Fractures, Dislocations, and Fracture-Dislocations of the Spine

Initial Spine Assessment

After the ABC (airway, breathing, and circulation) of the Advanced Trauma Life Support (ATLS) protocol has been completed, a thorough orthopaedic history should be obtained and full physical examination should be done. Important information includes the injury mechanism, preinjury functional level of the patient, patient report of weakness or sensory changes, signs of blunt head trauma, spine tenderness, spinal step offs, and interspinous widening. Findings of flaccidity in the extremities, incontinence, or penile erection may indicate spinal cord injury. A detailed neurologic examination, which includes motor function, sensory function, and rectal tone, recorded on the American Spinal Injury Association (ASIA) form and an assessment of mental status are part of this examination. The diagnostic imaging of a patient is inextricably linked to the neurologic examination. The initial spinal assessment of a trauma patient is to determine if the patient has a spinal cord injury. If an injury is found, all initial CT imaging, including that of the spine, is completed as rapidly as possible and treatment initiated. If a patient does not have a spinal cord injury, it should be determined if he or she meets the criteria to be considered asymptomatic with respect to the cervical spine. If the patient is found to be asymptomatic, then the cervical spine can be cleared clinically without the need for radiography. There are five specific criteria described in the National Emergency X-Radiography Utilization Study (NEXUS) that must be fulfilled to classify a patient as asymptomatic. The purpose of the study was to develop a decision rule that would reduce the number of radiographic examinations in trauma patients without missing significant injuries. The five specific criteria are noted in Table 41.1 .

Using these criteria, one third of the trauma patients evaluated in the 21 community emergency departments or level 1 trauma centers were found to be asymptomatic (range 14% to 58%). The determination of a patient’s level of alertness is the first step in the workup specifically for a spinal injury, which should begin immediately after the ABCs have been evaluated. If the patient is asymptomatic by the criteria of the NEXUS trial, no radiographs of the cervical spine are needed and the cervical spine may be “cleared” on clinical grounds, which significantly expedites care. Patients who are not alert or who do not meet the NEXUS trial criteria for other reasons require radiographic evaluation. The patient’s motor and sensory examination should be documented thoroughly; the International Spinal Cord Society (ISCoS) form is the accepted instrument that best serves this important function. For patients who are found to have a neurologic deficit, serial neurologic examinations are recorded using this form, which has proven to be useful in detecting clinical deterioration and guiding decisions on additional imaging or other interventions that may become necessary. For patients with neurologic deficits, ISCoS forms are completed every 4 to 6 hours usually for the first 24 hours after arrival, but this varies based on the patient’s course. If a patient is found to have a cervical spinal cord injury, then medical management and imaging workup will need to address this injury as the first priority in all but the most critically injured patients. In some patients immediate reduction of fractures or dislocations may be most appropriate, whereas other patients may benefit from MRI before proceeding with treatment.

Controversy persists about the optimal diagnostic imaging protocol for trauma patients as it relates to the spine. Our protocol is outlined in Figure 41.1 . There are several objectives for which there is general consensus among trauma surgeons and spine surgeons. First is the detection of any significant spinal injury that places the patient at risk for neurologic deterioration. This may be an osseous injury, a soft-tissue injury, including posterior ligament complex injuries and other important injuries such as disc disruptions, or a combination of the two. Second, make a determination that there is no significant injury as early as possible to allow discontinuation of cervical immobilization and lifting of spine precautions . This will help avoid the recognized morbidities of immobilization and to facilitate other aspects of the patient’s care. Rose et al. found that patients meeting the NEXUS criteria who had a “distracting injury” were correctly assessed on clinical grounds alone with 99% sensitivity and 99% negative predictive value. They concluded that the number of CT scans in this cohort of patients could be reduced by 61% and suggested that radiographic evaluation is unnecessary for safe clearance of the asymptomatic cervical spine in awake and alert blunt trauma patients with “distracting injuries.” Additionally, the imaging should assess for associated injuries, including vertebral artery injuries in the cervical region or visceral injuries involving the chest, abdomen, and pelvic areas when evaluating the thoracic, lumbar, and sacral spine regions. The initial evaluation of the noncervical spine (thoracic to sacrum) is best done using multidetector computed tomography (MDCT) with both sagittal and coronal reformatted images.

Spine Precautions

S pinal precautions often are mentioned but rarely described in publications regarding trauma to the spine. The following protocol is derived from our experience. Spinal immobilization has already been described as it pertains to transport of an injured patient, but, as mentioned, one of the goals of the initial assessment is to be able to remove the patient from the backboard quickly once hemodynamic stability has been obtained and CT evaluation completed. Even if a significant spinal injury at any level is found, the patient can be moved to a bed but maintained with a cervical collar in place on a pillow as needed to avoid cervical extension. Patients with ankylosing spondylitis may require several pillows to keep them more upright because of their rigid cervicothoracic kyphosis (see Chapter 38 ). If a patient is to be placed in cervical traction, the crossmember for the traction pulley is fixed to the bed frame such that the traction vector maintains neutral alignment and adjusts if the bed position is altered. With this level of precaution, a patient can be placed head up using the reverse Trendelenburg function of the bed. If the cervical CT is negative for injury and cervical immobilization is to be continued pending further evaluation, then the patient is allowed to be fully upright in a properly fitted rigid cervical orthosis until spinal clearance is possible unless other injuries prevent this. Prasarn et al. demonstrated in a cadaver study that a kinetic bed caused less cervical displacement through an injured segment than the traditional log roll maneuver.

Patients with unstable thoracic or lumbar injuries, such as fracture-dislocation or other injuries that will be treated with internal stabilization, are maintained flat in bed (using the reverse-Trendelenburg position to elevate the head) and log-rolled side-back-side every 2 hours while awake until the spine is stabilized. For patients with spinal cord injuries in whom operative stabilization will be delayed more than 24 to 48 hours a Roto-Rest (KCI, San Antonio, TX) type bed is preferred (also used in patients with cervical injuries). Once the thoracolumbar fracture is stabilized, or for those patients being treated in an orthosis, elevating the head of the bed 0 to 30 degrees is allowed without donning the orthosis. The orthosis is required when the head of the bed is above 30 degrees.

Keeping the head of the bed elevated is strongly encouraged if blood pressure, intracranial pressure, and other vital parameters permit, to reduce the risk of aspiration and to assist with pulmonary toilet. Once spinal stability is achieved, continued frequent turning of the patient or the use of a therapeutic air mattress is preferred as long as mobility is severely limited for any reason.

Diagnostic Imaging

Injuries that involve the thoracic, lumbar, or sacral regions of the spine generally can be diagnosed using CT, which has been established as the diagnostic imaging modality of choice in these areas. It usually is obtained as part of the primary workup by the trauma surgeons or the physicians in the emergency department. Additional evaluation with MRI in these areas or use of other modalities typically is not necessary, although there are circumstances in which obtaining an MRI is appropriate. Because CT studies are obtained routinely for other reasons, the specific indications for radiographic evaluations of the thoracic and lumbar spines and the sacrum have not been extensively studied. Additional attention is given to this topic in later sections dealing specifically with injuries to these areas.

