Spinal cord injury (SCI) constitutes a major cause of morbidity in the trauma patient. The United States has the highest incidence of traumatic spinal cord injuries, with approximately 40 new cases per 1 million persons per year ; moreover, the economic burden of SCI exceeds $14 billion each year. After prehospital stabilization and resuscitation, effective management of these patients relies on rapid, accurate clinical assessment and diagnosis of SCI and associated spine trauma, treatment tailored to the specific injury, prevention of complications, and early mobilization. The following chapter focuses on the evidence-based management of acute traumatic SCI in the intensive care unit (ICU).
Pathophysiology of Spinal Cord Injury
SCI occurs in two distinct phases. Primary injury is the immediate result of the initial traumatic insult—namely, sheer, compressive, or distractive forces that cause disruption of axons and blood vessels, leading to immediate neurologic dysfunction. Secondary injury evolves over the hours to days after the traumatic event and results from tissue hypoxia and ischemia—either at the cellular or systemic levels—inflammation, and neuronal hyperexcitability.
SCI is frequently observed in conjunction with trauma to the spinal column itself—bony fracture and/or dislocation, disc disruption, and ligamentous compromise—and may significantly contribute to neurologic dysfunction by exerting the aforementioned forces on the neural elements. Rapid diagnosis of spinal injuries is critical in the management of these patients. The absence of radiographic evidence of bony or ligamentous injury does not exclude spinal cord trauma, and the patient with unexplained poor findings from a neurologic examination may have underling cord injury. Furthermore, the same principle holds in patients who have abnormal results from a neurologic examination out of proportion to the degree of spinal column injury. Patients with severe cervical stenosis, diffuse idiopathic skeletal hyperostosis (DISH), rheumatoid arthritis, and baseline spinal instability are at the greatest risk.
A detailed neurologic assessment is necessary to properly categorize the severity of the injury, guide management, and facilitate communication among practitioners. Although several neurologic assessment scales have been published in the literature, the American Spinal Injury Association (ASIA) Classifications Standards/International Standards for Neurological Classification of Spinal Cord Injury (ISNICSCI) is the most heavily validated clinical neurologic assessment scale and is considered the gold standard by many practitioners for the clinical evaluation of acute SCI patients. There is currently class II evidence demonstrating significant interrater agreement, making this particular assessment scale appealing in terms of documenting and communicating serial examinations among practitioners. The ASIA Impairment Scale (AIS) ( Table 80-1 ) synthesizes the detailed neurologic assessment contained in the ASIA/ISNCSCI.
|A-Complete injury||No motor or sensory function below the level of injury, including the sacral segments|
|B-Sensory incomplete injury||Preservation of sensory function, including sacral segments, below the level of injury, but no motor function|
|C-Motor incomplete injury||Preservation of motor function below the level of injury, including sacral segments, with more than half of muscle groups below the level of injury with muscle grades 0-2|
|D-Motor incomplete injury||Preservation of motor function below the level of injury, including sacral segments, with more than half of muscle groups below the level of injury with muscle grades 3-5|
|E-Normal||Normal neurological motor and sensory exam, including sacral segments|
Radiographic Assessment and Collar Clearance in the Critically Ill Trauma Patient
Trauma patients with mechanisms of injury at risk for spine and SCI should be evaluated clinically before radiographic assessment. There is clear class I evidence regarding the initial radiographic assessment of such patients. Patients who are awake, alert, asymptomatic, and neurologically intact require neither radiographic imaging nor external cervical immobilization with collar. Patients who have symptoms or whose mental status precludes clinical evaluation should be placed in a cervical collar and undergo high-quality computed tomography (CT) scanning or three-view plain radiographic imaging when CT scanning is unavailable. CT scanning is superior to plain radiographs in the detection of cervical spine trauma. The guidelines for further evaluation of these patients is less clear in the setting of normal CT or three-view radiographs, relying on class II and III evidence.
In the awake but symptomatic patient with normal CT or three-view radiographic imaging, some clinicians have advocated cervical spine clearance with normal dynamic flexion-extension films of the cervical spine or magnetic resonance imaging (MRI) to rule out ligamentous injury obtained within 48 hours of injury. The utility of either modality in identifying clinically significant cervical spine injury is controversial, though. A recent systematic review of the literature concluded that dynamic flexion-extension imaging is inferior to MRI at detecting ligamentous injury. Moreover, flexion-extension films are largely dependent on patient cooperation. Duane et al. reported a relatively high rate of incomplete films (20.5%) and lower sensitivity compared with MRI, calling into question the clinical utility of this modality. On the other hand, Schuster et al. found that in the setting of the neurologically intact patient with negative CT imaging, MRI did not detect clinically significant ligamentous injury. Hence, given these limitations, either continuation of cervical collar until the patient is symptom free or cervical spine clearance at the discretion of the practitioner is a reasonable alternative in the symptomatic patient with negative CT imaging.
