Cervical spine traumaRAD Magazine, 42, 499, 21-22

Richard PullicinoNeuroradiology fellow

Mark RadonConsultant neuroradiologist

The Walton Centre, Liverpool

Background and epidemiologyCervical spine injury is the second most commoninjury to the spine, comprising 28.6% of spinalinjuries1 and 55% of spinal cord injuries. C2 isthe most common fractured vertebra, accountingfor 24% of injuries.2 However, despite measuresto detect spinal injuries, delayed diagnosis or sub-optimal treatment remain common, with 3-25%of patients developing neurological complications.3

C-spine injury is less common in children andrequires a more cautious imaging strategy.

Computed tomography (CT) is currently themainstay of imaging in adult cervical spinetrauma. Radiography has a limited role due to ahigh inadequate study rate and poor sensitivitycompared with CT (39-52% vs 98%).4

Anatomy and biomechanics of the cervical spineThe neurocranium accommodates multiple sensory organs,notably vision and balance. These must be oriented in adirection of interest and this requires considerable flexibilityand stability to be provided by the components of thecervical spine.

There are numerous ligaments that support the cervicalspine and craniocervical junction. Figure 1 illustrates themost important ligamentous structures visible on the mid-sagittal section. The alar and transverse ligaments arevisible on parasagittal sections.

Assessing the stability of the cervical spineIn a simple definition, stability of the spine is the ability ofvertebrae to maintain their relationship and limit their rel-ative movements during physiological postures or loads.Various methods have been developed to assess stability inthe subaxial (C3 and caudal) spine based on radiography(Denis), mechanism (Allen-Ferguson) and CT/MRI (SLICsystem).

Francis Denis developed the three-column spine model asa method to determine stability of spinal injury by radio-logical imaging, dividing the spine into three columns5 (table1). In general, if two or more of these columns are damagedthen the spine is considered unstable. Although developedfor thoracolumbar spine injuries, the Denis model is ade-quate in many cases of subaxial cervical trauma.

The SLIC system uses a composite score based on mor-phology of the fracture, the state of the discoligamentouscomplex and the neurological status of the patient, whichcan be used to advise on surgical or conservativemanagement.6

Clinical algorithms for imagingThe NEXUS study in 1998 developed a set of rules whichaimed to identify those at minimal risk of serious injury forwhom imaging could be omitted. The study identified fiverisk factors (focal neurological deficit; midline spinal tender-

ness; altered level of consciousness; intoxication; and dis-tracting injury). If all were absent, then radiography wasnot necessary.

Canadian C-Spine Rules In 2001, the Canadian C-Spine Rules (CCR) were introducedand use both high and low risk factors. The high-risk mark-ers are first checked and if any are present then imagingis recommended immediately. Otherwise, low-risk markersare checked to see whether a safe assessment of range ofmovement can be performed. If all low-risk markers areabsent, imaging is performed. If one or more are present,then the patient is assessed for ability to actively rotate theneck 45 degrees left and right, with imaging performed ifactive motion is not possible.7

The Canadian C-Spine Rules are more sensitive than theNEXUS criteria8 and reduce the need for imaging by 44%(versus 33% for NEXUS). The CCR are the specified algo-rithm in the most recent NICE guidance.

NICE guidelines The latest NICE guidelines (www.nice.org.uk/guidance/ng41)issued in February 2016 provide changed imaging advice.In adults, CT should be performed as first line imaging, andas part of whole body CT in a major trauma scenario. MRIis performed if, after CT, there are signs and symptoms sug-gestive of spinal cord injury. In children MR imaging is thefirst choice imaging modality to be used if there is strongclinical suspicion. X-rays are recommended in children whodo not satisfy the clinical criteria for MR imaging but sus-picion remains after repeated assessment.

Patterns of injury Craniocervical junction injuriesCraniocervical dissociationCraniocervical dissociation (CCD) is the disruption of thecraniocervical junction and its ligamentous complex, dueusually to hyperextension or hyperflexion-distraction injuries.While rare, it carries high mortality and morbidity.9 Survivalrates in children are better. As primarily a ligamentousinjury, approaches based on measurement have been devel-oped to aid diagnosis (figure 2). The Power’s ratio is a use-ful way to assess for this type of injury. In normalindividuals the ratio of BC:OA should be <1. If >1 then thisis suggestive of atlanto-occipital dislocation since the atlasand axis move posteriorly in relation to the cranium.