Patients who have cervical spine symptoms require imaging evaluation, and the recommendations for this process have changed in recent years. The standard radiographic evaluation of the cervical spine for trauma patients until relatively recently has been anteroposterior, lateral, and open-mouth odontoid radiographs. This three-view protocol has proven reliable when technically adequate images are obtained but has been documented to fail in demonstrating a small number of significant cervical injuries. Because the incidence of cervical injury in trauma patients is between 2% and 6%, a very high sensitivity is required to optimally evaluate symptomatic patients. In a series of 32,117 patients, Davis found 34 missed injuries. As has also been documented in numerous other studies, the most common reason the injury was missed in Davis’ series was failure to obtain adequate radiographs of the injured level (23 patients). Eight patients in the series had incorrect readings of adequate films, and only one patient was documented to have adequate radiographs that did not demonstrate the injury even in retrospect. Most studies on this topic have found that the occipitocervical junction and the cervicothoracic junction are the areas where injuries are most likely to be missed. Several studies have provided level I evidence that the negative predictive value of an adequate three-view series is from 93% to 98%; however, in these same studies the sensitivity was only 62% to 84%. Assuming a series of 100 patients, 6% of whom have cervical injury, five of six cervical injuries could be detected as abnormal on a three-view radiographic series and one truly injured patient would not be distinguished radiographically from the 94 correctly identified true negative series. This deficiency of plain radiographs is not improved with the addition of oblique films for a five-view series.

With greater availability of MDCT there has been a transition to using this modality for the primary evaluation of the cervical spine in trauma patients. In a large multicenter, level II study of patients failing to meet NEXUS criteria, Inaba et al. found that the sensitivity of CT was 98.5%, specificity was 91.0%, and negative predictive value was 99.97% for clinically significant injury. They also found that, for patients with neurologic injuries, CT alone missed a small number of clinically significant injuries and therefore recommended MRI in this group. Combining the cervical CT scan with the head-chest, abdomen, and pelvic scan, which often is ordered for these patients, has resulted in a lower cost than if the cervical study is done separately. Also, because the patient is already in the scanner and the scan times are much faster with MDCT compared with conventional CT, it actually takes less time to obtain an MDCT than it would for a three-view series of plain radiographs. When the relatively high proportions of technically inadequate studies that require CT are factored in, the MDCT has been found to be cost effective relative to plain radiographs. Despite these advantages, the higher radiation dose to the patient remains a concern with MDCT. Although comparisons between MDCT with coronal plane and sagittal plane reconstructions and plain radiographs have found higher sensitivity in detecting injuries with MDCT, several studies comparing autopsy findings with injuries noted on CT before death found that not all injuries present at autopsy were demonstrated by CT. Molina et al. found significant injuries in a small number of patients that were not demonstrated on the CT images. This indicates that CT may not be the “gold standard” by which to judge all other diagnostic imaging techniques. The role of MRI in the evaluation of the cervical spine in symptomatic patients to supplement MDCT continues to develop as the deficiencies of MDCT are better understood. A significant number of studies demonstrate improved diagnostic sensitivity with the use of MRI. Sarani et al. retrospectively found injuries on MRI in 42 of 164 (26%) trauma patients. All 164 patients had negative CT scans, and treatment was altered in 74% of these patients either with surgery or continuation of immobilization. In the subset of patients who could not be examined because of altered mental status, Sarani et al. found injuries on MRI in 5 of 46 (11%) patients who had negative CT scans, and 80% of these patients required surgery. Pourtaheri et al. found in a subset of patients with cervical fractures and altered mental status that MRI was very useful; MRI found additional injuries in 48% that changed treatment for 39%. This treatment change was from nonoperative to operative treatment 24% of the time. The clearest indication for MRI in a trauma patient is for the evaluation of an unexplained neurologic deficit at any spinal level. MRI has a higher sensitivity for detecting soft-tissue injuries, which are not well demonstrated on CT. MRI can detect a missed spinal column injury or neural compressive pathologic processes, such as disc fragments, epidural hematoma, or the presence of significant canal stenosis from other causes. For patients with a demonstrated injury and neurologic deficit at a corresponding level, MRI usually offers little additional information for that injury. However, noncontiguous injuries occur in up to 15% of patients. Because of this high rate of additional injuries, patients with cervical injuries demonstrated on MDCT at our institution are evaluated with MRI primarily to assess for soft-tissue injuries. This practice has resulted in alteration of the treatment plan for a significant percentage of patients when additional injuries are detected. Using MRI to assist in the “clearance” of the cervical spine remains controversial at this time. Although many of these additional injuries are significant and do alter treatment, some of the injuries are not clinically significant, so specific indications for obtaining MRI need to be defined. Determining which MRI findings correlate with clinical instability also needs to be better defined. Schoenfeld et al. found in a propensity-matched cohort that the addition of MRI to CT identified 8% more injuries than CT alone. Only 1% required surgery. The number needed to treat (NNT) in order to change patient management was 50 and the NNT for surgery was 167. Clearly MRI is not needed as a routine imaging modality. If a patient has abnormal findings on CT suggesting soft-tissue injury, such as soft-tissue density anterior to the midbody of C3 greater than 5 mm, a widened disc space (>1 to 2 mm) at one level relative to adjacent disc levels particularly if there is an anterior osteophyte avulsion at that level ( Fig. 41.2 ), or excessive widening of the interspinous distance posteriorly, MRI should be obtained. An additional confounding issue is the timing of the MRI. Because MRI is most effective for evaluating soft-tissue injury, either by showing discontinuity of anatomic structures such as the ligamentum flavum and annulus fibrosus or hemorrhage and edema associated with tissue disruption, the timing of the study is very important. If the MRI is obtained within the first 48 hours after injury, the sensitivity for hemorrhage and edema is optimal. The ability of MRI to identify injury after 48 hours is dependent on the direct demonstration of tissue disruption or subluxation of the spine. Emery et al. found that MRI done an average of 11 days from injury failed to demonstrate known soft-tissue injuries in 2 of 19 patients. Evaluation of the available literature revealed level III evidence to support the “clearance” of the cervical spine in a symptomatic patient if CT and MRI done within 48 hours of injury are found to be normal. Our process is to obtain an MRI in the obtunded patient within this 48-hour window if the patient is stable enough to undergo the study. A patient is not considered obtunded if the mental status is altered because of the presence of substances that will only transiently impair the patient. In this case the patient has CT examination and remains in a rigid orthosis with repeat examinations until the impairment has resolved and a determination is made to either “clear” the cervical spine on clinical grounds or proceed with MRI within 48 hours. If the condition of the patient does not allow the MRI to be completed within 48 hours and the patient remains obtunded, an MRI is obtained as soon as the patient can safely undergo the study and any identified injuries are treated. However, if the delayed MRI does not directly demonstrate an injury, the patient is kept in a rigid orthosis for up to 6 weeks as treatment for presumed soft-tissue injury or until his or her mental status improves and he or she can be cleared on clinical grounds by meeting the NEXUS criteria. This protocol has been effective in avoiding neurologic deterioration because of missed injuries. Although there has been occasional morbidity such as decubitus ulcers attributable to the orthosis, this is very rare. Skin breakdown on the posterior scalp above the orthosis results from improper fit of the orthosis, not keeping the patient upright, and not turning the patient adequately. No more serious morbidities from immobilization have occurred, although nursing care, especially tracheostomy care, is somewhat more difficult. Thus, the primary indications for cervical MRI are unexplained neurologic deficit, identified cervical injury, CT findings suggestive of soft-tissue injury, or a patient with altered mental status after intoxicants are metabolized. When possible, MRI is performed within 48 hours.