In the patient who is obtunded or comatose with negative CT imaging or three-view radiography, MRI or dynamic flexion-extension films are again available for further diagnostic evaluation and to assist in cervical spine clearance. As in awake, symptomatic patients, though, the marginal clinical value of these modalities is questionable. Prospective studies have demonstrated that flexion-extension adds little diagnostic value to CT imaging or plain radiographs in identifying clinically significant cervical spine injury. Again, as in the awake patient, the high rate of inadequate films either due to poor imaging quality or incomplete motion may limit the interpretation of this study in the obtunded patient. There are conflicting data within the literature regarding the utility of MRI in detecting clinically significant cervical spine injury. Multiple meta-analyses have demonstrated the merits of MRI in detecting occult cervical spine injury with a normal CT scan, although each of these analyses included studies with heterogeneous cohorts, somewhat limiting their conclusions. Panczykowksi et al. performed a meta-analysis comprising studies largely examining obtunded/intubated patients and showed that CT scanning alone could detect unstable cervical spine injury when compared with additional imaging modalities. Interestingly, Stelfox et al. demonstrated that intubated trauma patients who had collar clearance by CT scan alone had fewer complications, were ventilator dependent for fewer days, and had shorter ICU and hospital lengths of stay. Again, as in the case of the awake patient, cervical spine clearance in the setting of a normal CT may be deferred until the patient’s mental status improves or may be cleared with physical examination on the basis of a normal CT alone. Halpern et al. showed that in patients who are likely to be cleared clinically in a relatively short period of time (2 weeks), it is safe and cost-effective to leave these patients in a cervical collar as opposed to obtaining an MRI.
Frequently Encountered Injury Patterns
Although an extensive account of fracture/dislocation patterns and their management is beyond the scope of this text, we briefly review some of the more common fracture patterns of the spine.
Axial spine injuries extend from the occiput to C2. Occipital condyle fractures usually result from axial loading injuries (comminuted or linear type) and are usually stable; avulsion fractures of the condyle may result from distractive forces and should raise suspicion for atlanto-occipital dislocation (AOD) (see following discussion).
Primarily seen in high-velocity, high-impact trauma, AOD ( Fig. 80-1 ) results from distraction injury, causing disruption of ligamentous structures that stabilize the occipital-cervical junction. These patients may initially have normal results from a neurologic examination and progressive or fluctuating deficits, and this diagnosis can be easily missed. Moreover, concomitant traumatic brain injury is common among patients with AOD. Patients without a reliable examination or those with an unstable examination with a concerning mechanism should be evaluated for AOD. CT imaging, and in particular the condyle-C1 interval, is highly sensitive for the detection of AOD; other radiographic clues include prevertebral swelling; skull base or high cervical epidural, subdural, or subarachnoid hemorrhage; and occipital condyle avulsion fractures. MRI may be useful for direct visualization of the ligamentous structures and potential injury to the spinal cord. Ultimately, patients with AOD require surgical stabilization, usually with occipital-cervical fusion, frequently supplemented by external orthosis. Traction is not recommended in these patients and is associated with a tenfold increase in neurologic worsening compared with patients with subaxial cervical spine injuries. Before the definitive surgical fixation, cervical spine immobilization must be ensured, particularly during transfers or turns.
C1 fractures involve two-point fractures in the anterior arch, the posterior arch, or a combination of both; four-point fractures give rise to the classic Jefferson fracture ( Fig. 80-2 ). These fractures result from axial loading injuries, and stability depends on the integrity of the transverse ligament. If lateral displacement of the fracture fragments is significant, then these patients may require surgical fixation; otherwise, external immobilization is sufficient.
Bilateral C2 pars interarticularis fracture, or Hangman’s fracture, usually result from axial loading and flexion injuries ( Fig. 80-3 ). The stability of these fractures depends on the degree of C2-3 displacement or angulation. Fractures with minimal displacement generally heal well with external immobilization, whereas patients with significant displacement or angulation may signify disc disruption and may require reduction and surgical fixation.
