Vertebral Artery Injury in Cervical Trauma
Overview & Epidemiology
Vertebral artery injury (VAI) is one half of blunt cerebrovascular injury (BCVI), blunt trauma to the carotid or vertebral arteries. It matters because it is a potentially preventable cause of posterior-circulation (vertebrobasilar) stroke: the diagnosis is made on imaging, not symptoms, and early antithrombotic therapy markedly reduces both stroke and death.
The stroke risk, read correctly. The widely quoted 10–40% untreated stroke risk is for BCVI as a whole, carotid and vertebral pooled, and carotid injury carries the heavier share. Isolated vertebral injury does better, because of the collateral circulation described below: in a 156-patient isolated-VAI series only three strokes occurred. The pooled figure must not be presented as the risk facing a patient with an isolated vertebral injury.
The timing does not change with the vessel. Stroke risk is highest in the first 7 days and peaks within the first 24 hours, and in the largest isolated vertebral series every stroke had already happened before treatment could start. That is why screening is urgent even though the absolute risk for an isolated vertebral injury is lower than the pooled number suggests.
Pathophysiology & Anatomy
The four segments. V1 (pre-foraminal) runs from the subclavian to C6. V2 (foraminal) ascends through the transverse foramina from C6 to C2. V3 (extraspinal) runs from C2, looping around the atlas to pierce the dura, and V4 (intradural) joins its fellow to form the basilar artery.
Why V2 and V3 are vulnerable. V2 runs a bony, enclosed course through the transverse foramina, and the V3 loop at the craniocervical junction is mobile. A fracture that breaches the foramen transversarium, a facet dislocation or subluxation, a distraction injury or an upper cervical fracture can stretch, compress or tear the vessel, causing intimal injury, dissection, thrombosis, pseudoaneurysm or occlusion.
Why one vertebral artery can usually be lost. The two vertebral arteries unite to form the single basilar artery, and the posterior communicating arteries link the posterior (vertebrobasilar) circulation to the anterior (carotid) circulation through the Circle of Willis. If one vertebral artery occludes, the contralateral artery and the carotid system can still perfuse the territory, which is why an isolated unilateral VAI often does not cause a stroke.
When that tolerance fails. The tolerance is conditional, and the exceptions are what make an injury dangerous. The injured vessel may be the dominant vertebral artery: one VA, often the left, is usually larger, and the contralateral may be hypoplastic or end in PICA without reaching the basilar. Both vertebral arteries may be injured. Or thrombus may embolise distally into the basilar/PICA territory rather than simply occluding locally.
An isolated, non-dominant grade IV occlusion is therefore often tolerated on aspirin, whereas a dominant-artery or bilateral injury, or any embolic propagation, is far more dangerous. Antithrombotics prevent embolism, so they matter even when flow looks preserved.
Clinical Presentation
Silent at first. More than half of patients have no neurological signs at presentation, so the injury is found on screening imaging rather than on examination. A normal neurological examination does not exclude VAI, and the window to prevent the stroke is in the first hours.
When it declares. VAI produces a posterior-circulation (vertebrobasilar) stroke or TIA: vertigo, ataxia, diplopia, dysarthria, visual-field loss or a depressed conscious level, often a day or more after the injury. The local clues are neck pain, a cervical bruit or thrill, or an expanding neck haematoma.
The strokes to recognise. When VAI does cause a posterior-circulation infarct, expect a recognisable pattern rather than vague "dizziness":
- Lateral medullary (Wallenberg) syndrome, the classic VA/PICA-territory stroke. Ipsilateral facial pain and temperature loss, Horner's syndrome, ataxia, vertigo/nystagmus and bulbar signs (dysphagia, hoarseness), with contralateral body pain and temperature loss: a "crossed" sensory pattern.
- Basilar artery occlusion, the catastrophic end. Fluctuating or reduced consciousness, quadriparesis and cranial-nerve signs, the locked-in syndrome: a time-critical neurological emergency.
- Cerebellar infarction. Vertigo, ataxia and vomiting. The danger is that a large cerebellar infarct swells, compressing the fourth ventricle and brainstem and causing obstructive hydrocephalus, which can require suboccipital decompressive craniectomy with or without an external ventricular drain.
Investigations: Screening & Grading
Screen by pattern. Because BCVI is frequently clinically silent at first, screening is based on injury patterns and risk factors (modified Denver/Memphis-type criteria). Detection rises sharply when a screening protocol is used (odds ratio ~4.7), and is far higher among high-risk cervical injuries than low-risk ones (odds ratio ~12.7). Screen for VAI/BCVI in patients with:
- Cervical spine fractures, especially through the foramen transversarium, C1–C3 fractures, facet (sub)luxation/dislocation or fracture-subluxation, and distraction injuries
- Severe head injury (low GCS), basilar skull fracture, Le Fort II/III facial fractures
- Cervical bruit or thrill, expanding neck haematoma, seatbelt sign over the neck, or near-hanging/strangulation
- Focal neurology unexplained by brain imaging, or a stroke/TIA pattern
The pragmatic rule. Screen any patient whose injuries would otherwise prompt a CT of the neck or chest, which is also the EAST practice guideline position. Liberal screening catches the silent majority.
