12.38 Emergency management of the traumatized cervical spine

A patient with spinal cord injury and neck deformity would usually benefit from a general realignment effort and subsequent protective immobilization prior to transport.

In children, placement on a conventional backboard can lead to undesirable cervical kyphosis as paediatric heads are disproportionately larger than those of adults. Using a specialized backboard with a cutout for the head or placing a folded blanket under the victim’s torso helps to avoid this.

Exceptions to the principle of supine positioning of trauma patients on a hard board for retrieval purposes are rare. Neurologically intact patients who present with significant painful neck deformity, such as patients with ankylosing spondylitis, are preferably immobilized in situ with sandbags and folded towels in a semi-upright position to accommodate their deformity, provided that their vital signs are stable. It may also be helpful to ask the patient about their pre-injury posture and neck alignment as well as position of comfort, since patients with pre-existent neck deformity may be subject to further cord impingement and subsequent neurological deterioration attempts from ill-advised manipulation.

Unconscious patients should receive an emergency head computed tomography (CT) scan as soon as possible to exclude intracranial lesions and caudal extension of this scan to include the cervical spine. This allows visualization of the entire cervical spine and the cervicothoracic junction.

Log-rolling is performed with proper spine precautions maintained by a team approach. The cervical collar is released during palpation of the neck. Strict spine precautions should be maintained in all trauma victims until appropriate radiographic and clinical clearance have been completed; however, positioning on a rigid board for more than 4h increases the risk of developing sacral or occipital decubiti.

Neurological assessment starts with determination of mental status and cranial nerve function and is followed by systematic spinal cord functional evaluation following the principles established by the ASIA system (American Spinal Injury Association 1992) (Figure 12.38.1). The effects of sedatives, muscle-relaxing medications, and limb injuries on rating motor strength should be considered during this evaluation. Formal neurological status evaluation for trauma patients routinely includes evaluation of the anal sphincter region for perianal sensation, spontaneous sphincter tone, quality of voluntary contraction, and presence or absence of bulbocavernosus reflex. For patients with neurological deficits, an attempt is made to establish the level of sensory and motor injury. Further differentiation of complete and incomplete neurological injuries is of importance with regard to timing of treatment and patient prognosis. Repeat neurological evaluations should be obtained and recorded to exclude neurological deterioration which may occur, for instance, in the event of a developing epidural haematoma or malreduced spine. The presence of a bulbocavernosus reflex has been described as implying absence or resolution of spinal shock. Similar to neurological deficits secondary to spinal shock, patients with incomplete spinal cord injuries may recover some neurological function. Sacral sparing can present with signs of preserved voluntary anal sphincter contractility, perianal sensation, and voluntary long toe flexor function.

The ABC principles as formulated in the Advanced Trauma Life support algorithm had recommended three initial trauma radiographs, consisting of anteroposterior (AP) chest, AP pelvis, and lateral cervical spine views for several decades. The more ready availability of rapid image acquisition CT-scan technology, such as offered with helical CT, allows for comprehensive imaging of head, cervical spine, and torso with pelvis regions and has largely replaced conventional imaging in the emergency department setting. Despite this technological advance, a lateral cervical spine radiograph still remains the single most relevant diagnostic entity for trauma in most centres (Figure 12.38.2).

Significant injuries to these transition zones of the cervical spine (occipitocervical and cervicothoracic spine) are among the most common regions of missed cervical spine injuries due to limitations placed on conventional radiographic imaging. Helical CT has been consistently shown to be more sensitive, less time consuming, and more cost effective compared to the traditional work-up, with increased patient radiation exposure the only major downfall. Practitioners need to remember that certain cervical spine injuries are not readily seen on axial CT images, such as type II odontoid fractures and distractive injuries. Reformatted views in sagittal and coronal planes are integral to establishing CT as the sole imaging modality of trauma patients. Specific fine cut CT-scans in defined regions may be helpful to detect specific injuries such as craniocervical dissociations.