Additional Imaging

It is unusual for additional imaging to be required beyond that described. We have not found dynamic studies to be useful acutely to evaluate the cervical spine. There is a high rate of inadequate studies for a variety of reasons, foremost of which is inadequate range of motion. In obtunded patients, there have been reports of major neurologic injury caused by obtaining dynamic images. If a nonobtunded patient has adequate motion for flexion and extension lateral radiographs, typically clearance can be done on clinical grounds using the NEXUS criteria without further imaging.

On rare occasions a patient may have findings on MDCT suggestive but not definitive of a soft-tissue injury. Typically, an MRI study would be obtained, but in certain patients this is contraindicated (e.g., if the patient has a pacemaker). In these instances a “stretch test” as described by White, Southwick, and Panjabi is done to more completely assess the stability of the spine. This test allows measurement of the displacement within a motion segment under controlled conditions to identify soft-tissue injuries. Gardner-Wells tongs are applied before this test is performed. A head halter can be used but is less desirable because of the amount of weight that potentially may be used. The possible end points for the test are a change in neurologic status, an increase of 1.7 mm between adjacent vertebrae at any level, an angulatory change of 7.5 degrees at any disc level, or reaching one third of body weight or 65 lb, whichever is less. A prerequisite to performing a “stretch test” is that the patient must be alert and able to provide a consistent feedback for neurologic examination ( Box 41.1 ). Resuscitation should be complete, and the patient should be hemodynamically stable. Head CT should confirm no fracture near the planned cranial pin sites.

Neurologic Assessment

To properly direct the diagnostic imaging necessary for a patient, the neurologic examination findings play a key role. Assessment of mental status using the Glasgow Coma Scale (GCS) ( Table 41.2 ) determines the level of consciousness. If the GCS score is not 15, then imaging will be required as outlined earlier. Clearly document the motor and sensory examination, including the function of the rectal sphincter and the presence of perianal sensation. We have used the ASIA form from ISCoS. Using the ISCoS form, sensation is recorded for light touch and pinprick in 28 dermatomal distributions on each side of the body ( Fig. 41.3 ). Pinprick testing is done using a sterile needle rather than a pinwheel. A score of 2 (normal), 1 (altered), or 0 (absent) is determined for each dermatome, and specific “key” areas are identified on the diagram within each dermatome as optimal test locations. In addition, the presence of sensation for deep anal pressure is made to help determine if a spinal cord injury is complete or incomplete. Important dermatomal landmarks are the nipple line (T4), xiphoid process (T7), umbilicus (T10), inguinal region (T12, L1), and perianal region (S4 and S5). Motor function is scored 0/5 to 5/5 in each of 10 specific myotomes per side ( Table 41.3 ). Also, the presence or absence of voluntary anal sphincter contraction is recorded. In some circumstances, the designation of “NT” for not testable or 5∗/5 (weakness as expected, considered normal strength because of inhibiting factors such as fractures) are most appropriate. Before making a definitive determination of injury type the patient must be out of spinal shock. This usually occurs within 24 to 48 hours but can take substantially longer and is indicated by the return of the bulbocavernosus reflex and anal wink ( Figs. 41.4 and 41.5 ). The ISCoS document lists the requirements for each motor grade along with the definitions of the ASIA Impairment Scale and a flow chart to properly interpret it. Using the AIS, a neurologic level of injury (NLI) determination is made to classify the spinal cord–injured patient.

The NLS is defined by the most caudal myotome with at least 3/5 function and 5/5 function at all higher levels that also has normal sensory function and normal sensation at all higher levels. At levels without key myotomes, the level is determined by the sensory level. Type A patients are motor complete and sensory complete, with no motor or sensory function more than three segments caudal to the named injury level. Function within the zone of partial preservation should be recorded because a change by even a single level can be very significant, especially in the cervical region. Type B patients are motor complete but sensory incomplete (incomplete sensory loss but complete motor loss with no motor function more than three segments caudal to the named injury level); sensory sparing may be only light touch, pinprick in the perianal segments, or deep anal pressure. Type C patients have either voluntary sphincter contraction or voluntary motor function more than three segments below the named injury level with sacral sensory sparing. This motor sparing can be in non-key myotomes according to the standard at this time. More than half of functioning key myotomes are graded less than 3/5. Type D patients have at least half of functioning key myotomes greater than or equal to grade 3/5. Type E patients have a spinal cord injury that improves to normal. This type is not used to describe a patient without a spinal cord injury initially. This examination should allow the clinician to distinguish spinal cord injuries from isolated nerve root or nerve plexus type injuries.

The initial neurologic examination should be completed as soon as possible after the arrival of the patient to establish the correct baseline for the patient to which all subsequent examinations will be compared. It is our practice to complete serial neurologic assessments on patients with spinal cord injuries or unstable spinal column injuries every 4 to 6 hours for at least the first 24 hours and continue less frequent reassessments thereafter based on the patient’s clinical course. This regimen is derived from experience in a busy level I trauma center but is not evidence based, and it is unlikely that evidence-based practices could be used to examine how frequently optimal evaluations should be done. In addition to the motor and sensory examinations, it is important to include examination of the deep tendon reflexes. Acute spinal cord injury results in flaccid paralysis and areflexia. The presence of pathologic reflexes such as a Babinski or Hoffmann reflex or clonus indicates a more chronic process, which may be acutely worsened by trauma such as a central cord injury in the setting of chronic cervical stenosis. The purpose for serial examinations is to detect any neurologic change and institute management changes to improve the patient’s ultimate neurologic outcome. Deterioration of neurologic function can be caused by intracranial processes such as hemorrhage, metabolic processes such as acidosis, or spinal pathologic processes. Bony malalignment causing spinal cord compression, hypotension, expanding epidural hematoma, spinal cord infarction, inadequate immobilization, or improper movement of a patient are some of the reasons for deterioration that must be considered by the orthopaedic surgeon in collaboration with other consultants so treatment can be adjusted appropriately. Likewise, if a patient is noted to improve, management may need to be altered as well with regard to planning of spinal stabilization or nonoperative spinal interventions.