Alternatives include the basion-dental interval (BD in figure 2) and basion-axial interval, both of which should be<12mm. However, the normal range may be different on CT,and the reliability of the basion-axial interval is question-able.10

Occipital condyle fracturesAnderson and Montesano described a classification of theserare injuries comprising three types11 (table 2). The type IIIfracture should be recognised as it is potentially unstable(figure 3).

C1 fracturesThese fractures are usually classified into three types basedon morphology (table 3). The atlanto-dens interval (ADI)can give an approximate indication of the ligamentousintegrity. ADI is usually <3mm in adults and <5mm in chil-dren. In adults, a measure between 3-5mm, suggests trans-verse ligament disruption, but preservation of the alar andapical ligaments. If the ADI is >5mm the alar and apicalligaments have also been disrupted. On coronal imaging,misalignment of the lateral masses with C2 may also bepresent.

C2 fracturesThe Anderson and D’Alonzo classification, based on fracturesite is widely used12 (table 4). The majority are type II frac-tures which are unstable. They most commonly occur in theelderly and have a high rate of non-union. Type I and typeIII fractures are uncommon but are less likely to beunstable.

Subaxial cervical spinal injuriesExtension teardrop fracture The term “teardrop” is a description of an anterior-inferiorcorner fracture of a vertebral body. This can cause confusionas it can be seen in two distinct injuries. In the context ofsevere hyperextension, as may occur in sudden decelerationin RTAs, the teardrop represents an avulsion of the anteriorlongitudinal ligament from the vertebral body (figure 5).

Flexion teardrop fractureIn severe hyperflexion and compression, a similar shapedfracture fragment may result. Often, there is retropulsionor collapse of the vertebral body. This feature is not usuallypresent in extension teardrop fractures and can cause ante-rior cord injury. Three-column ligamentous injury can result,and commonly there is injury to the posterior ligaments, vis-ible as widening of the interspinous spaces and facet joints.13

Facet joint fracture dislocationThe facet joints are uniformly superimposed and stronglysupported by multiple ligaments and the joint capsule. Inthe case of a flexion-rotation injury, there can be disruptionof the posterior ligament complex and the unilateral sub-luxation of a superior facet over the inferior facet of theinvolved joint, with or without associated fracture. The sta-bilising effect of the ligamentous complex can result inrealignment of the spine, even in the context of significantinjury (figure 6).

Bilateral facet dislocation is a very unstable injury andusually results from hyperflexion resulting in disruption ofthe ligamentous complexes of all three columns. The facetjoints become locked and there is anterior vertebral bodydisplacement resulting in severe narrowing of the spinalcanal and therefore severe neurological sequelae.

Special circumstancesIn ankylosing spondylitis and similar conditions, ossificationof numerous ligaments occurs. The excess rigidity of thespine impairs energy absorption and, as the bone is oftenweakened, fractures may occur even with minimal trauma.Further, these fractures tend to be highly unstable, andtreatment is difficult. An extremely low threshold for imag-ing should be applied in these patients.

Vascular trauma can occur up to around 13% of patientswith blunt head and neck trauma.14 This is more commonin patients with facet subluxation or dislocation and in thosewho have fracture lines that involve the foramen transver-sarium or internal carotid canal. This can be easily assessedwith CT angiography.

ConclusionCervical spine trauma is common and an important causeof morbidity and mortality. In adults, CT should be themainstay of imaging the acute phase, with MRI used to clar-ify the extent of injury, especially when there is neurologicaldeficit. There are anatomical differences between the cranio-cervical junction and the subaxial cervical spine which leadto different patterns of injury with different prognosis.

References1, Hasler R M et al. Epidemiology and predictors of spinal injury in adult

major trauma patients: European cohort study. Eur Spine J 2011;20:2174-80.

2, Hoffman J R, Wolfson A B, Todd K, Mower W R. Selective cervical spineradiography in blunt trauma: Methodology of the national emergencyx-radiography utilization study (NEXUS). Ann Emerg Med 1998;32:461-69.

3, Sciubba D M, Petteys R J. Evaluation of blunt cervical spine injury. SouthMed J 2009;102:823-28.

4, Brohi K et al. Helical computed tomographic scanning for the evaluationof the cervical spine in the unconscious, intubated trauma patient. JTrauma 2005;58:897-901.

5, Denis F. Spinal instability as defined by the three-column spine conceptin acute spinal trauma. Clin Orthop Relat Res 1984;65-76.doi:10.1097/00003086-198410000-00008

6, Vaccaro A R et al. The subaxial cervical spine injury classification system:A novel approach to recognize the importance of morphology, neurology,and integrity of the disco-ligamentous complex. Spine (Phila Pa 1976).2007;32:2365-74.