Emergency magnetic resonance imaging (MRI) scanning is indicated in patients with cervical spinal cord injury, unexplained neurological deficit, discordant skeletal and neurological levels of injury, and worsening neurological status (Figure 12.38.3). Barring contraindication to MRI, such as presence of stimulators, pacemakers or ferromagnetic clips and other foreign bodies, urgent indications for scanning present predominantly in patients with non-osseous encroachment of the spinal cord. Emergency CT myelography should be considered if there are contraindications to MRI, or if MRI is not readily available. Although sagittal MRI sequences such as T₂-weighted images can be helpful in depicting discoligamentous injuries, this imaging study can usually be obtained on a non-emergency basis.

Dynamic radiographs, such as flexion–extension lateral radiographs of the cervical spine, are effective in assessing stability. However, in order to obtain flexion–extension radiographs, the patient must present without cognitive impairment and have a normal, entirely non-focal neurological examination with normal plain radiographs. Recent studies have suggested that follow-up flexion–extension radiographs are not efficacious when a negative CT has been performed in blunt trauma without neurological findings. Recent evidence has also suggested that initial CT imaging identified all unstable cervical spine injuries in obtunded trauma patients. Subsequent upright radiographs did not identify any additional injuries but significantly delayed spine clearance.

In the setting of a hectic Emergency Department we recommend against routine acute flexion–extension radiographs because of the risk of incurring neurological deterioration by failure to satisfy all the criteria mentioned earlier. Furthermore, a patient with an acutely sprained neck is unlikely to provide a useful neck motion effort in the acute postinjury phase, thus rendering the study less meaningful. We recommend putting the patient into a neck brace and then performing flexion–extension radiographs at 1 to 2 weeks.

Injuries to the vertebral arteries (VAI) should be suspected in any patient with cervical spine trauma presenting with altered mental status. Injuries with significant vertebral displacement and displaced fractures involving the transverse processes should also alert the clinician to evaluate the patient for possible vertebral artery injuries. Any fracture extending into the foramen transversarium or with significant displacement, as in the setting of jumped facets, may be associated with overt or occult vertebral artery injury. Catheter angiography has been the gold standard for the diagnosis of VAIs; however, new 16-slice CT angiography seems to have sensitivity and specificity which is as good. Vertebral angiography can also be performed at the same time as aortic angiography (Figure 12.38.4). Magnetic resonance angiography and Doppler evaluation are non-invasive techniques for assessing cranial flow abnormalities with as yet unclear clinical implications. In those patients with identified vertebral artery injury, serial monitoring with transcranial Doppler over a 3-day period is used to assess for emboli. Treatment options include observation, antiplatelet agents, anticoagulation, and endovascular treatments. Although some authors have advocated antithrombotic therapy for most asymptomatic VAIs, there is a lack of class I evidence to support any strong guidelines for treatment.

After a brief initial period of hypertension, spinal shock results in flaccid paralysis, decrease of systemic vascular resistance, and subsequent hypotension. In patients with thoracic spinal cord injury above the T6 level, loss of sympathetic input to the heart can lead to bradycardia and hypotension. Neurogenic shock may mask other important causes of trauma-related hypotension. Patients with neurogenic shock will not usually respond to continued intravenous fluid substitution. Excessive fluid resuscitation efforts may lead to pulmonary oedema and congestive heart failure. Administration of intravenous epinephrine (adrenaline) by bolus or with an intravenous drip for instance with dopamine is recommended to reverse the deleterious haemodynamic effects of neurogenic shock.

Some units give glucocorticoids to patients with spinal injuries. The rationale is based on animal experiments which show stabilization of neural membranes, prevention of uncontrolled intracellular calcium influx, decrease of lysozymal enzyme effects, and decrease of swelling and inflammation, thus limiting the effects of the secondary injury to the spinal cord.

Two multicentre studies have shown that high doses of methylprednisolone, using a loading dose of 30mg/kg over 1h, followed by 5.4mg/kg/h over 23h, is beneficial for adult patients presenting within 8h of spinal cord injury. However, there has been mounting concerns about the methodology of data analysis of these studies and the occurrence of side effects such as gastric bleeding and infections. This has led to the use of intravenous steroids for the treatment of spinal cord injury to be considered a ‘treatment option’. There is near universal agreement, however, that methylprednisolone has no influence on recovery of neurological deficits for patients with spinal nerve root injuries or ballistic trauma.