Neurogenic And Spinal Shock

Neurogenic shock refers to hemodynamic instability that occurs with rostral cord injuries related to the loss of sympathetic tone to the peripheral vasculature and heart, the consequences of which are bradycardia, hypotension, and hypothermia caused by absent thermoregulation. The combination of hypotension and bradycardia should alert the clinician to this cause of shock rather than hemorrhagic shock, which may coexist, particularly in patients with other injuries. Aggressive treatment of hypotension of any cause is a priority in patients with spinal cord injury. Spinal shock refers to a temporary dysfunction of the spinal cord, with a loss of reflexes and sensorimotor function caudal to the level of injury. It is manifested by absence of anal wink and bulbocavernosus reflexes and by flaccid paralysis. It is a temporary phenomenon and recovers usually in 24 to 48 hours even in severe injuries but can persist for weeks or rarely, months. There is no specific treatment for spinal shock.

For patients with a spinal cord injury, rapid diagnosis and institution of measures to minimize secondary spinal cord injury may be the most important interventions possible to improve ultimate neurologic and functional recovery. The controversy concerning the timing of surgery is centered on the concept of minimizing the secondary injury. Numerous studies such as the Surgical Timing in Acute Spinal Cord Injury Study (STACIS) have attempted to determine the optimal timing of surgical decompression and stabilization. At present, this remains somewhat of an open question, but evidence is mounting in favor of early decompression to enhance neurologic outcomes. Often this decompression is most rapidly accomplished by placing the patient in skeletal traction. This maneuver can be done much more quickly than operative treatment in most circumstances. In addition, multiple studies provide level III evidence that earlier decompression and stabilization are associated with shorter hospital stays and lower overall treatment costs for these patients. In a clinical study with direct measurements of spinal cord pressure and spinal cord perfusion, Werndle et al. found that spinal realignment and stabilization did not lead to improved spinal cord perfusion. This was attributed to spinal cord swelling within the inelastic dura mater.

The secondary injury cascade refers to the additional neurologic injury that results from cord ischemia, leading to electrolyte shifts with cell membrane alterations and accumulation of neurotransmitters and inflammatory mediators including free radicals that further injure neural tissue. A detailed discussion of these mechanisms is beyond the scope of this text; however, it must be recognized that proper medical management of a patient with a spinal cord injury is an important component in the overall care. The secondary mechanisms follow the initial or primary mechanical injury caused by compression, distraction, shear, or laceration of the spinal cord. The secondary injury cascade occurs over a period of hours to days, depending on the severity of injury and other injuries that may be present. Based on a number of animal models and level III evidence, it appears that the injury caused by ischemia of the spinal cord is the central feature of this secondary injury process. Avoiding or minimizing ischemia of the spinal cord appears to improve neurologic outcome. Spinal cord ischemia results in changes locally, with loss of autoregulation of spinal cord blood flow and changes to the systemic vasculature. These systemic alterations include cardiac rhythm irregularities, bradycardia, decreased mean arterial pressure (MAP), decreased cardiac output, and decreased peripheral vascular resistance. All of these abnormalities have the effect of a positive feedback loop to worsen the cord ischemia and thus worsen hemodynamic parameters. All of these hemodynamic parameters tend to be worse with more severe and more rostral injuries. Respiratory insufficiency or failure often accompanies spinal cord injury because of weakness of the respiratory muscles resulting in hypoxemia, which, in turn, worsens the spinal cord ischemia. Early detection and treatment of cardiopulmonary dysfunction does reduce the morbidity and mortality caused by these mechanisms. The goal for optimal blood pressure management is a MAP of 85 to 90 mm Hg with maintenance of 100% oxygen saturation. This is based on clinical observations and level III evidence, which remains the best guidance available to date. To properly treat these patients, arterial lines and central venous access or even Swan-Ganz catheters may be needed. Initially, hypotension should be treated as hemorrhagic in origin and fluid resuscitation should be with a balanced solution (e.g., lactated Ringer solution). After adequate crystalloid volume replacement, blood transfusion may be needed. If hypotension has not responded after fluid resuscitation and transfusion with normal central venous pressure, pressor agents should be administered to maintain the MAP in the desired range. Agents such as dobutamine, dopamine, or norepinephrine, with both alpha- and beta-agonist properties, are preferred over pure alpha agonists such as phenylephrine that can lead to reflex bradycardia. The duration of pressure support to maintain the median arterial pressure has been somewhat arbitrarily stated to be 7 days, but there is no evidence to support either a longer or shorter period of time. Supplemental oxygen should be administered and ventilator settings adjusted to keep oxygenation at or near 100% during this period as well.

Immediate Cervical Spinal Reduction

The primary objective for rapid cervical reduction and stabilization is to improve spinal cord blood flow and thus minimize the harmful effects of ischemia. In animal models, rapidly relieving spinal cord compression has been shown to be beneficial. The short period of time from injury to decompression determined in these studies to be optimal has not been clinically achievable. One intervention that can be accomplished in some patients to relieve spinal cord compression and improve cord blood flow is to reduce fractures and dislocations using skeletal traction. If the injury is recognized and the patient is emergently taken to the radiology suite, often the reduction can be achieved within the first 1 to 2 hours after the patient arrives at the hospital. To be effective this must be done absolutely as soon as possible even if the initial workup has not been completed. However, limited evidence exists as to how beneficial this may be, and there is some risk from other undetected injuries in this setting. Closed reduction usually can be accomplished significantly faster than can be achieved by operative means, and completion of the evaluation in a hemodynamically stable patient can usually safely follow the reduction. Closed reduction is not always possible and is not appropriate to attempt, for example, in patients with distraction type injuries at other levels, in obtunded patients, in patients with certain cranial fractures, or if the patient becomes hemodynamically unstable.

A great deal of controversy exists regarding timing of cervical reductions and the need for cervical MRI, particularly in the context of a patient with unilateral or bilateral facet fractures or dislocations. The controversy has been centered on whether there is a need to obtain prereduction MRI to determine if there is a disc herniation. The value of this information compared with the risk of the increased time to reduction has not been established. Consideration must be given to several pieces of information when treating these patients. The first is that dislocation of the spine with spinal cord compression is definitely associated with neurologic injury. Rizzolo et al. reported that in 55% of patients with facet injuries, disc herniations or disruptions occurred and that often the disc material displaced into the canal. The importance of this is not clear as it relates to spinal reduction. Vaccaro et al. documented by MRI that more disc herniations were present after reduction than before reduction, but disc displacement did not correlate with neurologic deterioration in a small series of patients. Grauer et al. noted the significant variability of using MRI in the setting of cervical dislocations among spine surgeons based on their primary specialty. The second important fact is that only rarely has closed spinal reduction been associated with neurologic worsening if the patient is awake and alert at the time of reduction. Although there is no level I evidence on this topic, it appears that the important issue is whether the patient is awake and alert at the time of reduction, not the presence of a disc injury. Many clinical series that were reported over a period of decades found only 11 of 1200 awake patients (<1%) who developed permanent neurologic worsening after closed reduction. At least two were root level injuries. Additionally, one or two patients had transient worsening that returned to baseline. Reduction was accomplished in 80% of patients, which should allow for better spinal cord perfusion. Thus, the risk of causing additional harm in an awake and alert patient with a cervical facet fracture or dislocation and a significant neurologic deficit is very low. In an awake and alert patient with a cervical fracture or dislocation with a significant neurologic deficit, we recommend expeditious reduction without obtaining an MRI.

