Evaluation of Cervical Spine Trauma

Cross table lateral cervical spine in collar.

Sagittal reconstructed CT of the cervical spine of the patient above demonstrating the dens fracture.

The evaluation of spinal injury traditionally has involved physical examination of the neck, augmented by plain radiography. In a recent major survey of North American hospitals, about one-third relied solely upon the cross table lateral view of the neck for initial radiological evaluation of the cervical spine. Approximately two-thirds augmented the cross table lateral view with a AP and open mouth odontoid view, as well as occasional oblique views. Unfortunately, this traditional pattern of screening is far from perfect, and it has been estimated that up to one-third of cervical spine injuries escape initial clinical and/or radiological detection. Of greater concern is the fact that approximately 10% of patients are neurologically intact at the initial examination, and subsequently develop neurological deficits as a result of manipulation during resuscitation. Fortunately, there have been recent major advances in imaging technology, which, if utilized, could reduce the rate of delayed diagnosis in cervical injuries to near zero.

Although the physical examination is an important component of screening, the ability of the physician to predict the presence of a cervical spine injury based on history and physical examination alone is only 50%. Many emergency room patients have an altered level of consciousness, which significantly compromises the physical exam. Thus, the absence of signs and symptoms of cervical trauma is not enough to exclude the possibility of a significant cervical injury in patients who have a significant mechanism of injury. There is some academic debate as to whether patients with serious cervical injury ever are truly asymptomatic. This is, however, a moot point, because in the reality of clinical practice, a sizeable minority of patients with serious cervical injury will not have an impressive physical exam. One recent review of the 179 patients who had cervical fractures found that 13% of the patients were asymptomatic and that more than two-thirds of these had unstable bony or ligamentous injuries. Even more alarming, however, is that about one-third of this subset of patients were alert and not intoxicated, with no other clinical evidence of cervical injury other than a known mechanism for cervical injury.

The other pillar of traditional screening is conventional plain film radiography. Cervical plain films have been regarded as the gold standard in the past for cervical trauma; however, significant limitations exist. With a complete cervical series of well-positioned and optimally exposed radiographs of the cervical spine, the sensitivity for cervical fracture is reported to be 95% under ideal conditions. Thus, the false negative rate of a full cervical spine series monitored by a radiologist is still 5%. As computed tomography has been documented to detect twice as many fractures as plain films, I would argue that even this number is optimistic. Nonetheless, an optimal full cervical spine series will detect the majority of fractures. There is growing awareness that the limited trauma series obtained in collar suffers from significant limitations. Cross table lateral views of the spine have been shown to miss over 25% of cervical fractures.

One large-scale review of cervical trauma patients found that the three-view collar cervical trauma series fails to detect 61% of fractures and 36% of subluxation and dislocations. Although some of these limited three-view studies were recognized as being suspicious for injury, a full 23% of these studies were still reported as normal in patients with significant cervical trauma, half of whom had unstable cervical injuries.

In Hawaii, many of our patients are obese, which markedly compromises the exposure and positioning of these patients with the three-view collar cervical spine series. Frequently, the lower cervical vertebra can only faintly be visualized on the lateral or swimmer's projections, which makes it extremely difficult to exclude a fracture in this type of patient. In view of the fact that approximately 31% of injuries occur at the cervicothoracic junction, I would caution the clinical staff to avoid a false sense of security when the three-view study in collar is reported as "negative".

3D CT reconstruction of the cervical spine.

Computed tomography has shown great promise in accurately delineating spinal fractures. Its utilization in the past has been limited because of concerns about costs, the relatively slow imaging time, and the difficulty of obtaining emergency studies quickly. Additionally, the axial CT imaging alone does not allow for optimal visualization of spinal alignment, and could occasionally be insensitive to fractures parallel to the plane of imaging, such as small avulsion fractures. Despite these limitations, older generation axial CT scanning were 98% sensitive for cervical spine injury. Recently, a retrospective review of spiral CT was preformed in 88 patients with proven cervical fractures. Of this number, 32 patients had cervical fractures that were not revealed, or were incompletely demonstrated at radiography. One-third of these patients had either clinically significant or unstable fractures. Our experience at Hilo Medical Center with spiral CT is similar. We frequently detect fractures on CT which cannot be seen in retrospect on plain X-rays.

Newer ultrahigh speed, high-resolution spiral computed tomography scanning platforms are now installed at many clinical locations in the United States. This advance in imaging technology allows acquisition of a high resolution three-dimensional spiral data set which can be projected or reformatted in any plane, including high resolution surface renderations which can be rotated in space at a computer work station. These advances allow CT to detect spinal mal-alignment as well as transverse fractures with great accuracy (thus additional plain films are often not required to specifically address these issues). These high speed units can scan up to a 180 cm length in a single acquisition in about 70 seconds. Admittedly, utilization of these technologically advanced imaging systems will increase the initial cost of screening for cervical spine trauma. However, these modalities have a very low false positive rate compared to plain spine radiographs, which have been reported as having a false positive rate in the range of 18-63%, thereby mandating additional studies. The higher cost in screening is also dwarfed by the personal and financial cost to the patient and medical center of a delay in spinal cord diagnosis. As a secondary consideration, full appropriate use of these modalities would significantly decrease medico-legal exposure.

3D surface rendition of the cervical spine demonstrating C4 fracture and widening of the facets.

In conclusion, those patients who are alert and clinically at a relatively low risk of cervical injury can be imaged by plain cervical x-rays. If patients are felt to be at high risk of cervical injury, either by symptoms or mechanism of injury, the CT should be strongly considered. Plain films can be obtained as clinically desired, but they are not a substitute for CT scanning. Computed tomography should also be considered as an addition to the initial imaging evaluation when the patient's altered mental status does not allow a sufficient physical examination, in patients with multiple injuries in whom time is of the essence, and those patients who have a large body habitus. Computed tomography should also be performed whenever the initial plain film series is positive or felt to be suspicious of injury. Additional studies may be of value in selected cases. Patients who are suspected to have ligamentous instability should have flexion and extension radiographs, even if no evidence of fracture is present. In patients with a neurological deficit, MRI should be considered in order to evaluate for the presence of possible cord injury, and to exclude the present of traumatic disk herniation.

REFERENCES:

Woodring JH, Lee C. Limitations of cervical radiography in the evaluation of acute cervical trauma. Journal of Trauma 1993;vol 34,1:32-39

Jacobs LM, Schartz R. Prospective analysis of acute cervical spine injury: A methodology to predict injury. Annuals Emerg Med 1986;15,1: 85-90

Borock EC, Gabram SG, et al. A prospective analysis of a two-year experience using computed tomography as an adjunct for cervical spine clearance. Journal of Trauma 1991:vol31,7: 1001-1006

Mace SE. Emergency evaluation of cervical spine injuries: CT versus plain radiographs. Annuals Emerg Med 1985; 14,10:973-975

Vandermark Rm. Radiology of the cervical spine in trauma patients: Practice pitfalls and recommendations for improving efficiency and communication. AJR 1990;155: 465-472

Djang WT. Radiology of acute spinal trauma. Critical Care Clinics 1987; vol 3,3: 495-518

Nunez DB, Zuluaga A, et al. Cervical spine trauma: How much more do we learn by routinely using helical CT. Radiographics 1996;16: 1307-1318