The diagnosis of craniocervical injury should be suspected in all blunt trauma patients. After resuscitation, patients are evaluated for spinal tenderness and by neurologic examination. Because cranial trauma and other injuries are common, distracting injury or impaired consciousness preclude adequate evaluation in many patients. Patients who have spinal symptoms, who are obtunded, or who have distracting injuries should undergo radiographic evaluation. If CT is being performed, then the inclusion of cervical spine CT is timely and cost-effective. Further, the diagnostic sensitivity (especially at the craniocervical junction) is significantly better than plain radiographs. However, the potential for harmful effects of radiation from diagnostic CT should be considered, especially in children. As an alternative, a three-view cervical spine series can be performed.

The physical examination is critical and is performed by palpating the spine from the occiput to sacrum for tenderness. The patient should then be rotated and the neck and back visually inspected. Pain and tenderness in the upper cervical spine will increase likelihood of a craniocervical injury, although it may not be present due to cognitive impairment of the patient. Neurologic examination should include reflexes, level of consciousness (Glasgow Coma Score), the cranial nerves, and motor and sensory function of all four extremities and perineum.

Occipital condyle fractures are rarely visualized on plain radiographs and almost always diagnosed from head or cervical spine CT. Reformations of the occipitoatlantal articulation in coronal and sagittal planes is essential to determine their stability. Displacement indicates rupture of the internal craniocervical ligaments and instability.

Several methods have been described to identify craniocervical injury. Powers described calculating the ratio of distances from the basion to dorsal arch of C1 and opisthion to the ventral arch of the C1 (Fig. 45.2A and 45.2) (11). This ratio should be less than 1.0 and ratios greater than 1.0 are associated with atlanto-occipital dislocation. Harris described the “Rule of Twelve” based on intervals between the basion to the dens (BDI) and the basion to the posterior axial line (Fig. 45.2B and C) (12). Normally both of these are less than 12 mm. If either of the distances is greater than 12 mm, then atlanto-occipital dislocation is likely present. The condylar-C1 interval (CCI) is the distance between the occipital condyles and C1 lateral masses (Fig. 45.2D) (7,13). The space is congruent on both frontal and lateral projections but can be evaluated only on CT or MRI. Another ratio used in children is the ratio of the interspinous intervals between the C1-C2 and C2-C3 described by Sun et al. (14) (Fig. 45.2E). When this is greater than 2.5, a craniocervical injury should be considered. Lee proposed the X-line method (Fig. 45.2F). This test is positive if the line from basion to the C2 spinolaminar point does not intersect the dens or the line from the dorsolateral corner of C2 to the opisthion does not intersect C1 (15).

Several investigators have evaluated the reliability of these criteria to diagnose craniocervical displacement based on modern imaging (Table 45.1). Dziurzynski et al. (16) compared five methods using both CT and plain radiographs. Plain radiographs were only readable in 40% to 80% of cases while CT was in all cases. The condylar-C1 distance and the basilar-dens interval had 100% sensitivity and 90% specificity. The Powers ratio and Harris rule had poor diagnostic parameters. Pang et al. (7) also compared
five methods and similarly found that the condylar-C1 interval had 100% sensitivity. He, similar to Dziurzynski, reported poor sensitivities and specificities for the Powers ratio, Sun ratio, and Harris rule.

MRI imaging is highly effective for identifying ligamentous injury to the tectorial membrane and alar ligaments and displacement of the craniocervical articulations. Pang critically evaluated 16 patients with craniocervical injury and documented associated findings including ventral prevertebral edema, subarachnoid hemorrhage along the brainstem, epidural and subarachnoid hemorrhage in the upper cervical spinal cord, transverse ligament rupture,
and distraction of the atlantoaxial joints (7). This latter injury is common and highlights the integral relationship between the occipitoatlantal and atlantoaxial joints, both being prevented from vertical displacements by the alar ligaments and tectorial membrane. In 13 cases, the injury occurred bilaterally, while in three patients the injury was unilateral. The use of coronal views and fat suppression techniques may increase the likelihood of identifying craniocervical ligament injury. However, MRI may be overly sensitive and identify many insignificant areas of hemorrhage or edema.