New promising drug-based therapeutic approaches include regenerative strategies to neutralize myelin-mediated neurite outgrowth inhibition, neuroprotective strategies to reduce apoptotic triggers, the targeting of cationic/glutamatergic toxicity, anti-inflammatory strategies, and the use of approaches to stabilize disrupted cell membranes. All of these substances, however, are experimental in nature and far from routine clinical administration.

Of a number of other measures, hypothermia has received a great deal of more recent attention due to its purported benefits in the treatment of some high-profile individuals, despite previous studies failing to show benefits. Formal repeat review through larger spine societies led to rejection of the stated benefits of hypothermia as treatment for acute spinal cord injury. Although there is some hope for advancements in our understanding of modulating acute spinal cord injury, current intervention results have been modest or unclear at best. Therefore pharmacological measures currently remain focused on blood pressure normalization, correction of anaemia, and hypoxia to maximize cord perfusion.

Hyperextension of the cervical spine in adults should be avoided due to its risk of increasing spinal canal stenosis. In adults, infolding of the ligamentum flavum may increase canal occlusion by 19%.

In spinal cord injuries, timing of the removal of any ongoing mass effect acting upon the spinal cord may contribute to potential neurological recovery. Animal studies have shown a brief time window of 6–8h during which removal of spinal cord compression can lead to reversal of spinal cord injury. In clinical reality such time frames are highly unrealistic to achieve and commonly are incompatible with other vital patient resuscitation efforts. In the cervical spine, however, indirect spinal cord decompression with cervical traction can facilitate effective decompression of a compromised spinal cord.

Application of cervical traction with reduction of displacement can decompress the spinal cord by reducing bony canal encroachment and flattening the ligamentum flavum. Reversal of physiological cervical canal occlusion caused by compression using cranial traction improves canal clearance by 12% in intact human cadaveric specimens. Similarly, the effect of ligamentotaxis can be used to reduce spinal canal compromise caused by burst fractures of the cervical spine and facet dislocations.

Skeletal cervical traction is unsuitable for patients with distractive injury patterns or specific conditions such as fractures in ankylosing spondylitis. Relative contraindications to cervical traction also present in patients with skull fractures, in distractive cervical spine injuries such as occipitocervical dissociations, and in combative patients who cannot be pharmacologically sedated. Reduction of a dislocated cervical spine in unconscious or anesthetized patients without spinal cord injury may endanger the spinal cord, and should preferably be performed with spinal cord monitoring or after a disc herniation is excluded with neural imaging studies.

Skeletal cranial traction should be applied by means of suitable devices such as Gardner–Wells tongs or a halo ring (Figure 12.38.5). These devices should be MRI compatible in order not to impair neural imaging in spinal cord injury patients. In trauma patients use of cranial sling setups such as a head halter device is not desirable due to the risk of airway encroachment, aspiration, and limitations of weight that can be safely applied (Box 12.38.4).

Conventional Gardner–Wells tongs are usually not appropriate in paediatric patients, for whom a halo ring is usually more suitable. Success of a closed reduction effort is related to timing following injury, monitored muscle relaxation and analgesia and by providing a controlled setup with shoulder pulldown and fluoroscopy unit. Reduction by manipulation is generally discouraged. Serial neurological re-evaluations are strongly encouraged to minimize the risk of a secondary traction-induced neural deterioration being missed.

Halo-ring placement is more complicated than tong placement and requires at least one assistant or a dedicated halo positioning board (ACE-Fisher). Open-back halo rings have become increasingly popular due to greater ease of application and some improvement in patient comfort. Closed reduction of cervical dislocations is performed with the patient supine on a hard cushioned stretcher with both upper extremities pulled caudally by straps or surgical tape attached to the shoulders. We recommend intravenous analgesics, muscle relaxants, and nasal oxygen in responsive patients. Monitoring of vital signs such as automated blood pressure, pulse rate, and oxygen saturation is very helpful in titrating sedatives adequately. It is important that the patient remain awake in order to provide neurological feedback during the closed reduction.

