Handbook of Neurosurgery 7th Ed

28. Spine injuries

20% of patients with a major spine injury will have a second spinal injury at another level, which may be noncontiguous. These patients often have simultaneous but unrelated injuries (e.g. chest trauma). Injuries directly associated with spinal cord injuries include arterial dissections (carotid and/or vertebral arteries).



A conceptual definition of clinical stability from White and Panjabi1: the ability of the spine under physiologic loads to limit displacement so as to prevent injury or irritation of the spinal cord and nerve roots (including cauda equina) and, to prevent incapacitating deformity or pain due to structural changes.

Biomechanical stability refers to the ability of the spine ex vivo to resist forces.

For models of stability for cervical spine injuries see page 969, for thoracolumbar fractures see page 986.


There is disagreement over what should be defined as “the level” of a spinal cord injury. Some use the lowest level of completely normal function. However, most sources define the “level” as the most caudal segment with motor function that is at least 3 out of 5 and if pain and temperature sensation is present.


Incomplete lesion

Definition: any residual motor or sensory function more than 3 segments below the level of the injury2. Look for signs of preserved long-tract function.

Signs of incomplete lesion:

• sensation (including position sense) or voluntary movement in the LEs

• “sacral sparing”: preserved sensation around the anus, voluntary rectal sphincter contraction, or voluntary toe flexion

• an injury does not qualify as incomplete with preserved sacral reflexes alone

Types of incomplete lesion:

• central cord syndrome: see page 948

• Brown-Séquard syndrome (cord hemisection): see page 950

• anterior cord syndrome: see page 950

• posterior cord syndrome: rare, see page 951

Complete lesion

No preservation of any motor and/or sensory function more than 3 segments below the level of the injury. About 3% of patients with complete injuries on initial exam will develop some recovery within 24 hours. Recovery is essentially zero if the spinal cord injury remains complete beyond 72 hours.


This term is often used in two completely different senses:

1. hypotension (shock) that follows spinal cord injury (SBP usually ≈ 80 mm Hg). See Hypotensionpage 935 for treatment. Caused by multiple factors:

A. interruption of sympathetics: implies spinal cord injury above T1

1. loss of vascular tone (vasoconstrictors) below level of injury

2. leaves parasympathetics relatively unopposed causing bradycardia

B. loss of muscle tone due to skeletal muscle paralysis below level of injury results in venous pooling and thus a relative hypovolemia

C. blood loss from associated wounds → true hypovolemia

2. transient loss of all neurologic function (including segmental and polysynaptic reflex activity and autonomic function) below the level of the SCI34 → flaccid paralysis and areflexia lasting varying periods (usually 1-2 weeks, occasionally several months and sometimes permanently), the resolution of which yields the anticipated spasticity below the level of the lesion. A poor prognostic sign. Spinal cord reflexes immediately above the injury may also be depressed on the basis of the Schiff-Sherrington phenomenon

28.1. Whiplash-associated disorders

“Whiplash” was initially a lay term, which is currently defined as a traumatic injury to the soft tissue structures in the region of the cervical spine (including: cervical muscles, ligaments, intervertebral discs, facet joints…) due to hyperflexion, hyperextension, or rotational injury to the neck in the absence of fractures, dislocations, or intervertebral disc herniation5. It is the most common non-fatal automobile injury6. Symptoms may start immediately, but more commonly are delayed several hours or days. In addition to symptoms related to the cervical spine, common associated complaints include headaches, cognitive impairment, and low back pain.

A proposed clinical classification system of WAD is shown in Table 28-1. A consensus regarding diagnosis and management of these injuries is shown in Table 28-2 and Table 28-3. Keep in mind that conditions such as occipital neuralgia may occasionally follow whiplash type injuries and should be treated appropriately (see page 804).

Table 28-1 Clinical grading of WAD severity




no complaints, no signs*



neck pain or stiffness or tenderness, no signs


above symptoms with reduced range of motion or point tenderness


above symptoms with weakness, sensory deficit, or absent deep tendon reflexes


above symptoms with fracture or dislocation*

* the definition of whiplash excludes these patients5

Table 28-2 Evaluation of WAD

Grade 1 patients with normal mental status and physical exam do not require plain radiographs on presentation

Grade 2 & 3 patients: C-spine x-rays, possibly with flexion-extension views. Special imaging studies (MRI, CT, myelography…) are not indicated

Grade 3 & 4: these patients should be managed as suspected spinal cord injury (see Initial management of spinal cord injury below, and sections that follow)



In a study of 117 patients < 56 years of age having WAD due to automobile accidents (excluding those with cervical fractures, dislocations, or injuries elsewhere in the body) conducted in Switzerland8 (where all medical costs were paid by the state and there was no opportunity for litigation and no compensation for pain and suffering, although there was the possibility of permanent disability), the recovery rate was as shown in Table 28-4. Of the 21 patients with continued symptoms at 2 yrs, only 5 were restricted with respect to work (3 reduced to part-time work, 2 on disability). Patients with persistent symptoms were older, had more varied complaints on initial exam, had a more rotated or inclined head position at the time of impact, had a higher incidence of pretraumatic headaches, and had a higher incidence of certain pre-existing findings (such as radiologic evidence of cervical osteoarthritis). The amount of damage to the automobile and the speed of the cars has little relationship to the degree of injury, and outcome was not influenced by gender, vocation, or psychological factors.

Table 28-4 Recovery of patients with WAD

Time (mos)

Percent recovered









28.2. Pediatric spine injuries

Spinal cord injury is fairly uncommon in children, with the ratio of head injuries to spinal cord injuries being ≈ 30:1 in pediatrics. Only ≈ 5% of spinal cord injuries occur in children. Due to ligamentous laxity together with a high head to body weight ratio, immaturity of paraspinal muscles and the underdeveloped uncinate processes, these tend to involve ligamentous rather than bony injuries (see SCIWORA, page 974). The cervical spine is the most vulnerable segment (with subaxial injuries being fairly uncommon), with 42% of injuries occurring here, 31% thoracic, and 27% lumbar. The fatality rate is higher with pediatric spine injuries than with adults (opposite to the situation with head injury), with the cause of death more often related to other severe injuries than to the spinal injury9.



Level II10: C-spine x-rays are not indicated in pediatric trauma patients who are:

• alert & conversant

• neurologically intact

• without posterior midline cervical tenderness (with no distracting pain)

• and who are not intoxicated

Level III10: for pediatric trauma patients who are: nonconversant or have altered mental status, neurologic deficit, neck pain or a painful distracting injury, are intoxicated, or have unexplained hypotension:

• patients < 9 yrs: AP & lateral C-spine x-rays

• patients ≥ 9 yrs: open-mouth odontoid view in addition to the above

• supplement these x-rays with additional thin cut CT through areas of suspicion or areas not visualized on plain x-ray

• flexion-extension C-spine x-rays or fluoroscopy may be considered to R/O ligamentous instability if there is still a suspicion of instability after the above x-rays are obtained

• consider: C-spine MRI to R/O cord or nerve root compression, evaluate ligamentous integrity, or provide information for neurologic prognosis


Level III10:

• children < 8 yrs age: immobilize with thoracic elevation or an occipital recess (allows more neutral alignment due to the relatively large head)

• children < 7 yrs age with injuries of the C2 dentocentral synchondrosis (see page 137): closed reduction and halo immobilization

• consider: primary operative treatment for isolated C-spine ligamentous injuries with associated deformity


For pediatric C-spine anatomy see page 137. In the age group ≤ 9 yrs, 67% of cervical spine injuries occur in the upper 3 segments of the cervical spine (occiput-C2)11.

SYNCHONDROSES (see page 137)

Normal synchondroses may be mistaken for fractures (especially the dentocentral synchondrosis of the atlas (see page 137) which may be mistaken for an odontoid fracture). Conversely, actual fractures may occur through synchondroses1213. Recommended treatment for fractures through synchondroses: the tendency for synchondroses to fuse suggests that emergency reduction followed by external immobilization be attempted. Internal immobilization/fusion should be reserved for persistent instability13.


Pseudospread of the atlas (defined as > 2 mm total overlap of the two C1 lateral masses on C2 on AP open-mouth view) is present in most children 3 mos to 4 yrs age. Prevalence is 91-100% during the second year of life. Youngest example at 3 mos, oldest at 5.75 yrs. Normal total offset is typically 2 mm during the first year, 4 mm during the second, 6 mm during the third, and decreasing thereafter. The maximum is 8 mm. Trauma is not a contributing factor.

Pseudospread is probably a result of disproportionate growth of the atlas on the axis. This could be misdiagnosed as a Jefferson fracture (see page 958), which rarely occurs prior to the teenages (owing to lower weight of children, more flexible necks, increased plasticity of skull, and shock absorbing synchondroses of C1).

Neck rotation can also sometimes simulate the appearance of a Jefferson fracture.

When suspicion of fracture is high: CT scan through C1 can resolve the issue of whether or not there is a fracture.


Either anterior displacement of C2 (axis) on C3 and/or significant angulation at this level. Seen in children (up to age 10 yrs) on lateral C-spine x-ray after trauma. Up to age 10 yrs, flexion and extension are centered at C2-3; this moves down to C4-5 or C5-6 after age 10. C2 normally moves forward on C3 up to 2-3 mm in peds15. When the head is flexed, displacement is expected; may be exacerbated by spasm16. Does not represent pathological instability. Fractures and dislocations are unusual in children, and when they do occur, they resemble those in adults.

10 cases reported between ages 4-6 yrs17: pain was not uncommon. In each case, either the head or neck was flexed (sometimes minimally); the pseudosubluxation corrected when x-ray was repeated with head in true neutral position.

Recommendation: treat patient for soft-tissue injury and not for subluxation.

28.3. Initial management of spinal cord injury

The major causes of death in spinal cord injury (SCI) are aspiration and shock4. Initial survey under ATLS protocol: assessment of airway takes precedence, then breathing, then circulation & control of hemorrhage (“ABC’s”). This is followed by a brief neurologic exam.

NB: other injuries (e.g. abdominal injuries) may be masked below the level of SCI.

Any of the following patients should be treated as having a SCI until proven otherwise:

1. all victims of significant trauma

2. trauma patients with loss of consciousness

3. minor trauma victims with complaints referable to the spine (neck or back pain or tenderness) or spinal cord (numbness or tingling in an extremity, weakness)

4. associated findings suggestive of SCI include

A. abdominal breathing

B. priapism (autonomic dysfunction)

The orientation of the management differs based on the patient’s situation as follows:

1. no history of significant trauma, completely alert, oriented and free of drug or alcohol intoxication with no complaints referable to the spine: most may be cleared clinically without the need for C-spine x-rays

2. significant trauma, but no strong evidence of spine or spinal cord injury: the emphasis here is in ruling-out a bony lesion and preventing injury

3. patients with neurologic deficit: the emphasis here is to define the skeletal injury and to take steps to prevent further cord injury and loss of function and minimize or reverse the present deficit. The pros and cons of the high-dose methylprednisolone protocol (see page 936) should be weighed if a neurologic deficit is identified


To date, there has not been a case of a significant occult cervical spine injury1819 in a trauma patient who met all of the criteria in Table 28-5A.


A. although reports of bony or ligamentous abnormalities have be described as possibly occurring in these patients, there has been no report of a patient who had neurologic injury as a result of these abnormalities


Table 28-5 Clinical criteria for cervical spine stability

1. awake, alert, oriented (no mental status changes, including no alcohol or drug intoxication)

2. no neck pain (with no distracting pain)

3. no neurologic deficits

Table 28-6 NATA helmet removal guidelines*

 NB: do not remove the helmet in the field.

• most injuries can be visualized with the helmet in place

• neurological exam can be done with the helmet in place

• the patient may be immobilized on a spine board with the helmet in place

• the facemask can be removed with special tools to access the airway

• hyperextension must be avoided following removal of the helmet and shoulder pads

In a controlled setting (usually after x-rays) the helmet and shoulder-pads are removed together as a unit to avoid neck flexion or extension

Possible indications for removal of helmet

• face mask cannot be removed in a reasonable amount of time

• airway cannot be established even with face mask removed

• life threatening hemorrhage under the helmet that can be controlled only by removal

• helmet & strap do not hold head securely so that immobilizing the helmet does not adequately immobilize the spine (e.g. poor fitting or damaged helmet)

• helmet prevents immobilization for transportation in an appropriate position

• certain situations where the patient is unstable (M.D. decision)

* for more details, see http://www.nata.org


1. immobilization prior to and during extrication from vehicle and transport to prevent active or passive movements of the spine. For possible C-spine injuries in football players, see Table 28-6 for the National Athletic Trainers’ Association (NATA) guidelines for helmet removal. When CPR is necessary it takes precedence. Caution with intubation (see below)

A. log-roll patient to turn

B. place patient on back-board

C. sandbags on both sides of the head with a 3 inch strip of adhesive tape from one side of the back-board to the other across the forehead immobilizes the spine as well as a rigid orthosis20 but allows movement of the jaw and access to the airway

D. a rigid cervical collar (e.g. Philadelphia collar) may be used to supplement

2. maintain blood pressure (see below under Hypotension)

A. pressors: treats the underlying problem (essentially a traumatic sympathectomy). Dopamine is the agent of choice, and is preferred over fluids (except as necessary to replace losses) (see Cardiovascular agents for shockpage 22 for pressors).  Avoid phenylephrine (see below)

B. fluids as necessary to replace losses

C. military anti-shock trousers (MAST): immobilizes lower spine, compensates for lost muscle tone in cord injuries (prevents venous pooling)

3. maintain oxygenation (adequate FIO2 and adequate ventilation)

A. if no indication for intubation: use NC or face mask

B. intubation

1. indications: may be required for

a. airway compromise

b. hypopnea:

i. from paralyzed intercostal muscles

ii. from paralyzed diaphragm (phrenic nerve = C3, 4 & 5)

iii. or from depressed LOC

2. caution with intubation with uncleared C-spine

a. use chin lift (not jaw thrust) without neck extension

b. nasotracheal intubation may avoid movement of C-spine but patient must have spontaneous respirations

c. avoided tracheostomy or cricothyroidotomy if possible (may compromise later anterior cervical spine surgical approaches)

4. brief motor exam to identify possible deficits (also to document delayed deterioration); ask patient to:

A. move arms

B. move hands

C. move legs

D. move toes


Basic phases of management with respect to the spine:

1. stabilization (medical & spinal), preliminary evaluation & treatment

2. evaluation of spinal stability

3. subsequent (definitive) treatment


Clinical Assessment

Level III21: the ASIA international standards for neurological and functional assessment of spinal cord injury (SCI) (see page 945) is recommended

Functional outcome assessment

Level II21: the Functional Impairment Measure™ (FIM™) (see page 1184) is recommended

Level III21: the modified Barthel index (see page 1183) is recommended


Level III22: monitor patients with acute SCI (especially those with severe cervical level injuries) in an ICU or similar monitored setting

Level III22: cardiac, hemodynamic & respiratory monitoring after acute SCI is recommended

Level III23: hypotension (SBP < 90 mm Hg) should be avoided or corrected ASAP

Level III23: maintain MAP at 85-90 mm Hg for the first 7 days after SCI to improve spinal cord perfusion


1. immobilization: maintain backboard/head-strap (see above) to facilitate transfers to CT table, etc. Once studies are completed, remove patient from backboard by logrolling (early removal from board reduces risk of decubitus ulcers)

2. hypotension (spinal shock): maintain SBP ≥ 90 mm Hg. Spinal cord injuries cause hypotension by a combination of factors (see page 930) which may further injure spinal cord24 or other organ systems

A. pressors if necessary: dopamine is agent of choice ( avoid phenylephrine: non-inotropic and possible reflex increase in vagal tone with bradycardia)

B. careful hydration (abnormal hemodynamics → propensity to pulmonary edema)

C. atropine for bradycardia associated with hypotension

3. oxygenation (see above)

4. NG tube to suction: prevents vomiting and aspiration, and decompresses abdomen which can interfere with respirations if distended (paralytic ileus is common, and usually lasts several days)

5. indwelling (Foley) urinary catheter: for I’s & O’s and to prevent distension from urinary retention

6. DVT prophylaxis: see below

7. temperature regulation: vasomotor paralysis may produce poikilothermy (loss of temperature control), this should be treated as needed with cooling blankets

8. electrolytes: hypovolemia and hypotension cause increased plasma aldosterone which may lead to hypokalemia

9. more detailed neuro evaluation (see ASIA (American Spinal Injury Association) motor scoring systempage 945). Patients may be stratified using the ASIA impairment scale (see Table 28-13page 947)

A. focused history: key questions should center on:

1. mechanism of injury (hyperflexion, extension, axial loading…)

2. history suggestive of loss of consciousness

3. history of weakness in the arms or legs following the trauma

4. occurrence of numbness or tingling at any time following the injury

B. palpation of the spine for point tenderness, a “step-off”, or widened interspinous space

C. motor level assessment

1. skeletal muscle exam (can localize dermatome)

2. rectal exam for voluntary anal sphincter contraction

D. sensory level assessment

1. sensation to pinprick (tests spinothalamic tract, can localize dermatome): be sure to test sensation in face also (spinal trigeminal tract can sometimes descend as low as ≈ C4)

2. light (crude) touch: tests anterior cord (anterior spinothalamic tract)

3. proprioception/joint position sense (tests posterior columns)

E. evaluation of reflexes

1. muscle stretch reflexes: usually absent initially in cord injury

2. abdominal cutaneous reflexes

3. cremasteric reflex

4. sacral

a. bulbocavernosus: see footnote, page 946

b. anal-cutaneous reflex

F. examine for signs of autonomic dysfunction

1. altered patterns of perspiration (abdominal skin may have low coefficient of friction above lesion, and may seem rough below due to lack of perspiration)

2. bowel or bladder incontinence

3. priapism: persistent penile erection

10. radiographic evaluation: see below

11. medical management specific to spinal cord injury:

A. methylprednisolone (see below)

B. experimental/investigational drugs: none of these agents shown to have un-equivocal benefit in man: naloxone, DMSO, Lazaroid®. Tirilazad mesylate (Freedox®) was less beneficial than methylprednisolone25



image Still highly controversial even among experts26-28

Level III: treatment with methylprednisolone for 24 or 48 hrs after SCI is an option that should be undertaken only with the knowledge that the evidence suggesting harmful side effects is more consistent than any demonstrated clinical benefit

It has been asserted that beneficial (sensory and motor) effects at 6 weeks, 6 months and 1 year are seen (for both complete and incomplete injuries) when methylprednisolone (MP) is administered as shown below, but only if given within 8 hours of injury (NB: outcome is possibly worse at 1 year if the drug is started after 8 hrs from injury)2930. See Critique below.

Exclusionary criteria from the study (these patients were not studied, and no determination was made whether the drug was helpful or not, or safe or not):

1. cauda equina syndrome (see page 446)

2. gunshot wounds (GSW) to the spine: a retrospective study showed no benefit and increased risk of complications with steroids with GSW31

3. life-threatening morbidity

4. pregnancy

5. narcotic addiction

6. age < 13 years

7. patient on maintenance steroids


1. concentration: in the following protocol, all solutions are mixed as 62.5 mg/ml (e.g. by diluting 16 gm methylprednisolone with bacteriostatic water to 256 ml)

2. bolus: 30 mg/kg initial IV bolus over 15 minutes, infused as shown in Eq 28-1 with an IV controller (this delivers 0.48 ml/kg of solution in 15 minutes):


3. followed by a 45 minute pause

4. maintenance infusion: then 5.4 mg/kg/hr continuous infusion as shown in Eq 28-2 (infusion is maintained during any necessary surgery if possible)


duration of maintenance infusion: when therapy is initiated ≤ 3 hrs after injury, the infusion is administered for 23 hrs. If therapy is started between 3 and 8 hrs of injury, there may be an incremental benefit in 47 hrs of infusion, with slightly higher risk of infection and pneumonia25


A metaanalysis32 of the literature could not identify any study that replicated the results of the original studies. At 1 year the MP group only showed a slight sensory advantage over the placebo group. Furthermore, high-dose MP may cause acute corticosteroid myopathy33 (ACM) which might indicate that some patients that improved after MP were actually recovering from their ACM. ACM and it’s associated complications (prolonged ventilator dependency…) should be added to the list of potential complications of high-dose MP (hyperglycemia, pneumonia, sepsis).


The position statement of the joint sections of the AANS and the CNS is that there is not enough evidence to recommend for or against local or systemic hypothermia for acute SCI, and that it should be noted that systemic hypothermia is associated with medical complications in TBI34.


Also see Thromboembolism in neurosurgerypage 42. Incidence of DVT may be as high as 100% when 125I-fibrinogen is used35. Overall mortality from DVT is 9% in SCI patients.


Level I36: prophylactic treatment of thromboembolism in patients with severe motor deficits due to SCI. Choices include:

• LMW heparin, rotating beds, adjusted dose heparin, or some combination of these measures

• or, low-dose heparin + pneumatic compression stockings or electrical stimulation

Level II36:

 not recommended: low-dose heparin used alone

 not recommended: oral anticoagulation alone

Level III36:

• duplex doppler ultrasound, impedance plethysmography & venography are recommended as diagnostic tests for DVT in patients with SCI

• vena cava interruption filters for patients who do not respond to, or are not candidates for, anticoagulation


Study of 75 patients found titrating dose of SQ heparin q 12 hrs to a PTT of 1.5 times control resulted in lower incidence of thromboembolic events (DVT, PE) than “mini-dose” heparin (5000 U SQ q 12 hrs) (7% vs. 31%)37. Heparin can cause thrombosis, thrombocytopenia and chronic therapy may produce osteoporosis (see heparinpage 39).



