Clinical manifestations

Presentation and course

Neuroanatomical location of injury.

Extracranial vascular injuries. The majority of patients with extracranial vascular injuries can be initially asymptomatic but can potentially present with neck pain, headache, a pulsatile mass lesion, bruit, airway compromise, hemodynamic instability, or a fixed, progressive, or transient neurologic deficit. Due to the variability in symptoms and the propensity for these patients to present asymptomatically, a high index of suspicion is needed to promptly and accurately identify these injuries and, if necessary, intervene before they result in subsequent morbidity or death.

Intracranial vascular injuries. Intracranial vascular injuries can present with headache, subarachnoid hemorrhage (photophobia, meningismus, headache, altered consciousness), altered mental status, coma, signs and symptoms of skull base fractures (otorrhea, rhinorrhea, cranial nerve palsies, hemotympanum, Battle’s sign), or even asymptomatically (04; 56; 58).

Carotid artery injury. Most patients with symptomatic carotid artery injuries present with focal ischemic problems. These are typically unilateral and can include pain, hemiparesis, aphasia, visual loss, transient ischemic attacks in an estimated 64% of cases, or Horner syndrome (89).

Vertebral artery injury. Most injuries to the vertebral artery remain asymptomatic, with only 9% to 20% of unilateral vertebral artery injuries presenting with ischemic signs or symptoms, though bilateral vertebral artery injury stroke risk can be as high as 50% (58; 89). However, these injuries have been associated with precipitous neurologic decline in some patients, with little clinical warning, as well as the possibility of delayed stroke in a small subset of cases. When symptomatic, vertebral artery injuries can result in ischemic symptomatology of the brainstem, cerebellum, and occipital lobes, with a wide range of clinical findings. These include lateral medullary infarction (Wallenberg syndrome), hemiatrophy, ataxia, dysmetria, tremor associated with nystagmus, tinnitus, deafness, dysarthria, dysphasia, cranial nerve palsies, cortical blindness, and coma. Vertebral artery injury following cervical spine injury is uncommon, but 70% of all traumatic injuries of the vertebral artery are associated with a cervical spine fracture (42). Clinical judgment and radiographic assessment are important, as routine screening is not fruitful or cost-effective.

Basilar artery injury. Injuries to the basilar artery and the intradural vertebral artery (V4) can rupture intracranially and present with signs and symptoms of subarachnoid hemorrhage in addition to the symptoms listed (04; 56; 58).

Type of injury.

Extradural traumatic pseudoaneurysms. Extradural traumatic pseudoaneurysms represent dilations of a weakened vessel wall; some are asymptomatic, but the sidewall outpouching can promote stagnant flow and intraluminal thrombus formation. Intradural traumatic pseudoaneurysms often become symptomatic through rupture (subarachnoid, intraparenchymal, or subdural hemorrhage) or via mechanical compression as they enlarge (eg, cranial nerve palsies). Pseudoaneurysms can present as early as a few hours after initial injury to months or even years later (28). Traumatic pseudoaneurysms following blunt trauma of the extracranial carotid artery and vertebral artery have a reported incidence of 15% to 23% and 4% to 8%, respectively (45).

Traumatic carotid-cavernous fistulas. Traumatic carotid-cavernous fistulas result from a tear in the wall of the cavernous carotid artery, allowing direct arterial flow into the cavernous sinus. Traumatic carotid-cavernous fistulas typically present with headaches, subjective bruit, and ipsilateral eye findings, which can include diminished vision, chemosis, excessive lacrimation, proptosis, ophthalmoplegia, and, in rare cases, brainstem edema due to reflux into a superior petrosal sinus draining the lateral mesencephalic vein (103). Symptoms often present immediately but can be delayed for weeks after the initial injury, with delayed onset of up to 8 weeks in some cases (09; 53; 94).

Penetrating injuries to the neck can result in complete or partial disruption of a vessel or no significant vascular injury at all. Patients with significant vessel injury may present with hemodynamic instability, arterial hemorrhage, airway compromise, or expanding hematomas. These patients are taken emergently to the operating room. It is often difficult to rule out a coexisting neurologic deficit in patients with severe blood loss and hypotension. Less severe penetrating injuries can present with or without focal neurologic deficits. Physical examination alone has been shown to be a reliable indicator of vessel injury. Azuaje and colleagues reported that physical examination alone had a sensitivity of 93% and a negative predictive value of 97% when evaluating potential vascular injuries from penetrating injuries. With the widespread availability of CT angiography, some have suggested that there is no longer a need for more invasive algorithms (08; 95; 119).

Following blunt trauma, it is often difficult to attribute symptoms specifically to either a vertebral or carotid artery injury unless the injury is isolated. Blunt trauma patients commonly present with coexisting injuries to the brain, spinal cord, or other organ systems. When evaluating posttraumatic patients with neurologic deficits lacking an obvious or immediate source, a vascular cause should be ruled out. Clinical and radiographic findings associated with carotid or vertebral artery injuries include cervical spine fractures extending into the foramen transversarium, cervical spine facet joint dislocations, cervical spine vertebral body subluxation, spinal cord injuries, craniofacial and skull base fractures, and significant craniocervical injuries (04; 56; 58).

Prognosis and complications

The natural history for dissections of the intracranial arteries is much worse than that for dissections of the extracranial carotid and vertebral arteries. Intracranial vertebral artery dissections have a mortality rate approaching 50% in untreated cases (102).

The rate of mortality resulting from carotid and vertebral artery injury is 10%. A study on functional outcomes after blunt cerebrovascular injury found an 11% rate of in-hospital deaths (35). Thirteen percent of patients had a stroke at some point during their hospitalization; 59% of the patients with ischemia had radiologic findings of ischemia on arrival to the hospital. In a retrospective study of 786 patients with blunt cerebrovascular injury, the risk of stroke in patients was 11.5% compared to controls, increasing up to 24.1% and 30% in severe carotid and vertebral artery injury, respectively (130). Approximately 75% of patients who suffer a stroke from a carotid or vertebral artery injury eventually have a good recovery (44). Most patients could return to work at their 3-year follow-up after blunt cerebrovascular injury (35). In general, most extracranial cervical artery dissections have a fairly benign clinical course (see Management section). Underlying arteriopathies are often suspected in the pathogenesis of spontaneous dissection of the internal carotid and vertebral arteries. However, in patients presenting with spontaneous dissection, the rate of recurrent dissection in an additional artery not involved in the initial presentation was only about 1% per year (112).