Over-distraction should be avoided during cervical traction. We recommend starting with low initial weights of 2–5 kg (5–10 pounds), and checking a lateral radiograph for focal widening beyond 1.5mm between each cervical segment. Particular attention must be directed toward the occipital–cervical junction to assess for an undiagnosed craniocervical dissociation. If there is no over-distraction, traction weight is increased in 2- to 5-kg increments accompanied by clinical and radiological assessment (Figure 12.38.6). Older patients and patients with injuries of the upper cervical spine usually require lower weights to achieve the desired reduction results. Skeletal cervical traction weights usually should not exceed half to two-thirds body weight or approximately 45kg. Changes in the pull-angle that place the neck in a more flexed position can help in disengaging locked facets. Flexion is generally needed to reduce flexion type injuries in order to unlock the facets and allow them to translate back to their native position; in-line traction cannot accomplish this. Possible causes of unsuccessful closed-reduction efforts are inadequate muscle relaxation, presence of a fracture dislocation with comminuted lateral masses, or dislocations of the cervicothoracic junction.

Continued controversy surrounds the timing and technique of reduction of cervical spine dislocations. Reduction carries with it a theoretical risk of the intervertebral disc displacing into the spinal canal with potential for subsequent secondary neurological deterioration. Neural imaging, preferably in form of a MRI scan, has been suggested to rule out presence of a potentially cord compromising mass prior to embarking on closed reduction efforts (see Figure 12.38.3). However, the time taken for such imaging studies introduces a potentially critical delay in realigning and thereby indirectly decompressing the spinal cord. There is also a certain degree of risk in transferring a patient with an unstable dislocated spine from a trauma stretcher onto an imaging table.

The overwhelming majority of studies addressing this controversy of timing with a defined patient cohort have found immediate closed reduction to be a safe and effective intervention with successful reduction achieved in over 80% of patients and few, if any, cases of secondary neurological deterioration. While the incidence of visualized disc herniation on postreduction MRI has been described to range from 10–25% of patients, the vast majority of these lesions did not compromise the cord. Not surprisingly, the incidence of these disc herniations appears to be somewhat higher in patients with unilateral compared with bilateral facet dislocations. Based on these experiences the following protocol has become widely accepted: patients with neurological injury preferably receive emergent reduction using a formal skeletal traction sequence prior to receiving an MRI scan. In contrast, patients who are neurologically intact can be considered for prereduction MRI as long as this imaging modality is readily available. For cognitively impaired patients a best faith judgement call as to likelihood of presence of concurrent spinal cord injury can be used to guide the decision of emergent traction reduction. In general, however, the risks of leaving a patient with a dislocated spine appear to outweigh the risks associated with expediently performed closed skeletal reduction. Should a traumatic disc herniation with cord compression be identified on neuroimaging, anterior surgical decompression and fusion are recommended instead of pursuing closed reduction. Similarly, patients with failed attempts at closed reduction should be considered for emergent surgical reduction and stabilization.

For patients with unstable cervical spine injuries unsuitable for early surgical stabilization traction may be considered as a temporizing alternative. In order to minimize the risk of thromboembolic disease, decubital ulcers, and pulmonary deterioration during prolonged recumbence spinal injury patients who are unsuitable for early surgical stabilization are preferably placed in dedicated hospital beds, which maintain spinal alignment while allowing for some horizontal mobilization such as afforded by rotating trauma beds (ROTOREST).

Of several contraindications to the application of cranial traction, patients with distractive injuries to the cervical spine are particularly concerning due to their propensity for further displacement and subsequent neurological deterioration (see Figure 12.38.1). Temporary reduction under gentle compression can be attempted by application of a halo ring and halo vest or placement of sandbags around the patient’s head and securing it with tape placed across the head. For craniocervical dissociative injuries or similar upper cervical spine distraction injuries a reverse Trendelenburg position will likely cause further distraction. Postural reduction with a reverse Trendelenburg position can be achieved with the head protected by circumferential sand sacks; however, this positioning is unsustainable over a longer period due to increase in intracranial pressure and progressive cardiopulmonary compromise.