Asymptomatic trauma patients

Level I38 & Level II39: radiographic studies are not indicated* in patients who have:

• no mental status changes (and no evidence of alcohol or drugs)

• no neck pain or posterior midline tenderness (and no distracting pain)

• no focal neurologic deficit (on motor or sensory exam)

• and who do not have significant associated injuries that detract from their general evaluation

Trauma patients who are symptomatic, or obtunded or have unreliable exam (altered mental status, distracting injuries…)

Level II39: the primary screening modality is thin-cut axial CT from the occiput to T1 with sagittal and coronal reconstructions

Level II39: plain x-rays add no information and should not be done§

* these are basically the 5 NEXUS criteria from Hoffman et al.40: no midline cervical tenderness, no focal neuro deficit, normal alertness, no intoxication, no painful distracting injury. The Canadian C-Spine Rule (CCR) was found to be more sensitive & specific41, but the EAST has not embraced it as of this writing39

 altered mental status can include GCS ≤ 14; disorientation to person, place, time or events; inability to remember 3 objects at 5 minutes; delayed response to external stimuli

 evidence of alcohol or drugs includes information from the history, physical findings (slurred speech, ataxia, odor of alcohol on the breath) or positive blood or urine tests

§ unless CT scan is not available or contraindicated


A cervical collar is not needed in trauma patients who meet these criteria

• awake, alert, without neurologic deficit or distracting injury who have no neck pain or tenderness and full ROM of the cervical spine (Level II39)

• penetrating brain trauma: unless the trajectory suggests direct cervical spine injury (Level III39)

• Level III42 & Level III39: awake with neck pain or tenderness and normal cervical CT scan after either*

1. normal & adequate dynamic flexion-extension C-spine x-rays

2. or a normal cervical MRI is obtained

In obtunded patients with normal cervical CT scan and gross movement of all 4 extremities

•  flexion-extension C-spine x-rays should not be performed (Level II39)

• options:

1. maintain cervical collar until a clinical exam can be performed39

2. remove the collar on the basis of the normal CT scan alone39§

3. obtain cervical MRI: the risk and benefit of cervical MRI in addition to CT is unclear, and must be individualized (Level III39). If the MRI is normal, the collar may be safely removed (Level II39)

* these tests are performed in the absence of an identifiable fracture or obviously unstable dislocation to rule-out ligamentous or other soft-tissue injury that might be occult and unstable

 AANS/CNS guidelines from 2002 recommended getting the MRI within 48 hours42

 MRI is usually employed in this setting when the patient is unable to cooperate for flex-ext x-rays. For MRI findings and issues related to timing, etc., see page 941

§ the incidence of ligamentous injury with negative CT is < 5%, and the incidence of clinically significant injury is unknown but is much < 1%39

Cervical immobilization

Cervical collars should be removed as soon as it can be determined that it is safe to do so. The benefits from early collar removal include: reduction of skin breakdown43, fewer days of mechanical ventilation44, shorter ICU stays44, reduction of ICP4546.

Minimal radiographic evaluation

There is controversy regarding what constitutes a minimal radiographic evaluation of the cervical spine in multiple trauma patient. No imaging modality is 100% accurate.

Asymptomatic patients (meeting criteria outlined in PRACTICE GUIDELINE 28-6) may be considered to have a stable cervical spine and no radiographic studies of the cervical spine are indicated3942. Factors associated with increased risk of failing to recognize spinal injuries include: decreased level of consciousness (due to injury or drugs/alcohol), multiple injuries, technically inadequate x-rays47 (also, see Delayed cervical instabilitypage 982).

2009 recommendation of the Eastern Association for the Surgery of Trauma (EAST) is that the primary screening modality is thin-cut axial CT from the occiput to T1 with sagittal and coronal reconstructions39 which is more accurate than plain radiographs.

When CT scan is not appropriate as the initial screening exam, the following guidelines are offered:

See X-rays, C-Spine on page 135 for normal vs. abnormal findings. Table 28-7 lists some indicators that should alert the reviewer that there may be significant C-spine trauma (they do not indicate definite instability by themselves).

1. cervical spine: must be cleared radiographically from the cranio-cervical junction down through and including the C7-T1 junction (incidence of pathology at C7-T1 junction may be as high as 9%48):

A. lateral portable C-spine x-ray while in rigid collar: this study by itself will miss ≈ 15% of injuries50

B. if all 7 cervical vertebra AND the C7-T1 junction are adequately visualized and are normal, and if the patient has no neck pain or tenderness and is neurologically intactA, then remove the cervical collar and complete the remainder of the cervical spine series (AP and open-mouth odontoid (OMO) view). Lateral, AP, and OMO views together detect essentially all unstable fractures in intactA patients51(although the AP view rarely provides unique information52). In a severely injured patient, limitation to an AP and lateral view usually suffices for the acute (but not complete) evaluation53

C. if the above studies are normal, but there is neck pain, tenderness or neurologic findings (there may be a spinal cord injury even with normal plain films), or if the patient is unable to reliably verbalize neck pain or cannot be examined for neurologic deficit, then further studies are indicated, which may include any of the following:

1. oblique viewsB: demonstrates the neural foramina (may be blocked with a unilateral locked facet (see page 973)), and helps assess the integrity of the articular masses and lamina (the lamina should align like shingles on a roof)53

2. flexion-extension views: see below

3. CT scan: helpful in identifying bony injuries. However, CT cannot exclude significant soft-tissue or ligamentous injury54

4. MRI: utility is limited to specific situations (see page 941) and the accuracy has not been determined

5. polytomograms: becoming less available

6. pillar view: devised to demonstrate the cervical articular masses en face (reserved for cases suspected of having articular mass fracture)55: the head is rotated to one side (requires that the upper cervical spine injury has been excluded by previous radiographs), the x-ray tube is off centered 2 cm from midline in the opposite direction and the beam is angled 25° caudad, centered at the superior margin of the thyroid cartilage

7. if subluxation is present at any level and is ≤ 3.5 mm and the patient is neurologically intactA, obtain flexion-extension films (see below)

8. if no pathologic movement, may discontinue cervical collar

9. even if no instability is demonstrated, may need delayed films once pain and muscle spasms have resolved to reveal instability

D. if lower C-spine (and/or cervical-thoracic junction) are not well visualized

1. repeat lateral C-spine x-ray with caudal traction on the arms (if not contraindicated based on other injuries, e.g. to shoulders)

2. if still not visualized, then obtain a “swimmer’s” (Twining) view: the x-ray tube is positioned above the shoulder furthest from the film, and aimed towards the axilla closest to the film with the tube angled 10-15° toward the head while the arm is elevated above the head

3. if still not visualized: CT scan through non-visualized levels (CT is poor for evaluating alignment and for fractures in the horizontal plane, thin cuts with reconstructions ameliorates this shortcoming)

E. for questions regarding stability of the subaxial spine, see page 969

F. patients with C-spine fractures or dislocations should have daily C-spine x-rays during initial traction or immobilization

2. thoracic and lumbosacral spine: AP and lateral x-rays for all trauma patients who:

A. were thrown from a vehicle, or fell ≥ 6 feet to the ground

B. complain of back pain

C. are unconscious

D. are unable to reliably describe back pain or have altered mental status preventing adequate exam

E. have an unknown mechanism of injury, or other injuries that cast suspicion of spine injury

3. reminder: when abnormalities of questionable vintage are identified, a bone scan may be helpful to distinguish an old injury from an acute one (less useful in the elderly; in an adult, a bone scan will become “hot” within 24-48 hrs of injury, and will remain hot for up to a year; in the elderly, the scan may not become hot for 2-3 weeks and can remain so for over a year)

4. CT scan through area of bony abnormality or level of neurologic deficit (see below)


A. neurologically intact implies patient is alert, not drugged/intoxicated, & able to report pain reliably

B. some authors include oblique views in a “minimal” evaluation53, others do not51


Table 28-7 Radiographic signs of C-spine trauma (modified49)

Soft tissues

• retropharyngeal space > 7 mm, or retrotracheal space > 14 mm (adult) or 22 mm (peds) (see Table 6-10page 137 for details)

• displaced prevertebral fat stripe

• tracheal deviation & laryngeal dislocation

Vertebral alignment

• loss of lordosis

• acute kyphotic angulation

• torticollis

• widened interspinous space (flaring)

• axial rotation of vertebra

• discontinuity in contour lines (see page 135)

Abnormal joints

• ADI: > 3 mm (adult) or > 4 mm (peds) (see Table 6-9page 136 for details)

• narrowed or widened disc space

• widening of apophyseal joints


Purpose: to disclose occult ligamentous instability.

It is possible to have a purely ligamentous injury involving the posterior ligamentous complex without any bony fracture (see Hyperflexion sprainpage 971). Lateral flexion-extension views help detect these injuries, and also evaluate other injuries (e.g. compression fracture) for stability. For patients with limited flexion due to paraspinal muscle spasm (sometimes resulting from pain), a rigid collar should be prescribed, and if the pain persists 2-3 weeks later56 the flexion-extension films should be repeated.


1. the patient must be cooperative and free of mental impairment (i.e. no head injury, street or prescription drugs, alcohol…)

2. there should not be any subluxation > 3.5 mm at any level on cross-table C-spine x-rays (which is a marker for possible instability, see page 972)

3. patient must be neurologically intact (if there is any degree of spinal cord injury, proceed instead first with imaging studies, e.g. MRI)

4. F/E x-rays are no longer recommended in obtunded patients due to a low yield, poor cost-effectiveness, and they may be dangerous39


The patient should be sitting, and is instructed to flex the head slowly, and to stop if it becomes painful. Serial x-rays are taken at 5-10° increments (or followed under fluoro with spot films at the end of movement), and if normal, the patient may be encouraged to flex further. This is repeated until evidence of instability is seen, or the patient cannot flex further because of pain or limitation of motion. The process is then repeated for extension.


Normal flexion-extension views demonstrate slight anterior subluxation distributed over all cervical levels with preservation of the normal contour lines (see Figure 6-2page 135). Abnormal findings include: “flaring” of the spinous processes - exaggerated widening (see page 137).


Obtained through levels of abnormality identified on plain films or myelogram, or at level of neurologic deficit in patients with normal films, or at levels inadequately evaluated by plain films. Thin cuts (1.5-3 mm) through the level of suspicion are required. Assesses bony anatomy in detail. When combined with intrathecal contrast (i.e. after myelogram), also delineates any neural impingement.


Indications for emergent MRI in spinal cord injury (SCI) are listed below.

When MRI cannot be performed a myelogram is required (employing intrathecal water soluble contrast with CT to follow)  Caution: cervical myelogram in patients with cervical spine injuries usually requires C1-2 puncture to achieve adequate dye concentration in the cervical region without dangerous extension of the neck or tilting of the patient as required when dye is injected via LP. Furthermore, pressure shifts from LP exacerbates deficit in 14% of cases with complete block57.


1. incomplete SCI with normal alignment: to check for soft tissue compressing cord

2. neurologic deterioration (worsening deficit or rising level) including after closed reduction

3. neurologic deficit not explained by radiographic findings, including:

A. fracture level different from level of deficit

B. no bony injury identified: further imaging is done to R/O soft tissue compression (disc herniation, hematoma…) that would require surgery

C. always consider possible arterial dissection in this setting (see page 1160)


MRI may be used to identify potentially unstable occult ligamentous or soft tissue injury. Note: abnormal signal on MRI is not always associated with instability on flexion-extension x-rays58. It has been recommended that this MRI should be done within 48 hours42 or 72 hrs59 of injury. MRI is not reliable for identifying osseous injury. Indications for non-emergent MRI (modified60):

1. inconclusive cervical spine radiography, including questionable fractures

2. significant midline paraspinal tenderness and patient unable to have flexion-extension x-rays

3. obtunded or comatose patients

T2WI and FLAIR are the most helpful sequences. Significant abnormal findings:

1. ventral signal abnormalities with prevertebral swelling

2. dorsal signal abnormalities. Abnormal signal limited to the interspinous is probably not as unstable as when it extends into the ligamentum flavum60. These patients were treated with rigid collars or Minerva jackets for 1-3 months, and one that was felt to be very unstable underwent fusion

3. disc disruption indicated by abnormal signal intensity within the disc, increased disc height, or frank disc protrusions



To reduce fracture-dislocations, maintain normal alignment and/or immobilize the cervical spine to prevent further spinal cord injury. Reduction decompresses the spinal cord and roots, and may facilitate bone healing.


Level III61

• early closed reduction of C-spine fracture/dislocation injuries with craniocervical traction to restore anatomic alignment in awake patients

 not recommended: closed reduction in patients with an additional rostral injury

• patients with C-spine fracture-dislocation who cannot be examined during attempted closed reduction, or before open posterior reduction, should undergo cervical MRI before attempted reduction*. The presence of a significant herniated disc in this setting is a relative indication for anterior decompression before reduction

• cervical MRI is also recommended for patients who fail attempts at closed reduction*

* NB: prereduction MRI will show disrupted or herniated discs in 33-50% of patients with facet subluxation. These findings do not seem to significantly influence outcome after closed-reduction in awake patients; image the usefulness of prereduction MRI in this setting is uncertain


1. the rapidity with which reduction should be done4

2. whether MRI should be done prior to attempted closed reduction (see footnote to PRACTICE GUIDELINE 28-8)

A. in intact patients, to R/O a condition that might cause worsening of neurologic condition with reduction (e.g. traumatic disc herniation) - must be balanced against risks of transferring patients to MRI

B. in patients with neurologic deficit (complete or partial SCI)


1. atlatooccipital dislocation (see page 951): traction may worsen deficit. If immobilization with tongs/halo is desired, use no more than ≈ 4 lbs

2. types IIA or III hangman’s fracture: see page 960

3. skull defect/fracture at anticipated pin site: may necessitate alternate pin site

4. use with caution in pediatric age group (do not use if age ≤ 3 yrs)

5. very elderly patients:

6. demineralized skull: some elderly patients, osteogenesis imperfecta…

7. patients with an additional rostral injury

8. patients with movement disorders: constant motion may cause pin erosion through the skull

Application of tongs or halo ring

Supplies: gloves, local anesthetic (typically 1% lidocaine with epinephrine), beta-dine ointment. Optional equipment: razor or hair clipper, scalpel.

Choice of device: a number of cranial “tongs” are available. Crutchfield tongs require predrilling holes in the skull. Gardner-Wells tongs are the most common tongs in use. If later use of halo-vest immobilization is anticipated after acute stabilization, a halo ring may be used with an adapter for the initial cervical traction, and then converted to vest traction at the appropriate time (e.g. post-fusion).

Preparation: placed with patient supine on a gurney or bed. Option: shave hair around proposed pin sites (see below). Betadine skin prep, then infiltrate local anesthetic. Option: incise skin with scalpel (prevents pins from driving in surface contaminants).

Gardner-Wells tongs: Pin sites: the pins are placed in the temporal ridge (above the temporalis muscle), 2-3 finger-breadths (3-4 cm) above pinna. Place directly above external acoustic meatus for neutralposition traction; 2-3 cm posterior for flexion (e.g. for locked facets); 2-3 cm anterior for extension. One pin has a central spring-loaded force-indicator. Tighten pins until the indicator protrudes 1 mm beyond the flat surface. Retighten the pins daily until indicator protrudes 1 mm for 3 days only, then stop.

Halo ring: Supplies (in addition to above): optional paddle AKA “spoon” to support the head beyond the edge of the bed. Read all of this (including pointers) before starting

1. ring size: choose an appropriately sized ring that leaves a ≈ 1-2 cm gap between the scalp and the ring all the way around

2. ring position: generally placed at or just below the widest portion of the skull (the “equator”), but the front should be ≈ 1 cm above the orbital rim and the back should be ≈ 1 cm above the pinna62. The ring is usually stabilized with temporary pins that have plastic discs where they contact the skull

3. pin sites: choose the threaded holes in the ring that place the pins as perpendicular to the skull as possible as follows

A. anterior pins: above the lateral two thirds of the orbit

B. posterior pins: just behind the ears

C. in pediatrics, additional pins may be placed to further distribute the load on the thinner skull

4. pin insertion: the pins are gradually brought close to the scalp which is then anesthetized with local anesthetic. Pins are then sequentially tightened, starting with any pin then going to the kitty-corner pin, then a third pin and finally its opposite. Most halos provide some type of torque wrench to permit approximately 8 in-lb of torque for most adults; 2-5 in-lb for peds

5. placement pointers

A. the cervical collar is left in place until traction/immobilization is established

B. try to place the halo as level from left to right as possible. While a skewed placement can be compensated for when attaching the vest, it looks bad

C. prior to penetrating the forehead skin for anterior pins, have the patient close their eyes and hold them closed as the pins are advanced (this avoids “pinning the eyes open”)

D. avoid placing pins in the temporalis muscle or the temporal squamosa

E. do not place pins above the medial third of the orbit to avoid the supraorbital and supratrochlear nerves, and to reduce theorist of penetrating the relatively thin anterior wall of the frontal sinus

Post-placement care: For traction, transfer to a bed with ortho headboard with tongs or halo ring in place. Tie a rope to tongs/halo and feed through a pulley at the head of bed. Slight flexion or extension is achieved by changing the height of the pulley relative to the patients long axis.

X-rays: lateral C-spine x-rays immediately after application of traction and at regular intervals and after every change in weights and every move from bed. Check alignment and rule-out overdistraction at any level and atlanto-occipital dislocation (BDI should be ≤ 12 mm, see page 953).

Weight: if there is no malalignment and traction is being used just to stabilize the injury and to compensate for ligamentous instability, use 5 lbs for the upper C-spine or 10 lbs for lower levels. To reduce locked facets, see page 973. May remove cervical collar once patient is in traction with adequate reduction or stabilization.

Pin tightening: pins are retorqued in 24 hours. Some authors do one additional tightening the day after that. Avoid further tightenings which can penetrate the skull

Pin care: clean (e.g. half strength hydrogen peroxide), then apply povidone-iodine ointment. Frequency: in hospital: q shift. At home following discharge: twice daily.

Alternatively, simple cleaning with soap and water twice daily is acceptable.

Application of halo vest

For vest placement (i.e. patients not remaining in traction) once the halo ring is placed (see above) it needs to be attached to the vest by posts. The mechanism varies between manufacturers. If possible, have the patient in a cotton T-shirt prior to placing the vest (this may require cutting the neck opening to accommodate the ring).

The vest should be snug, but too tight so as to restrict respirations. Shoulder straps should be contacting the shoulders (the vest will tend to ride up when the patient is sitting). Most vests come with a wrench that is taped to the vest for emergency removal.

Reduction of locked facets

See page 973.


1. skull penetration by pins. May be due to:

A. pins torqued too tightly

B. pins placed over thin bone: temporal squamosa or over frontal sinus

C. elderly patients, pediatric patients, or those with an osteoporotic skull

D. invasion of bone with tumor: e.g. multiple myeloma

E. fracture at pin site

2. reduction of cervical dislocations may be associated with neurologic deterioration which is usually due to retropulsed disc63 and requires immediate investigation with MRI or myelogram/CT

3. overdistraction from excessive weight (especially with upper cervical spine injuries), may also endanger supporting tissues

4. caution with C1-C3 injury, especially with posterior element fracture (traction may pull fragments in towards canal)

5. infection:

A. osteomyelitis in pin sites: risk is reduced with good pin care

B. subdural empyema: rare6465 (see page 356)


Caution: laminectomy in the face of acute spinal cord injury has been associated with neurologic deterioration in some cases. When emergency decompression is indicated, it is usually combined with a stabilization procedure.

Modified recommendations of Schneider66

In patients with complete spinal cord lesions, no study has demonstrated improvement in neurologic outcome with either open decompression or closed reduction67. In general, surgery is reserved for incomplete lesions (possibly excluding central cord syndrome, see page 948) with extrinsic compression, who, following maximal possible reduction of subluxation show:

1. progression of neurologic signs

2. complete subarachnoid block by Queckenstedt test or radiographically (on myelography or MRI)

3. myelogram, CT, or MRI shows bone fragments or soft tissue elements (e.g. hematoma) in the spinal canal causing spinal cord compression

4. necessity for decompression of a vital cervical root

5. compound fracture or penetrating trauma of the spine

6. acute anterior spinal cord syndrome (see page 950)

7. non-reducible fracture-dislocations from locked facets causing spinal cord compression


Figure 28-1 Relationship between spinal cord, nerve roots, and bony spine

Contraindications to emergent operation

1. complete spinal cord injury ≥ 24 hrs (no motor or sensory function below level of lesion)

2. medically unstable patient

3. central cord syndrome (controversial, see page 948)

28.4. Neurological assessment

Evaluation of the level of the lesion requires familiarity with the following concepts about the relationship between the bony spinal canal and the spinal cord and nerves (see Figure 28-1).

1. since there are 8 pairs of cervical nerves and only 7 cervical vertebra

A. cervical nerves 1 through 8 exit above the pedicles of their like-numbered vertebra

B. thoracic, lumbar and sacral nerves exit below the pedicles of their like-numbered vertebra

2. due to disproportionately greater growth of the spinal column than the spinal cord during development, the following relationships of the spinal cord to the vertebral column exist:

A. to determine which segment of the cord underlies a given vertebra:

1. from T2 through T10: add 2 to the number of the spinous process

2. for T11, T12 and L1, remember that these overlie the 11 lowest spinal segments (L1 through L5, S1 through S5, and Coccygeal-1)

B. the conus medullaris in the adult lies at about L1 or L2 of the spine


The following tables are for rapid assessment (see Table 24-4page 787 and Table 24-6page 788 for detailed tables of motor innervation).



A system6869 that may be rapidly applied to grade 10 key motor segments using the MRC Grading Scale (Table 24-1page 786) from 0-5 on the left and the right, for a total score of 100 possible points (see Table 28-8). NB: most muscles receive innervation from two adjacent spinal levels, the levels listed in Table 28-8 are the lower of the two. The standard considers a segment intact if the motor grade is fair (≥ 3). For additional information, see www.asia-spinalinjury.org.

Table 28-9 Axial muscle evaluation70



Action to test



tidal volume (TV), FEV1, and vital capacity (VC)





upper abdominals

lower abdominals

use sensory level, abdominal reflexes, & Beevor’s sign (see below)





ASIA standards68

28 key points identified in Table 28-12 are scored separately for pinprick and light touch on the left & right side using the grading scale shown in Table 28-11, for a maximum possible total of 112 points for pinprick (left & right) and 112 points for light touch (left & right).

NB: regarding the “C4 cape” AKA “bib” region across the upper chest and back: sensory segments “jump” from C4 to T2 with the intervening levels distributed exclusively on the UEs (see Figure 5-13page 94). The location of transition is inconstant from person to person.


1. external anal sphincter is tested by insertion of the examiner’s finger

A. perceived sensation is recorded as present or absent. Any sensation felt by the patient indicates that the injury is sensory incomplete

B. note resting sphincter tone and any voluntary sphincter contraction

2. bulbocavernosus reflex (BC)see footnote, page 946

Table 28-11 Sensory grading scale






impaired (partial or altered appreciation)




not testable

Table 28-12 Key sensory landmarks




occipital protuberance


supraclavicular fossa


top of acromioclavicular joint


Lateral side of antecubital fossa


thumb, dorsal surface, proximal phalanx


middle finger, dorsal surface, proximal phalanx


little finger, dorsal surface, proximal phalanx


medial (ulnar) side of antecubital fossa


apex of axilla


third intercostal space (IS)


fourth IS (nipple line)


fifth IS (midway between T6 & T8)


sixth IS (xiphoid process)


seventh IS (midway between T6 & T8)


eighth IS (midway between T6 & T10)


ninth IS (midway between T8 & T10)


tenth IS (umbilicus)


eleventh IS (midway between T10 & T12)


inguinal ligament at mid-point


half the distance between T12 & L2


mid-anterior thigh


medial femoral condyle


medial malleolus


dorsum of foot at 3rd MTP joint


lateral heel


popliteal fossa in the mid-line


ischial tuberosity


perianal area (taken as 1 level)


The following elements are considered optional but it is recommended that they be graded as absent, impaired or normal:

1. position sense: test index finger and great toe on both sides

2. awareness of deep pressure/deep pain


The ASIA impairment scale*68 is shown in Table 28-13 (a modified Frankel Neurological Performance scale71).