Vertebral artery injury alone is estimated to occur in 0.09% to 0.63% of all blunt trauma cases presenting to the emergency department, with 70% to 78% occurring in conjunction with cervical spine trauma. Although the majority of these lesions remain neurologically silent, sequelae such as cortical blindness, stroke, quadriplegia, and death occur in 0% to 24% (05; 73).

Patients with penetrating carotid artery injury at more than one level had higher mortality and morbidity than patients with blunt carotid artery injuries (93). Patients with penetrating injuries to the extracranial cerebral vasculature were noted to have an overall mortality of 21.2%, with a stroke rate in surviving patients of 15.1% (37). Patients with penetrating brachiocephalic artery injuries had the highest mortality (38.1%), whereas internal carotid artery injuries were associated with the highest stroke rate (22.7%). Poor outcomes were associated with hypovolemic shock, internal carotid artery injury, complete vessel transaction, and arterial ligation.

Clinical vignette

A 61-year-old previously healthy female was walking her dog when she was struck by a deer on the left side. Per EMS reports, the patient did have a loss of consciousness and was GCS 14 at the scene. On arrival to the Emergency Department, she was neurologically intact with complaints of a headache. She was monitored closely for a small epidural hematoma underlying a temporal bone fracture and then was discharged home 3 days after the trauma. One week after discharge, she developed double vision with left lateral gaze, orbital swelling, proptosis and worsening left eye scleral injection. She was evaluated by neuro-ophthalmology, who noted an increased intraocular pressure of 26; imaging was concerning for a carotid-cavernous fistula. She underwent a diagnostic angiogram and was found to have a direct carotid-cavernous fistula. Subsequent embolization with a pipeline stent was completed without any complications. Follow-up angiogram at 6 months showed complete radiographic resolution of the fistula. The patient also had resolution of her proptosis, injection, and diplopia in that timeframe.

Biological basis

Etiology and pathogenesis

Neurovascular injury to the extracranial or intracranial arteries (vertebral, carotid, or basilar) can occur spontaneously, after mild or severe blunt trauma, following surgical or endovascular interventions on or near these vessels (iatrogenic), or following penetrating injuries.

Dissections.

Spontaneous arterial dissection. One percent to 15% of spontaneous arterial dissections result from an intimal tear and blood “dissecting” between the vessel wall layers, which can lead to an intraluminal intimal flap and secondary vessel wall dilation termed an arterial pseudoaneurysm. Spontaneous dissections have been reported in association with a number of conditions, including Marfan syndrome, Ehlers-Danlos syndrome type IV, cystic medial necrosis, fibromuscular dysplasia, atherosclerosis, autosomal dominant polycystic kidney disease, and in patients using oral contraceptives and other medications and drugs. In these patients, there is an abnormality in the vessel itself, making it either more prone to spontaneous intimal layer disruption or more susceptible to damage following minor trauma. However, spontaneous dissections have also been reported in patients with no significant medical or traumatic history. Spontaneous arterial dissections tend to occur more commonly extracranially than intracranially and are twice as likely to affect the carotid artery as the vertebral artery. Bilateral involvement and a higher probability of reoccurrence are seen with spontaneous vertebral artery dissection (67; 44; 104; 129).

Extracranial arterial dissection. Extracranial arterial dissection can occur either spontaneously or following trauma. Spontaneous arterial dissections can occur in patients with the conditions listed above or in patients with no significant medical history. The arterial tunica intima is the most vulnerable to injury. Once there is a disruption in the intimal layer (either following trauma or due to an inherent vessel problem), intraluminal blood gains access to the subintimal space. As this blood is propelled under pressure, it forms an intramural hematoma, which can enlarge and cause a progressive dissection along the course of the vessel due to the laminar flow of blood. Once damaged, the internal elastic lamina will never reattach, but neointimal proliferation due to new smooth muscle and collagen synthesis can allow for healing within 2 months. Alternatively, direct hemorrhages from the vasa vasorum of the tunica media into the vessel wall may also form hematomas, which can dissect along this same plane. Arterial dissections can extend into the subadventitial layer, causing the formation of pseudoaneurysms, which, if intracranial, can rupture and cause subarachnoid hemorrhages (40; 44; 120).

Intracranial artery dissection. Intracranial artery dissections are rare and tend to be spontaneous in origin rather than traumatic, typically presenting with stroke symptoms in young adults. Intracranial artery dissections can be caused by either subintimal weakness as a result of luminal stenosis (potentially leading to ischemia) or by subadventitial weakness (potentially causing pseudoaneurysm formation and compression of nearby structures or subarachnoid hemorrhage in the event of rupture). The intracranial internal carotid artery is relatively fixed at the base of the skull, making it less prone to trauma. Also, differences in the intracranial arterial vessel anatomy (fewer elastic fibers in the tunica media and thinner adventitia, as well as the lack of external elastic lamina) may also account for spontaneous intracranial dissections and their tendency to bleed. Intracranial arterial dissections occur more frequently in the vertebral and basilar arteries than in the internal carotid artery. The supraclinoid internal carotid artery is the most commonly involved portion of the internal carotid artery, and dissection can extend from this region into the middle cerebral artery and anterior cerebral artery. The most common symptom is headache, which usually immediately precedes ischemic or stroke symptoms. Previous reports of intracranial arterial dissections were frequently associated with massive strokes and a high mortality rate (approximately 75%); however, more recent series have reported significantly better outcomes, which may be related to improved diagnostic imaging modalities and treatment options. Mortality rates are now estimated to be between 0.01% and 16.7% without subarachnoid hemorrhage and 13.8% to 62.5% with subarachnoid hemorrhage (27; 120).