Odontoid fractures are the most common C2 fracture with the pattern through the body of the dens the most frequent subtype ( Fig. 80-4 ). Younger patients may heal well with external orthosis with a halo vest. Significant displacement or angulation of the dens may require surgical stabilization. Elderly patients generally do not fair well either with external immobilization because of respiratory and swallowing issues or with surgical intervention because of higher risk profile, which poses a significant clinical dilemma. Surgical options include posterior C1-2 fusion or anterior odontoid screw placement.
Because of its mobility, the subaxial cervical spine (C3-T1) is prone to injury from various forces. Axial load forces can lead to compression deformities or more serious burst fractures. A combination of rotational, flexion, or distractive forces can lead to fracture-dislocation injuries with resultant SCI ( Fig. 80-5 ); these injuries require emergency evaluation by a trained specialist.
The thoracic spine is structurally well reinforced by the rib cage; hence, significant forces are required to produce injuries. As such, when these injuries occur, they can be devastating ( Fig. 80-6 ). The thoracolumbar junction is particularly prone to injury because it is the transition point between the thoracic spine and the relatively more mobile lumbar spine ( Fig. 80-7 ). In general, the lower lumbar vertebrae are less mobile and less prone to injury.
Surgical Decision Making
Surgical decision making and management rely on classification of injury, determination of an injury’s stability and degree of compression of neural elements, and the accurate assessment of neurologic function. The goals of neurosurgical management of SCI consist of decompression of neural elements and stabilization of the spine, the timing of which depends on the patient’s neurologic function.
Early decompression and stabilization in patients with incomplete injuries has become the prevailing trend. Even with complete injuries, early surgical intervention results in earlier mobilization, reduced pulmonary complications, fewer days for patients to undergo mechanical ventilation, and shorter length of stay. Rapid decompression of neural elements can be achieved with certain cervical spine fractures before definitive surgical treatment with closed reduction with cervical traction using Gardner-Wells tongs performed under fluoroscopy; this procedure should be performed only by a skilled specialist. An exhaustive account of the various surgical approaches is beyond the scope of this text, and we refer readers to references dedicated to the surgical treatment of spinal trauma.
Several grading scales have been proposed to classify spinal injury and guide in surgical decision making. The Subaxial Cervical Spine Injury Classification (SCLICS) and the Thoracolumbar Injury Classification and Severity Scale (TLICS) are commonly used and ideal in that they include metrics for ligamentous integrity and neurologic function. Of note, greater emphasis is placed on incomplete neurologic injury, suspected spinal cord over nerve root injury, and injury patterns that suggest a high degree of instability (e.g., distraction, subluxation). Both the TLICS and SCLICS have shown to be safe and effective guides for surgical intervention in prospective studies. Interestingly, the introduction of the TLICS score led to greater adoption of nonsurgical intervention. The TLICS score has been validated and is reliable whereas the SCLICS has been performed less well with regards to interrater variability.
Acute Traumatic Central Cord Syndrome
Acute traumatic central cord syndrome (ATCCS) is a heterogeneous clinical diagnosis that usually results from a hyperextension injury in which there is preferential damage to the medial anterior and posterior columns of the spinal cord ( Fig. 80-8 ). Because of the somatotopic organization of corticospinal tracts, this typically leads to weakness in the upper extremities with relative sparing of the lower extremities, although this clinical presentation can be varied. In general, for the diagnosis of ACTSS to be made, the ASIA motor score in the upper extremities must be 10 points less than the corresponding score in the lower extremities. This injury pattern may or may not be associated with bony or ligamentous disruption.
Patients with ATCCS are especially prone to further SCI secondary to hypotension; as a result, maintaining adequate perfusion pressure is imperative in the acute postinjury and perioperative periods. Because ATCCS frequently occurs in elderly patients, baseline pressures may be elevated relative to younger patients with SCI, and as a result, these patients may require higher mean arterial pressure (MAP) to maintain adequate spinal cord perfusion.
The timing of surgical intervention for ATCCS is somewhat controversial, although recent studies have demonstrated that surgery can be safely performed within 24 hours after injury and may lead to improved outcomes compared with delayed intervention. The surgical approach is variable and, in addition to surgeon preference, is guided by spinal alignment, the number of levels involved, the location of pathology, and the presence of other bony or ligamentous injuries.