Which test. CT angiography (CTA) of the neck is the screening modality of choice: widely available and fast. Digital subtraction angiography (DSA) remains the reference standard and is used in selected or equivocal cases, or for intervention. Follow-up imaging is often used to track how a lesion evolves.
Grading. Injuries are graded by the Biffl/Denver scale, which guides treatment and prognosis.
- Lesion
- Intimal irregularity / dissection with under 25% luminal narrowing
- Lesion
- Dissection or intramural haematoma with 25% or more narrowing, intraluminal thrombus, or a raised intimal flap
- Lesion
- Pseudoaneurysm
- Lesion
- Complete occlusion
- Lesion
- Transection with active extravasation (or AV fistula in some schemes)
Management
Antithrombotics are the mainstay. The goal is stroke prevention. Aspirin or therapeutic anticoagulation (heparin) has proven safety in trauma patients and reduces both stroke and mortality (EAST: stroke odds ratio 0.20, mortality odds ratio 0.17). Treatment is guided by the grade of injury and started as early as the overall injury burden safely allows, given the early stroke window.
Choosing the agent and the timing. Both are individualised by balancing stroke risk against bleeding risk from a concomitant TBI, solid-organ injury or planned surgery. For isolated vertebral artery injury the stroke rate is low and aspirin is generally effective, particularly for grades I and IV.
Endovascular and surgical treatment. These are reserved for selected lesions: an enlarging or symptomatic pseudoaneurysm (grade III), or a grade V transection, which needs endovascular (stent, coil/occlusion) or surgical vessel control, often by occluding the vessel. The EAST guideline recommends against routine stenting as an adjunct to antithrombotics for grade II–III injuries.
- Typical management
- Antithrombotic (often aspirin); usually resolves
- Typical management
- Antithrombotic; follow-up imaging (can progress)
- Typical management
- Antithrombotic; endovascular if enlarging/symptomatic
- Typical management
- Antithrombotic (aspirin); monitor for posterior-circulation stroke
- Typical management
- Endovascular/surgical control (often vessel occlusion)
In a high-risk cervical injury (facet dislocation, foramen-transversarium fracture), closed reduction or operative manipulation can propagate a dissection. Screen with CTA where feasible and weigh that small risk, without unduly delaying urgent reduction of a cord-threatening injury, and coordinate antithrombotic timing with the spinal surgical plan and any associated injuries.
Complications & Prognosis
Stroke. Posterior-circulation stroke is the dominant complication; its risk and its timing are set out in the overview.
Lesion progression. Grade II dissections and haematomas can progress despite therapy, and grade III pseudoaneurysms may enlarge, which is why grade II and III lesions are followed up with imaging.
Prognosis by grade. A grade IV occlusion does not recanalise early, but isolated VAI still has a low stroke rate on aspirin. Grade V is the most dangerous.
Iatrogenic Injury: The Vertebral Artery in Cervical Spine Surgery
Every mechanism above is something that happens to your patient before they reach you. This one is the mechanism you control. The artery a foramen-transversarium fracture threatens is the same artery a C2 screw threatens, and the anatomy that makes it vulnerable is variant in roughly a third of people on the most inclusive definition, and invisible without a preoperative CT.
The high-riding vertebral artery. Its transverse foramen sits so far medial and cranial within the C2 lateral mass that the bone corridor for a pars or transarticular screw is too thin to take one safely. It is defined on the preoperative CT by an isthmus thickness under 5 mm or an internal height of the isthmus under 2 mm. On a series of 100 consecutive cervical CTs it was present in 32% of patients, 13 left-sided, 9 right-sided and 10 bilateral, and it is frequently unrecognised because nobody measured.
Why other sources say 10%, or 18 to 23%. The figures are not in conflict; they measure different things, and the definition above is the most inclusive in use. Mandel's cadaveric study of 205 C2 vertebrae applies a single 5 mm floor to isthmus height and width and puts about 10% at risk. Paramore reconstructed CT along the planned screw trajectory and judged 18 to 23% unsuitable on at least one side. Adding the internal-height criterion captures the shallow-but-wide isthmus the other two miss and reaches 32%: a wider net, not a more dangerous population. Whenever you quote a number, say which criterion produced it, and see /topics/occipitocervical-fusion for how the choice changes the construct you plan.