Sequela of arterial dissection. Arterial dissections are associated with a high risk of ischemic sequelae and strokes. The risk of rebleeding amongst patients presenting with subarachnoid hemorrhage is estimated at 22% to 58% within the first 24 hours; however, it decreases significantly with time and proper medical management. Mural hematomas resulting from arterial dissections and pseudoaneurysms can cause vessel narrowing or occlusion. In addition, intimal injury can result in the formation of intimal flaps. These flaps are thrombogenic and can result in the formation of emboli that can embolize downstream, causing vessel occlusion distal to the site of dissection. Thrombus formation is further encouraged because dissections and pseudoaneurysms also disrupt the normal flow of blood, both at the site of injury and distal to the site of injury. Of note, traumatic and spontaneous dissections are distinct entities and have different clinicopathologic courses, with a less favorable outcome in the traumatic dissections (85; 68; 40; 44; 120).

Injuries.

Penetrating vascular injuries are primarily secondary to gunshot wounds. Other sources include stab wounds and accidents (lacerations or impalement). Penetrating injuries to the neck can result in complete or partial disruption of a vessel and are estimated to have morbidity and mortality as high as 80%. Due to the hemodynamic consequences of major arterial bleeding, these injuries are frequently explored, and the vessel is ligated, repaired, or bypassed. Unexplored partial injuries and occult injuries can result in the formation of arteriovenous fistulae as well as dissections or pseudoaneurysms, which represent a subacute stage that is amendable to curative therapy. If left untreated, pseudoaneurysms can cause distal thromboembolism, hemorrhagic rupture, or neurologic deficits from mass effect. In a retrospective study, Levy and colleagues found that 3.2% of patients who presented with subarachnoid hemorrhage after a gunshot wound to the head harbored a traumatic aneurysm (76; 96).

Blunt trauma. Blunt trauma to the head and neck (most commonly due to motor vehicle accidents) can also result in carotid and vertebral artery injury. The carotid artery is fixed at the base of the skull in the petrous canal. Sudden hyperextension with rotation causes the internal carotid artery to be pulled against the lateral mass of either C1 or C2. The vertebral artery is most susceptible to injury at the V2 segment, where it becomes fixed in the bony orifices of the foramen transversarium, and at the V3 segment, between its exit of the foramen transversarium and its entrance into the atlanto-occipital membrane. Cervical spine distraction, with flexion or hyperextension, and lateral flexion are the most frequent mechanisms of vertebral artery injury; cervical spine fracture is the only independent predictor (118). The most frequent radiological findings associated with vertebral artery injury include cervical spine fractures extending into the foramen transversarium, facet joint dislocations, significant vertebral subluxation, and complete spinal cord injury. Skull base fractures near or involving the carotid canal, the petrous bone, and Le-Fort II or III fractures and patients with significant craniocervical and upper cervical spinal cord injuries warrant further imaging studies to evaluate the carotid artery (50; 58; 73).

Carotid-cavernous sinus fistulas. Carotid-cavernous sinus fistulas are connections between the intracavernous carotid artery, or one of its branches, and the cavernous sinus. These can occur spontaneously but are most common following trauma, representing 75% of cases. Aneurysms of the cavernous carotid artery can also occur posttraumatically. When associated with trauma, a carotid-cavernous sinus fistula is usually a direct, high-flow lesion (type A) that produces a distinctive clinical syndrome characterized by pulsating exophthalmos and a bruit. Other less common presentations have been reported, including massive epistaxis, ophthalmoplegia, conjunctival infection, and otorrhagia. Morbidity associated with a missed diagnosis of a carotid-cavernous sinus fistula is significant, risking hemorrhage, blindness, cranial nerve palsy, and stroke. The cavernous sinus is invested with dura, and any rupture here rarely extends into the subarachnoid space (09; 53; 66; 135).

Iatrogenic injuries. Iatrogenic injuries to the carotid and vertebral arteries have been reported in association with multiple procedures: carotid endarterectomy, neck dissections for tumors, sinus surgery, anterior and posterior spine surgery, oral surgery, otological surgery, craniotomy, endovascular procedures, internal jugular vein cannulation procedures, irradiation to the neck, and many others. For many of these procedures, it is difficult to calculate the actual incidence of vessel injury because the natural history of most small, inadvertent minor vessel injuries is to heal without consequence, typically within 3 to 6 months with medical management (104). In studies of patients undergoing mechanical thrombectomy for stroke, the incidence of iatrogenic dissection varied from 1% to 3.9% and was found to be more common in smokers (44% vs. 19%) in one cohort of 866 patients (Goeggel Simonetti et a 2019). The incidence of unintentional carotid artery puncture during attempted cannulation of the internal jugular vein (using landmarks, in a nonemergent setting) is 5.9%. This rate is much higher in emergent and pediatric settings but can be reduced by 10% using ultrasound guidance (106; 105; 36). Despite this, there are several reports of strokes and serious complications. In spine surgery, vertebral injuries are fairly rare in anterior cervical approaches, with an incidence of 0.2% to 0.5%. For posterior surgeries, the incidence depends on the approach and ranges from 4.1% to 8.2% in C1/C2 transarticular fixation to 0% for subaxial lateral mass fixation (97). Anomalies of the vertebral artery course also pose a risk, including entrance into the transverse foramina at levels other than C6, tortuous artery course, or midline migration (78; 91). Cervical spinal manipulative therapy (eg, chiropractic manipulation) has also been shown to be an independent risk factor for vertebral artery dissection (122) but remains controversial due to the confounding factors, such as underlying connective tissue and musculoskeletal disorders (86).

Epidemiology

The incidence of extracranial internal carotid artery dissection is estimated to be two to three per 100,000 per year, and the incidence of extracranial vertebral artery dissection is estimated to be 1 to 1.5 per 100,000 per year (111). According to another study by Lee and colleagues in 2006, the incidence of internal carotid artery dissection is between 2.6 to 2.9 per 100,000; however, reliable data for vertebral artery dissection were unavailable (75).