Table 28-13 ASIA impairment scale




Complete: no motor or sensory function preserved


Incomplete: sensory but no motor function preserved below the neurologic level (includes sacral segments S4-5)


Incomplete: motor function preserved below the neurologic level (more than half of key muscles below the neurologic level have a muscle strength grade < 3)*


Incomplete: motor function preserved below the neurologic level (more than half of key muscles below the neurologic level have a muscle strength grade ≥ 3)


Normal: Sensory & motor function normal

* for muscle strength grading see Table 24-1page 786


* NB: this scale indicates the completeness of spinal cord injury and is separate from the other ASIA grading scales that appear earlier in this chapter.


28.5. Spinal cord injuries

28.5.1. Complete spinal cord injuries

See page 930 for definition of complete vs. incomplete spinal cord injury.

In addition to loss of voluntary movement, sphincter control and sensation below the level of the injury, there may be priapism. Hypotension and bradycardia (spinal shock, see page 930) may also present.


Occurs as a result of spinal cord injury at or above ≈ C3 (includes SCI from atlanto-occipital and atlantoaxial dislocation). Bulbar-cervical dissociation produces immediate pulmonary and, often, cardiac arrest. Death results if CPR is not instituted within minutes. Patients are usually quadriplegic and ventilator dependent (phrenic nerve stimulation may eventually allow independence from ventilator).

28.5.2. Incomplete spinal cord injuries


image Key concepts:

• disproportionately greater motor deficit in the upper extremities than lower

• usually results from hyperextension injury in the presence of osteophytic spurs

• surgery is often employed for ongoing compression, usually on a non-emergency basis except for rare cases of progressive deterioration

Central cord syndrome (CCS) is the most common type of incomplete spinal cord injury syndrome. Usually seen following acute hyperextension injury in an older patient with pre-existing acquired stenosis as a result of bony hypertrophy (anterior spurs) and infolding of redundant ligamentum flavum (posteriorly), sometimes superimposed on congenital spinal stenosis. Translational movement of one vertebra on another may also contribute. A blow to the upper face or forehead is often disclosed on history, or is suggested on exam (e.g. lacerations or abrasions to face and/or forehead). This often occurs in relation to a motor vehicle accident or to a forward fall, often while intoxicated. Younger patients may also sustain CCS in sporting injuries (see burning hands syndrome,page 980). CCS may occur with or without cervical fracture or dislocation72. CCS may be associated with acute traumatic cervical disc herniation. CCS may also occur in rheumatoid arthritis.


Theory: the centermost region of the spinal cord is a vascular watershed zone which renders it more susceptible to injury from edema. Long tract fibers passing through the cervical spinal cord are somatotopically organized such that cervical fibers are located more medially than the fibers serving the lower extremities (see Figure 5-12page 93).


The clinical syndrome is somewhat similar to that seen in syringomyelia.

1. motor: weakness of upper extremities with lesser effect on lower extremities

2. sensory: varying degrees of disturbance below level of lesion may occur

3. myelopathic findings: sphincter dysfunction (usually urinary retention)

Hyperpathia to noxious and non-noxious stimuli is also common, especially in the proximal portions of the upper extremities, and is often delayed in onset and extremely distressing to the patient74. Lhermitte’s sign occurs in ≈ 7% of cases.


There is often an initial phase of improvement (characteristically: LEs recover first, bladder function next, UE strength then returns with finger movements last; sensory recovery has no pattern) followed by a plateau phase and then late deterioration75. 90% of patients are able to walk with assistance within 5 days76. Recovery is usually incomplete, and the amount of recovery is related to the severity of the injury and patient age77.

If CCS results from hematomyelia with cord destruction (instead of cord contusion), then there may be extension (upward or downward).


Findings: young patients tend to have disc protrusion, subluxation, dislocation or fractures76. Older patients tend to have multi-segmental canal narrowing due to osteophytic bars, discs, and inbuckling of ligamentum flavum76.

C-spine x-rays: may demonstrate congenital narrowing, superimposed osteophytic spurs, traumatic fracture/dislocation. Occasionally, AP narrowing alone without spurs may be seen72. Plain x-rays will fail to demonstrate canal narrowing due to: thickening or inbuckling of ligamentum flavum, hypertrophy of facet joints, and poorly calcified spurs72.

Cervical CT scan: also helpful in diagnosing fractures and osteophytic spurs. Not as good as MRI for assessing status of discs, spinal cord and nerves.

MRI: discloses compromise of anterior spinal canal by discs or osteophytes (when combined with plain C-spine x-rays, it increases the ability to differentiate osteophyte from traumatic disc herniation). Also good for evaluating ligamentum flavum. T2WI may show spinal cord edema acutely78, and can detect hematomyelia. MRI is poor for identifying fractures.



Level III77

• because of possible cardiac, pulmonary & BP disturbances, many* of these patients may require management in an ICU or other monitored setting (for cardiac, hemodynamic & respiratory monitoring), especially patients with severe neurologic deficits

• maintain MAP 85-90 mm Hg (use BP augmentation if necessary) for the 1st week after injury to improve spinal cord perfusion

• early reduction of fracture-dislocation injuries is recommended

• surgical decompression, particularly for focal and anterior spinal cord compression that is approached anteriorly, seems* to be of benefit in selected patients

* all italics added by the editor

The indications, timing and best treatment method for CCS remains controversial. Initial management options include the methylprednisolone spinal-cord injury protocol for patients seen within 8 hours of the time of injury (see page 936).

Indications for surgery:

1. continued compression of the spinal cord79 (p 1010) that correlates with the level of deficit with any of the following:

A. persistent significant motor deficit following a varying period of recovery (see Timing of surgery below)

B. deterioration of function

C. continued significant dysesthetic pain

2. instability of the spine

Improvement has been shown in short and long-term follow-up with subacute decompression of the offending lesion76. Nonsurgical treatment results in a longer period of pain and weakness in many cases.

Timing of surgery: A perennial point of controversy. Classic teaching was that early surgery for this condition is contraindicated because this may worsen the deficit. In the absence of spinal instability, traditional management consisted of bed rest in a soft collar for ≈ 3-4 weeks, with consideration for surgery after this time, or else gradual mobilization in the same collar for an additional 6 weeks. It is presently felt that there is no solid evidence that earlydecompressive surgery (without cord manipulation) is actually harmful, but there is also no evidence that it is helpful, either. There may be good justification for early surgery in the rare patient who is improving and then deteriorates80, however, great restraint must be used in avoiding what would be an inappropriate operation in many patients81. Surgery may improve the rate and degree of recovery in selected patients82. Surgery has been recommended for patients with gross spinal instability or for patients with significant persistent cord compression (e.g. by osteophytic spurs) who fail to progress consistently after an initial period of improvement78, often within 2-3 weeks following the trauma. Better results occur with decompression within the first few weeks or months rather than very late (e.g. ≥ 1-2 years)79 (p 1010).


There is no role for surgery without ongoing compression or instability. The rare patient with ongoing compression undergoing documented progressive deterioration should be decompressed ASAP. Improving patients should be followed and decompression can be done electively for ongoing compression. There is controversy regarding timing of surgery for stable CCS and ongoing compression: while Class I or II data are lacking, there seems to be a trend to decompress these patients as soon as they are medically stable without an arbitrary waiting period.

Technical considerations: The most rapid procedure to decompress the cord is often a multi-level laminectomy. This is frequently accompanied by dorsal migration of the spinal cord which may be seen on MRI75. With myelopathy, fused patients fare better than those that are just decompressed without fusion. Fusion may be accomplished posteriorly (e.g. with lateral mass screws and rods) at the time of decompression, or anteriorly (e.g. multi-level discectomy, or corpectomy with strut graft and anterior cervical plating) at the same sitting as the laminectomy or staged at a later date.


In patients with cord contusion without hematomyelia, ≈ 50% will recover enough LE strength and sensation to ambulate independently, although typically with signifi-cant spasticity. Recovery of UE function is usually not as good, and fine motor control is usually poor. Bowel and bladder control often recovers. Elderly patients with this condition generally do not fare as well as younger patients, with or without surgical treatment (only 41% over age 50 become ambulatory, versus 97% for younger patients83).


AKA anterior spinal artery syndrome. Cord infarction in the territory supplied by the anterior spinal artery. Some say this is more common than central cord syndrome.

May result from occlusion of the anterior spinal artery, or from anterior cord compression, e.g. by dislocated bone fragment, or by traumatic herniated disc.


1. paraplegia, or (if higher than ≈ C7) quadriplegia

2. dissociated sensory loss below lesion:

A. loss of pain and temperature sensation (spinothalamic tract lesion)

B. preserved two-point discrimination, joint position sense, deep pressure sensation (posterior column function)84


It is vital to differentiate a non-surgical condition (e.g. anterior spinal artery occlusion) from a surgical one (e.g. anterior bone fragment). This requires one or more of: myelography, CT, or MRI.


The worst prognosis of the incomplete injuries. Only ≈ 10-20% recover functional motor control. Sensation may return enough to help prevent injuries (burns, decubitus ulcers…).


Spinal cord hemisection. First described in 1849 by Brown-Sequard85.

Classical findings (rarely found in this pure form):

• ipsilateral findings:

image motor paralysis (due to corticospinal tract lesion) below lesion

image loss of posterior column function (proprioception & vibratory sense)

• contralateral findings: dissociated sensory loss

image loss of pain and temperature sensation inferior to lesion beginning 1-2 segments below (spinothalamic tract lesion)

image preserved light (crude) touch due to redundant ipsilateral and contralateral paths (anterior spinothalamic tracts)


Usually a result of penetrating trauma, it is seen in 2-4% of traumatic spinal cord injuries86. Also may occur with radiation myelopathy, cord compression by spinal epidural hematoma, large cervical disc herniation87-89 (rare), spinal cord tumors, spinal AVMs, cervical spondylosis, and spinal cord herniation (see page 514).


This syndrome has the best prognosis of any of the incomplete spinal cord injuries. ≈ 90% of patients with this condition will regain the ability to ambulate independently as well as anal and urinary sphincter control.


AKA contusio cervicalis posterior. Relatively rare. Produces pain and paresthesias (often with a burning quality) in the neck, upper arms, and torso. There may be mild paresis of the UEs. Long tract findings are minimal.

28.6. Cervical spine fractures

Also see C-Spinepage 135 for cervical x-rays. One system of classifying cervical spine fractures identifies the following subatlantal injuries90 (see below for occiptoatlantoaxial injuries, see page 955 for atlantoaxial subluxation/dislocation):

1. hyperextension fracture-dislocations

A. posterior fracture-dislocation of the dens

B. traumatic spondylolisthesis of the axis (hangman’s fracture, see page 959)

C. hyperextension sprain (momentary dislocation) with fracture

D. hyperextension fracture-dislocation with fractured articular pillar

E. hyperextension fracture-dislocation with comminution of the vertebral arch

2. hyperflexion fracture-dislocations

A. anterior fracture-dislocation of the dens (see page 963)

B. hyperflexion sprain: rare. Occurs when posterior ligaments are disrupted but locking of articular facets does not occur (see page 971)

C. locked articular facets with fracture (see page 972)

D. “teardrop” fracture-dislocation (see page 970)

28.6.1. Occiptoatlantoaxial injuries Atlanto-occipital dislocation

See Occiptoatlantoaxial-complex anatomy on page 91 for relevant anatomy.

Atlanto-occipital dislocation (AOD) AKA craniocervical junction dislocation. Disruption of the stability of the craniocervical junction (which results from ligamentous injuries). Probably underdiagnosed, may be present in ≈ 1% of patients with “cervical spine injuries”91 (definition of cervical spine injuries not specified), found in 8-19% of fatal cervical spine injury autopsies9293. More than twice as common in pediatrics as adults, possibly owing to the flatter (i.e. less cupped) condyles in peds, the higher ratio of cranium to body weight, and increased ligamentous laxity. Patients usually either have minimal neurological deficit or exhibit bulbar-cervical dissociation (BCD) (see page 948). Some may exhibit cruciate paralysis (see page 1196). Most mortality results from anoxia due to respiratory arrest as a result of BCD.


For classification, see Figure 28-2. Combinations (e.g. anterior-distracted AOD95) may also occur.

Type I anterior dislocation of occiput relative to the atlas

Type II longitudinal dislocation (distraction)

Type III posterior dislocation of occiput


Figure 28-2 Classification of atlanto-occipital dislocation



Level III96

• lateral C-spine x-ray

• if it is desired to employ a radiologic method of measurement, the BAIBDI method (see page 953) is recommended

• upper cervical prevertebral soft-tissue swelling on an otherwise nondiagnostic plain lateral C-spine x-ray should prompt additional imaging

• if there is clinical suspicion of AOD, and plain x-rays are nondiagnostic, CT or MRI is recommended, especially for non-Type II dislocations


Level III96

• internal fixation & arthrodesis using one of a variety of methods

•  CAUTION: traction may be used in the management of AOD, but it is associated with a 10% risk of neurologic deterioration


1. may be neurologically intact, therefore must be ruled-out in any major trauma

2. bulbar-cervical dissociation: see page 948

3. may have lower cranial nerve deficits (as well as VI palsies) ± cervical cord injury

4. worsening neurologic deficit with the application of cervical traction: check lateral C-spine films immediately after applying traction (see page 943)


Numerous methodologies have been devised to radiographically diagnose AOD. Most utilize surrogate markers for the end-point of interest: instability of the occipitalcervical junction. None are completely reliable97. Measurements on CT scans are more accurate than plain radiographs (landmarks are easier to identify, no magnification or rotation error…) - however, normal values may differ from plain radiographs. Some methods are shown in Table 28-15 (recommended: BAI-BDI method & the AOI method).

Technical suggestions

1. radiographs: verify that the film is a true lateral (e.g. check alignment of the two mandibular rami as well as of the posterior clinoids)

2. CT: sensitivity, specificity, and positive/negative predictive values of most of the these measures improves when applied to sagittal CT reconstructions98 (relevant landmarks could be identified in > 99% of CTs, vs. 39-84% on x-ray)

Powers’ ratio91: distance BC (basion to posterior arch of atlas) is divided by distance AO (opisthion to anterior arch of atlas), see Table 28-15. Interpretation is shown in Table 28-14.

 Cannot be used with any fracture involving the atlas or the foramen magnum, or with congenital anatomic abnormalities. Applies only to anterior AOD (i.e. not for posterior or distracted AOD).

Table 28-14 Powers’ ratio

Ratio BC/AO



< 0.9


1 standard deviation below the lowest case of AOD

≥ 0.9 and < 1

“grey zone” (indeterminate)

included 7% of normals and no cases of AOD

≥ 1


encompassed all AOD cases



A helpful rule-of-thumb: the inferior tip of the clivus should point directly to tip of dens (this may be obscured on x-ray).

Additional CT clues: there may be blood in the basal cisterns (an indirect sign). On thin-cut axial CT, there may be one or more slices showing no bone (abnormal).


Initial management: If AOD is suspected, immediately immobilize the neck with halo orthosis or with sandbags.  Do not apply cervical traction in an attempt to reduce AOD because of risk of neurologic deterioration (option: 2-4 lbs may be used early only for immobilization in adults, not for reduction).

Subsequent management: Controversial whether operative fusion vs. prolonged immobilization (4-12 months) with halo brace is required. However, posterior occipito-cervical fusion is usually recommended (see atlantooccipital dislocation injuriespage 952). Horn et al.99 suggest that patients be grouped and then managed as shown in Table 28-16.

In infants: reduce in the OR and fuse (usually with transarticular screws).

Table 28-16 Grading & management of AOD99





no abnormal CT criteria* with only moderately abnormal MRI (high signal in posterior ligaments or occipoalatlantal joints)

external orthosis (halo or collar)


≥ 1 abnormal CT criteria* or grossly abnormal MRI findings in occipoalatlantal joints, tectorial membrane, or alar or cruciate ligaments

surgical stabilization

* CT criteria used: Power’s ratio, BAI-BDI, X-line


The most important predictor of outcome is the severity of neurologic injuries at the time of presentation99. Among AOD patients who survived the initial injury, those with severe TBI and brainstem dysfunction or complete bulbar-cervical dissociation all had poor outcome99. Those with incomplete SCI or nonsevere TBI may improve. Occipital condyle fractures

image Key concepts:

• uncommon (0.4% of trauma patients)

• may present with lower cranial nerve deficits which may be delayed in onset (e.g. hypoglossal nerve palsy), mono-, para-, or quadriparesis or plegia

• W/U: image CT scan with reconstructions (rarely detected on plain x-rays)

• Tx: usually treated with rigid collar. Indications for occipitocervical fusion or halo immobilization: craniocervical misalignment (occipital-C1 interval > 2.0 mm)



Level II114

• clinical suspicion should be raised by the presence of ≥ 1 of the following: blunt trauma with high-energy craniocervical injuries, altered consciousness, occipital pain or tenderness, impaired cervical movement, lower cranial nerve palsies, or retropharyngeal soft-tissue swelling

• CT* is recommended for establishing the diagnosis

Level III114: MRI is recommended to assess the integrity of the craniocervical ligaments


Level III114: external cervical immobilization

* with reconstructions

 Maserati et al.110 recommend occipitocervical fusion for craniocervical misalignment (retrospective review)

Rare. Incidence: 0.4% (in a series of 24,745 consecutive trauma patients surviving to the E/R110). Occipital condyle fractures (OCF) were first described in 1817 by Bell111.


A widely used system is that of Anderson & Montesano113 which is shown in Table 28-17.

Maserati et al.110 classified patients simply on the basis of whether craniocervical misalignment was present or absent on CT with reconstructions (craniocervical misalignment (they defined as an occipital condyle-C1 interval > 2.0 mm). They felt other classification systems were superfluous as they did not affect outcome in their retrospective review (see Treatment below).

Table 28-17 Anderson & Montesano classification of occipital condyle fractures




comminuted from impact: may occur from axial loading


extension of linear basilar skull fracture112


avulsion of condyle fragment (traction injury): may occur during rotation, lateral bending, or a combination of mechanisms. Considered unstable by many


Controversial. Lower cranial nerve deficits often develop in untreated cases of OCF, and may resolve or improve with external immobilization. Anderson & Montesano Types I & II have been treated with or without external immobilization (cervical collar or, occasionally, halo) without obvious difference. External immobilization x 6-8 weeks is suggested for Type III fractures because of the higher risk of delayed deficits.

In a retrospective review of 100 patients with OCF110, 3 patients underwent occipitocervical fusion (see page 179) for craniocervical misalignment (2) or unrelated C1-2 fracture (1). The remainder (without craniocervical misalignment) were treated with a rigid collar and delayed clinical & radiographic follow-up. None of their unoperated patients had neurologic deficit, and none developed delayed instability, malalignment, or neurologic deficit (regardless of their classification on the other systems in use).

28.6.2. Atlantoaxial subluxation/dislocation

Lower morbidity and mortality than atlantooccipital dislocation115. See Occiptoatlantoaxial-complex anatomy on page 91 for relevant anatomy.

Types of atlantoaxial subluxation:

1. rotatory: (see below) usually seen in children after a fall or minor trauma

2. anterior: more ominous (see below)

3. posterior: rare. Usually from erosion of odontoid. Unstable. Requires fusion


image Key concepts:

• typically seen in children

• associations: trauma, RA, respiratory tract infections in peds (Grisel syndrome)

• often present with cock-robin head position (tilt, rotation, sl. flexion)

• classification: Fielding & Hawkins (Table 28-18)

• Tx: early traction often successful. Treat infection in Grisel syndrome. Subluxation unreducible in traction may need transoral release then posterior fusion

Rotational deformity at the atlanto-axial junction is usually of short duration and easily corrected. Rarely, the atlantoaxial joint locks in rotation (AKA atlantoaxial rotatory fixation116). Usually seen in children. May occur spontaneously (with rheumatoid arthritis117 or with congenital dens anomalies), following major or minor trauma (including neck manipulation or even with neck rotation while yawning116), or with an infection of the head or neck including upper respiratory tract (known as Grisel syndrome118: inflammation may cause mechanical and chemical injury to the facet capsules and/or transverse atlantal ligament (TAL)). The Fielding and Hawkins classification116is shown in Table 28-18.

The dislocation may be at the occipito-atlantal and/or the atlanto-axial articulations119. The mechanism of the irreducibility is poorly understood. With an intact TAL, rotation occurs without anterior displacement. If the TAL is incompetent as a result of trauma or infection, there may also be anterior displacement with more potential for neurologic injury. Posterior displacement occurs only rarely116.

The vertebral arteries (VA) may be compromised in excessive rotation, especially if it is combine with anterior displacement.


Clinical findings

Patients are usually young. Neurologic deficit is rare. Findings may include: neck pain, headache, torticollis (characteristic “cock robin” head position with ≈ 20° lateral tilt to one side, 20° rotation to the other, and slight (≈ 10°) flexion - see page 541 for DDx), reduced range of motion, and facial flattening116. Although the patient cannot reduce the dislocation, they can increase it with head rotation towards the subluxed joint with potential injury to the high cervical cord.

Brainstem and cerebellar infarction and even death may occur with compromise of circulation through the VAs120.

Radiographic evaluation

X-rays: Findings (may be confusing) include:

1. pathognomonic finding on AP C-spine x-ray in severe cases: frontal projection of C2 with simultaneous oblique projection of C1121 (p 124). In less severe cases, the C1 lateral mass that is forward appears larger and closer to the midline than the other

2. asymmetry of the atlantoaxial joint that is not correctable with head rotation, which may be demonstrated by persistence of asymmetry on open mouth odontoid views with the head in neutral position and then rotated 10-15° to each side

3. the spinous process of the axis is tilted in one direction and rotated to the other (may occur in torticollis of any etiology)

CT scan: Demonstrates rotation of the atlas119.

MRI: May assess the competence of the transverse ligament.


Grisel’s syndrome: appropriate antibiotics for causative pathogen with traction (see below) and then immobilization for the subluxation as follows118: Fielding (see Table 28-18) Type I: soft collar, Type II: Philadelphia collar or SOMI, Type III or IV: halo. After 6-8 weeks of immobilization, check stability with flexion-extension x-rays. Surgical fusion for residual instability

Traction: If treated within the first few months122 the subluxation can usually be reduced with gentle traction (in children start with 7-8 lbs and gradually increase up to 15 lbs over several days, in adults start with 15 lbs and gradually increase up to 20). If the subluxation is present > 1 month, traction is less successful. Active left-right neck rotation is encouraged in traction.

If reducible, immobilization in traction or halo is maintained x 3 months116 (range: 6-12 weeks).