Significant neurologic injury in our protocol has been determined to mean less than grade 3/5 in more than one half of the key myotomes caudal to the level of injury (ASIA Impairment Scale A, B, or C). By using this regimen, most awake and alert patients have reductions before obtaining an MRI. These patients do have MRI after reduction but before definitive treatment to assist in surgical planning. For the rare patient with a bilateral facet injury, or more likely a unilateral facet injury, and more than half of the key myotomes caudal to the injury level grade 3/5 or higher, an MRI is obtained before reduction even if the patient is awake and alert. The rationale is that if a patient’s neurologic function is grade 3/5 or higher initially, there is more potential for harm with immediate reduction and less benefit. If during the process of reduction worsening of neurologic deficit occurs, the attempt at reduction is terminated. Immediate MRI is obtained, and operative treatment is undertaken, depending on the pathologic process present. If the patient is obtunded, closed reduction cannot be undertaken safely and immediate reduction is not attempted. For patients in whom closed reduction is attempted but not successful, MRI is completed to help guide the surgical approach.

Spinal Cord Injury Treatment

At this time there remains no effective treatment to reverse spinal cord injury that has been established by level I evidence. Many patients do improve neurologically, and in some the improvement is very dramatic. The measures that have been established to date are those detailed earlier that reduce the secondary injury. These include rapid realignment of the spine when appropriate, maintaining MAP at 85 to 90 mm Hg, and maintaining 100% oxygen saturation. The use of maintaining MAP in the range of 85 to 90 mm Hg continues to be evaluated. Hawryluk et al. evaluated minute-by-minute data on 100 patients with spinal cord injuries and found a correlation between maintaining a MAP of 85 to 90 mm Hg and better neurologic outcomes; intermittent lapses below the target range negatively affected outcomes. Also, the effect appeared most important during the first 3 days after injury. An extensive literature review of cervical spinal cord injuries by the Congress of Neurological Surgeons also recommends maintaining MAP between 85 and 90 mm Hg during the first 7 days after injury. There has been extensive research into various interventions to discover any possible clinical benefit that may aid patients with spinal cord injury. One such intervention that initially gained clinical acceptance was the use of high-dose methylprednisolone using the National Acute Spinal Cord Injury Study (NASCIS) II and then the NASCIS III protocols. Subsequent evaluations of these studies found significant flaws in the data analysis, and the claimed benefits of corticosteroid use have not been realized. There is now level I and level II evidence showing that these high-dose protocols do not improve SCI recovery and are associated with significant harm. These protocols are generally not recommended as treatment options to patients because significant complications are associated with these very high corticosteroid doses, which outweigh any benefit.

Spinal Cord Syndromes

When evaluating patients with spinal cord injuries, incomplete injuries must be distinguished from those that are complete because treatment decisions are based in part on this determination. If a complete spinal cord injury exists, the patient may regain some function within the zone of partial preservation but needs to understand that functional recovery at a more caudal level is not to be expected. This determination cannot be made until spinal shock has resolved and a reliable detailed neurologic examination is possible. In the case of an incomplete spinal cord injury, there are several recognized syndromes. If the injury can be categorized as one of these syndromes, prognostic information can be provided to the patient in general terms, but determination of specific functional recovery remains impossible at this time. There are, however, some generalizations that help inform the patient: (1) the greater the sparing of motor and sensory function is caudal to the injury, the greater is the expected recovery; (2) the earlier that recovery appears and the more rapidly it progresses, the greater is the expected recovery; (3) patients younger than 50 years have a better prognosis than older patients with the same deficit; and (4) recovery can occur over 12 to 15 months, but once progress ceases further recovery should not be expected. The most recognized syndromes are central cord syndrome, Brown-Séquard syndrome, anterior cord syndrome, posterior cord syndrome, conus medullaris syndrome, and cauda equina syndrome. There are some injuries that do not fit well into these described syndromes, and prognostic information cannot be given for these mixed syndromes.

Central cord syndrome is the most common. It consists of injury to the central area of the spinal cord, including gray and white matter ( Fig. 41.7B ). The centrally located upper extremity motor neurons in the corticospinal tracts are the most severely affected, and the lower extremity tracts are affected to a lesser extent. Generally, patients have a tetraparesis involving the upper extremities to a greater degree than the lower extremities with greater dysfunction distally in the extremities than proximally. Sensory sparing varies, but usually sacral pinprick sensation is preserved. These patients frequently show early partial recovery and may have preexisting cord compression and may not have spinal instability. Prognosis varies, but more than 50% of patients have return of bowel and bladder control, become ambulatory, and have improved hand function. This syndrome usually results from a hyperextension injury in an older individual with preexisting osteoarthritis of the spine. The spinal cord is pinched between the vertebral body anteriorly and the buckling ligamentum flavum and lamina posteriorly ( Fig. 41.7A ). It also may occur in younger patients with flexion injuries.

Management of acute traumatic central cord syndrome (ATCCS) remains controversial with regard to operative or nonoperative treatment superiority. A recent systematic review by the Spinal Cord Society and the Spine Trauma Study Group found important questions have not been answered. The study by Karthik Yelamarthy et al. did find evidence indicating that for ATCCS patients with instability caused by fractures, disc disruptions, or dislocations and persistent cord compression, outcomes are improved with early surgery (within 24 hours). For ATCCS patients without instability, there is no clear evidence that surgery or conservative treatment is superior. Both groups generally improve, and 70% to 80% of patients can expect to regain bladder control and the ability to ambulate with improvement in hand function relative to immediately after injury, although full recovery is not common. Because of the clinical ambiguity, our approach in these patients is initially to manage the spinal cord injury medically and monitor neurologic improvement. Once improvement plateaus after several weeks, a decision is made regarding surgery based on the functional level of the patient. For patients with instability, early surgery to achieve stability and relieve residual compression is recommended.

Brown-Séquard syndrome is an injury to either side of the spinal cord ( Fig. 41.7C ) and usually is the result of a unilateral laminar or pedicle fracture, penetrating injury, or rotational injury resulting in a subluxation. It is characterized by motor weakness with loss of proprioception on the side of the lesion and the contralateral loss of pain and temperature sensation. Prognosis for recovery is good, with significant neurologic improvement often occurring. Most of the recovery occurs in the first few months after injury, but improvement can occur over 2 years. Gait usually recovers within 6 months. Pollard and Apple noted that only central cord and Brown-Séquard syndromes were statistically associated with improved recovery at 2 years after injury. In carefully selected patients, nerve transfers may be of benefit.