The incidence of traumatic extracranial carotid artery dissection following craniocervical blunt trauma is estimated to be 0.3% to 0.67% in adults and 0.03% in children (41; 70; 77; 25). The incidence of vertebral artery injury from blunt trauma is estimated to be from 0.20% to 0.77% (58). Traumatic intracranial aneurysms comprise less than 1% of all intracranial aneurysms (71), and the incidence of carotid-cavernous sinus fistula formation is estimated to be approximately 0.2% of all head trauma patients. In addition to trauma, spontaneous extracranial arterial dissections (1% to 15% of all dissections) have been reported in association with a number of conditions (see Etiology section) and in association with a number of risk factors, which include hypertension, diabetes, smoking, hyperlipidemia, and oral contraceptive medications. There are fewer than 100 cases of spontaneous intracranial internal carotid artery dissection reported in the English literature (27).

Prevention

Primary prevention. Primary prevention of neurovascular injuries can be accomplished by advocating appropriate safety precautions such as the use of seat and safety belts, obeying traffic laws and regulations, avoiding risky behavior when driving a motor vehicle (eg, cellular phones, texting), and proper personal weapon safety. In addition, the use of appropriate protective gear (eg, helmets, body armor) when riding bicycles and motorcycles can help prevent and reduce the severity of neurovascular injuries caused by blunt and penetrating trauma.

National firearm mortality rates in the United States have been stable since 2000 (62). Comparing rates of firearm mortality across different states, some authors have argued that specific laws may reduce the incidence of gun violence mortality (61), but this analysis is controversial (02; 110).

Further means of prevention include proper management of treatable risk factors such as hypertension, hypercholesterolemia, and hyperhomocysteinemia (84).

Secondary and tertiary prevention. Secondary and tertiary prevention of sequela from vascular injuries involves proper and timely recognition and interventions for vascular injuries to reduce the potentially devastating consequences. These prevention strategies should be the focus of primary responders, emergency room physicians, and treatment teams of physicians and health professionals.

Differential diagnosis

In general, the differential diagnosis for neurologic sequela that can be seen with neurovascular injuries includes the following:

Diagnostic workup

In the last couple of decades, the search for the ideal diagnostic imaging modality for suspected cerebrovascular injury has undergone intensive investigation, particularly with improvements in CT angiography (CTA) technologies and interactive reconstructive software programs. There is no level I evidence for screening and diagnosis of blunt cervical vascular injuries (17). Although there is level II evidence to recommend conventional cerebral angiography as the reference standard (17), magnetic resonance imaging (MRI), magnetic resonance angiography (MRA), and CTA have become well-established, noninvasive screening methods with sensitivities and specificities approaching that of conventional angiography (136; 74; 123; 82; 29; 12; 100; 18). Level III evidence suggests that thin-slice, multidetector CTA is comparable to conventional cerebral angiography as a screening tool (17). Therefore, in many centers, catheter angiography is generally reserved for cases where screening modality results are equivocal or in acute clinical scenarios where endovascular interventions are concurrently required (95; 99). Further, catheter angiography can be used for selective arteriography to limit contrast dose in patients with renal insufficiency. A publication suggests the use of angiography following blunt trauma only in the case of unexplained neurologic findings or expanding hematoma of the neck (108). However, guidelines suggest imaging based on mechanism and severity of injury rather than just clinical findings (65).

The choice of noninvasive screening modality (CTA or MRA) appears to be influenced by institutional preference, ease of obtaining the study (including length of study time), and ability to couple the study with other desired imaging studies. MRA can be combined with advanced MRI methods, such as diffusion-weighted and perfusion-weighted imaging, to identify acute ischemic lesions and areas of hemodynamic compromise (44). CTA is relatively rapid and allows for the evaluation of multiple sites of injury, including those outside of the head and neck, on a single scanner; however, there remain reports of clinically silent neurovascular compromise that evades screening techniques. Guidelines published at one institution suggest the use of whole-body CT screening following blunt trauma when clinically judged to have sufficient mechanism, as it has a lower probability of missing cerebrovascular injury and thereby lowers stroke risk (99; 18).

Hoffman and colleagues studied risk factors associated with the diagnosis of blunt cerebrovascular injury in mild to moderate traumatic brain injury using the National Trauma Data (55). Blunt carotid artery injury is associated in ranked order with cervical spine fracture, mandible fracture, and basilar skull fracture. Blunt vertebral artery injury was associated mostly with cervical spine fracture, followed by spinal cord injury and neck contusion, and had a higher injury severity score. Stroke was more common in blunt carotid artery injury, but blunt vertebral artery injury had a higher mortality rate.

There is level II evidence to suggest that ultrasound is not adequate for screening blunt cervical vascular injuries (17). However, ultrasound is an excellent modality for serial imaging when the dissection is clearly demonstrated in the sonographic studies. For internal carotid artery dissections, ultrasound is sensitive for high-grade stenosis and occlusion, but it is less sensitive for low-grade stenosis and intracranial carotid artery dissections within the subpetrous segment and carotid canal (44). Because of the length of the vertebral artery and its surrounding bony structures, extracranial and transcranial color Doppler ultrasound fails to detect vertebral artery pseudoaneurysms; however, it may be a fast, reliable, and noninvasive method for follow-up imaging on vertebral artery dissections (133).

One study examined the use of transcranial Doppler emboli detection as a tool for stroke risk stratification in patients with internal carotid artery injury, suggesting early Doppler quantification of intracranial microemboli may aid in identifying patients at high risk for impending stroke. In their series of patients with carotid injury, transcranial Doppler studies positive for at least one microembolus predicted future stroke after adjusting for carotid artery injury grade. Of the none patients experiencing stroke more than 24 hours after injury, five had a positive transcranial Doppler microemboli study compared to 46 positive studies in the 248 patients without stroke. These findings were not seen with vertebral artery injuries (16).

As in the adult population, the screening guidelines for blunt cervical trauma in pediatric patients are not well established. Two major areas of interest include the realms of contact sports and motor vehicle accidents. Although the “return to play” guidelines for concussions experienced during contact sports are well-known, little is reported regarding cerebrovascular compromise, which can prove deadly if left untreated. Attention has been given to pediatric management following motor vehicle accidents, in particular to a cervical “seatbelt” sign. Reports in the literature have demonstrated that the benefit of CTA in the setting of a cervical “seat-belt” sign is minimal, as it is not commonly associated with cerebrovascular injury (34; 32).

