rate. Recently, Spaniolas and colleagues⁴⁵ reviewed the National Trauma Data Bank entries for 2001 to 2005 and identified 574 VAIs among the 761,385 blunt trauma admissions (0.075% overall incidence). The investigation found no improvement in the modified functional independence measure (FIM) score during this time despite progressively increasing rates of detection over that period (annual incidence .053% in 2001 to 0.1% in 2005 (P < 0.001)). There are definite dangers associated with anticoagulation administration such as hemorrhage, worsening neurologic function, and increasing infarction size.¹⁹ Routine anticoagulation administration also precludes surgical intervention that may be necessary to restore spinal stability and to relieve neural compression. A potential treatment choice may be the use of oral antiplatelet therapy (i.e., aspirin following surgery or nonoperative care for 3 months in the asymptomatic patient), although this has not been supported with any clinical trials to date.

Because of the risk of rupture and embolic source, a pseudoaneurysm of the VA is usually aggressively treated most commonly with coil embolization or stenting.

Transection of the VA is a life-threatening emergency due to hemorraghic shock and/or airway compromise. Treatment is airway management, resuscitation, and emergent interventional neuroradiology embolization.⁴⁶ An additional reason to routinely evaluate for the presence of a vertebral vessel disruption in an asymptomatic patient is to ensure safety to the intact vessel if surgical intervention with implants that may put the vertebral vasculature at risk are to be used.