Surgical fusion: Subluxation that cannot be reduced or that recurs following immobilization should be treated by surgical arthrodesis after 2-3 weeks of traction to obtain maximal reduction. The usual fusion is C1 to C2 (see page 183) unless other fractures or conditions are present116. Fusion may be performed even if the rotation between C1 & C2 is not completely reduced. For irreducible fixation, a staged procedure can be done with anterior transoral release of the atlantoaxial complex (the exposure is taken laterally to expose the atlantoaxial joints which must be done carefully to avoid injury to the VAs, soft tissue is carefully removed from the joints and the atlantodental interval, no attempt at reduction was made at the time of this 1st stage) followed by gradual skull traction and then a second stage posterior C1-2 fusion122


One third of patients have neurologic deficit or die. Subluxation may be due to:

1. disruption (rupture) of the transverse (atlantal) ligament (TAL): the atlantodental interval (ADI) (see below) will be increased

A. attachment points of the TAL may be weakened in rheumatoid arthritis (see page 495)

B. trauma: may cause anatomic or functional ligament disruption (see below)

2. incompetence of the odontoid process: ADI will be normal

A. odontoid fracture

B. congenital hypoplasia (e.g. Morquio syndrome, see page 494)


Both CT & MRI are recommended to evaluate fractures, TAL & its bony attachments.


1. TAL disruption: see below

2. odontoid fractures with intact TAL: managed as outlined on page 964


For relevant anatomy, see Occiptoatlantoaxial-complex anatomy on page 91.

Classification of TAL disruption123

Type I: anatomic disruption. Tear of TAL itself. Rare (the odontoid usually fractures before the TAL tears). Unlikely to heal. Requires surgical stabilization

Type II: physiologic disruption. Detachment of the tubercle of C1 to which TAL is attached (see Figure 5-11page 92) as may occur in comminuted C1 lateral mass fractures. 74% chance of healing with immobilization (halo recommended123)

“V” shaped pre-dens space124: Widening of the upper space between the anterior arch of C1 and the odontoid seen on lateral C-spine flexion x-ray. It is not known if this increased mobility represents elongation or laxity of the transverse ligament and/or the posterior ligamentous complex. This may also be a normal finding in flexion in peds.

Assessing the integrity of the transverse ligament

1. disruption of the TAL may be inferred indirectly from

A. Rule of Spence: on open-mouth odontoid x-ray, if the total overhang of both C1 lateral masses on C2 is ≥ 7 mm (see page 136 for details)

B. atlantodental interval (ADI): > 3 mm in adults, > 4 mm in peds (see page 136 for details)

2. MRI may image TAL directly. Findings of disruption (axial MRI): high signal within TAL on gradient-echo MRI, loss of continuity of TAL, blood at insertion site125

3. CT demonstrates bony injuries in regions of TAL insertion on C1 tubercles


One approach is to fuse all Type I TAL injuries, and those Type II TAL injuries that are still unstable after 3-4 months of immobilization123. Fusion is also recommended with irreducible subluxations. If C1 is intact, a C1-2 fusion is usually adequate. For situations involving C1 fractures, see below.

28.6.3. Atlas (C1) fractures

Acute C1 fractures account for 3-13% of cervical spine fractures126. 56% of 57 patients had isolated C1 fractures; 44% had combination C1-2 fractures; 9% had additional non-contiguous C-spine fractures. 21% had associated head injuries126.



Type I: fractures involving a single arch (31-45% of C1 fractures)

Type II: burst fracture (37-51%): the classic Jefferson fracture (see below)

Type III: lateral mass fractures of the atlas (13-37%)

Jefferson fracture

Described by Sir Geoffrey Jefferson128. Classically a four-point (burst) fracture of the C1 ring129, but the term is now often used to include the more common three or two-point fractures130, the latter through the C1 arches (thinnest portion). Usually from axial load (a “blow-out” fracture). 41% chance of an associated C2 fracture.

In pediatrics, it is critical to differentiate a C1 fracture from the normal synchondroses (see page 137), and from pseudospread of the atlas (see page 933). A fracture may also occur through the unfused synchondroses.

Unstable, usually no neurologic deficit if isolated (due to large canal diameter at this level, plus tendency for fragments to be forced outwards away from spinal cord).


To reiterate: stability of the occiptoatlantoaxial complex is primarily due to ligaments, with little contribution from bony articulations (see Occiptoatlantoaxial-complex anatomy on page 91). image Integrity of transverse ligament (TAL) is the most important determinant of stability (see Assessing the integrity of the transverse ligament above).


Neurologic deficit is rare. 3 of 25 patients with Jefferson fractures sustained neurologic injuries (1 complete injury, 2 central cord syndromes) in one series.


Complete C-spine series and thin section high-resolution CT from C1 through C3 to delineate details of C1 fracture and to assess for associated C2 injury.

MRI to assess TAL.




Level III131: for isolated atlas fractures:

• if the transverse ligament is intact: cervical immobilization alone

• if the transverse ligament is disrupted*: either

A. cervical immobilization alone

B. or, surgical fixation and fusion

* disruption of the TAL may be anatomic or physiologic (see text for details)

Treatment options depend heavily on status of TAL, and are delineated in Table 28-19131. When external immobilization is employed, it is used for 8-16 weeks (mean = 12).

Fusion options when surgery is indicated123:

1. unilateral ring or anterior C1 arch fractures: C1-2 fusion

2. multiple ring fractures or posterior C1 arch fractures: occipital-cervical fusion

Table 28-19 Treatment options for isolated C1 fractures

Fracture type

Treatment options

anterior or posterior arch

collar or SOMI

anterior AND posterior arch (burst)

stable (TAL* intact)

unstable (TAL disrupted)

collar or SOMI, halo

halo, C1-2 stabilization & fusion

lateral mass fractures

comminuted fracture

transverse process fracture

collar or SOMI, halo collar or SOMI

* abbreviations: TAL = transverse atlantal ligament

Surgical operations the do not involve arthrodesis include: posterior C1 screw placement, anterior transoral screw/plate placement.


In many series126132, treatment without surgery results in satisfactory outcome when the TAL is not disrupted.

28.6.4. Axis (C2) fractures

Acute fractures of the axis represent ≈ 20% of cervical spine fractures. Neurological injury is uncommon, and occurs in < 10% of cases. Most injuries may be treated by rigid immobilization.

Steele’s rule of thirds: each of the following occupies one third of the area of the canal at the level of the atlas: dens, space, spinal cord133.

Types of C2 fractures

1. odontoid fractures (see page 963): type II is the most common injury of the axis

2. hangman’s fracture: see below

3. miscellaneous C2 fractures: see page 967 Hangman’s fracture

image Key concepts:

• bilateral fracture through the pars interarticularis of C2 with traumatic subluxation of C2 on C3, most often due to hyperextension + axial loading

• most are stable with no neurologic deficit

• classification: Levine system (Table 28-20). Critical dividing line: disruption of C2-3 disc (Types II and higher) which may render the fracture unstable

• W/U: image cervical CT with sagittal & coronal recons for allimage Cervical MRI to assess C2-3 disc disruption (Levine II). image CTA for dissection if fx passes thru foramen transversarium (consider for all C2 fractures - see Table 28-36page 983)

• most do well with non-halo immobilization x 8-14 weeks (exceptions: severe/unstable fractures (see page 961) or those that do not remain aligned in brace)

AKA traumatic spondylolisthesis of the axis.


Figure 28-3 Anatomy of axis (C2). The pars interarticularis is shown in dark blue

Description: bilateral fracture through the pars interarticularis (isthmus) of the pedicleA of C2 (see Figure 28-3). There is often anterior subluxation of C2 on C3.


A. the configuration of C2 is unique, and the distinction between the pars and the pedicle is ambiguous


The term “hangman’s fracture” (HF) was coined by Schneider et al.134 although the mechanism of most modern HFs (hyperextension and axial loading, from MVAs or diving accidents) differs from that sustained in judicial hangings (where submental placement of the knot results in hyperextension and distraction135). Some cases may be due to forced flexion or compression of the neck in extension.

Pediatrics: rare in children < 8 years old where the forces tend to fracture the incompletely fused odontoid (epiphyseal fracture - see page 137). In pediatrics, consider pseudosubluxation in the differential diagnosis (see page 933).

Usually stable. Deficit is rare. Nonunion is rare. 90% heal with immobilization only. Operative fusion is rarely needed. Fractures of C2 that do not go through the isthmus are not true hangman’s fractures and may require different management (see Miscellaneous C2 fracturespage 967).


Levine/Effendi classification: The system of Effendi et al.136 as modified by Levine137 and others (see Table 28-20) is widely used in grading adult HF (not applicable to peds). Angulation is measured as the angle between the inferior endplates of C2 and C3. Anterior subluxation of C2 on C3 > 3 mm (Type II) is a surrogate marker for C2-3 disc disruption which can be evaluated more directly with cervical MRI.


Grading system of Frances et al: The grading system140 is shown in Table 28-21. The methodology of measurements is depicted in Figure 28-4.

Levine/Francis correlation: In a series of 340 axis fractures141, the most common fracture type was Type I in the Levine system (72%) and Grade I in the Francis system (65%); and there was a close correlation as follows:

Levine Type I ≈ Francis Grade I

Levine Type III ≈ Francis Grade IV

Other fracture types: Not all fractures fit into one or both of these classification systems142. Example: coronally oriented fracture extending through the posterior C2 vertebral body.

Table 28-21 Francis Grading* system for hangman’s fracture


Angulation θ



< 11°

d < 3.5 mm


> 11°


< 11°

d > 3.5 mm and d/b < 0.5


> 11°



disc disruption

see Figure 28-4 for definitions


Figure 28-4 Grading system of Francis


Most (≈ 95%) are neurologically intact, those few with deficits are usually minor (paresthesias, monoparesis…) and many recover within one month140. Almost all conscious patients will have cervical pain usually in the upper posterior cervical region, and occipital neuralgia is not uncommon143. There is a high incidence of associated head injury and there will be other associated C-spine injuries (e.g. C1 fracture (see above) or clay shoveler’s fracture (see page 969)) in ≈ one third, with most occurring in the upper 3 cervical levels. There are usually external signs of injury to the face and head associated with the hyperextending and axial force.


Cervical CT: with sagittal & coronal reconstructions should be done to fully assess the fracture.

CTA: should be done to evaluate the vertebral arteries if fracture extends through foramen transversarium (especially Levine Type IA) and in patients with symptoms suggestive of stroke. Some recommend CTA for all C2 fractures - see Table 28-36page 983). Alternatively, Angiography or MRA may be done as an alternative to CTA.

image MRI: cervical MRI should be done to look for C2-3 disc disruption (a marker for instability (Levine grade II) which usually requires surgical stabilization). Findings may include abnormal increased signal intensity on sagittal FLAIR or T2WI.

X-Rays: lateral C-spine x-rays show the fracture in 95% of cases. Also demonstrates C2 angulation and/or subluxation. Most fractures pass through the pars or the transverse foramen140, 7% go through the body of C2 (also see Miscellaneous C2 fracturespage 967). Instability can usually be identified as marked anterior displacement of C2 on C3 (guideline140: unstable if displacement exceeds 50% of the AP diameter of C3 vertebral body), excessive angulation of C2 on C3, or by excessive motion on flexion-extension films.

Patients suspected of having Levine Type I fractures and are neurologically intact should have physician-supervised flexion-extension x-rays to rule out a reduced type II fracture.



Level III144

• hangman’s fractures may initially be managed with external immobilization in most cases (halo or collar)

• surgical stabilization should be considered in cases of:

A. severe angulation of C2 on C3 (Francis II & IV, Levine II)

B. disruption of the C2-3 disc space (Francis V, Levine II)

C. or inability to establish or maintain alignment with external immobilization

Nonsurgical management produces adequate reduction in 97-100% and results in a fusion rate of 93-100%115145146 if the external immobilization is adequately maintained for 8-14 weeks147 (average time for healing is ≈ 11.5 weeks140). Specific treatment depends on the reliability of the patient and the degree of stability as described below. Most cases do well with non-halo immobilization146.


Treat with immobilization (Aspen or Philadelphia collar148 (p 2326) or cervicothoracic orthosis (CTO) (e.g. SOMI) is usually adequate) x 3 months138. Halo-vest may be needed in unreliable patients or for combination C1-C2 fractures. Schneider reported 50 cases of Type I fracture treated with non-halo fixation, only 1 was taken to surgery and was found to already be fused.


Levine Type II

Reduce with gentle cervical traction (most reduce with ≤ 30 lbs138) with the head in slight extension (preferably in halo ring) under close x-ray monitoring to prevent “iatrogenic hanging” in cases with ligamentous instability140. Place in halo vest x 3 months. Follow patients with serial x-rays. Stabilize surgically if fracture moves.

Type II fractures with ≤ 5 mm of subluxation & angulation < 10°: once reduced, apply halo-vest and begin to mobilize (usually within 24 hrs of injury). Verify that immobilization is adequate in the halo with upright lateral C-spine x-ray, operate if inadequate. After 8-12 weeks, change to Philadelphia collar or CTO until fusion is definitely complete (usually 3-4 months).

Type II fractures with > 5 mm subluxation or ≥ 10° of angulation: Surgical fusion in these patients is recommended because of the following concerns:

1. risk of settling if immediately mobilized in halo-vest

2. healing with significant angulation may result in chronic pain

3. if not reduced, the gap may be too large for bony bridging using traction alone

Alternatively, cervical traction can be maintained for ≈ 4 weeks and then reduction should be reassessed 1 hour after removing weight from traction, and if stable, again 24 hours after mobilizing in a halo vest. If unstable, return to traction and repeat trial at 5 & 6 weeks. If still unstable at 6 weeks, surgical fusion is recommended138.

Levine Type IIA

 Traction will accentuate the deformity138. Fractures should be reduced by immediate placement in halo vest (bypassing traction) with extension and compression applied. Halo-vest immobilization x 3 months produces ≈ 95% union rate.

Levine Type III

 Reduction with traction may be dangerous with locked facets. ORIF is recommended115. MRI prior to surgery is recommended to assess the C2-3 disc. Can follow ORIF with halo-vest for the fracture, or can fuse at the same time as ORIF.



Few patients have indications for surgical treatment of HF, and include those with:

1. inability to reduce the fracture (includes most Levine Type III & some Type II)

2. failure of external immobilization to prevent movement at fracture site

3. traumatic C2-3 disc herniation with compromise of the spinal cord149

4. established non-union: evidenced by movement on flexion-extension film140, see Flexion-extension cervical spine x-rayspage 940 (all failures of nonoperative treatment had displacement > 4 mm115)

Hangman’s fractures likely to need surgery141:

1. Levine Type II or III

2. or Francis grade II, IV or V

3. or if either:

A. anterior displacement of C2 VB > 50% of the AP diameter of the C3 VB

B. or if angulation produces widening of either the anterior or posterior borders of the C2-3 disc space > the height of the normal C3-4 disc below

Surgical options

1. fusion techniques:

A. posterior approach: if the fracture is not transfixed (osteosynthesis - see below) then a C1-2 fusion is required. This depends on the integrity of the C2-3 disc and facet joint capsules, otherwise a C1-3 fusion is required. Occasionally the occiput is incorporated as well. Options for C1-2 fusion:

1. C1-2 wiring and fusion

2. C1-2 lateral mass screws/rods (see page 185)

B. anterior C2-3 discectomy140 with fusion. Optional anterior plating via a transverse anterior cervical incision midway between the angle of the jaw and the thyroid cartilage145149

1. preserves more motion by excluding C1

2. this approach is also recommended for established non-union140

3. not optimal for Levine Type III requiring ORIF for locked facets

4. also used when at least a partial reduction cannot be achieved

5. technique:

a. to maximize access at the upper end of the approach (i.e. C2):

i. the upper and lower teeth must be approximated, which means an ET tube cannot be used. NT intubation, or if the patient has one, a tracheostomy, may be used. An OG tube also cannot be used, an NG tube is employed. Halter traction may be used for distraction and to keep the jaw closed

ii. the head is rotated slightly away from the side of the approach to move the mandible out of the way

b. skin incision just under the angle of the mandible (check level on fluoro or x-ray prior to making incision)

c. intervertebral discectomy as per usual

d. if it is intact and the MRI does not show a herniated disc, then the posterior longitudinal ligament need not be opened

e. the end plates should be made parallel to each other

f. anterior plating is often used to supplement

g. halo immobilization post-op is usually employed

2. osteosynthesis: screw placement from posterior approach through the C2 pedicle across the fracture fragment138 (p 443). Reduction must be achieved before the screw holes are drilled. The technique for C2 pedicle screws (see page 187) is used. The posterior fracture fragment may be overdrilled with a 3.5 mm drill. A “top hat” is placed in the hole and a 2.7 mm drill is used to drill the VB. Screw length: 30-35 mm for average adults. Alternatively, a lag screw may be used (with 20 mm unthreaded)


Plain x-rays should show trabeculation across the fracture site or interbody fusion of C2 to C3. Flexion-extension lateral radiographs should show no movement at the fracture site. Odontoid fractures

image Key concepts:

• 10-15% of C-spine fx. Can occur in older patients with minor trauma (GLF), or in younger patients typically following MVA, falls froma height, skiing…

• may be fatal at time of injury, most survivors are intact. Neck pain is common

• classification: Anderson & D’Alonzo (Table 28-22). Type II is the most common

• Tx: surgery is considered for: Type II if age > 50 yrs,, Type IIA, or Type II & III if displacement ≥ 5 mm or if alignment cannot be maintained with halo

Significant force is required to produce an odontoid fracture in a young individual, and is usually sustained in an motor vehicle accident (MVA), a fall from a height, a skiing accident, etc. In patients > 70 years age, simple ground level falls (GLF) with head trauma may produce the fracture. Odontoid fractures comprise ≈10-15% of all cervical spine fractures150. They are easily missed on initial evaluation, especially since significant associated injuries are frequent and may mask symptoms. Pathologic fractures can also occur, e.g. with metastatic involvement (see page 743).

Flexion is the most common mechanism of injury, with resultant anterior displacement of C1 on C2 (atlantoaxial subluxation). Extension only occasionally produces odontoid fractures, usually associated with posterior displacement.

Signs and symptoms

The frequency of fatalities at the time of the accident resulting directly from odontoid fractures is unknown, it has been estimated at 25-40%151. 82% of patients with Type II fractures were neurologically intact, 8% had minor deficits of scalp or limb sensation, and 10% had significant deficit (monoparesis to quadriplegia)152. Type III fractures are rarely associated with neurologic injury.

Common symptoms are high posterior cervical pain, sometimes radiating in the distribution of the greater occipital nerve (occipital neuralgia). Almost all patients with high posterior cervical pain will also have paraspinal muscle spasm, reduced range of motion of the neck, and tenderness to palpation over the upper cervical spine. A very suggestive finding is the tendency to support the head with the hands when going between the up-right and supine position. Paresthesias in the upper extremities and slight exaggeration of muscle stretch reflexes may also occur. Myelopathy may develop in patients with non-union (see page 965).


The classification system of Anderson and D’Alonzo153 is shown in Figure 28-5 and Table 28-22.


Figure 28-5 Major types of odontoid fractures (AP view)

Table 28-22 Classification of odontoid fractures





through tip (above transverse ligament), rare



through base of neck, the most common dens fracture (may be best seen on AP x-ray)

usually unstable


similar to type II, but with large bone chips at fracture site154, comprise ≈ 3% of type II odontoid fractures. Diagnosed by plain radiographs and/or CT

usually unstable


through body of C2 (usually involves marrow space). May involve superior articular surface

usually stable

* controversial, see text

Type I fractures are due to avulsion of the attachment of the alar ligament. They are very rare. Although long considered to be a stable injury, they may not occur as an isolated fracture and may be a manifestation of atlanto-occipital dislocation155. Also, it may be a marker for possible disruption of the transverse ligament156 which may result in atlanto-axial instability.

image Imaging pearl: a type III odontoid fracture may be misinterpreted as type II on sagittal CT reconstructions because the fracture appears to lie above the VB. Always check the coronal reconstruction which more readily demonstrates the relationship of the fracture to the VB.



Level II144isolated Type II odontoid fractures in adults ≥ 50 years age should be considered for surgical stabilization & fusion

Level III144

• Type I, II & III fractures may be managed initially with external cervical immobilization

• Type II & III: consider surgical fixation for:

A. fracture displacement ≥ 5 mm

B. or Type IIA fracture (comminution)

C. or inability to maintain alignment with external immobilization


So rare that meaningful analysis is difficult. Due to possible associated atlanto-axial instability, surgical fusion may at times be necessary.


Treatment remains controversial. No agreement has been reached after many attempts to identify factors that will predict which type II fractures are most likely to heal with immobilization and which will require operative fusion. Critical review of the literature reveals a paucity of well designed studies. A wide range of nonunion rates with immobilization alone (5-76%) is quoted: 30% is probably a reasonable estimate for overall nonunion rate, with 10% nonunion rate for those with displacement < 6 mm157. Possible key factors in predicting nonunion include:

1. degree of displacement: probably the most important factor

A. some authors feel that displacement > 4 mm increases nonunion153158

B. some authors use ≥ 6 mm as the critical value, citing a 70% nonunion rate147 in these regardless of age or direction of displacement

2. age:

A. children < 7 yrs old almost always heal with immobilization alone

B. some feel that there is a critical age above which the nonunion rate increases, and the following ages have been cited: age > 40 yrs (possibly ≈ doubling the nonunion rate)158, age > 55 yrs159, age > 65 yrs160, yet others do not support increasing age as a factor157

Indications for surgery

Given the above, there can be no hard and fast rules. The following is offered as a guideline (also, see isolated odontoid fractures above).

image Surgical treatment (instead of external immobilization) is recommended for odontoid Type II fractures in patients ≥ 7 years age with any of the following:

1. displacement ≥ 5 mm

2. instability at the fracture site in the halo vest (see below)

3. age ≥ 50 years: increases nonunion rate (with halo) 21-fold161

4. nonunion (see Table 28-24 for radiographic criteria) including firm fibrous union162, especially if accompanied by myelopathy143

5. disruption of the transverse ligament: associated with delayed instability123

Surgical options

1. odontoid compression screw: appropriate for acute type II fractures with transverse ligament intact and attached (see page 181)

2. C1-2 arthrodesis: see page 184 for options including wiring/fusion, transarticular screws, halifax clamps…


For those not meeting surgical indications above, 10-12 weeks of immobilization as suggested in Table 28-23 is recommended. There is no Class I medical evidence comparing immobilization options.

Halo vest: fusion rate = 72%157, appears superior to a SOMI. If a halo is used, obtain supine and upright lateral C-spine x-rays in the halo. If there is movement at the fracture site, then surgical stabilization is recommended.

Rigid collar157163: fusion rate = 53%.

In patients who are poor surgical candidates, there is theoretical and anecdotal rationale to consider calcitonin therapy (see page 993) in conjunction with a rigid cervical orthosis164.