Imaging findings that are characteristic of intracranial arterial dissection are areas of irregular stenosis, the “string sign,” the “string and pearl sign,” vessel occlusion, intimal flap, the double lumen sign, and the crescent sign (on MRI). Arterial dissections often change appearance in serial imaging studies. Traumatic pseudoaneurysms have different imaging characteristics from traditional or berry aneurysms. Intracranial pseudoaneurysms are typically found along the course of secondary or distal branches of the major arteries, have a more irregular dome contour, usually do not have a neck, and can exhibit delayed filling and emptying of the aneurysm (28).

Penetrating cervical vascular injury is evaluated by examination of the injury site in conjunction with a complete neurologic evaluation. Possible airway compromise should be assessed, and chest x-ray should be performed to rule out pneumothorax and hemothorax. A lateral cervical spine (c-spine) x-ray or c-spine CT scan allows assessment of injury trajectory and is important to rule out cervical subluxation, which can be associated with vertebral artery injury. Patients with penetrating injuries and hemorrhagic shock, airway compromise, expanding hematomas, or definite signs of significant vascular injury are typically taken straight to the operating room without preoperative imaging. In stable patients with penetrating injuries through the platysma and otherwise unexplained neurologic deficits or physical exam findings suggestive of vascular injury, further diagnostic imaging (eg, CTA) is recommended. Clinical exam sensitivity is relatively high for vascular lesions and airway compromise at 81% and 77%, respectively, but low for esophageal injury at 34%, whereas CT imaging sensitivity is 90%, 83%, and 53%, respectively (95; 63).

All patients with signs and symptoms of blunt cerebrovascular injury should be screened. Patients with major risk factors for cerebrovascular intracranial injury without contraindications to heparin therapy should be screened within 12 hours. Patients with major risk factors but with contraindications to heparin should be screened when stable (13). Disagreement exists as to what qualifies as a major risk factor and which injuries warrant screening in blunt cervical injuries without obvious vascular injury. Screening recommendations range from aggressive – screening of all patients with cervical spine injuries – to more conservative – screening for specific high-risk injuries such as foramen transversarium fractures, facet joint dislocations, or high cervical injuries (C1 to C3) (14; 58; 42). In 2002, the section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons issued a joint guideline recommending screening in all patients following blunt cervical spine trauma who have complete cervical spinal cord injuries, fracture through the foramen transversarium, facet dislocation, or vertebral body subluxation (50). The American College of Surgeons Advanced Trauma Life Support manual proposes screening for suspected traumatic vertebral artery injury in patients with C1-C3 vertebral fractures, cervical fracture subluxations, and foramen transversarium fractures (118). Bromberg and colleagues reviewed the medical literature and found no level I evidence for recommendations on screening (17). There was level II evidence for screening patients with head or neck injuries and unexplained neurologic abnormalities or with head and neck injuries with arterial bleeding. Level III evidence was identified for screening patients after significant blunt head injuries, which included Glasgow Coma Scale (GCS) scores less than or equal to 8, a petrous bone fracture, diffuse axonal injury, cervical spine fractures (C1 to C3, involving foramen transversarium or with subluxation or rotation), and Lefort II or III facial fractures (17).

Practice guidelines recommend screening with computed tomography angiography to detect blunt cerebrovascular injury in patients with blunt polytrauma (65; 72). Additional recommendations include screening adult patients with high-risk cervical spine injuries with CTA to detect blunt cerebrovascular injuries. Inclusion of head CTA in routine screening is beneficial and leads to a change in clinical management in a minority of cases (03). There is level III evidence that reliance on clinical screening alone (using expanded Denver criteria for blunt cerebrovascular injury) increases the likelihood (approximately 37%) of missing cerebrovascular injuries (52). The Eastern Association for the Surgery of Trauma 2020 practice management guideline on the evaluation and management of blunt cerebrovascular injury states that CTA is recommended for all patients with high-risk cervical spine injuries, with a conditional recommendation for CTA in low-risk injuries (65). All patients with upper cervical spine (C1-C3) fractures, subluxation, and cervical spine fractures extending into the transverse foramens should be considered as having high-risk injuries. More liberal CTA screening may be of benefit as it identifies blunt vascular injuries missed by standard risk criteria (52).

Positive imaging findings are graded to guide management and stratify stroke risk. The Denver grading scale, developed initially by Biffl and colleagues for arteriogram findings in blunt carotid artery injury, categorizes vascular injuries into 4 grades of increasing occlusion status and a fifth grade representing transections with extravasation (15). With blunt carotid artery, stroke incidence was found to increase from injury grades 1 to 4. Grading of blunt vertebral artery injury, however, evaded such correlation. Biffl and colleagues speculated that because of collateral circulation within the vertebral artery, embolic infarcts resulting from subtotal occlusive intimal injuries (grades 1 to 3) are the most likely source in blunt vertebral artery injury-related strokes.

Blunt cerebrovascular injuries graded 1 to 3 have been found to change over time. Follow-up imaging 7 to 10 days after initial injury with MR, CTA, or duplex ultrasound has been suggested to be clinically useful in deciding whether to discontinue intervention (ie, anticoagulation) in cases of interval improvement or to pursue new treatment strategies in cases of grade elevation (42; 133). However, there are no level I or III recommendations on how one should monitor response to therapy, and the only level II recommendation is that follow-up conventional cerebral angiography be performed 7 days postinjury to reduce the incidence of angiography-related complications (17). Due to the high risk of complications, conventional cerebral angiography is not recommended in children unless concurrent endovascular treatment is required or there is doubt about the diagnosis. Instead, MRA and CTA are recommended for first-line screening in children (25).