7, Stiell I G et al. The Canadian C-spine rule for radiography in alert andstable trauma patients. JAMA 2001;286:1841-48.

8, Stiell I G et al. The Canadian C-spine rule versus the NEXUS low-riskcriteria in patients with trauma. N Engl J Med 2003;349:2510-18.

9, Mueller F J et al. Incidence and outcome of atlanto-occipital dissociationat a level 1 trauma centre: a prospective study of five cases within 5 years.Eur Spine J 2013;22:65-71.

10, Rojas C A, Bertozzi J C, Martinez C R, Whitlow J. Reassessment of thecraniocervical junction: Normal values on CT. Am J Neuroradiol2007;28:1819-23.

11, Anderson P S, Montesano XP. Morphology and treatment of occipitalcondyle fractures. Spine 1988;13:731-836.

12, Anderson L D, D’Alonzo R T. Fractures of the odontoid process of the axis.J Bone Joint Surg Am 1974;56:1663-74.

13, Kim K S, Chen H H, Russell E J, Rogers L F. Flexion teardrop fractureof the cervial spine: Radiographic characteristics. Am J Roentgenol1989;152:319-26.

14, Delgado Almandoz J E et al. Multidetector CT angiography in the evalu-ation of acute blunt head and neck trauma: A proposed acute craniocervicaltrauma scoring system. Neuroradiology 2010;254:236-44.

TABLE 2Anderson and Montesano classification of occipitalcondyle fractures.

Type I Comminution of the occipital condyle due to compres-sion between the atlanto-odontoid joint. Stable injury

Type IILinear skull fracture that extends into the occipitalcondyles mostly because of a direct blow to the skull.Stable injury

Type IIIAvulsion fracture of the condyle at the insertion of thealar ligament (forced contralateral bending). Unstabledue to craniocervical ligamentous disruption (figure 3).

TABLE 3Classification of C1 fractures.

Type I Isolated fracture of the anterior or posterior arch.

Type IIBilateral fracture of the anterior AND posterior arches(Jefferson Burst fracture). Stability can be determined bythe integrity of the transverse ligament.

Type III Fracture of a unilateral C1 mass. Again stability isdetermined by the integrity of the transverse ligament.

TABLE 4Anderson and D’Alonzo classification of C2 fractures.

Type I Fracture at the apex of the odontoid peg.

Type II Fracture through the base of the odontoid and body ofthe axis (unstable).

Type III Fracture through the body of the axis.

TABLE 1Denis three column model and its components.

Anterior columnAnterior longitudinal ligamentAnterior annulus fibrosusAnterior half of the vertebral body

Middle columnPosterior annulus fibrosusPosterior half of the vertebral bodyPosterior longitudinal ligaments

Posterior column

Facet jointLigamentum flavumInterspinous ligamentSupraspinous ligament

B

Figure 1(A) schematic drawing of ligamentsand relations of the craniocervicaljunction; (B) typical appearances onMRI.

Basion

Opisthion

C2

Superior crus oftransverse ligament

Posterior atlanto-occipital

membraneAnterior atlantooccipital membrane

Anterior atlanto-dental ligament

Posterior longitudinalligament

Anterior longitudinalligament

Anterior arch of G1

Apical ligament

A

Figure 2Measurements at the craniocervical junction. (A)anterior arch of C1, (B) basion, (C) posterior arch ofC1, (D) tip of odontoid, (O) opisthion.

Figure 3Type III occipital condyle fracture (arrow) indicat-ing avulsion of the alar ligament.

Figure 4Sagittal image showing the typical location of thefracture line in a type II odontoid fracture. In thiscase alignment is preserved, but this fracture isfrequently unstable.

Figure 5Extension teardrop fracture. Sagittal CT (A) andMRI STIR (B) images. Disruption of the anteriorlongitudinal ligament (arrow) is visible on MRI, butthere is no subluxation or disruption of the inter-spinous ligaments. This fracture is distinguishedfrom the more serious flexion teardrop injury byabsence of widening of the interspinous distancesor interspinous ligamentous oedema. Figure 6

Unifacet fracture. The CT (A-C) shows only a smallfracture of the right C6 superior articular facetwith normal alignment. However, the MRI showsthat the injury is more extensive, including the disc.A defect in the ligamentum flavum is also visible,together with some interspinous oedema. The spinalcord is also compressed by haematoma (arrow). TheDenis three column model can be misleading incases such as this.

A B

A B

C D