Table 28-23 Immobilization for odontoid fractures

Fracture Type


Type I

collar, halo

Type II*

halo, collar*

Type IIA*


Type III*

collar, halo*

* consider surgery for these, use indicated brace when surgery not deemed appropriate

Table 28-24 Radiographic criteria of nonunion of odontoid fractures

1. defect in the dens with contiguous sclerosis of both fragments (vascular pseudarthrosis)

2. defect in the dens with contiguous resorption of both fragments (rarefying osteitis or atrophic pseudarthrosis)

3. defect in the dens with definite loss of cortical continuity

4. movement of dens fragment demonstrated on flexion-extension x-rays


The most common symptom of nonunion is continued high posterior cervical pain beyond the time that the brace is removed. Late myelopathy can develop in as many as 77% of mobile nonunions151165 as a result of motion and soft tissue proliferation around the unstable fracture site. The radiographic criteria for nonunion are shown in Table 28-24.


Early surgery is recommended for all type IIA fractures154.


≈ 90% heal with external immobilization (and analgesics) if adequately maintained for 8-14 weeks147. Halo-vest brace is probably best163, fusion rate ≈ 100% in 1 series157. Rigid collar: fusion rate = 50-70%; if used, monitor the patient with frequent C-spine x-rays to rule-out nonunion.


See Atlantoaxial fusion (C1-2 arthrodesis) on page 183 and Anterior odontoid screw fixation on page 181 for surgical options and operative details.




Level III171

• recommended: the following plain C-spine x-rays: AP, open-mouth odontoid, lateral (static & flexion-extension)

• consider: tomography (CT or plain) and/or MRI of craniocervical junction


Level III171

• patients without neurologic signs or symptoms (even with C1-2 instability) may be followed with clinical & radiographic surveillance

• those with neurologic signs or symptoms and C1-2 instability

A. may be managed with posterior C1-2 internal fixation and fusion

B. options:

1. posterior wiring & fusion. Post-op halo immobilization is recommended following these procedures

2. C1-2 transarticular screw fixation and fusion: successful screw placement seems to obviate the need for post-op halo

• consider occipitocervical fusion with or without C1 laminectomy for patients with irreducible cervicomedullary compression and/or evidence of associated occipitoatlantal instability

• consider transoral decompression in patients with irreducible ventral cervicomedullary compression

A separate bone ossicle of variable size with smooth cortical borders separated from a foreshortened odontoid peg, occasionally may fuse with the clivus. May mimic Type 1 or 2 odontoid fracture. Etiology is debated with evidence to support both of the following (diagnosis & treatment do not depend on which etiologic theory is correct):

1. congenital: developmental anomaly (nonunion of dens to body of axis). However, does not follow known ossification centers (see page 137), and has been demonstrated in 9 patients with previously normal odontoid processes166

2. acquired: postulated to represent an old nonunion fracture or injury to vascular supply of developing odontoid166167

True os odontoideum is rare. Ossiculum terminale: nonunion of the apex at the secondary ossification center, is more common.

Two anatomic types:

1. orthotopic: ossicle moves with the anterior arch of C1

2. dystopic: ossicle is functionally fused to the basion. May sublux anterior to the C1 arch


Main groups identified in the literature168:

1. occipitocervical/neck pain

2. myelopathy: further subdivided166

A. transient myelopathy: common following trauma

B. static myelopathy

C. progressive myelopathy

3. intracranial signs or symptoms: from vertebrobasilar ischemia

4. incidental finding

Most patients are neurologically intact and present with atlantoaxial instability which may be discovered incidentally. Many symptomatic and asymptomatic patients have been reported with no new problems over many years of follow-up169. Conversely, cases of precipitous spinal cord injury after seemingly minor trauma have been reported170.


The natural history is variable, and predictive factors for deterioration, especially in asymptomatic patients, have not been identified171.


For diagnosis, see os odontoideum below.

It is critical to R/O C1-2 instability. However, myelopathy does not correlate with the degree of C1-2 instability. An AP canal diameter < 13 mm does correlate with the presence of myelopathy.


Regardless of whether os odontoideum is congenital or an old non-union fracture, immobilization is unlikely to result in fusion. Therefore, when treatment is elected, surgery is required (usually atlantoaxial arthrodesis, see page 183). Miscellaneous C2 fractures

Comprise ≈ 20% of C2 fractures115. Includes fractures of spinous process, lamina, facets, lateral mass or C2 vertebral body. Fractures of spinous process or lamina may be treated with Philadelphia collar or cervicothoracic orthosis (CTO). Fractures which compromise the anterior or middle columns (i.e. fractures of facets, C2 body, or lateral mass) requires CTO or halo-vest if nondisplaced, or halo if displaced.



Level III144: external immobilization is recommended for isolated axis body fractures

28.6.5. Combination C1-2 injuries

Combination C1-2 injuries are relatively common and may imply more significant structural and mechanical injury than isolated C1 or C2 fractures. The frequency of C2 fractures in C1-2 combination injuries is shown in Table 28-25. 5-53% of patients with Type II or III odontoid fractures and 6-26% of hangman’s fractures have an associated C1 fracture172.

Table 28-25 Accompanying C2 injuries



Type II dens fracture


Type III dens fracture


hangman’s fracture






Level III172

• recommended: base treatment primarily on the type of C2 injury

• recommended: external immobilization of most C1-2 fractures

• consider surgical stabilization* for these situations:

A. C1-Type II odontoid combination fractures with an ADI ≥ 5 mm

B. C1-hangman’s combination fractures with C2-3 angulation ≥ 11°

* loss of integrity of the C1 ring may necessitate modification of the surgical technique

 these injuries are potentially unstable (see Axis (C2) fractures on page 959)

Treatment options172 are summarized in Table 28-26.

Table 28-26 Treatment options for combination C1-C2 injuries


Treatment options

C1 + hangman’s



collar, halo, surgery*

untable (C2-3 angulation ≥ 11°)

halo, surgery

C1 + Type II odontoid fracture


stable (ADI* < 5 mm)

collar, halo, surgery

unstable (ADI ≥ 5 mm)

halo, surgery

C1 + Type III odontoid fracture


C1 + miscellaneous C2

collar, halo

* abbreviations: ADI = atlantodental interval; surgery = surgical fixation & fusion


Only 1 nonunion (C1-Type II odontoid, treated initially with halo). No new neuro deficits.

28.6.6. Subaxial (C3 through C7) injuries/fractures

Classification systems

Various systems have been proposed to help assess stability and/or guide management. A descriptive system by Allen, et al (see below) is based on the mechanism of injury. Attempts at quantifying biomechanical stability include the White and Panjabi system (see page 969) and the recently proposed subaxial injury classification (SLIC) (see below). Measurements for spine injuries are based on methods outlined by Bono et al.173.


A system adapted from Allen et al.176 divides cervical spine fracture/dislocations into 8 major groups based on the dominant loading force and neck position at the time of injury as shown in Table 28-27. Grades of severity within each group are described, and any of these fractures may also be associated with damage from rotatory loads.

Details on some of these fracture types are provided in following sections.



The subaxial injury classification (SLIC)174 (Table 28-29) includes injuries to the disco-ligamentous complex (DLC) in addition to neurologic and bony injuries. This system demonstrates moderate reliability.

DLC integrity174: The DLC includes: anterior longitudinal ligament (the strongest component of the anterior DLC), posterior longitudinal ligament, ligamentum flavum, facet capsule (the strongest component of the posterior DLC), interspinous and supraspinous ligaments. The DLC is the hardest SLIC parameter to evaluate. Largely inferred indirectly from MRI findings. Healing is less predictable than bone healing in the adult. More data needs to be accrued before this parameter can be reliably quantified.

Management guidelines

Table 28-28 Management based on total SLIC score

SLIC score



non surgical


not specified

≥ 5


Table 28-29 Subaxial injury classification (SLIC)174

Injury (rate the most severe injury at that level)



No abnormality


Simple compression (compression fx, endplate disruption, sagittal or coronal plane VB fx.)


Burst fracture


Distraction (perched facet, posterior element fx.)


Rotation/translation (facet dislocation, teardrop fx., advanced compression injury, bilateral pedicle fx., floating lateral mass (page 976)…). Guidelines: relative axial rotation ≥ 11°175 or any translation not related to degenerative causes


Discoligamentous complex (DLC)



Indeterminate (isolated interspinous widening with < 11° relative angulation & no abnormal facet alignment, ↑ signal on T2WI MRI in ligaments…)


Disrupted (perched or dislocated facet, < 50% articular apposition, facet diastasis > 2 mm, widened anterior disc space, ↑ signal on T2WI MRI through entire disc…)


Neurologic status



Root injury


Complete spinal cord injury


Incomplete spinal cord injury


• Continuous cord compression with neuro deficit


Description using the SLIC

A given injury can be described as follows:

1. spinal level

2. SLIC morphology (from Table 28-29): use the most severe injury type at this level

3. description of bony injury: e.g. fracture or dislocation of transverse process, pedicle, endplate, superior or inferior articular process, lateral mass…

4. SLIC DLC status (from Table 28-29) with descriptors: e.g. herniated disc…

5. SLIC neurologic status (from Table 28-29)

6. confounders: e.g. presence of ankylosing spondylitis, DISH, osteoporosis, previous surgery, degenerative disease…


Guidelines for determination of clinical instability (see page 930) of the subaxial cervical spine published by White and Panjabi1 (p 314) are shown in Table 28-30. In general, compromise of anterior elements produces more instability in extension, whereas compromise of the posterior elements produces more instability in flexion (important in patient transfers and immobilization). NB: certain conditions such as ankylosing spondylitis may cause an otherwise stable injury to be unstable (see page 502).

Stretch test: The cervical stretch test may be helpful in cases where stability is difficult to determine based on other factors. It may also be useful in detecting instability in cases such as an athlete with no obvious bony or ligamentous disruption. It is performed by applying graduated cervical traction with the patient lying supine on an x-ray table. Serial neurologic exams and lateral radiographs are performed as outlined in the footnote of Table 28-30.


Avulsion of spinous processes (usually C7) first described in Perth, Australia (path-omechanics: during the throwing phase of shovelling, clay may stick to the shovel jerking the trapezius and other muscles which are attached to cervical spinous processes)178. Can also occur with: whiplash injury179, injuries that jerk the arms upwards (e.g. catching oneself in falling), neck hyperflexion, or a direct blow to the spinous process.

This fracture is stable, and by itself poses little risk. If the patient is intact, they should have further study (flexion-extension C-spine x-rays or CT scan through the affected level) to R/O other occult fractures. A rigid collar is used PRN pain.


In order to apply a purely compressive force to the cervical spine, reversal of the normal cervical lordosis is required, as may occur in a slightly flexed posture. Burst fractures are the most common result, with the possibility of retropulsion of bone into the spinal canal.


Constitutes up to 15% of cervical spine trauma. Common causes include: MVAs, falls from a height, and diving into shallow water180.

Table 28-30 Guidelines for diagnosing clinical instability of the mid & lower C-spine1



anterior elements destroyed or unable to function


posterior elements destroyed or unable to function


positive stretch test


spinal cord damage


nerve root damage


abnormal disc narrowing


developmentally narrow spinal canal, either

• sagittal diameter < 13 mm, OR

• Pavlov ratio§ < 0.8


dangerous loading anticipatedΔ


Radiographic criteria

• neutral position x-rays

sagittal plane displacement > 3.5 mm or 20%

relative sagittal plane angulation > 11°





• flexion-extension x-rays

sagittal plane translation > 3.5 mm or 20%

sagittal plane rotation > 20°



Unstable if total ≥ 5

* if there is inadequate information for any item, add half of the value for that item to the total

 in the C-spine, posterior elements = anatomic components posterior to the posterior longitudinal ligament

 stretch test: apply incremental cervical traction loads of 10 lbs q 5 mins up to 33% body wt. (65 lbs max). Check Xray and neuro exam after each Δ. Positive if Δ in separation > 1.7 mm or Δ angle > 7.5° on x-ray or change in neuro exam. This test is contraindicated if obvious instability

§ Pavlov ratio = the ratio of (distance from the midlevel of the posterior VB to the closest point on the spinolaminar line): (the AP diameter of the middle of the VB)

Δ e.g. heavy laborers, contact sports athletes, motorcyclists


The classic diving injury is the prototypical example. Posterior element fractures occur in up to 50% of compression flexion injuries181. Although flexion-compression injuries do distract the posterior elements to some degree, most do not produce posterior ligamentous injuries. Sub-types include: teardrop fractures (see below), quadrangular fractures (see page 971).

Treatment: mild cervical compression fractures without neurologic deficit or retropulsion of bone into the spinal canal are usually treated with a rigid orthosis until x-rays show healing has occurred (usually 6-12 wks). Stability is assessed with flexion-extension views (see page 940) before discontinuing the brace. More severe compression fractures heal in a halo brace with ≈ 90% rate of ankylosing fusion.


Originally described by Schneider & Kahn182. Results from hyperflexion or axial loading at the vertex of the skull with the neck flexed (eliminating the normal cervical lordosis)183 (often mistakenly attributed to hyperextension because of the retrolisthesis). There are varying degrees of severity. In its most severe form, the injury consists of complete disruption of all of the ligaments, the facet joints and the intervertebral disk184. An important feature is displacement of the inferior margin of the fractured vertebral body posteriorly into the spinal canal182. Usually unstable.

Seen in ≈ 5% of patients in a large series with x-ray evidence of cervical spine trauma90. Patients are often quadriplegic, although some may be intact and some may have anterior cervical cord syndrome (see page 950). Possible associated injuries and radiographic findings include184185:

1. a small chip of bone (the “teardrop”) just beyond the anterior inferior edge of the involved vertebral body (VB) on lateral cervical spine film

2. often associated with a fracture through the sagittal plane of the VB (sagittal split) which can usually be seen on AP view. Thin cut CT scan is more sensitive

3. a large triangular fragment of the anterior inferior VB

4. other fractures through the vertebral body may also occur

5. the fractured vertebrae is usually displaced posteriorly on the vertebra below (easily appreciated on oblique x-rays, see Figure 28-6page 972). However, cases without retrolisthesis are also described181

6. the fractured body is often wedged anteriorly (kyphosis), and may also be wedged laterally

7. disruption of the facet joints which may be appreciated as separation of the joints on lateral x-ray, often unmasked by cervical traction

8. prevertebral soft-tissue swelling (see page 136, for measurements)

9. narrowing of the intervertebral disc below the fracture (indicating disruption)

Distinguishing between teardrop fracture and avulsion fracture

Rationale: Teardrop fractures must be distinguished from a simple avulsion fracture which may also result in a small chip of bone off the anterior inferior VB, usually pulled off by traction of the anterior longitudinal ligament (ALL) in hyperextension. Although there may be disruption of the ALL in these cases, it does not usually cause instability.

Methodology: In a patient with a small bone chip off of the inferior anterior VB, a “tear-drop” fracture needs to be ruled-out. Determine if the following criteria are met:

• neurologically intact (because of the need for cooperation, this includes mental status, and excludes the inebriated or concussed patient)

• size of bone fragment is small

• no malalignment of vertebral bodies

• no evidence of VB fracture in sagittal plane on AP C-spine x-rays or on CT

• no posterior element fracture on x-ray or CT

• no prevertebral soft tissue swelling at level of fragment (see page 136)

• and no loss of vertebral body height or disc space height

If the above criteria are met, obtain flexion-extension C-spine x-rays (see Flexion-extension cervical spine x-rayspage 940). If no abnormal movement, discharge patient in rigid collar (e.g. Philadelphia collar), and repeat the films in 4-7 days (i.e. after the pain has subsided to be certain that alignment is not being maintained by cervical muscle spasm from pain), D/C collar if 2nd set of films is normal.

If the patient does not meet the above criteria, treat them as an unstable fracture and obtain a CT scan through the fractured vertebra to evaluate for associated fractures (e.g. sagittal plane fracture that may not be apparent on plain x-ray).

MRI assesses the integrity of the disc and gives some information about the posterior ligaments.

Treatment of teardrop fracture

If the disc and ligaments are intact (determined by MRI) then an option is to employ a halo brace until the fragment is healed (perform flexion-extension x-rays after removing the halo to rule-out persistent instability). Alternatively, surgical stabilization may be performed, especially if ligamentous or disc injury is seen on MRI. When the injury is primarily posterior due to disruption of the posterior ligaments and facet joints, and if there is no anterior compromise of the spinal canal, then posterior fusion suffices (e.g. see page 179). Severe injuries with canal compromise often require a combined anterior decompression and fusion (performed first) followed by posterior fusion using either a modified Bohlman triple-wire technique or lateral mass screws and rods.


Four features:

1. oblique vertebral body (VB) fracture passing from anterior-superior cortical margin to inferior end plate

2. posterior subluxation of superior VB on the inferior VB

3. angular kyphosis

4. disruption of disc and anterior and posterior ligaments

Treatment: May require combined anterior and posterior fusion.


Ranges from hyperflexion sprain (mild, see below) to minor subluxation (moderate) to bilateral locked facets (severe, see below). Posterior ligaments are injured early and are usually evidenced by widening of the interspinous distance (see page 137).


A purely ligamentous injury that involves disruption of the posterior ligamentous complex without bony fracture. May be missed on plain lateral C-spine x-rays if they are obtained in normal alignment (requires flexion-extension views, see Flexion-extension cervical spine x-rayspage 940). Instability may be concealed when films are obtained shortly after the injury if spasm of the cervical paraspinal muscles splints the neck and prevents true flexion187. For patients with limited flexion, a rigid collar should be prescribed, and if the pain persists 1-2 weeks later the films should be repeated (including flexion-extension).

Radiographic signs of hyperflexion sprain188 (x-rays may also be normal):

1. kyphotic angulation

2. anterior rotation and/or slight (1-3 mm) subluxation

3. anterior narrowing and posterior widening of the disc space

4. increased distance between the posterior cortex of the subluxed vertebral body and the anterior cortex of the articular masses of the subjacent vertebra

5. anterior and superior displacement of the superior facets (causing widening of the facet joint)

6. fanning (abnormal widening) of the interspinous space on lateral C-spine x-ray, or increased interspinous distance on AP (see Interspinous distancespage 137)


Cadaver studies have shown that horizontal subluxation > 3.5 mm of one vertebral body on another, or > 11° of angulation of one vertebral body relative to the next indicates ligamentous instability189190 (see Table 28-30page 970). Thus, if subluxation of ≤ 3.5 mm on plain films is seen, and there is no neuro deficit, obtain flexion-extension films (see Flexion-extension cervical spine x-rayspage 940). If no abnormal movement, remove cervical collar.


Severe flexion injuries can result in locked facets (AKA “sprung” facets AKA “jumped” facets) with reversal of the normal “shingled” relationship between facets (normally the inferior facet of the level above is posterior to the superior facet of the level below). Involves disruption of facet capsule. Facets that have not completely locked but have had significant ligamentous disruption allowing distraction just short of the point of locking are known as “perched facets”.

Flexion + rotation → unilateral locked facets. Hyperflexion → bilateral locked facets.

Unilateral locked facets:

25% of patients are neurologically intact, 37% have root deficit, 22% have incomplete cord injuries, and 15% are complete quadriplegics191.

Bilateral locked facets:

Occurs with disruption of ligaments of apophyseal joints, ligamentum flavum, longitudinal and interspinous ligaments, and the anulus. Rare. Most common at C5-6 or C6-7. 65-87% have complete quadriplegia, 13-25% incomplete, ≤ 10% are intact. Adjacent fractures (VB, facet, lamina, pedicle…) occur in 40-60%176192. Nerve root deficits may also occur.


Figure 28-6 Unilateral locked facets (left C4 on C5) & C5 tear-drop fracture (see page 970) 60° LAO C-spine x-ray on left, and schematic on right (sagittally oriented VB fracture through C5 seen on CT scan, not shown). Note the anterior subluxation of C4 on C5, and the slight retrolisthesis of C5 on C6


C-spine x-rays: both unilateral (ULF) and bilateral locked facets (BLF) will produce subluxation (ULF → rotatory subluxation).

BLF (bilateral locked facets): usually produces > 50% subluxation on lateral C-spine xray.

ULF (unilateral locked facets):

1. AP: spinous processes above the subluxation rotate to the same side as the locked facet (with respect to those below)

2. lateral: “bow-tie sign” (visualization of left & right facets at the level of the injury instead of the normal superimposed position191). Subluxation may be seen. Disruption of the posterior ligamentous complex may produce widening of the interspace between spinous processes

3. oblique (see Figure 28-6): may demonstrate the locked facet which will be seen blocking the neural foramen (use ≈ 60° LAO for left locked facet, 60° RAO for right)

CT: “naked facet sign”: the articular surface of the facet will be seen with the appropriate articulating mate either absent or on the wrong side of the facet (see Figure 28-7). With ULF, CT also demonstrates the rotation of the level above anteriorly on the level below on the side of the locked facet.

MRI: the best test to rule-out traumatic disc herniation (found in 80% of BLF)193.


Figure 28-7 Locked facet (left C4-5). (CT scan). Note the rotation of the C4 vertebral body on C5 (curved arrow)



Level III177

• initial treatment: closed or open reduction is recommended

• subsequent treatment

A. rigid external immobilization, anterior arthrodesis with plate fixation, or posterior arthrodesis with plate or rod or interlaminar clamp fixation

B. prolonged bedrest in traction if the above treatment options are not available

* excluding facet dislocations, see page 973 for those injuries

Closed reduction of locked facets:  Contraindicated if traumatic disc herniation is demonstrated on MRI. Patients who cannot be assessed neurologically may be done using SSEP monitoring. Two methods of closed reduction:

1. traction: more commonly employed in the U.S.

A. initial weight (in lbs) ≈ 3 x cervical vertebral level, increase in 5-10 lb increments usually at 10-15 minute intervals until desired alignment is attained (assess neurologic exam (or SSEP/MEP) and lateral C-spine x-ray or fluoroscopy after each Δ to avoid overdistraction)

B. end points (i.e. stop the procedure):

• usually do not exceed 10 lbs per vertebral level (some say 5 lbs/level). This is just a guideline to avoid overdistraction

• distraction of perched/locked facet or desired reduction is achieved

• if occipitocervical instability develops

• if any disc space height exceeds 10 mm (overdistraction)

• if any neurologic deterioration or deterioration of SSEP/MEP

C. with unilateral locked facets, one may add gentle manual torsion towards the side of the locked facets. With bilateral locked facets, one may add gentle manual posterior tension (e.g. with a rolled towel under the occiput)

D. once the facets are perched or distracted, gradual reduction of the weights will usually result in reduction - verify with x-ray (placing the neck in slight extension, e.g. with small shoulder roll, may help maintain the reduction)

2. manipulation: less commonly accepted191. Involves manually applying axial traction and sagittal angulation sometimes with rotation and direct pressure at the fracture level under fluoroscopy

Paraspinal muscle relaxation (but not enough to cause obtundation) may assist in reduction. Use IV diazepam (Valium®) and/or narcotic. General anesthesia may be used in difficult cases (with SSEP/MEP monitoring).