Using the Denver grading scale of blunt cerebrovascular injuries, Scott and colleagues have published the results of The Parkland Carotid and Vertebral Artery Injury Survey. Separate subgroup analyses of grade I/II carotid, grade III/IV carotid, grade I/II vertebral, and grade III/IV vertebral artery injuries have been published (113; 114; 115; 116). They found that low-grade (I/II) carotid injuries generally remained stable with follow-up imaging, and those that had radiographic worsening did not have worse clinical outcomes. High-grade (III/IV) carotid injuries carried the highest stroke risk among blunt cerebrovascular injuries at 7%. Furthermore, grade III aneurysms were found to have a high risk of persistent pseudoaneurysm formation (89%). Low-grade (I/II) vertebral artery injuries had a low rate of cerebral infarction and did not change significantly when treated with antiplatelet or anticoagulation. High-grade (III/IV) vertebral artery injuries were found to mostly remain stable on follow-up imaging. Radiological worsening did not correlate with adverse outcomes. The only cerebral infarctions seen were associated with grade IV vertebral artery injuries and 100% mortality.

Management

Treatment of vascular injuries is focused primarily on preventing the development or progression of neurologic sequelae. This is accomplished by restoring compromised blood flow and preventing (or reducing) thrombus formation and distal embolization (44). To date, there have been no prospective randomized controlled trials to guide the treatment of blunt cerebrovascular injuries.

Bromberg and associates reviewed the medical literature for evidence-based treatment recommendations for blunt cerebrovascular injury and could not identify any level I evidence (17). They reported that there was level II evidence for treating patients with intimal irregularity with less than 25% narrowing or a dissection or intramural hematoma with more than 25% narrowing with medical antithrombotic agents (anticoagulation or antiplatelet medications in patients with no contraindications). Although anticoagulation is advocated as the treatment of choice in many studies and reviews, there is no level I evidence in support of this approach; however, there is level III evidence that antiplatelet and anticoagulation therapy are similar (79; 17). There is also level 3 evidence recommending use of antithrombotic therapy (65). In a cohort of 370 patients with intracranial and extracranial carotid and vertebral artery dissection, antiplatelet and anticoagulant therapies were found to have similar rates of transient ischemic attack, new or recurrent stroke, and hemorrhage; either approach is recommended to prevent thromboembolic complications (33). A retrospective review of 301 blunt trauma patients with vertebral injury also found anticoagulant and antiplatelet agents equally effective in reducing stroke risk (118). In another study of 725 patients, Hanna and colleagues found that compared with antiplatelet agents, anticoagulants are associated with a lower stroke rate in the first 6 months postdischarge (51). One large study involving around 13,500 patients by Catapano and associates showed that treatment with acetylsalicylic acid alone was efficacious and safe in managing extracranial blunt cerebrovascular injury (24). Interestingly, isolated vertebral artery injury is associated with a very low risk of stroke, and medical treatment with antithrombotics does not have a significant impact on stroke rate (101), although, in general, antithrombotics are associated with a lower incidence of stroke and death in all blunt cerebrovascular injury categories (65; 109). On the other hand, patients with injury to multiple intracranial vessels are at much higher risk of stroke and require medical versus surgical treatment (07). The general consensus holds that anticoagulation is preferable in severe stenosis, arterial occlusion, or pseudoaneurysm, whereas antiplatelet therapy is preferable in large infarcts or local compression without cerebral ischemia (84). With respect to the timeframe of initiation of medical therapy, the highest risk of stroke is within 72 hours after injury, encouraging early initiation of therapy (20). Guidelines suggest follow-up of vascular injuries that are medically managed at approximately 1 week with discontinuation of therapy if imaging suggests resolution of findings. Otherwise, the antithrombotic treatment can be continued for 3 to 6 months if findings persist, and worsening findings may be managed by endovascular treatment (92).

With respect to endovascular stenting, there continues to be controversy in the literature, with stenting being reserved for higher-grade lesions. A study published 10-year results to assess the efficacy of anticoagulation as opposed to endovascular stent placement in the prevention of stroke following blunt cerebrovascular trauma. Though the study concluded that anticoagulation was superior in stroke risk reduction, total enrollment numbers were weak, and conclusions were based on one less stroke event in the anticoagulation group (19). There is level III evidence that patients who present with neurologic deficits and have an accessible carotid artery lesion causing significant flow abnormalities should be treated with endovascular or surgical interventions to restore flow. There is also level III evidence that pseudoaneurysms rarely resolve with observation or medical therapy, and endovascular or surgical intervention is warranted; patients put on warfarin therapy should have an international normalized ratio goal of 2 to 3 for 3 to 6 months. There is no evidence supporting an ideal partial thromboplastin time goal during heparin therapy or the length of antithrombotic therapy. Although many authors have recommended that recurrences and cases of medical “failure” be treated via surgical or endovascular interventions, no definitive studies support these recommendations (44; 17). In fact, there is level 3 evidence against the routine use of endovascular stents as an adjunct to antithrombotic therapy in grade 2 or 3 injuries (65). In a study of 168 patients initially treated medically, Ahlhelm and colleagues found that only 10% of patients eventually required endovascular treatment by their protocol (01). A retrospective registry showed benefit with stent placement in high-grade stenosis or expanding pseudoaneurysm (127).

There are multiple arguments for and against immediate anticoagulation in certain high-risk patients. Engelter and associates list clinical features that would argue against immediate anticoagulation, such as severe strokes with National Institute of Health Stroke Scale (NIHSS) scores greater than 15 (due to high risk of hemorrhagic conversion of stroke), patients who have an accompanying intracranial dissection or hemorrhage, patients for whom no brain imaging is available, and patients who have concomitant diseases, with an increased risk of bleeding (40). Nanda and associates reviewed their series of patients with both blunt and penetrating carotid artery injuries and found that in patients who had preexisting intracerebral hemorrhages, there was evidence of worsening of the hemorrhage in two of three patients on anticoagulation (93). There are no formal recommendations for or against the use of recombinant tissue plasminogen activator (rt-PA) in acute strokes secondary to arterial dissection, and it is important to note that dissections were not an exclusion criterion in the National Institute of Neurological Disorders and Stroke (NINDS) study (48).