The arcuate foramen. The same study found a fully formed arcuate foramen (ponticulus posticus) in 14%, with a further 24% partially formed. The ponticulus can be mistaken on exposure for a broad C1 posterior arch, and a C1 lateral mass screw started on what is actually the roof of the arcuate foramen enters the artery. The two anomalies coexist on the same side in about 5%.
- What it threatens
- C2 pars screw and, above all, C1-C2 transarticular screw - the corridor passes directly over the artery
- What to do instead
- C2 pedicle screw if the pedicle itself is adequate; otherwise a C2 laminar (translaminar) screw, which avoids the artery altogether, or a unilateral construct with the screw on the safe side only
- What it threatens
- C1 lateral mass screw started too superiorly - the bony ring is mistaken for arch
- What to do instead
- Identify the true posterior arch and start the screw on the lateral mass proper; do not judge the entry point by feel alone once a ponticulus is known to be present
- What it threatens
- Any bilateral transarticular construct
- What to do instead
- Translaminar screws bilaterally, or an alternative fixation strategy - accept that both sides are unavailable rather than taking one 'carefully'
- What it threatens
- Raises the consequence of any injury from tolerable to potentially fatal
- What to do instead
- Weigh the whole construct again; an injury to a dominant or solitary vessel is the scenario in which posterior-circulation infarction actually occurs
Exposing C1 posteriorly. The V3 segment leaves the C2 transverse foramen, runs upwards and laterally to the C1 transverse foramen, then turns medially across the superior surface of the posterior arch of C1 in a groove before piercing the dura. That horizontal course is what a subperiosteal exposure of the C1 arch runs into.
How far laterally you can go. A cadaveric study measuring the landmarks directly found the distance from the posterior tubercle of C1 to the groove cradling the vertebral artery to be 19.1 ± 3.8 mm. Taking the lower end of that spread, staying within roughly 15 mm of the midline on the C1 arch is the practical rule. The variability is the point: a standard deviation of nearly 4 mm means the artery is appreciably closer to the midline in some patients than the average suggests.
Blind dissection is what injures the artery. That is the authors' more important warning. Work subperiosteally, stay on bone, and do not sweep laterally into the soft tissue above the arch looking for a plane.
Do not chase it. Blind attempts to clip or cauterise in a bleeding field injure the artery further, and may injure the other structures around it.
- Tamponade immediately: direct pressure with haemostatic packing controls almost all of these.
- Do not explore the vessel blindly, and do not attempt bipolar coagulation of a vessel you cannot see.
- Complete or abandon the instrumentation deliberately. A construct left half-finished in a patient who now needs angiography is its own problem. Where a screw on the opposite side was planned, reconsider it: bilateral vertebral artery injury is the situation to avoid at all costs.
- Postoperative angiography to define the lesion, grade it and decide on antithrombotic therapy, the same decision tree as for a blunt injury, complicated by fresh surgical haemostasis.
- Wake and assess neurologically as a priority, and be alert for delayed posterior-circulation signs.
Reported vertebral artery injury during C1-C2 transarticular screw placement is 1.3-4.1%, and a high-riding artery is the commonest setting. Preoperative CT assessment of the isthmus is how that number is kept low.
Viva practice
Practise clinical reasoning and management decisions out loud
“A patient with a cervical facet dislocation and a fracture through the foramen transversarium is neurologically intact. Why are you concerned about the vertebral artery, who would you screen, and how?”
“CTA confirms a grade II vertebral artery dissection with 30% narrowing. How would you manage it, and how does treatment vary by grade? What if you also need to reduce the cervical injury?”
Anatomy & risk
- V1 pre-foraminal, V2 foraminal (transverse foramina), V3 atlas loop, V4 intradural to basilar
- V2/V3 most at risk; injures posterior (vertebrobasilar) circulation
- At-risk injuries: foramen transversarium fracture, facet (sub)luxation, distraction, C1-C3 fracture
Why it matters
- Stroke in 10-40% if untreated - but that is BCVI overall; isolated vertebral injury is lower
- More than half SILENT initially; stroke risk peaks first 24h (highest first 7 days)
- Screen early - don't wait for symptoms
Screen & grade
- CTA neck = screening modality of choice (DSA reference standard in select cases)
- Screen Denver-type criteria + anyone needing CT neck/chest
- Biffl/Denver grades I-V guide treatment
Management
- Antithrombotics (aspirin or anticoagulation) = mainstay, safe in trauma; grade-guided
- Isolated VAI: low stroke rate, aspirin effective (esp. I, IV); follow-up imaging for II
- Endovascular for enlarging pseudoaneurysm (III)/grade V; not routine stenting II-III; consider vessel before cervical reduction
Evidence
Blunt carotid arterial injuries: implications of a new grading scale
- Derived the Biffl/Denver I-V grading scale from 76 patients with 109 blunt arterial injuries, with prognostic and therapeutic implications.