Once reduction is achieved, the patient is left in 5-10 lbs of traction for stabilization.

Disadvantage of closed reduction

1. fails to reduce ≈ 25% of cases of BLF

2. risks overdistraction at higher levels or worsening of other fractures

3. neurologic worsening following closed reduction may occur with traumatic disc herniation192194 and should be evaluated immediately with MRI and if confirmed treated with prompt discectomy

4. adds time and potentially pain to the patient’s care, especially since many will go on to have surgical fusion anyway

Following closed reduction, the need for internal (operative) stabilization vs. external stabilization (i.e. bracing) may be addressed (see Stabilization below).

Open reduction and fixation is usually required if reduction is not achieved. Closed reduction is often more difficult with bilateral locked facets than with unilateral.

Open reduction of locked facets:

1. posterior approach: the most common approach. Although rare, still subjects the patient to risk of deterioration from traumatically herniated disc. Therefore a pre-op MRI should be done if possible. Often requires drilling of the superior aspect of the articular facet of the level below. A foraminotomy is recommended when there are root symptoms to visualize and decompress the root

2. anterior approach: by removing the disc at the subluxed level and exploring the anterior epidural space, the risk of worsening deficit due to a traumatic herniated disc is theoretically reduced. Reduction may be achieved by adding simultaneous manual traction

3. combined anterior/posterior (360°) approach: using anterior plate and posterior lateral mass screws/rods eliminates need for post-op external immobilization

Stabilization: Surgical fusion is commonly performed after successful closed reduction, failed closed reduction, or following open reduction.

If there are fracture fragments about the articular surfaces, there may be satisfactory healing with halo vest immobilization (for 3 months) once closed reduction is achieved195. Frequent x-rays are needed to rule-out redislocation196. Flexion-extension x-rays are obtained upon halo removal and surgery is required for continued instability. Up to 77% of patients with unilateral or bilateral facet dislocation (with or without facet fracture fragments) will have a poor anatomic result with halo vest alone (although late instability was uncommon), suggesting that surgery should be considered for all of these patients197. Surgical fusion is more clearly indicated in cases without facet fracture fragments (ligamentous instability alone may not heal) or if open reduction is required.

If surgery is indicated, an MRI should be done beforehand if possible. A posterior approach is preferred if there are no anterior masses (such as traumatic disc herniation or large osteophytic spurs), if subluxation of the bodies is > one third the VB width (suggesting severe posterior ligamentous injury), or for fractures of the posterior elements. A posterior approach is mandatory if there is an unreducible dislocation. Options for posterior approach: see Choice of posterior techniquepage 978.



Extension injuries can produce spinal cord injury (SCI) without evidence of bony injury. Injury patterns include central cord syndrome (see page 948) usually in an older adult with cervical spondylosis, and SCIWORA (see below) usually in young children. Middle aged adults with hyperextension dislocations that reduce spontaneously immediately may present with SCI and no bony abnormality on x-ray, but there may be rupture of the anterior longitudinal ligament and/or intervertebral disc on MRI or autopsy. Extension forces may also be associated with carotid artery dissections (see page 1163).


Although spinal cord injuries are uncommon in children, there is a subgroup of these in which no radiographic evidence of bony or ligamentous disruption can be demonstrated (including on dynamic flexion-extension x-rays). This is attributed to the normally increased elasticity of the spinous ligaments and paravertebral soft-tissue in the young population199 and has been dubbed SCIWORA (an acronym for “Spinal Cord Injury Without Radiographic Abnormality”). The age range of children with SCIWORA is 1.5-16 yrs, it has a much higher incidence in age ≤ 9 yrs11. The spinal cord may undergo contusion, transection, infarction, stretch injuries, or meningeal rupture. Additional etiologies include: blunt abdominal trauma with disruption of blood flow from the aorta or segmental branches, traumatic disc herniation. There may be an increased risk of SCIWORA among young children with asymptomatic Chiari I malformation200.

54% of children had a delay between injury (at which time some children experience transient numbness, paresthesias, Lhermitte’s sign, or a feeling of total body weakness) and the onset of objectivesensorimotor dysfunction (“latent period”) ranging from 30 minutes to 4 days.



Level III198

• recommended: plain C-spine x-rays and spinal CT through the suspected level of injury to R/O occult fractures

• MRI of the region of suspected injury may provide useful diagnostic information

• consider: plain x-rays of the entire spinal column

 not recommended: spinal angiography or myelography


Level III198

• recommended: external immobilization until stability is confirmed with flexion-extension x-rays

• consider: external immobilization of the injured spinal segment for up to 12 weeks

• consider: avoidance of “high-risk” activities for up to 6 months after SCIWORA


Level III198

• MRI through the region of neurologic injury may provide useful information about neurologic outcome after SCIWORA

Radiographic evaluation

In addition to plain films and flexion-extension films (to identify overt instability which would require surgical fusion), should include MRI which may show increased signal within the spinal cord parenchyma on T2WI. There were no intraspinal space occupying lesions in 13 patients studied with myelography/CT199.


Surgical intervention, including laminectomy, has shown no benefit in the few cases where it has been tried202.

Due to a 20% rate of repeat injury (some due to trivial trauma, and some without identifiable trauma) within 10 weeks of the original trauma when treated with only a rigid collar and restriction of contacts sports (both for 2 months), more aggressive measures were initially recommended (see Table 28-31).

Table 28-31 Treatment protocol for SCIWORA (modified201)

• admit patient to hospital (helps emphasize seriousness of injury)

• BR with rigid cervical collar until flexion-extension films are normal

• MRI of cervical spine to document presence of spinal cord injury

• detailed discussion with patient and family about seriousness of injury and rationale for treatment outlined here

• immobilization in Guilford brace for 3 months*

• prohibition of contact and noncontact sports

• regular follow-up visits for monitoring condition and compliance

• liberalize activities at 3 months if flexion-extension films are normal

* this represents a conservative recommendation, a less restrictive recommendation is immobilization for 1-3 weeks202 (see PRACTICE GUIDELINE 28-19 SCIWORA)


Includes spinous process and lamina fractures. By themselves, are stable.


This is the most common mechanism of lateral mass/facet fractures (see below).


Classification of cervical lateral mass and facet fractures

4 patterns identified in lateral mass and facet fractures203 are shown in Table 28-32.

Anterior subluxation of the fractured vertebra was observed in 77% of whole lateral-mass fractures203.

Horizontal facet or separation fracture of the articular mass: Extension combined with compression and rotation may produce fracture of one pedicle and ipsilateral lamina which permits the detached articular mass (“floating” lateral mass) to rotate forward to a more horizontal orientation204 (horizontalization of the facet) (see Table 28-32). May be associated with rupture of the anterior longitudinal ligament (ALL) and fissure of the disc at one or two levels. Neuro deficit is common. Unstable.


Failure of nonoperative treatment

A study of CT scans of 26 unilateral cervical facet fractures205 identified the risk factors shown below for failure of nonoperative treatment (see Figure 28-8 for illustration of the measurement definitions): where the fracture fragment (FF) height was defined as the maximum tip-to-tip cephalocaudal height on sequential sagittal reconstructions.

Nonoperative management is likely to fail if FF is:

1. > 1 cm, or

2. > 40% of LM (the height of the intact contralateral lateral mass at the same level, defined as the maximum tip-to-tip cephalocaudal height on sequential sagittal reconstructions)


Figure 28-8 Facet fracture fragment measurements. Sagittal reconstructed CT. FF = fracture fragment height, LM = lateral mass height (measured on the contralateral side at the same level as the fracture, (not as shown here which just illustrates the technique used to measure LM)

Surgical treatment of cervical lateral mass & facet fractures

Most cases can be treated with a posterior approach using fixation screws (lateral mass screws or pedicle screws203) and rods extending at least 1 level above and below the level of fracture (usually omitting a screw on the side of the fracture at the index level). Simultaneous neural decompression is performed when needed. Additional treatment with an anterior approach may be required for release of rigid deformity or for additional anterior column support203. Some separation fractures may be candidates for osteosynthesis (to preserve motion) using a cervical pedicle screw203 that traverses the fracture.

An anterior approach is an alternative. Advantage: usually only 1 level needs to be fused. Disadvantages: decompression of compressing fragments cannot always be accomplished and requires disrupting an area that may not be compromised (if there is subluxation, the anterior column is probably compromised). Treatment of subaxial cervical spine fractures


Management of some types of C-spine fractures is covered in the preceding sections. For injuries not specifically addressed, general management principles are as follows1:

1. immobilize and reduce externally (if possible): may use traction x 0-7 days

2. determine if there is an indication for decompression as soon as practical (clinical conditions permitting), and decompress if needed. Although controversial, the following are generally accepted indications for acutedecompression in patients without complete spinal cord injury:

A. radiographic evidence of bone or foreign material in the spinal canal with associated spinal cord symptoms

B. complete block on CT, myelogram or MRI

C. clinical judgement: e.g. a progressive incomplete spinal cord injury where the surgeon believes that decompression would be beneficial

3. ascertain stability of the injury (see page 969)

A. stable fractures: treat in non-halo orthosis for 1-6 weeks (see page 979)

B. unstable fractures: all of the following choices are appropriate, with little evidence to recommend one scheme over another in most cases

1. traction x 7 weeks, followed by orthosis x 8 weeks

2. halo x 11 weeks, followed by orthosis x 4 weeks

3. surgical fusion, followed by orthosis x 15 weeks

4. surgical fusion with internal immobilization (lateral mass screws & rods…) ± orthosis for short period of time (≈ several weeks)



Operating on a patient with a complete cord injury does not result in significant recovery of neurologic function147. However, aggressive non-surgical reduction of traumatic subluxation should be pursued.

The primary goal of surgery in this setting is spinal stabilization, allowing the patient to be placed in a sitting position for improved pulmonary function, for psychological benefit, and to allow initiation of rehabilitation. Although the spine will fuse spontaneously in many cases (taking ≈ 8-12 weeks), surgical stabilization expedites the mobilization process and reduces the risk of delayed kyphotic angulation deformity. Early surgery may lead to further neurological injury, and should be delayed until the patient has stabilized medically and neurologically. In most cases, performing surgery within 4-5 days (if the patient is otherwise stable) is probably early enough to help reduce pulmonary complications.


Patients with incomplete cord injuries who have compromise of the spinal canal (by bone, disc, unreducible subluxation or hematoma) and either do not improve with nonoperative therapy or deteriorate neurologically should undergo surgical decompression and stabilization147. This may facilitate some further return of spinal cord function. An exception may be the central cord syndrome (see page 948).


The choice of technique depends to a large degree on the mechanism of injury, as the treatment should tend to counteract the instability, and ideally should not compromise structures that are still functioning. Instrumentation (wires/cables, lateral mass screws & rods, clamps…) immobilize the area of instability while bony fusion is occurring. In the absence of bony fusion, all mechanical devices will eventually fail, and so it becomes a “race” between fusion and instrument failure. Extensive injuries (including teardrop fractures (see page 970) and compression burst fractures) may require a combined anterior and posterior approach (staged, or in a single sitting; anterior decompression precedes posterior fusion).


Indications: The procedure of choice for most flexion injuries. Useful when there is minimal injury to the vertebral bodies and in the absence of anterior compression of the spinal cord and nerves. Including: posterior ligamentous instability, traumatic subluxation, unilateral or bilateral locked facets, simple wedge compression fractures.

The most common technique consists of open or closed reduction, followed by lateral mass screws & rods (see page 179). Interlaminar Halifax clamps are an alternative206. Although successes have been reported using methylmethacrylate207, it does not bond to bone and weakens with age, and thus its use in the setting of traumatic injury is discouraged208.

Choice of posterior technique: If the anterior column is capable of weight-bearing and the posterior elements are not damaged or absent, wiring and fusion provides adequate stability. If the anterior weight-bearing column is significantly damaged, or if there is absence or compromise of the lamina or spinous processes, then either a combined anterior-posterior approach is needed or posterior rigid instrumentation (e.g. lateral mass screw-plate or rod fixation) with fusion is recommended209.


Does not depend on integrity of posterior elements to achieve stability.


1. fractured vertebral body with bone retropulsed into spinal canal (burst fracture)

2. most extension injuries

3. severe fractures of posterior elements that preclude posterior stabilization and fusion

4. may be used for traumatic subluxation of the cervical spine

Usually consists of:

1. corpectomy: decompresses the neural elements (if necessary) and removes fractured and structurally compromised bone

A. decompression usually requires wide corpectomy, at least ≈ 16 mm (palpate anterior surface of vertebral body to determine width; note position of vertebral arteries on pre-op CT). NB: it is suggested to take the corpectomy no wider than 3 mm lateral to the medial edge of the longus coli muscle, this leaves ≈ 5 mm margin of safety to the foramen transversarium210

B. if decompression is not needed, ≈ 12 mm corpectomy suffices (i.e. about the width of a half-inch cottonoid)


2. strut graft fusion: replaces the involved body or bodies with either:

A. bone (usually iliac crest, rib or fibula, either homologous or cadaveric)

B. or synthetic cage (e.g. titanium or PEEK)

3. usually accompanied with compression plates

4. usually followed with external immobilization


1. hardware problems

A. wire failure

1. improper wire gauge for type of fracture

2. improper wire-handling

3. inadequate post-operative immobilization

a. improper brace selected

b. poor patient compliance with immobilization device

B. problems with plating

1. screw pull-out, loosening or breakage

2. fatigue fracture of plate

3. screw injury: nerve root, spinal cord or vertebral artery

2. failure of graft to take (nonunion)

3. judgmental error

A. failure to incorporate all unstable levels

B. improper surgical approach



Soft (sponge rubber) collar: does not immobilize the cervical spine to any significant degree. Its function is primarily to remind the patient to reduce neck movements.

Rigid cervical collars

Inadequate for stabilizing upper and mid-cervical spine and for preventing rotation. Common rigid collars:

• Miami J collar & Aspen collar: have removable pads

• Philadelphia collar: no removable pads. Feels hotter to wear


Distinguished from cervicothoracic orthoses (see below) by the lack of straps under the axilla. Includes the four poster brace. Generally good for preventing flexion at midcervical levels.


Cervicothoracic orthoses (CTO) incorporate some form of body vest to immobilize the cervical spine. The following are presented in increasing degree of immobilization.

Guilford brace: essentially a ring around the occiput and chin connected by two posts to anterior and posterior thoracic pads.

SOMI brace: acronym for Sternal Occipital Mandibular Immobilizer. Good for bracing against flexion (especially upper cervical spine). Inadequate for hyperextension type injuries because of weak occipital support. Has special forehead attachment to allow patient to eat comfortably without mandibular support.

Yale brace”: a sort of extended Philadelphia collar. The most effective CTO for bracing against flexion-extension and rotation. Major shortcoming is poor prevention of lateral bending (only ≈ 50% reduced).


Can immobilize the upper or lower cervical spine, not very good for mid-cervical spine (due to snaking of the midcervical spine). Unable to provide adequate distraction support following vertebral body resection when patient assumes upright position (i.e. it is not a portable cervical traction device). Overall reduction of flexion/extension as well as lateral bending is ≈ 90-95%, rotation is reduced by 98%. For placement, see page 942.


After initial management (surgical or nonsurgical) of cervical spine problems (stable or unstable) the follow-up schedule shown in Table 28-33 is suggested to permit recognition of problems in time for treatment1.

Table 28-33 Sample follow-up cervical spine clinic visit schedule

Time post-op


7-10 d

(for post-op patients only) wound check, D/C sutures/staples if used

6 wks

AP & lateral C-spine x-ray in brace

3 months

• AP & lateral C-spine x-rays with flexion/extionsion views out of brace

• if x-rays look good and patient is doing well, begin weaning brace

6 months

• AP & lateral C-spine x-rays with flexion/extionsion views

• some surgeons release patients at this time if they are doing well

1 year (optional)

• AP & lateral C-spine x-rays with flexion/extionsion views

• release patient if they are doing well

28.6.7. Sports related cervical spine injuries

Any of the previously described injuries can be sports related. This section considers some injuries peculiar to sports. Also see page 850 for sports-related head injuries.

Bailes et al.211 classified sports-related spinal cord injures (SCI) as shown in Table 28-34. Type II injuries include spinal concussion, spinal neuropraxia (see below), and the burning hands syndrome (see below), all in the absence of radiographic abnormalities and all with complete resolution of symptoms. Patients should be carefully evaluated, and return to competition should not be allowed in the presence of neurologic deficit, radiographically demonstrated injury, certain congenital C-spine abnormalities, and possibly for “repeat offenders” (see Return to play and pre-participation guidelines below). Type III injuries are the most common. Unstable injuries should be treated appropriately (see page 977).

Table 28-34 Sports-related spinal cord injuries




permanent SCI


transient SCI without radiographic abnormality


radiologic abnormality without neurologic deficit


Football players with suspected C-spine injury should not have their helmet removed in the field (see page 934). The following terms probably originated as locker-room jargon for various cervical spine-related injuries usually sustained in playing football. Medical definitions have subsequently been retro-fitted to them. As a result, the precise definitions may not be uniformly agreed upon. Although the semantics may differ, it is more important from a diagnostic and therapeutic standpoint to distinguish nerve root injuries, brachial plexus injuries, and spinal cord injuries.

1. cervical cord neuropraxia 212 (CCN): sensory changes which may involve numbness, tingling or burning. May or may not be associated with motor symptoms of weakness or complete paralysis. Typically lasts < 15 mins (although may persist up to 48 hrs), involves all 4 extremities in 80%. Narrowing of the sagittal diameter of the cervical spinal canal is felt to be contributory. With resumption of contact activities, recurrence rate is ≈ 56%, with higher risks of recurrence among those with narrower canals. Evaluation should include cervical MRI. Torg212 feels that uncomplicated cases of CCN (no spinal instability and no MRI evidence of cord defect or edema) have a low risk of permanent injury and does not recommend activity restrictions

2. “stinger” or “burner”: distinct from the burning hands syndrome. Unilateral. Burning dysesthetic pain radiating down one arm from the shoulder, sometimes associated with weakness involving the C5 or C6 nerve roots. Usually follows a tackle. May result from downward traction on the upper trunk of the brachial plexus or by direct nerve root compression in the neural foramina (not a SCI)

3. burning hands syndrome 213: similar to a stinger, but bilateral. Probably represents a SCI (possibly a mild variant of a central cord syndrome, see page 948)

4. other neurologic injuries include: vascular injury to carotid or vertebral arteries. Usually related to intimal dissection (see page 1160) following a direct blow to the neck or by extreme movements. Symptoms are those of a TIA or stroke

Spear tackler’s spine

Rule changes in 1976 banned spearing (the practice of using the football helmet as a battering ram to tackle an opponent) and resulted in a reduction of the number of foot-ball-related occurrences of cervical spine fractures and quadriplegia214.

Four characteristics of spear tackler’s spine:

1. cervical spinal stenosis

2. loss of normal cervical lordosis

3. evidence of pre-existing traumatic abnormalities

4. documented spear-tackler’s technique

Suggested management:

The athlete is removed from competition until the cervical lordosis returns and the player learns to use other tackling techniques.


Return to play (RTP) and pre-participation evaluation guidelines related to the cervical spine are shown in Table 28-35 (modified215). These are just guidelines, and do not insure safety. Clinical judgement must always be employed.

Table 28-35 C-spine-related contraindications for participation in contact sports*




1. odontoid abnormalities (serious injury may result from atlanto-axial instability)

A. complete aplasia (rare)


B. hypoplasia (seen in conjunction with achondroplasia and spondyloepiphyseal dysplasia)


C. os odontoideum (probably of traumatic origin)


2. atlanto-occipital fusion (partial or complete fusion of atlas to occiput): sudden onset of symptoms & sudden death have been reported


3. Klippel-Feil anomaly (congenital fusion of 2 or more cervical vertebrae)Δ

A. Type I: mass fusion of C-spine to upper T-spine


B. Type II: fusion of only 1 or 2 interspaces

1. associated with limited ROM, occipitocervical anomalies, instability, disc disease or degenerative changes


2. associated with full ROM and none of the above



1. cervical spinal stenosis

A. asymptomatic


B. with one episode of cord neuropraxia


C. cord neuropraxia + MRI evidence of cord defect or edema


D. cord neuropraxia + ligamentous instability, symptoms or neurologic findings > 36 hrs, or multiple episodes


2. spear tackler’s spine (see text)


3. spina bifida occulta: rare, incidental x-ray finding


Post-traumatic upper cervical spine

1. atlantoaxial instability (ADI > 3 mm adults, > 4 mm peds)


2. atlantoaxial rotatory fixation (may be associated with disruption of transverse ligament)


3. fractures

A. healed, pain-free, full ROM, & no neurologic findings with any of the following fractures: nondisplaced Jefferson fracture; odontoid fracture; or lateral mass fracture of axis


B. all others


4. post-surgical atlantoaxial fusion


Post-traumatic subaxial cervical spine

1. ligamentous injuries: > 3.5 mm subluxation, or > 11° angulation on flexion-extension views


2. fractures

A. healed, stable fractures listed here with normal exam: VB compression fracture without posterior involvement; spinous process fractures


B. VB fractures with sagittal component or posterior bony or ligamentous involvement


C. comminuted fracture with displacement into spinal canal


D. lateral mass fracture producing facet incongruity


3. intervertebral disc injury

A. healed herniated disc treated conservatively


B. S/P ACDF with solid fusion, no symptoms, normal exam and full pain-free ROM


C. chronic herniated disc with pain, neuro findings or ↓ ROM, or acute herniated disc


4. S/P fusion

A. stable one-level fusion


B. stable two-level fusion


C. fusion > 2 levels


* organized contact sports includes215: boxing, football, ice hockey, lacrosse, rugby & wrestling

 also see page 851 for cranial-related (and craniocervical) conditions (e.g. Chiari I malformation…)

 C.I. = contraindications, classified as absolute, relative (i.e. uncertain) or none

§ congenital abnormalities may have particular relevance to Special Olympics

Δ NB: Klippel-Feil may be associated with abnormalities in other organ systems (e.g. cardiac) which may impact on participation in contact sports (see page 253)

 Pavlov ratio (see page 489) has a low positive predictive value for injuries in contact sports and is therefore not a useful screening test (i.e. an asymptomatic Pavlov ratio < 0.8 is not a contraindication to participation)

28.6.8. Delayed cervical instability

Definition (adapted216): cervical instability that is not recognized until beyond 20 days after the injury. The instability itself may be delayed, or the recognition may be delayed. Reasons for delayed cervical instability:

1. inadequate radiologic evaluation47

A. incomplete studies (e.g. must see all the way to C7-T1 junction)

B. suboptimal studies: motion artifact, incorrect positioning… Etiologies include: poor patient cooperation as a result of agitation/intoxication, portable films, poor technique…

2. abnormality missed on x-ray

A. overlooked fracture, subluxation

B. injury failed to be demonstrated despite sufficiently adequate x-rays216A


A. see page 939 for recommendations of extent of radiologic workup


1. type of fracture not demonstrated on the radiographs obtained

2. patient positioning (e.g. supine) may reduce some malalignment

3. spasm of cervical muscles may reduce and/or stabilize the injury

4. microfractures

3. inadequate models: some findings may be judged to be stable using certain models, but in the long-run may be unstable (there is no perfect model for instability)

Further studies or repeat x-rays several weeks post-trauma should be considered with neurologic deficit, persistent pain, significant degenerative changes when the original films were suboptimal, subluxations < 3 mm, or when surgery is contemplated217.