The Cervical Artery Dissection in Stroke Study (CADISS) was completed, comparing anticoagulation versus antiplatelet agents in the medical therapy for carotid and vertebral artery dissection (23). This multicenter study is the only randomized trial comparing the two most common medical treatments. As postulated before the trial’s completion, it was underpowered, with only 250 patients enrolled instead of the over 1400 necessary (80). The primary endpoints of the trial were ipsilateral stroke or death within 3 months of randomization; secondary endpoints were major bleeding, any stroke or transient ischemic attack, and the presence of residual stenosis at 3 months. No difference in efficacy was seen between the groups, with stroke being very rare in both groups (1% to 2%). This is corroborated by the previously reported results of the nonrandomized arm of the trial (64).

Surgical intervention for extracranial vascular injuries is possible but can be difficult due to the length and location of carotid artery dissections (more distal location and longer length of lesions when compared to typical atherosclerotic carotid artery lesions). In addition, surgical approaches to the vertebral artery can be difficult, are associated with high morbidity and mortality, and are not a commonly performed operation for most surgeons. The largest surgical series for extracranial internal carotid artery dissections was 49 patients who underwent 50 carotid endarterectomies (90). All of these patients were operated on after failing conservative therapy (mean time of operation after initiation of conservative therapy was 9 months). Perioperative cranial nerve deficits were present in 58% of the patients, though most were transient, and six patients suffered perioperative strokes (one fatal). Postoperatively, severe stenosis persisted in 13 cases, and a pseudoaneurysm was identified in 27 cases. A surgical alternative to carotid endarterectomy is ligation with in situ or extracranial-intracranial bypass. Morgan and Sekhon reported their series of six cases treated with extracranial-intracranial bypass, none of which suffered a subsequent stroke (87).

Over the last 2 decades, significant advances have been made in neuroendovascular therapies. Endovascular treatment of carotid and vertebral artery dissections, including angioplasty or stent placement, has, for the most part, replaced surgical intervention for dissections and is recommended by many authors as the intervention of choice in symptomatic dissections not responding to medical therapy and in dissections with significant occlusion with or without pseudoaneurysms (11; 81; 21; 88; 60; 84). In addition, pseudoaneurysms (with or without coexisting dissections) can be successfully treated using stent with or without coiling techniques. However, there has been no well-designed trial to definitively support these endovascular recommendations. Although there is no clear consensus regarding therapy for many traumatic dissections, for patients with carotid artery dissection and embolic intracranial large vessel occlusions, endovascular thrombectomy advised by time or perfusion imaging standards is now recommended based on a meta-analysis (57). Endovascular stenting for extracranial dissection is situational and remains a level IIb recommendation. Potential indications for endovascular treatment of both spontaneous and traumatic dissections include contraindications to medical therapy, persistent symptoms despite medical therapy, enlarging pseudoaneurysm, contralateral vessel involvement, and significant ongoing ischemic symptoms secondary to high-grade arterial stenosis (86).

Kadkhodayan and associates reported on a series of 26 patients treated endovascularly for carotid artery dissections with or without pseudoaneurysms (nine traumatic, eight spontaneous, and nine iatrogenic; all but one were symptomatic) (60). They successfully reduced dissection-related stenosis from 71% to no significant stenosis in 20 of 21 patients. Mean follow-up time was 14.6 months. They had no direct procedural-related strokes, but the transient ischemic attack rate was 10.3%. Two patients were found to have vessel occlusion on follow-up imaging (one symptomatic, 22 days after procedure), but both of these occlusions occurred before the institution of clopidogrel therapy for stent placement. The 30-day death and occlusion rate was two of 29 procedures (6.9%), and there were no transient or permanent cranial nerve palsies. In patients with persistent carotid pseudoaneurysms following blunt cerebrovascular injury, medical therapy may be preferred due to the risks associated with carotid artery stenting. Specifically, patients with stents placed experienced a 45% occlusion rate compared to 5% in patients treated with antithrombotic agents alone (31).

Moon and colleagues published their experience of 10 years of endovascular treatments for carotid and vertebral artery dissections (86). Their 116 patients included spontaneous, traumatic, and iatrogenic dissections. Among their 116 patients, most were treated by stenting. They found that these interventions had a low stroke rate and overall low permanent morbidity mortality (3.4%).

Kremer and associates prospectively studied the natural history of spontaneous carotid artery dissections with persistent stenosis or pseudoaneurysms (69). They found the annual risk of ipsilateral stroke to be 0.7% and the risk of any stroke to be 1.4% in this group (all but three presented symptomatically). They compared this group to a group of patients with transient stenosis (complete recanalization or less than 50% restenosis). The transient stenosis group had an ipsilateral stroke annual risk of 0.3% and a risk of 0.6% for any stroke. Although the risk of stroke was low in both groups, which confirms that carotid artery dissection has a relatively benign long-term course even in symptomatic patients, the risk of stroke was approximately doubled in the persistent stenosis group. Therefore, the decision to intervene, even in patients with symptomatic carotid artery dissections or persistent stenosis, should not be taken lightly and should occur only after the risks, benefits, indications, and alternatives are discussed with patients or family.

Hernandez-Duran and Ogilvy reported a metaanalysis of clinical outcomes among vertebral dissections managed with various endovascular procedures, including balloon-assisted coiling, balloon occlusion, stenting, stent-assisted coiling, internal trapping, and proximal occlusion (54). Conclusions indicate no significant advantage among the subgroups; however, proximal occlusion techniques were associated with higher mortality, though the authors suggest some of this effect might be inherent to the lesion itself rather than the treatment modality.

Newer devices have also been attempted; Cohen and colleagues have described the use of a flow diverter in the treatment of traumatic vertebral artery dissection (30). Theoretical benefits of flow diverter use include reduced in-stent thrombosis and associated reduction of distal emboli. The largest series to date of craniocervical arterial dissections with flow diverters has been by Brzezicki and colleagues, who treated 13 carotid dissections in 11 patients with the Pipeline Embolization Device (22). Larger studies and long-term follow-up will be needed.

There are few studies evaluating the ideal method of treatment for cervical artery dissection in children. Antiplatelet medications are generally preferred to antithrombotics for initial medical management, and endovascular interventions are reserved for cases of medical failure (25).