Anterior cord syndrome usually is caused by a hyperflexion injury in which bone or disc fragments compress the anterior spinal artery and cord. It is characterized by complete motor loss and loss of pain and temperature discrimination below the level of injury. The posterior columns are spared to varying degrees ( Fig. 41.7D ), resulting in preservation of deep touch, position sense, and vibratory sensation. Prognosis for significant recovery in this injury is poor. Posterior cord syndrome involves the dorsal columns of the spinal cord and produces loss of proprioception and vibratory sense while preserving other sensory and motor functions. This syndrome is rare and usually is caused by tumors but can occur with an extension injury.

Conus medullaris syndrome, or injury of the sacral cord (conus) and lumbar nerve roots within the spinal canal, usually results in areflexic bladder, with urinary retention with overflow incontinence, areflexic bowel with fecal incontinence, and lower extremity weakness with increased tone that initially may be flaccid. Most of these injuries occur between T11 and L2 and result in flaccid paralysis in the lower extremities and loss of all bladder and perianal muscle control. The irreversible nature of this injury to the sacral segments is evidenced by the persistent absence of the bulbocavernosus reflex and the perianal wink. Motor function in the lower extremities between L1 and L4 may be present if nerve root sparing occurs.

Cauda equina syndrome, or injury between the conus and the lumbosacral nerve roots within the spinal canal, also can result in an areflexic bladder, bowel, and lower limbs. With a complete cauda equina injury, all peripheral nerves to the bowel, bladder, perianal area, and lower extremities are lost and the bulbocavernosus reflex, anal wink, and all reflex activity in the lower extremities are absent, indicating absence of any function in the cauda equina. The cauda equina injuries are lower motor neuron injuries, and there is a possibility of return of function of the nerve rootlets if they have not been completely transected or destroyed. Most often, cauda equina syndrome manifests as a neurologically incomplete lesion.

Radiographic Evaluation Protocol

The helical CT scan is the imaging modality of choice for the diagnosis of cervical fractures and dislocations. Axial images, sagittal reconstructions, and coronal plane reconstructions each provide optimal visualization for particular injuries. Having a systematic and methodical routine for viewing these series is required to detect injuries. Beginning with the sagittal reconstructions, three images are of particular value. These are the midline image and each of the parasagittal plane images through the occipital condyle-C1 joint and the facet joints on each side. These parasagittal images should be evaluated specifically for (1) congruity of the occipital condyle-C1 joint, which should be concentric and should not be more than 2 mm wide laterally, (2) intact isthmus at the C2 level, and (3) a normal relationship at each facet joint and intact lateral masses. The midline image should be evaluated specifically for (1) relation of Wackenheim’s line to the dens (normally tangential to the posterior aspect of the dens), (2) widening of the atlantodens interval (normal <3 mm; abnormal >5 mm), (3) soft-tissue swelling at the C3 midbody (normal <5 mm), (4) bony integrity of the dens, (5) anterior vertebral body alignment; (6) posterior vertebral body alignment, (7) alignment of the spinolaminar line, and (8) assessment for excess angulation or widening of each disc space.

The coronal reconstructions are best for evaluating the occiput-C1 joints, the C1-C2 joints, and the bony integrity of the dens.

The continuity of the posterior bony arch at each cervical level and the occiput is best determined on the axial images. Fractures involving the body, pedicle, foramen transversarium, lateral mass, lamina, and spinous process can be seen at individual levels ( Fig. 41.8 ).

Halo Vest Immobilization and Cervical Orthoses

Cervical immobilization is a mainstay of treatment for many cervical injuries. There is extensive clinical experience covered in the orthopaedic literature over many years regarding cervical immobilization. This literature base is mostly level III and level IV evidence studies. Unfortunately, controlled randomized prospectively collected data on specific means of immobilization for specific injuries are not available. It is unlikely such data will become available given the difficulty of devising an ethical study that could appropriately collect this information. A cadaver study of various spinal orthoses, including rigid collar, sternal occipital mandibular immobilizer (SOMI), and halo vest, by Holla et al. found that generally the reduction in flexion/extension occurred at C0-C1 level and the reduction in rotation occurred at C1-C2. Additionally, they found that restriction in motion was progressively increased in order from rigid collar to SOMI to halo vest.

The first modern halo vest was developed at Ranchos Los Amigos and described by Perry and Nickel in 1959. Numerous modifications have been made to the halo vest, and other orthoses for the cervical spine have been developed. These orthoses generally have been designed to serve one of two purposes: immobilization during extrication and transport procedures or adjunctive or definitive treatment for unstable cervical injuries. The adjunctive role is either as temporary immobilization preoperatively or to provide immobilization after surgical stabilization. The goals of stable fixation and early mobilization are appropriate with spinal injuries, but often a short period of external support is recommended after surgery.

Extrication-type collars are not appropriate for treatment because they are too restrictive and would cause skin breakdown with prolonged use. They should be exchanged or removed if immobilization is not needed after initial assessment of the patient. The most commonly used types of orthoses for the cervical spine include a soft collar, a two-piece “rigid” collar, a SOMI, a Minerva (similar to a SOMI with some forehead control), and a halo vest. Several authors have compared the relative ability of these devices to limit motion in the cervical spine. Studies comparing limitation of motion in normal volunteers using devices of the same basic type usually have not found statistically significant differences between devices within the same class. These studies generally have shown progressively more limitation of motion by the orthosis type in the sequence they are listed above. These studies usually measure global motion of the cervical region and are limited in that the study participants do not have cervical injuries and as such their spinal biomechanics may be different than patients. Other authors have used cadaver models to assess the effectiveness of different orthoses in limiting motion after instability is created at a specific cervical level. Richter et al. studied an odontoid fracture model and found the halo vest to be more effective than a two-piece collar or a Minerva type brace. In another cadaver study, Horodyski et al. found that a two-piece rigid collar did not limit motion effectively after severe C5-C6 instability was created. Other studies have found atypical motion, such as “snaking,” at individual levels that is caused by orthoses, especially the more restrictive types, during activities of daily living. Further studies are needed to evaluate the effect of these devices with mastication, swallowing, and oral hygiene, although these devices have been shown to affect these activities.

The halo vest has been studied more than other types of braces, and several findings have been determined. The halo vest is the most effective brace for limiting motion within the cervical spine. This appears true for the craniocervical junction, subaxial region, and cervicothoracic junction. Motion is allowed to a greater extent in the junctional areas than in the midcervical region in the halo vest. However, it is clear that motion remains throughout the cervical spine even with a halo vest properly applied. Despite this persistent motion, the halo vest has proven effective in the management of many types of cervical injuries, especially bony injuries involving the craniocervical junction. As surgical methods have improved, the halo vest has remained useful in part because for many upper cervical injuries normal motion can be preserved after fracture union. This region is responsible for a large portion of the normal cervical spine, and this motion is often permanently sacrificed with operative stabilization.

The use of halo vest immobilization does have significant associated complications. Recently, several studies have examined the morbidity and mortality associated with immobilization in a halo vest in elderly blunt trauma victims; however, no high-quality studies have prospectively evaluated this subgroup of patients. Retrospective studies in the trauma literature have noted an increased mortality rate in elderly trauma patients with cervical fractures treated with immobilization with a halo vest compared with those treated operatively or with a collar.