28.7. Blunt cerebrovascular injuries

Cerebrovascular injuries may be classified as follows:

1. penetrating injury: see page 998

2. nonpenetrating injury: discussed in this chapter (see below)

A. blunt trauma: dissection is the usual type of injury

B. stretch: e.g. dissection from spinal manipulation or neck hyperextension

C. occlusion by bone: more common with vertebral arteries

1. kinking: e.g. with facet dislocation

2. compression by bone fragments: e.g. by fractures through foramen transversarium

3. iatrogenic injuries may be related to angiography catheters

The material that follows deals with nonpenetrating cerebrovascular injuries (mostly blunt trauma).

Incidence of blunt cerebrovascular injury (BCVI) in the literature: 1-2% of blunt trauma patients218 (among those who stayed > 24 hrs in a trauma hospital the incidence was 2.4%218). Optimal screening, diagnostic and treatment methods are controversial. Improving experience with CTA & endovascular techniques is resulting in a reworking of paradigms. A 13% mortality rate is considered low. Nearly one-third of cases are not treatable.

Risk factors

Risk factors for BCVI are shown in Table 28-36. However, BCVI can occur even in the absence of identifiable risk factors218.


Signs and symptoms of BCVI are shown in Table 28-37.

Table 28-36 Risk factors for BCVI219

• high energy transfer mechanism associated with:

• displaced mid face fracture (LeForte fracture type II or III - see page 890)

• basilar skull fracture involving carotid canal

• TBI consistent with DAI and GCS < 6

• cervical vertebral body or transverse foramen fracture, subluxation, or ligamentous injury at any level

• any fracture involving C1-3

• near hanging with anoxic brain injury

• clothesline-type injury or seat belt abrasion with significant cervical swelling, pain, or mental status changes

Table 28-37 Signs & symptoms of BCVI219

• arterial hemorrhage from neck/nose/mouth (? go to O.R.)

• cervical bruit in pt. < 50 yrs old

• expanding cervical hematoma

• focal neurologic deficit: TIA, Horner’s syndrome, hemiparesis, VBI (see page 1158)

• neurologic deficit inconsistent with head CT

• stroke on CT or MRI


The following is a summary of the Western Trauma Association219 guidelines. Their recommendations are based on observational studies and expert opinion (no Class I data was available).

Evaluation of patients with risk factors or signs/symptoms of BCVI

1. 16-slice multidetector CT angiography (MDCTA)A should be obtained as follows

A. emergently in patients with signs/symptoms of BCVI (see Table 28-37)

B. asymptomatic patients with risk factors (see Table 28-36) for BCVI:

1. if the presence of BCVI would alter therapy (e.g no contraindication to heparin) then MDCTA should be done within 12 hours if possible

2. if heparin is contraindicated due to associated injuries, timing of MDCTA is determined by patient stability

2. if the MDCTA is equivocal, or if it is negative but clinical suspicion remains high: a catheter arteriogram should be done (otherwise, if negative: stop)

3. grading: if the MDCTA or the arteriogram shows positive findings (listed on page 1162):

A. the injury is graded using the scale shown in Table 28-38225 (sometimes referred to as the “Denver grading scale”)

B. proceed with grade-based management (see below)

Table 28-38 BCVI grading scale225




luminal irregularity with < 25% stenosis


≥ 25% luminal stenosis or intraluminal thrombus or raised intimal flap






transection with free extravasation

Management of documented BCVI

1. grade specific therapy

• Grade I

1. heparin as outlined (see below)

2. stroke rate is so low with Grade I that surgery is not justified

• Grade II-IV:

1. administer heparin as outlined (see below)

2. if available, consultation with neuroendovascular interventionalist or vascular neurosurgeon to assess for intervention:

a. surgically accessible lesions in patients without complete stroke: pursue operative repair

b. inaccessible lesionsB: consider endovascular repair

• Grade V: highly lethal injury

1. accessible lesions should be considered for urgent surgical repair (anecdotal)

2. inaccessible lesions (the majority): incomplete transection may be amenable to endovascular stenting with concurrent antithrombotics; complete transections should be ligated (or occluded endovascularly)

2. repeat MDCTA or angiography 7-10 days post injury to assess healing226. Results:

A. lesion healed: discontinue anticoagulation

B. non-healed lesions:

1. consider endovascular stenting “with caution” for severe luminal narrowing or expanding pseudoaneurysm (controversial: results have been mixed - favorable227 and unfavorable228)

2. transition from heparin to aspirin (75-150 mg/d) alone

3. repeat MDCTA or angiography 3 months post injury (rationale: most heal with canalization in 6 wks). Results:

a. healed lesion: consider discontinuing aspirin

b. non-healed: optimal drug and duration is not known. Recommendation219: lifelong antiplatelet therapy with either aspirin or clopidogrel. Dual therapy is used for acute coronary syndromes and following angioplasty (± stenting) but is not recommended in patients who have had a stroke or TIA229


A. CTA on scanners with ≥ 16 detectors have an accuracy near 99%220 & equivalent predictive value to cerebral angiogram. MRA221222 and ultrasound223224 are not considered adequate for BCVI screening

B. most inaccessible lesions are at the skull base


Heparinization: When anticoagulation is not contraindicatedA, perform a baseline PTT and then begin heparin drip 15 U/kg/hr IV. Repeat PTT after 6 hours, and titrate to PTT = 40-50 seconds.


A. trauma contraindications to heparinization: patients that are actively bleeding, have potential for bleeding, or in whom the consequences of bleeding are severe. Specific examples include: liver and spleen injuries, major pelvic fractures, and intracranial hemorrhage


Dissection-related anticoagulation risks include: extension of the medial hemorrhage (with possible SAH), and intracerebral hemorrhage (conversion of pale infarct to hemorrhagic).


For general information related to cerebral arterial dissections and for spontaneous dissections, see page 1160. For evaluation and management, see above.

This section considers blunt (i.e. nonpenetrating) specifically related to ICA dissection. Neck hyperextension with lateral rotation is a common mechanism of injury, and is thought to stretch the ICA over the transverse processes of the upper cervical spine. In posttraumatic dissection, ischemic symptoms are the most common230. Post-traumatic dissections may follow minor injuries to already susceptible vessels, e.g. in a patient with fibromuscular dysplasia.


1. following MVAs: the most common etiology

2. attempted strangulation231

3. spinal manipulation therapy: VA dissections are more common than ICA

Most carotid dissections start ≈ 2 cm distal to the ICA origin.


The risk of stroke with various ICA dissection grades is shown in Table 28-39

Grade I injuries: 70% heal with or without heparin. 25% will persist. 4-12% will progress to more severe grade. Data suggests that anticoagulation reduces the risk of progression232.

Grade II: ≈ 70% progress to more severe grade even with heparin therapy.

Grade III & IV: most persist.

Initially, there may be no neurologic sequelae, however, progressive thrombosis, intramural hemorrhage or embolic phenomenon may develop in a delayed fashion. The distribution of time delays following trauma to time of presentation are shown in Table 28-40 (the majority are evident within the 1st 24 hours).

Table 28-39 Risk of stroke with ICA dissection



Stroke risk


stenosis < 25%



stenosis > 25%







uniformly lethal

* for grading, see Table 28-38

Table 28-40 Time to presentation after non-penetrating trauma



0-1 hours

6-10% of cases

1-24 hours


after 24 hours



See Management of documented BCVI on page 983.


Natural history is not well known. Many patients with minor symptoms may not present and presumably do well. In one series, 75% of patients returned to normal, 16% had a minor deficit, and 8% had a major deficit or died233.




Level III234: conventional angiography or MRA* after nonpenetrating cervical trauma in patients who have: complete SCI, fracture through the foramen transversarium, facet dislocation, and/or vertebral subluxation


Level III234

• recommended: anticoagulation with IV heparin for vertebral artery injury (VAI) with evidence of posterior circulation stroke

• recommended: either observation or treatment with anticoagulation for VAI with evidence of posterior ischemia

• recommended: observation for VAI with no evidence of either of above

* at the present time, MDCTA would probably be recommended over MRA for this indication (editor)

Blunt vertebral artery injury (BVI) is very rare, being found in 0.5-0.7% of patients with blunt injuries using aggressive screening235. It may produce vertebrobasilar insufficiency (VBI). Fractures through the foramen transversarium, facet fracture-dislocation, or vertebral subluxation are frequently identified in patients with BVI232236237 (overall incidence increases to 6% in the presence of cervical fracture or ligamentous injury235).


While motor vehicle accidents are the most common mechanism of injury, any trauma that can injure the C-spine can cause BVI (diving accidents, spinal manipulation…).

1. automobile accidents

2. spinal manipulation therapy (SMT): including chiropractic238 or similar, which comprise 11 of 15 case reports reviewed by Caplan, et al.239). VA dissections were independently associated with SMT within 30 days in multivariate analysis (odds ratio = 6.62, 95% CI 1.4 to 30)240

3. sudden head turning

4. direct blows to the back of the neck239

Stroke from BVI

Risk of stroke is shown in Table 28-41232. Unlike with carotid injuries, there is rarely a premonitory “warning” TIA. Time from injury to stroke: mean 4 days (range: 8 hours -12 days).


Indications: associated C-spine injuries are common and are the only independent predictor of BVI - no fracture pattern stands out as being more commonly associated. Grading: see Table 28-38.

When BVI is identified, it is critical to assess the status of the contralateral VA.

Table 28-41 Risk of stroke with VA injuries



Stroke risk


stenosis < 25%



stenosis > 25%








* for grading, see Table 28-38 (no grade V patients in this study)


Although there may be some differences from carotid dissections, the management outlined above (see page 983) is suggested. Strokes were more frequent in patients with BVI who were not treated initially with IV heparin despite an asymptomatic BVI232. However, based on historical controls, it is not clear if either screening or treatment improves overall outcome235.

Treatment options include endovascular stenting when amenable. This can restore near-normal flow, but long-term results are lacking241. Also, stenting requires ≥≈3 months of antiplatelet therapy which is contraindicated in some situations.


Overall mortality with BVI was 16% (7/44)235. Bilateral VA dissection appears highly fatal.

28.8. Thoracic & lumbar spine fractures


Thoracolumbar junction (TLJ): comprised of T11, T12 & L1: normally has a slight lordosis of ≤ 10°. 64% of spine fractures occur at the TLJ, usually T12-L1. 70% of these occur without immediate neurologic injury.

For measurements of normal kyphosis, see Sagittal balancepage 441.

28.8.1. Assessment and management decisions

A widely used model for thoracolumbar spine stability is the 3-column model of Denis (see below). A recently proposed TLICS system is presented on page 990.


Denis’ 3 column model of the spine (see Figure 28-9) attempts to identify CT criteria of instability of thoracolumbar spine fractures242. This model has generally good predictive value, however, any attempt to create “rules” of instability will have some inherent inaccuracy.

1. anterior column: anterior half of disc and vertebral body (VB) (includes anterior anulus fibrosus (AF)) plus the anterior longitudinal ligament (ALL)

2. middle column: posterior half of disc and vertebral body (includes posterior wall of vertebral body and posterior AF), posterior longitudinal ligament (PLL), & the pedicles

3. posterior column: posterior bony complex (posterior arch) with interposed posterior ligamentous complex (supraspinous and interspinous ligament, facet joints and capsule, and ligamentum flavum (LF)). Injury to this column alone does not cause instability


Figure 28-9 Three column model of the spine (TP = transverse process, see text for other abbreviations) (Adapted from Spine, Denis F, Vol. 8, pp. 317-31, 1983, with permission)



Involve only a part of a column and do not lead to acute instability (when not accompanied by major injures). Includes:

1. fracture of transverse process: usually neurologically intact except in two areas:

A. L4-5 → lumbosacral plexus injuries (there may be associated renal injuries, check U/A for blood)

B. T1-2 → brachial plexus injuries

2. fracture of articular process or pars interarticularis

3. isolated fractures of the spinous process: in the TL spine: these are usually due to direct trauma. Often difficult to detect on plain x-ray

4. isolated laminar fracture: rare. Should be stable


The McAfee classification describes 6 main types of fractures243. A simplified system with four categories follows (also see Table 28-42):

1. compression fracture: compression failure of anterior column. Middle column intact (unlike the 3 other major injuries below) acting as a fulcrum,

A. 2 subtypes:

1. anterior: most common between T6-T8 and T12-L3

a. lateral x-ray: wedging of the VB anteriorly, no loss of height of posterior VB, no subluxation

b. CT: spinal canal intact. Disruption of anterior end-plate

2. lateral (rare)

B. clinical: no neurologic deficit


2. burst fracture: pure axial load → compression of vertebral body → compression failure of anterior and middle columns. Occur mainly at TL junction, usually between T10 and L2

A. 5 subtypes (L5 burst fractures may constitute a rare subtype, see page 990)

1. fracture of both end-plates: seen in lower lumbar region (where axial load → increased extension, unlike T-spine where axial load → flexion)

2. fracture of superior end-plate: the most common burst fracture. Seen at TL junction. Mechanism = axial load + flexion

3. fracture of inferior end-plate: rare

4. burst rotation: usually midlumbar. Mechanism = axial load + rotation

5. burst lateral flexion: mechanism = axial load + lateral flexion

B. radiographic evaluation

1. lateral x-ray: cortical fracture of posterior VB wall, loss of posterior VB height, retropulsion of bone fragment from end plate(s) into canal

2. AP x-ray: increase of interpediculate distance (IPD), vertical fracture of lamina, splaying of facet joints: ↑ IPD indicates failure of middle column

3. CT: demonstrates break in posterior wall of VB with retropulsed bone in spinal canal (average: 50% obstruction of canal area), increase in IPD with splaying of posterior arch (including facets)

4. MRI or myelogram: anterior encroachment in spinal canal

C. clinical: depends on level (thoracic cord more sensitive and less room in canal than conus region), the impact at the time of disruption, and the extent of canal obstruction

• ≈ 50% intact at initial examination (half of these recalled leg numbness, tingling, and/or weakness initially after trauma that subsided)

• of patients with deficits, only 5% had complete paraplegia

3. seat-belt fractureA: flexion across a fulcrum anterior to the anterior column (e.g. seat belt) → compression of anterior column & distraction failure of both middle and posterior columns. May be bony or ligamentous

A. 4 subtypes

1. Chance fracture: one level, totally through bone

2. one level, through ligaments

3. two level, through bone in middle column, through ligament in anterior and posterior columns

4. two level, through ligament in all 3 columns

B. radiographic evaluation

1. plain x-ray: ↑ interspinous distance, pars interarticularis fractures, and horizontal split of pedicles and transverse process. No subluxation

2. CT: axial cuts are poor for this type (most of fracture is in plane of axial CT cuts). Sagittal and coronal reconstructions demonstrate this well. May demonstrate pars fracture

C. clinical: no neurologic deficit

4. fracture-dislocation: failure of all 3 columns due to compression, tension, rotation or shear → subluxation or dislocation

A. x-ray: occasionally, may be reduced when imaged. Look for other markers of significant trauma (multiple rib fractures, unilateral articular process fractures, spinous process fractures, horizontal laminar fractures)

B. 3 subtypes

1. flexion rotation: posterior and middle columns totally ruptured, anterior compressed → anterior wedging

a. lateral x-ray: subluxation or dislocation. Preserved posterior VB wall. Increased interspinous distance

b. CT: rotation and offset of VBs with ↓ canal diameter. Jumped facets

c. clinical: 25% neurologically intact. 50% of those with deficits were complete paraplegics

2. shear: all 3 columns disrupted (including ALL)

a. when trauma force directed posteriorly to anteriorly (more common) VB above shears forward fracturing the posterior arch (→ free floating lamina) and the superior facet of the inferior vertebra

b. clinical: all 7 cases were complete paraplegics

3. flexion distraction

a. radiographically resemble seat-belt type with addition of subluxation, or with compression of anterior column > 10-20%

b. clinical: neurologic deficit (incomplete in 3 cases, complete in 1)


A. some call this a flexion-distraction fracture, but that term is also used for a subtype of fracture-dislocation



In addition to the above, associated injuries include: vertebral end-plate avulsion, ligamentous injuries, and hip and pelvic fractures. Thoracolumbar fractures may be associated with hemodynamic instability as a result of hemothorax or aortic injury. Fractures of the transverse processes may be associated with abdominal trauma (e.g renal injuries at L4-5).


Minor injuries

Isolated thoracolumbar transverse process fractures (as demonstrated on spinal CT) do not require intervention or consultation of a spine service244245.

Major spine injuries

Denis categorized the instability as:

• 1st degree: mechanical instability

• 2nd degree: neurological instability

• 3rd degree: mechanical & neurological instability

Table 28-43 Treatment of stable anterior or middle column thoracolumbar spine injuries

• treat initially with analgesics and recumbency (bed-rest) for comfort x 1-3 weeks

• diminution of pain is a good indication to commence mobilization with or without external immobilization (corset or Boston brace or extension TLSO x ≈ 12 weeks) depending on the degree of kyphosis

• vertebroplasty (± kyphoplasty) may be an option (see page 994)

• serial x-rays to rule-out progressive deformity

Anterior column injury

Isolated anterior column injuries are usually stable and are treated as outlined in Table 28-43. The following exceptions may be unstable (1st degree) and often require surgery242246:


1. a single compression fracture with:

A. loss of > 50% of height with angulation (particularly if the anterior part of the wedge comes to a point)

B. excessive kyphotic angulation at one segmentA (various criteria are used, none are absolute. Values quoted: > 30°, > 40°)

2. 3 or more contiguous compression fractures

3. neurologic deficit (generally does not occur with pure compression fracture)

4. disrupted posterior column or more than minimal middle column failure

5. progressive kyphosis: risk of progressive kyphosis is increased when loss of height of anterior vertebral body is > 75%. Risk is higher for lumbar compression fractures than thoracic


A. usually indicates disruption of the posterior ligamentous complex


Middle column failure

Unstable (often requiring surgery) with the following exceptions which should be stable (stable injuries may be treated as outlined in Table 28-43):


1. above T8 if the ribs and sternum are intact (provides anterior stabilization)

2. below L4 if the posterior elements are intact

3. Chance fracture (anterior column compression, middle column distraction)

4. anterior column disruption with minimal middle column failure

Posterior column disruption

Not acutely unstable unless accompanied by failure of the middle column (posterior longitudinal ligament and posterior anulus fibrosus). However, chronic instability with kyphotic deformity may develop (especially in children).

Seat-belt type injuries without neurologic deficit

No immediate danger of neurologic injury. Treat most with external immobilization in extension (e.g. Jewett hyperextension brace or molded TLSO).


Unstable. Treatment options:

1. surgical decompression and stabilization: usually needed in cases with

A. compression with > 50% loss of height with angulation

B. or, kyphotic angulation > 40° (or > 25%)

C. or, neurologic deficit

D. or, desire to shorten length of time of bedrest

2. prolonged bedrest: an option if none of the above are present

When vertebral body resection (vertebral corpectomy) is performed, options to access: transthoracic or transabdominal approach (or combined), transpedicular (for thoracic spine), lateral (retroperitoneal/retropleural) approach. Fracture and compression usually occurs at the superior margin of vertebral body, thus start resection at the inferior disc interspace. Followed by strut graft (cage or bone: iliac crest or fibula or tibia). Posterior instrumentation is usually required (see Spinal instrumentationpage 991).

Burst fractures

Some burst fractures may eventually cause neurologic deficit (even if no deficit initially). Middle column fragments in canal endanger the neuro elements. Criteria have been proposed to differentiate mild burst fractures from severe ones as follows242247:


Surgery recommended for burst fracture with any of the following:

1. anterior vertebral body height ≤ 50% of the posterior height

2. residual canal diameter ≤ 50% of normal*

3. kyphotic angulation ≥ 20°

4. when the increased interpediculate distance usually present on the initial film widens further on AP x-ray when standing in brace/cast

5. neurologic deficit (incomplete)

6. progressive kyphosis

* retropulsed bone in the canal is often resorbed with either bracing or surgery and is therefore controversial as an isolated indication for surgery248249

For those not undergoing surgery (i.e. when surgery is not required or is contraindicated), an option is to treat with recumbency from 1-6 weeks247. Avoid early ambulation → further axial loading (even in cast). When appropriate, begin ambulation in an orthosis (e.g. molded thoracolumbar sacral orthosis (TLSO) or a Jewett brace) and follow patient for 3-5 months with serial x-rays to detect progressive collapse or angulation which may need further intervention. L5 burst fractures may be an exception to the usual management (see below).

L5 burst fractures: These fractures are extremely rare, and it is difficult for instrumentation to maintain alignment at this level250. Therefore, if neurologic deficit is absent or mild, conservative treatment should be considered250251. Regardless of treatment, patients will probably lose ≈ 15° of lordosis between L4 and the sacrum. Permanent neurologic loss may occur251.

Early reports of conservative management utilized ≈ 6-10 weeks of bed rest followed by mobilization in a brace. A more contemporary approach utilizes 10-14 days of bed rest. The patient should be fitted with a TLSO with a unilateral non-movable thigh cuff in 10° of flexion (on either side, to reduce motion at the fracture segment). Mobilization should be done very gradually as the pain allows. The brace should be worn ≈ 4-6 months, and serial x-rays should be performed to rule-out progressive deformity.

If surgical treatment is indicated, a posterior approach with fusion and fixation of L4-S1 may be performed utilizing pedicle screws.


The TLICS system has been proposed to simplify classification and discussion of thoracolumbar fractures252253. Points are assigned as shown in Table 28-44. The scores are summed, and management guidelines are given in Table 28-45.

Neurologic deficit, especially when partial, favors surgery.

As of this writing, the TLICS system has not been validated.