For the most part, traumatic carotid-cavernous sinus fistulas and cavernous carotid artery aneurysms are not considered life-threatening, so treatment strategies must carry a lower risk of complication than the natural history of disease. Carotid-cavernous sinus fistulas are abnormal arterial communications between the cavernous carotid artery and the cavernous sinus that can occur after closed-head injuries. Endovascular obliteration is the mainstay of therapy. Depending on the type of fistula (direct or indirect) and the microanatomy, the lesion is approached via either an arterial or venous source, and obliteration techniques using balloons, liquid embolic agents (eg, onyx), coils, or stents have all been described (47; 125; 06). In general, cavernous carotid artery aneurysms are treated conservatively. Carotid artery aneurysms can become symptomatic due to rupture or progressive mass effect. Signs and symptoms that warrant treatment are visual loss, pain, diplopia, sphenoid sinus erosion, aneurysm projection into the intradural or sphenoid sinus, and acute rupture. Endovascular therapy is the most common treatment modality (balloon-assisted and stent-assisted coil embolization or covered stent placement). There has been an increased interest in using flow-diversion devices, such as the Pipeline Embolization Device, to treat such lesions (132). In patients who are not considered endovascular candidates and require interventions, surgical procedures such as ligation/occlusion with or without bypass should be considered (39).

Patients with penetrating injuries to the neck who have signs of significant vessel injury and clinical instability (hemorrhagic shock, airway compromise, expanding hematomas) are taken directly to the operating room. Surgical options include vessel reconstruction (with or without a patch graft: synthetic or autologous), direct repair of lacerations and partial vessel injuries, placement of interposition grafts for complete vascular injuries, thromboendarterectomy for occlusive thrombi, vessel ligation, surgical trapping of the injured segment, or revascularization or bypass procedures (04). A detailed discussion of these procedures is beyond the scope of this article. In patients with penetrating injuries to the neck who are more clinically stable and who have otherwise unexplained neurologic deficits or physical exam findings suggestive of vascular injury, further diagnostic imaging is needed to identify the vascular injury and its severity. The timeframe within which neuroimaging is obtained and the type of neuroimaging depends on the stability of the patient, the access and resources of the institution, and the likelihood of a need for endovascular intervention. Once identified, treating a penetrating vascular injury in a stable patient is similar to that described earlier and is primarily medical or endovascular. Although it is highly institutional and physician-dependent, surgical interventions (in stable patients), in general, are reserved for those cerebrovascular injuries that are not amenable to endovascular therapy and those patients who are not candidates for long-term antithrombotic therapy, which is required after endovascular stent placement.

Special considerations

Pregnancy

Medical and therapeutic management of patients with pregnancy-related stroke symptoms is largely the same as that of nonpregnant patients, except that more considerations are given to maternal and fetal risks associated with medical and therapeutic interventions. Guidelines for stroke prevention in pregnancy stress the control of hypertension to reduce the long-term risks of stroke (83). Some antithrombotic therapies and therapeutic interventions can be performed fairly safely during pregnancy. It is critical that the risks and benefits of each of these medications and interventions and their potential impact on the pregnant women and fetus(es) be considered cautiously and be discussed in detail by both the treating physician and obstetrician with the patient (or patient’s family) before initiation of any therapy (10; 124).

Anesthesia

Anesthetic management significantly affects the prognosis of the patient who requires neurovascular interventions (surgically or endovascularly) on either the brain or cervical arteries. Several preoperative risk factors must be considered before proceeding with surgery to optimize clinical outcomes. A study suggests that carotid endarterectomy cases have no significant difference in overall outcome when performed under local versus general anesthesia (126). Advances in neuromonitoring during therapeutic interventions (including EEG, somatosensory evoked potentials, brainstem auditory evoked potentials, intracranial pressure, and cerebral blood flow) and the ability to perform some procedures with an “awake patient” can lead to improved neurologic outcomes following these procedures. Neuroprotective agents (eg, barbiturates, etomidate, and propofol) and hypothermia can reduce cerebral neuronal metabolism, and cytoprotective agents can also contribute to cerebral protection; both can extend neuronal tolerance to cerebral ischemia when given before periods of temporary arterial occlusion. Endovascular balloon test occlusion in “awake patients” with appropriate neuromonitoring can also yield substantial information, which can help reduce postintervention deficits and help plan the ideal therapeutic approach (117; 128).

Pediatric

Special considerations in the pediatric population that increase the risk of blunt cerebrovascular injury include age-related factors such as head-to-body ratio, immature neck vasculature, and ligamentous laxity that can lead to injury. Compared to adults, the lower incidence of arteriosclerotic changes is potentially protective (134). Based on a 2019 international multicenter analysis led by Weber and colleagues, who aimed to characterize blunt cerebrovascular injury in the pediatric population, 42 of 8128 (0.5%) pediatric (0-17 years old) trauma patients were identified with a blunt cerebrovascular injury (131). They found that the incidence of blunt cerebrovascular injury was related to severe and high-energy traumas, carotid artery injury was more common than vertebral artery injury, and there is generally less data for pediatric blunt cerebrovascular injury than for adults. Predictive factors for blunt cerebrovascular injury in children included cervical spine injuries, facial injuries, and basilar skull fractures. CT imaging (head/neck vs. whole body) is part of the diagnostic workup for the pediatric trauma population.

In a review, Galardi and associates studied injuries that result from blunt trauma in the pediatric population, stating that ischemic and hemorrhagic stroke can occur in this setting, especially within the first 2 weeks after trauma (46). Traumatic lesions (arterial, venous, and capillary lesions) can lead to ischemic and hemorrhagic stroke in various pathways, as outlined in their paper. They emphasized provider awareness of stroke as sequelae for pediatric patients with blunt cerebrovascular injury, with daily examinations that could identify stroke early in this population in addition to how stroke management plays into the context of other concomitant injuries and management (46).

Dunn and Burjonrappa studied the incidence, management (including surgical and multimodality), and outcomes in pediatric cerebrovascular trauma (38). This review encompassed both blunt and penetrating cerebrovascular injury in children. Their work reports an incidence rate of 0.4% and overall stroke rate of 3.1%, and penetrating injury was associated with a higher mortality rate compared to blunt injury; furthermore, mortality rate increases with multivessel injury.