In institutions with higher death rates in patients with cervical spine injuries, higher rates of respiratory complications and deep vein thrombosis also were noted, suggesting that this group may not have been mobilized as well as the other subgroups evaluated. In a more thorough but still retrospective evaluation, Bransford et al. did not find an increased death rate associated with use of a halo vest. This study, which was a retrospective review of all patients at a level I trauma center for 8 years, evaluated treatment outcomes, complications, injury type, and patient age. Successful treatment was reported in 85% of patients treated with halo vest immobilization, although 11% of patients had the time in a halo vest shortened because of complications such as pin site infections. Treatment success was defined as healing of the injury in satisfactory alignment without additional intervention or secondary neurologic deterioration. The adverse events encountered in this study included death, pin site problems, pulmonary deterioration, skin breakdown, dysphagia, neurologic deterioration, and other miscellaneous complications. Twenty-two of 311 patients died after halo vest immobilization was initiated, and 19 of these deaths were within 21 days of starting halo vest immobilization. Review by a seven-member panel as to the cause of death, contributing comorbidities, and specifically whether the halo vest immobilization was a contributing cause of death was done in each case. It was determined that all 22 patients died for reasons that were not attributable to halo vest immobilization. The most common region treated with halo vest immobilization was the occiput to C2, especially odontoid fractures, although about a third of patients had subaxial injuries. Also, there were a significant proportion of study patients with more than one injury.

Complications of halo vest immobilization are frequent, with some studies having complication rates as high as 59%, although most studies identify complications in about 35% of patients. The most common type of complication involves pin site infection or loosening, which accounts for about 40% of all complications. Most pin site infections respond well to oral antibiotics if started early. Local pin cleaning daily and close follow-up of these patients allow early detection of these problems. Occasionally, infections are more serious and require pin site change or early discontinuation of halo vest immobilization. The most serious infections, which rarely occur, can lead to intracranial abscess requiring debridement and possibly result in death. Other less common pin-related complications include dural penetration, loosening without infection, or even skull fracture at or near the pin site. Another common complication of halo vest immobilization is failure to maintain adequate fracture reduction and spinal alignment. Rates of persistent instability with halo vest immobilization are 30% to 35% in most series. Most of these complications are detected in the first 7 to 10 days if radiographic imaging at the time halo vest immobilization is started is compared with imaging obtained after mobilization has been accomplished. Conversion to an alternative treatment may be necessary if alignment is not maintained because of the increased probability of nonunion. Nonunion detected after adequate halo vest immobilization for 12 to 16 weeks also may require surgical stabilization. Neurologic deterioration secondary to persistent instability also is a concern, although this is not common with halo vest immobilization. More serious complications, such as pneumonia or respiratory insufficiency, can occur but most often are related to inadequate mobilization of the patient. If a determination is made that adequate stability will not be attained with halo vest immobilization to allow mobilization to the full extent that the patient’s other injuries would allow, then other treatment should be undertaken if possible. In this way, most of the serious complications can be avoided. Most of the later but less serious complications related to the pins are avoided by using care in applying the halo vest immobilization and by having appropriate follow-up.

Occipitocervical Dissociation Injury Patterns

Injuries to the craniocervical junction can occur at a variety of locations. Atlantooccipital dislocations, C1-C2 dislocations, or combinations of fractures and dislocations involving the occiput, atlas, and the axis also can disrupt soft tissues, such as the tectorial membrane, alar and apical ligaments, transverse atlantal ligament (TAL), and joint capsules at occiput-C1 or C1-C2 joints. Some injuries such as fracture of the occipital condyle or isolated joint capsule injuries may be stable. However, these injuries may occur as components of a more complex injury with occipital cervical instability, which can be fatal if not treated. Often these injuries result in fatalities before the patient is transported. The diagnosis of craniocervical junction injuries requires awareness of and suspicion for the expected injury patterns. The presence of cranial nerve (CN) VI, CN X, or CN XII palsies, subarachnoid hemorrhage at the craniocervical junction, or soft-tissue swelling anterior to the upper cervical spine should increase suspicion of a craniocervical injury. More severe deficits, including monoparesis, hemiparesis, tetraparesis, apnea, or other high cord symptoms, also have been reported with these injuries. Careful evaluation of the CT images, particularly the reconstruction images, is needed because these injuries often are dislocations and only the relative position of one bony structure to another may be abnormal without the presence of a fracture. Atlantooccipital dislocation has become recognized more frequently as awareness of the injury has increased and initial patient care has improved. The best method for the diagnosis of atlantooccipital dislocation has not been definitively determined.

Older methods based on lateral radiographs such as the Power’s ratio (basion to posterior arch distance/opisthion to anterior arch distance) have been described. Harris et al. described measuring the basion atlas interval and the basion dens interval (BAI-BDI), both of which should be less than 12 mm ( Fig. 41.10 ). The BAI-BDI method as described by Harris et al. is the most reliable method using lateral radiographs. With the use of helical CT scans, more detailed analysis of the bony relationships is possible. The method that we have used to diagnose the presence of atlantooccipital dislocation is the Harris method, and we evaluate each of the occipital condyle-C1 joints for congruity and concentricity. Normally, these joints measure 0.5 to 1 mm and should be concentric. In a radiographic study by Martinez del Campo et al., a joint space of 1.5 mm or more was highly sensitive for atlantooccipital dislocation type of injury. If both joints are normal, there is no atlantooccipital dislocation. In addition, the relationship of Wackenheim’s line to the dens is evaluated. If this relationship also is normal, there is no distraction injury between C1 and C2. The most commonly used classification system for atlantooccipital dislocation is the Traynelis system, which is described by direction of displacement, but it lacks treatment guidance. The Traynelis classification includes type I (anterior); type II (longitudinal); type III (posterior); and “other,” which includes lateral or multidirectional displacement. Review of the literature revealed that patients with occipitocervical displacement who were not initially diagnosed had neurologic worsening 73% of the time before the diagnosis was recognized and about half did not improve even to their baseline neurologic examination after treatment. Ten percent of patients placed in traction had neurologic worsening in a small number of reported cases. Also, patients treated definitively with external immobilization excluding traction had a 40% rate of neurologic worsening that necessitated stabilization. Another 27% who did not worsen neurologically failed to achieve stability even after up to 22 weeks of immobilization. Patients treated with halo vest immobilization temporarily while awaiting operative stabilization had 0% neurologic worsening preoperatively. This evidence is level III but has led us to recommend operative stabilization for all patients with unstable occipitocervical dislocations. Initial management is in a halo vest to provide provisional stabilization until the patient can undergo posterior occipitocervical fusion. Traction is not used under usual circumstances. Typically, fusion is from the occiput to C2 or C3, with multiple points of skull fixation and C1 lateral mass screws, C2 isthmus screws, and, when needed, C3 or lower lateral mass screws with autologous bone grafting. Some injuries to individual craniocervical structures without dislocation can be treated without operative stabilization.