Early surgical intervention is now felt to be generally safe and effective. It has consistently shown decreased pulmonary decompensation, and shortened intensive care and overall hospital stays.

Surgical intervention for patients with cervical spine injuries serves the purpose of decompressing the spinal cord, minimizing further trauma to the injured segment, realigning and stabilizing the spinal column, and mobilizing the patient expediently in order to maximize spinal cord recovery. Unlike cord injuries in the thoracic region, cervical spinal cord injuries offer a greater hope for recovery not only due to the propensity for root area functional return but also in terms of cord level recovery. Results of animal experiments and evolving clinical data support the concept of emergent surgical decompression and stabilization with current instrumentation techniques for cervical spine injuries with spinal cord compromise that are not reducible by closed means. A recent ongoing multicentre study also supported the hypothesis that patients with early surgical management have improved outcomes over those treated in a delayed fashion. There is also emerging evidence that surgery within 24h may reduce length of intensive care unit stay and reduce postinjury medical complications.

The role of routine emergent surgical intervention in spinal cord injury remains controversial. Fortunately, most displaced cervical spine injuries can be reduced with closed means using the protocol described earlier. Surgical intervention may be beneficial if indirect decompression of the spinal cord is unsuccessful. Patients with worsening neurological status in the presence of an unstable spine may also be considered for expedient surgical decompression and stabilization. Patients with conditions not amenable to indirect closed decompression, such as traumatic disc herniation, expanding epidural haematoma, depressed lamina fracture, or complex facet fracture dislocations, may similarly benefit from surgical decompression and stabilization (Figure 12.38.7). Patients with unusual distractive injuries, such as occipitocervical dissociations, or patients with acute injuries superimposed on pre-existent spine deformities, such as ankylosing spondylitis or diffuse idiopathic hyperostosis are also candidates for early surgical stabilization due to inability to maintain closed reduction in these patients.

From an anaesthetic point of view, atraumatic intubation and maintenance of normotension is desirable. In any emergent surgical undertaking, the potentially deleterious effects of a ‘second hit’ in form of hypotension on neural cell survival following initial resuscitation should be considered in preoperative planning. Spine instrumentation should be used with the goal to facilitate early mobilization and the minimum necessary bracing. In considering the various factors previously mentioned, there has been an international trend favouring anterior procedures with rigid locking instrumentation over posterior or combined surgeries for a majority of cervical spine injuries. Exceptions to this observation present in patients with ankylosing spinal disorders or patients with depressed lamina fractures.

Patients with VAI pose a different set of problems in management due to continued bleeding or intimal damage leading to thromboembolic events. Interventional angiography can allow for embolization or stenting of the lesion site as needed. Usually a period anticoagulant therapy is recommended for patients with emboli emanating from intimal vertebral artery injuries as long as spinal cord or unstable column injury has been excluded (Figure 12.38.4). Should the cervical spine injury require surgical care, it is desirable to expedite such surgery so that a suitable anticoagulant therapy for the concurrent VAI can be started as soon as deemed safe. Delay of spine surgery in a patient who has already been started on anticoagulant treatment is usually far more complicated.

From the perspective of spinal cord pathophysiology most primary cellular damage has occurred by 6h or 8h, with cord necrosis present at 24h post-trauma. Current treatment mainly aims to reduce secondary zone spinal cord injury by minimizing cord swelling and maximizing cord perfusion. To date, the most meaningful emergency intervention for cord-injured patients consists of the earliest possible closed reduction of cervical spine dislocations or burst fractures by controlled skeletal traction. While there are few indications for emergency surgical decompression and fusion, the preponderance of recent publications favour early intervention over delayed surgery in patients who have been adequately resuscitated.

Current literature has redefined the role of high-dose methylprednisolone in the management of spinal cord injuries as optional. Novel neuroprotective therapies, including rho antagonists, minocycline, and sodium/glutamate blockers are the main pharmacological agents undergoing investigation.

Preferably, care of patients with spinal cord injury should be concentrated at specifically designated centres at the earliest possible time to allow for comprehensive treatment and to enhance our understanding of these measures by using standardized intervention measures.