- Grade-specific behaviour: two-thirds of grade I healed regardless of therapy; grade II progressed in ~70% despite heparin; only 8% of grade III pseudoaneurysms healed with heparin but 89% resolved after stenting; grade IV occlusions did not recanalise; grade V transections were lethal.
- Stroke risk increased with injury grade - the basis for grade-guided treatment.
Evaluation and management of blunt cerebrovascular injury: A practice management guideline from the Eastern Association for the Surgery of Trauma
- GRADE-based systematic review/meta-analysis (23 studies): a screening protocol increased BCVI detection (OR 4.74), and high-risk cervical injuries had far higher detection than low-risk (OR 12.7).
- Antithrombotic therapy reduced stroke (OR 0.20) and mortality (OR 0.17) versus no antithrombotic therapy.
- Recommends screening protocols, CTA for high-risk cervical injuries, antithrombotic therapy for diagnosed BCVI, and AGAINST routine stenting as an adjunct to antithrombotics for grade II-III injuries.
Management of Blunt Cerebrovascular Injury
- Untreated BCVI causes stroke in 10-40% of patients, but more than half do not present with stroke symptoms initially; stroke risk is highest in the first 7 days (peak in the first 24 hours).
- CT angiography is the screening modality of choice (DSA in selected cases); screen all patients with injuries that would otherwise prompt CT of the neck or chest.
- Antithrombotic therapy is the mainstay and is safe in trauma patients; endovascular intervention benefits selected patients; treatment is guided by injury grade.
Blunt Traumatic Vertebral Artery Injuries: Incidence, Therapeutic Management, and Outcomes
- 156 isolated blunt vertebral artery injuries: most patients (135/156) were treated with aspirin alone; the risk of stroke after cervical vertebral artery injury was low.
- Aspirin prophylaxis was efficacious in grade I and grade IV injuries; data are limited for grades II and III.
- The three strokes that occurred were detected within 24 hours of admission, before treatment was started - reinforcing early screening and treatment.
Ipsilateral arcuate foramen and high-riding vertebral artery: implication on C1-C2 instrumentation
- 100 consecutive cervical CT scans. A high-riding vertebral artery - isthmus thickness under 5 mm or internal isthmus height under 2 mm - was present in 32% of patients: 13 left, 9 right and 10 BILATERAL.
- A fully formed arcuate foramen (ponticulus posticus) was present in 14%, with a further 24% partially formed.
- Ipsilateral arcuate foramen AND high-riding vertebral artery coexisted in 5%.
- The authors' conclusion is the practical one: both anomalies are common and are often not recognised, and identifying them on the preoperative CT is what allows the safest instrumentation technique to be chosen.
A review of complications associated with craniocervical fusion surgery
- 22 reports covering 2,274 craniocervical and upper cervical fusion procedures analysed for complications.
- Vertebral artery injury occurred in 1.3% to 4.1% of C1-C2 transarticular screw placements, MOST COMMONLY in the setting of a high-riding vertebral artery.
- The commonest complications overall were instrumentation failure after nonunion - up to 7% for occipitocervical and 6.7% for atlantoaxial fusion - alongside dural tears and wound infection.
- The authors attribute safety to preoperative assessment of anatomic variants and preparation for perioperative management of complications.
Simple identification of the third segment of the extracranial vertebral artery (ELITE approach)
- 16 cadaver heads dissected bilaterally with silicone-injected vessels, measuring the bony landmarks that locate the V3 segment.
- The distance from the posterior tubercle of C1 to the groove cradling the vertebral artery on the posterior arch was 19.1 plus or minus 3.8 mm.
- The V3 segment was identified successfully in every specimen using bony landmarks alone, and the artery was injured in none.
- The authors state the hazard plainly: blind dissection below the level of the foramen magnum can injure the vertebral artery.
The I-V grading scale comes from Biffl et al. 1999 (DOI); the screening and antithrombotic recommendations (and the detection/stroke/mortality odds ratios) from the EAST practice guideline (Kim et al. 2020, DOI); the stroke-risk figures and CTA-screening synthesis from Stone et al. 2018 (DOI); and the isolated-vertebral-artery outcomes from Zeineddine et al. 2022 (DOI). Note the one tension worth holding in mind: the 10-40% untreated stroke figure is BCVI-wide, pooling carotid with vertebral, while the isolated-vertebral series is markedly more reassuring. Both are on this page, and the difference is the vessel, not a disagreement. For the iatrogenic material, the prevalence of high-riding vertebral artery and arcuate foramen is from Elgafy et al. 2014 (DOI), the transarticular-screw injury rate from Lall et al. 2010 (DOI), and the C1 exposure distances from Wanibuchi et al. 2009 (DOI). The segmental vertebral-artery anatomy is standard, well-established reference. (See also our cervical-trauma topics.)