Occipital Condyle Fractures

Fractures of the occipital condyle are recognized more often now with increased use of screening CT with reformatted images ( Fig. 41.11 ). They occur in association with traumatic brain injuries in over half the cases, and frequently patients have additional cervical fractures. Dysfunction of cranial nerves is uncommon, but involvement of CN VI, CN IX, CN X, CN XI, and CN XII has been reported. Cranial nerve palsies most often are reported when fractures of the occipital condyle are untreated. These fractures do occur as isolated injuries but are most significant when they occur as part of a more severe craniocervical injury, such as occipital cervical dislocation. The occipital condyles articulate with the C1 lateral masses and are attached to the dens by the paired alar ligaments. The alar ligaments function to limit rotation of the occiput and atlas with respect to C2. The mechanisms for fractures of the occipital condyle usually are axial loading and lateral bending. Anderson and Montesano described the classification that is most commonly used: type I, impaction; type II, basilar skull fracture; and type III, avulsion fracture. Type I and type II fractures are usually stable and can be treated with a rigid orthosis for 6 to 12 weeks. About 6 to 8 weeks of immobilization in a rigid orthosis is usually recommended; there is no good evidence to support a specific treatment period. If instability is detected after a period of adequate immobilization, occiput to C2 fusion may be indicated. Type III fractures are potentially unstable, especially if displaced more than 2 mm, because of the avulsion of the alar ligament, which may be bilateral. Treatment in a halo vest for 12 weeks or surgical management may be needed for the rare unstable occipital condyle fracture. Occipital condyle fractures are most significant as indicators of high-energy blunt trauma to the head and neck. A large retrospective study by West et al. found the incidence to be 0.3% of the study population, but 30% of patients had associated intracranial injuries and 43.5% had significant other cervical injuries. Most were treated in a rigid orthosis, some with observation, and none required surgery.

Transverse Atlantal Ligament Rupture

Rupture of the TAL or cruciform ligament usually occurs from a force applied to the back of the head, such as occurs in a fall. Thus, injuries involving the TAL can be a purely ligamentous midsubstance tear of the ligament or can occur as the result of an avulsion of the insertion into the C1 lateral mass. Dickman et al. classified these injuries as type I, disruptions of the substance of the ligament, and type II, fractures and avulsions involving the tubercle insertion of the TAL on the lateral mass of C1. Treatment is based on classification type. According to Dickman et al., type I injuries are incapable of healing without internal fixation and they should be treated with early surgery. Type II injuries, which render the transverse ligament physiologically incompetent even if the ligament substance is not torn, should be treated initially with a rigid cervical orthosis. Dickman et al. had a 74% success rate with nonoperative treatment of type II injuries, reserving surgery for patients who had a nonunion and persistent instability after 3 to 4 months of immobilization. Conversely, 26% of type II injuries in this study failed to heal after immobilization, suggesting that close follow-up is needed to determine which patients require delayed operative intervention. Usually the anterior subluxation of the ring of C1 can be detected on flexion films and the instability can be reduced in extension ( Fig. 41.12 ). Sagittal midline reformatted CT views should be checked carefully for retropharyngeal swelling, which suggests an acute injury, and for small flecks of bone avulsed off the lateral masses of C1, which may indicate avulsion of the ligament. These avulsed fragments are best seen on coronal reformatted views. The primary indication of this injury is instability at C1-C2 on flexion and extension films. Anterior widening of the atlantodens interval of more than 3 mm on the midsagittal CT reconstruction or on a flexion view suggests that the transverse ligament is incompetent. MRI has become the standard imaging modality to evaluate the integrity of the TAL. Flexion and extension views should be made under the supervision of the physician, and the patient must be monitored closely for alterations in neurologic or respiratory function. As described by Dickman et al., mid-substance injury of the TAL will not heal with immobilization, and operative treatment is indicated. Posterior C1-C2 fusion using the fixation technique described by Harms with autologous bone graft is preferred. This technique is more rigid than wiring and has an advantage over wiring methods in that it can be used in the presence of fractures of the posterior ring of C1. An alternative fixation method is the Gallie method of wiring that creates a posteriorly directed force on C1 to reduce any atlantodens interval widening (see Fig. 41.12C ). An intact dens will prevent over-reduction of C1. A Brooks-Jenkins bone block technique should not be used because it cannot maintain the reduction as well. In 12 patients with ruptures of the transverse ligament, Levine and Edwards found an average loss of correction of 4 mm after bone block techniques and 1 mm after Gallie wiring. Isolated TAL injuries without associated atlas fractures are rare.

Atlas Fractures

The first description of a C1 fracture was by Cooper in 1822, and Jefferson published his case review adding four new cases in 1920. This paper contained his classification system, which has subsequently been revised by multiple authors, but his description of a burst fracture of the ring of C1 continues to carry the label of a “Jefferson fracture” ( Fig. 41.15 ). Spence et al. published their work in 1970 on injuries to the transverse ligament in association with C1 fractures in 10 cadaver specimens. They found that if the total lateral displacement of the lateral masses was 6.9 mm or more, then the transverse ligament was likely incompetent ( Fig. 41.16 ). This determination based on plain radiographs is referred to as the rule of Spence. Later, this was revised to 8.1 mm to account for magnification on plain radiographs. Dickman et al. studied 39 patients with injuries to the TAL with plain radiographs, thin-cut CT, and high-resolution MRI. MRI was found to be very sensitive in detecting rupture of the transverse ligament, and their classification of these injuries was described previously. These authors found that applying the rule of Spence would have missed 61% of the transverse ligament injuries.

Biomechanical studies by Panjabi et al. and Oda et al. have shown that axial loading is the primary force that leads to C1 fractures. Because the C1 lateral masses are wedge shaped, axial loading creates a hoop stress and bone failure occurs at the weakest points that are just anterior and posterior to the lateral masses. Less force is required if the head is in extension when force is applied. Even when the TAL is injured, the alar ligaments, joint capsules, and tectorial membrane are spared with axial-loading injuries. This is an important difference for TAL injuries associated with C1 fractures and those associated with more complex injuries of the craniocervical junction. Landells and Van Peteghem modified Jefferson’s classification into three fracture types, which is useful for treatment. Type I injuries include isolated anterior or posterior arch fractures, type II injuries involve the anterior and posterior portion of the ring, and type III injuries involve the lateral mass with or without a fracture of the ring. Gehweiler et al. also described a classification system that is perhaps the most useful. Type I is an isolated anterior arch fracture, type II is an isolated posterior arch fracture, type III is a combined anterior and posterior arch fracture, type IV is a fracture involving only the lateral mass usually with a sagittally oriented fracture, and a type V fracture is through the foramen transversarium of C1 ( Fig. 41.17 ).

Axis Fractures

The most common fractures of the axis are those involving the odontoid process. The remaining fractures are those involving the isthmus (hangman fracture), which are the next most common fracture patterns of the axis body. Although any of these fracture types can occur with concomitant cervical injuries, they frequently occur as isolated fractures and are discussed separately. Odontoid fractures are especially common in the elderly, and in this patient group the most common mechanism is a low-energy fall.