Table 28-44 Thoracolumbar injury classification & severity score (TLICS)




Radiographic findings

compression fx


burst component or lateral angulation > 15°


distraction injury


translational/rotational injury


Neurologic status



root injury


complete SCI


incomplete SCI


cauda equina syndrome


Integrity of posterior ligamantous complex





definite injury


TLICS = Total Points →

Table 28-45 Management based on TLICS



≤ 3

nonoperative candidates


“grey zone” may be considered for operative or nonoperative management

≥ 5

surgical candidates

28.8.2. Surgical treatment

Fragments within the canal: if the PLL is intact (may not be the case with middle column failure), distraction may be able to “pull” the fragments back into their normal position (ligamentotaxis) although this is not assured254. Ligamentotaxis has a better chance of succeeding if performed within 48 hours of injury. From a posterior approach with laminectomy: intraoperative ultrasound may demonstrate residual canal fragments255, and if needed the fragments may be impacted anteriorly out of the canal, e.g. using tamps such as Sypert spinal impactors.


The posterior approach is preferred when there is not a specific need to go from the front.


Anterior instrumentation of the lower lumbar spine is difficult, and is usually not recommended below ≈ L4.

Burst fractures

There is more than one solution to any given problem, there are a number of controversies, and the following is proposed as a recommendation.

Choice of approach: Surgical considerations: a posterior approach is preferred if there is a dural tear, whereas a burst fracture with partial deficit and canal compromise may be treated more effectively from an anterior approach243. A small progression in angular deformity may occur when posterior stabilization is performed alone (since the injury to the anterior column is not corrected), but by itself usually does not require intervention.

For a posterior approach: In ideal situation (good bone quality, pedicle screw placement goes well (i.e. no fracture, no breach), and non-smoking patient) then one can fuse/rod one above and one below the fracture (using pedicle screws; longer constructs are needed with laminar hooks). With a short segment fusion like this, approximately 10° of lordosis be lost with time, therefore, one should try to overcorrect a little to accommodate the anticipated settling. If the patient does not meet the above criteria (e.g. poor bone quality), an option is to “rod long, fuse short” (e.g. rod 2 levels above and below the fracture but fuse only 1 level above and below) and then to remove the hardware when the fusion is solid (e.g. at ≈ one year) - this avoids fusing a nonpathologic segment just to get a better anchor. Junctional deterioration to the point that further surgery is needed often occurs at 3 years when 4 segments are fused, whereas it occurs at 8-9 years when only 3 levels are fused. Fusing across critical levels (i.e. thoracolumbar junction with T11 or L1 compression fractures) requires that the fusion incorporate 2-3 levels on each side of the the junction (the forces of the long segment of the relatively immobile thoracic spine with the lumbar spine at the T-L junction increase the risk of nonunion).

Schanz screws work well for applying a lordotic force across the fracture. If there is bone in the canal, it is important not to overdistract to avoid neural injury. If lordosis is applied withthe rod locking screws loose, the connectors may move closer together which could increase bone within the canal (therefore, either tighten the rod screws or place blocks on the rods before applying lordotic force). If you have gone two above and two below the fracture, first, tighten all retaining nuts (rod and pedicle nuts) on the ends of both rods (8 nuts total). Then place pedicle screw nut tighteners over the pedicle screws in the levels immediately above and below the fracture. Cross the tighteners to reduce the angle of kyphosis, then tighten the nuts. Next distract For distraction, a rod-holder can be placed on the rod to distract against since the gap across the fracture segment (where no pin is placed) may be too long. The rod nuts are then tightened. After a final tightening of all nuts, the pedicle screws are cut off using the insitu cutter.

For thoracic fractures that are not severe and do not require decompression, an option is to place pedicle screws and rods (which can be done percutaneously) without placing any graft. The concept is that the ribs anteriorly and the screws/rods posteriorly provide adequate stabilization while the fractured VB heals. This is more commonly practiced in Europe than the U.S..

Wound infections

Postoperative wound infections with spinal instrumentation are usually due to Staph. aureus, and may respond to prolonged antibiotic administration without hardware removal243. Persistent infection may respond to debridement of devitalized tissue (e.g. onlay bone graft) and thorough washout without removal of instrumentation followed by antibiotics. If this inadequate, removal of instrumentation may occasionally be required.

28.9. Osteoporotic spine fractures

Osteoporosis is defined as a condition of skeletal fragility as a result of low bone mass, microarchitectural deterioration of bone, or both256. It is found most commonly in post-menopausal white females, and is rare prior to menopause. Lifetime risk of symptomatic vertebral body (VB) osteoporotic compression fractures is 16% for women, and 5% for men. There are ≈ 700,000 VB compression fractures per year in the U.S.

These patients are often found to have significant VB compression fractures on plain films after presenting with back pain following a seemingly minor fall. CT often shows an impressive amount of bone retropulsed into the canal.

Risk factors

Factors that increase the risk of osteoporosis include:

1. weight < 58 kg

2. cigarette smoking257

3. low-trauma VB fracture in the patient or a first degree relative

4. drugs

A. heavy alcohol consumption

B. AEDs (especially phenytoin)

C. warfarin

D. steroid use:

1. bone changes can be seen with 7.5 mg/d of prednisone for > 6 months

2. VB fractures occur in 30-50% of patients on prolonged glucocorticoids

5. postmenopausal female

6. males undergoing androgen deprivation therapy (e.g. for prostate Ca). Orchiectomy or ≥ 9 doses of gonadotropin-releasing hormone agonists had a 1.5 fold increase in risk of all fractures258

7. physical inactivity

8. low calcium intake

Factors that protect against osteoporosis include impact exercise and excess body fat.


To differentiate osteoporotic compression fractures from other pathologic fractures, see Pathologic fractures of the spinepage 1232.

Pre-fracture diagnosis

1. measuring bone fragility is not possible

2. the best correlate with bone fragility is radiographic measurement of bone mineral density (BMD) using DEXA scan (see below)

3. patients with low-trauma fractures or fragility fractures are considered osteoporotic even if their BMD are greater than these cutoffs

DEXA scan (dual energy x-ray absorptiometry): the preferred way to measure BMD

1. proximal femur: BMD in this location is the best predictor for future fractures

2. LS spine: best location to assess response to treatment (need AP and lateral views, since AP often overestimates BMD because of superimposition of overlying posterior elements and aortic calcifications)

3. forearm BMD may be used if hip or spine are unsuitable

Interpretation of DEXA scan results:

1. findings are reported as

A. T-score: norms for healthy young adults

B. Z-score: norms of subjects of same age and sex as the patient

2. diagnostic criteria: WHO definitionsA


A. with a normal distribution 1 SD below the mean is the lowest 25th percentile, 2 SD below is 2.5th %ile


A. normal: > –1 standard deviations (SD)

B. osteopenia: from –1 to –2.5 SD

C. osteoporosis: < than –2.5 SD259

Post-fracture considerations

1. other causes of pathologic fracture, especially neoplastic (e.g. multiple myeloma, metastatic breast cancer) should be ruled out

2. younger patients with osteoporosis require evaluation for a remediable cause of the osteoporosis (hyperthyroidism, steroid abuse, hyperparathyroidism, osteomalacia, Cushing’s syndrome)



High calcium intake during childhood may increase peak bone mass. Weight-bearing exercise in adulthood helps slow calcium loss from bones. Also effective: estrogen (see below), bisphosphonates (alendronate and risedronate), and raloxifene.


Drugs that increase bone formation include:

1. intermittent low-dose parathyroid hormone: still experimental

2. sodium fluoride: 75 mg/d increases bone mass but did not significantly reduce the fracture rate. 25 mg PO BID of a delayed-release formulation (Slow Fluoride®) reduced fracture rate but may make bone more fragile and could increase risk of hip fractures. Fluoride increases demand for Ca++, therefore supplement with 800 mg/d Ca++ and 400 IU/d vitamin D. Not recommended for use > 2 yrs

Drugs that reduce bone resorption are less effective on cancellous bone (found mainly in the spine and at the end of long bones261). Medications include:

1. estrogen: cannot be used in men. Studies of estrogen hormone replacement therapy (HRT) have shown increased vertebral bone mass by > 5% and decreased rate of vertebral fractures by 50%. Also relieves post-menopausal symptoms and reduces risk of CAD. However, because HRT increases the risk of breast cancer264 and of breast cancer recurrence265 as well as DVT, its use has diminished substantially

2. calcium: current recommendations are for 1,000-1,500 mg/d for postmenopausal women266 taken with meals

3. vitamin D or analogues: promote calcium absorption from the GI tract. Typically administered with calcium therapy. Vitamin D 400-800 IU/d is usually sufficient. If urinary Ca++ remains low, high dose vitamin D (50,000 IU q 7-10 d) may be tried. Since high-dose formulations have been discontinued in the U.S., analogues such as calcifediol (Calderol®) 50 μg/d or calcitriol (Rocaltrol®) up to 0.25 μg/d may be tried with Ca++ supplement. Serum levels of 25-hydroxyvitamin D [25(OH)D], AKA calcidiol is the best indicator of vitamin D status. The significance of vitamin D levels are shown in Table 28-46266. With high dose vitamin D or analogues, monitor serum and urinary Ca++

Table 28-46 Serum 25-hydroxyvitamin D levels




< 10-11

< 25-27.5

vit D deficiency → rickets (in peds) and osteomalacia (adults)


< 25-37.5

inadequate for bone and overall health

≥ 15

≥ 37.5

adequate for bone and overall health

consistently > 200

consistently > 500

potentially toxic → hypercalcemia & hyperphosphatemia

4. calcitonin: derived from a number of sources, salmon is one of the more common. Benefit in preventing fractures is less well established263

A. parenteral salmon calcitonin (Calcimar®, Miacalcin®): indicated for patients for whom estrogen is contraindicated. Expensive ($1,500-3,000/yr) and must be given IM or sub-Q. 30-60% of patients develop antibodies to the drug which negates its effect. Rx: 0.5 ml (100 U) of calcitonin (given with calcium supplements to prevent hyperparathyroidism) SQ q d

B. intranasal forms (Miacalcin nasal spray): less potent. 200-400 IU/d given in one nostril (alternate nostrils daily) plus Ca++ 500 mg/d and vitamin D

5. bisphosphonates: carbon-substituted analogues of pyrophosphate have a high affinity for bone and inhibit bone resorption by destroying osteoclasts. Not metabolized. Remain bound to bone for several weeks

A. etidronate (Didronel®), a 1st generation drug. Not FDA approved for os-1 ng/ml = 2.5 nmol/L teoporosis. May reduce rate of VB fractures, not confirmed on F/U. Possible increased risk of hip fractures due to inhibition of bone mineralization may not occur with 2nd & 3rd generation drugs listed below. Rx 400 mg PO daily x 2 wks followed by 11-13 weeks of Ca++ supplementation

B. alendronate (Fosamax®): can cause esophageal ulcers. Rx Prevention: 5 mg PO daily; treatment 10 mg PO daily; taken upright with water on an empty stomach at least 30 minutes before eating or drinking anything else. Once weekly dosing of 35 mg for prevention and 70 mg for treatment263267. Taken concurrently with 1000-1500 mg/d Ca++ and 400/d IU of vitamin D

C. risedronate (Actonel®): Rx Prevention or treatment: 5 mg PO daily, or 35 mg once/week267 on an empty stomach (as for alendronate, see above)

D. other drugs not FDA approved for osteoporosis: tiludronate (Skelid®), pamidronate (Aredia®) (some are used for Paget’s disease, see page 501)

6. estrogen analogues:

A. tamoxifen (Nolvadex®), an estrogen antagonist for breast tissue but an estrogen agonist for bone, has a partial agonist effect on uterus associated with an increased incidence of endometrial cancer

B. raloxifene (Evista®): similar to tamoxifen but is an estrogen antagonist for uterus268. Decreases the effect of warfarin (Coumadin®). Rx: 60 mg PO q d. SUPPLIED: 60 mg tablets

7. RANK ligand (RANKL) inhibitors: RANKL binds to RANK receptors and stimulates precursor cells to mature into osteoclasts and inhibits their apoptosis269. Agents undergoing investigation include denosumab (Prolia®) 60 mg SQ q 6 months appears more effective than alendronate270


Patients rarely have neurologic deficit. They are also usually fragile elderly women who usually do not tolerate large surgical procedures well, and the rest of their bones are also osteoporotic which are poor for internal fixation.

Management consists primarily of analgesics and bed rest followed by progressive mobilization, often in an external brace (often not tolerated well). Surgery is rarely employed. In cases where pain control is difficult to obtain or where neural compression causes deficit, limited bony decompression may be considered. Percutaneous vertebroplasty (see below) is a newer option.

Typical time course of conservative treatment:

1. initially, severe pain may require hospital or subacute care facility admission for adequate pain control utilizing

A. sufficient pain medication

B. bed rest for about 7-10 days (DVT prophylaxis recommended)

2. begin physical therapy (PT) after ≈ 7-10 days as patient tolerates (prolonged bed rest can promote “disuse osteoporosis”)

A. pain control as patient is mobilized may be enhanced by a lumbar brace which may work by reducing movement which causes repetitive “microfractures”

B. discharge from the hospital with lumbar brace for outpatient PT

3. pain subsides on the average after 4-6 weeks (range 2-12 weeks)


Percutaneous vertebroplasty (PVP): Transpedicular injection of polymethylmethacrylateA (PMMA) “(AKA methylmethacrylate) cement” into the compressed bone with the following goals:


A. PMMA injection is FDA approved for treatment of compression fractures due to osteoporosis or tumor, but not for trauma (PMMA would prevent healing of the fracture)


1. to shorten the duration of pain (sometimes providing pain relief within minutes to hours). Remember: the natural history is that pain will eventually diminish in essentially all of these patients. Mechanism of pain relief may be due to stabilization of bone and/or due to disruption of nerve pain transmission by heat released during the exothermic curing of the cement

2. to try and stabilize the bone: may prevent progression of kyphosis

Recent randomized studies found no benefit in vertebroplasty over a sham procedure at 1 month271 or at any time up to 6 months post-procedure272. NB: kyphoplasty (see below) was not studied; use with metastatic spine tumors was also not evaluated. Patient selection issues may make these results more or less generalizable to a specific patient.

Kyphoplasty: Similar to PVP, except first, a balloon is inserted into the compressed VB through the pedicle. The balloon is inflated and then deflated and removed. PMMA is injected into the thusly created defect. Potential benefits of this over vertebroplasty: there may be some restoration of height, and there may be less tendency for PMMA extravasation/embolization (due to the cavity creation and the thicker PMMA used). In the (industry sponsored) randomized non-blinded FREE study300 there was a significant positive difference in pain reduction and quality of life improvement in the kyphoplasty group compared to the nonoperated group at 1 month that diminished by 1-year post-op.


1. painful osteoporotic compression fractures:

A. usually do not treat fractures producing < 5-10% loss of height

B. severe pain that interferes with patient activity

C. failure to adequately control pain with oral pain medication

D. image pain localized to fracture level

E. acute fractures: procedure is not effective for healed fractures. In questionable cases, look for changes on STIR MRI (see below)

2. levels: FDA approved for use from T5 through L5, however has been used off-label (primarily for tumor, e.g. multiple myeloma) from T1 through sacrum, and has been described (for tumor) in the cervical spine from an anterior approach

3. vertebral hemangiomas that cause vertebral collapse or neurologic deficit as a result of extension into the spinal canal (not for incidental hemangiomas): see page 738

4. osteolytic metastases and multiple myeloma273: pain relief and stabilization

5. pathologic compression fractures274 from metastases: PVP does not give as rapid pain relief as with osteoporotic compression fractures (it may actually be necessary to increase pain meds for 7-10 days post PVP)

6. pedicle screw salvage when pedicle fractures or screws strip during pedicle screw placement


1. coagulopathy

2. completely healed fractures (no edema on MRI or cold on bone scan)

3. active infections: sepsis, osteomyelitis, discitis and epidural abscess

4. spinal instability

5. focal neurologic exam: may indicate herniated disc, retropulsed fragment in canal. Get CT or MRI to rule these out

6. relative contraindications:

A. fractures > 80% loss of VB height (technically challenging)

B. acute burst fractures

C. significant canal compromise from tumor or retropulsed bone

D. partial or total destruction of the posterior VB wall: not an absolute contraindication

7. iodine allergy: there is a small risk of a balloon rupturing with spill of the iodinated contrast used to fill the balloons prior to injecting the PMMA. Options include: iodine allergy prep (see page 124), use of gadolinium instead of iodinated contrast


Complication = rate: 1-9%. Lowest when used to treat osteoporotic compression fractures, higher with vertebral hemangiomas, highest with pathological fractures

1. methacrylate leakage:

A. into soft tissues: usually of little consequence

B. into spinal canal: symptomatic spinal cord compression is very rare

C. into neural foramen: may cause radiculopathy

D. into disc space

E. venous: can get into spinal venous plexus or vena cava with ≈ 0.3-1% risk of clinically significant methacrylate pulmonary embolism (PE)275

2. radiculopathy: 5-7% incidence. Some cases may be due to heat released during cement curing. Often treated conservatively: steroids, pain meds, nerve block…

3. pedicle fracture

4. rib fracture

5. transverse process fracture

6. anterior penetration with needle: puncture of great vessels, pneumothorax…

7. increased incidence of future VB compression fractures at adjacent levels

Management of some associated developments

1. chest pain

A. get rib x-rays

B. VQ scan if indicated

2. patient starts coughing during injection: fairly common. May be reaction to rib pain or to odor of PMMA, may also indicate solvent in lungs. Stop injecting

3. back pain: take x-ray to rule-out new fracture or PMMA in veins

4. neurologic symptoms: get CT scan

Pre-procedure evaluation

1. plain x-rays: minimum requirement, most practitioners get MRI or bone scan

2. CT: helps rule-out bony compromise of spinal canal which may indicate increased risk of leakage for PMMA into canal during procedure

3. MRI: not mandatory, may be helpful in some cases

A. short tau inversion recovery (STIR) images demonstrate bone edema indicative of acute fractures (not as good for differentiating pathology)276

B. MRI can also disclose neurologic compression by soft tissue (e.g. tumor)

4. patients with multiple compression fractures: consider getting bone scan and perform PVP in the VB near the level of pain that lights up the most (↑ activity on bone scan correlates strongly with good outcome from PVP)

Booking the case - kyphoplasty

Also see default values (page v).image

1. position: prone

2. anesthesia: may be done under general, or under MAC

3. equipment: 2 C-arms for biplane fluoro

4. implants:

A. kyphoplasty set

B. iodinated contrast from radiology to fill balloons

5. consent (in lay terms for the patient - not all-inclusive):

A. procedure: insertion of a needle into the fractured/abnormal bone, some-times getting a biopsy as well, and then inflating a balloon in the bone to try and bring it back to a more normal size and then to inject a liquid cement which will then harden inside the bone to strengthen it

B. alternatives: nonsurgical management, open surgery, in cases of tumor sometimes radiation therapy can be done

C. complications: leakage of cement which can compress nerves and may need to be removed surgically if possible, rib fracture (from positioning), injury to large blood vessel or lung by the needle, failure to achieve the desired pain relief


1. pain medication

A. remember, this procedure is done with the patient lying on their stomach and is usually performed on frail, elderly females who smoke. Therefore use caution to avoid oversedation and respiratory compromise

B. sedation and pain medication

C. use of local anesthetic during needle placement

D. additional pain medication just prior to injection

2. use biplane fluoro to pass needle through the pedicle to enter VB (see Percutaneous pedicle screwspage 192) and place tip ≈ 1/2 to 2/3 of the way through the VB

3. test inject with contrast (e.g. iohexol (Omnipaque 300) see page 122) (do digital subtraction study if equipment is available). For kyphoplasty, the balloon is inflated at this time

A. a little venous enhancement is acceptable

B. if you visualize vena cava

1. do not pull needle back (the fistula has already been created)

2. push needle in a little further, or

3. push some gelfoam (soaked in contrast) through the needle, or

4. inject a very small amount of PMMA under visualization and allow it to set to block the fistula

4. inject PMMA (that has been opacified with tantalum or barium-sulfate) under fluoroscopic visualization until:

A. 3-5 cc injected (minimal compression fractures accept more cement, some-times up to ≈ 8 cc). No correlation between amount of PMMA injected and pain relief273

B. PMMA approaches posterior VB wall. Stop if cement ever: enters disc space, vena cava, pedicle, or spinal canal


1. PVP is often an outpatient procedure, but sometimes overnight admission is used

2. watch for

A. chest or back pain (may indicate rib fracture)

B. fever: may be reaction to cement

C. neurologic symptoms

3. activity

A. gradual mobilization after ≈ 2 hours

B. ± physical therapy

C. ± short term use of external brace (most centers do not use)

4. institute medical treatment for osteoporosis: remember the patient with fragility fractures by definition has osteoporosis with risk of future fractures

28.10. Sacral fractures

Uncommon. Usually caused by shear forces. Identified in 17% of patients with pelvic fractures277 (image keep in mind that neurologic deficits in patients with pelvic fractures may be due to associated sacral fractures).

The sacrum below S2 is not essential to ambulation or support of the spinal column, but may still be unstable since pressure to the area may occur when supine or sitting.

Neurologic injuries occur in 22-60%277. Three characteristic clinical presentations based on zone of involvement277278 as shown in Table 28-47.



In one series279, all 35 fractures were treated without surgery, and only 1 patient with a complete cauda equina syndrome did not improve. Others feel that surgery may have a useful role277:

1. operative reduction and internal fixation of unstable fractures may aid in pain control and promote early ambulation

2. decompression and/or surgical reduction/fixation may possibly improve radicular or sphincter deficits

Some observations277:

1. reduction of the ala may promote L5 recovery with Zone I fractures

2. Zone II fractures with neurologic involvement may recover with or without surgical reduction and fixation

3. horizontal Zone III with severe deficit: controversial. Reduction & decompression does not ensure recovery, which may occur with nonoperative management

28.11. Gunshot wounds to the spine

Most are due to assaults with handguns. Distribution: cervical 19-37%, thoracic 48-64%, and lumbosacral 10-29% (roughly proportional to lengths of each segment). Spinal cord injury due to civilian GSWs are primarily due to direct injury from the bullet (unlike military weapons which may create injury from shock waves and cavitation). Steroids are not indicated (see page 937).

Indications for surgery:

1. injury to the cauda equina (whether complete or incomplete) if nerve root compression is demonstrated280

2. neurologic deterioration: suggesting possibility of spinal epidural hematoma

3. compression of a nerve root

4. CSF leak

5. spinal instability: very rare with isolated GSW to the spine

6. to remove a copper jacketed bullet: copper can cause intense local reaction281

7. incomplete lesions: very controversial. Some series show improvement with surgery282, others show no difference from unoperated patients

8. debridement to reduce the risk of infection: more important for military GSW where there is massive tissue injury, not an issue for most civilian GSW except in cases where the bullet has traversed GI or respiratory tract

9. vascular injuries

10. surgery for late complications:

A. migrating bullet

B. lead toxicity283 (plumbism): absorption of lead from a bullet occurs only when it lodges in joints, bursae, or disc space. Findings include: anemia, encephalopathy, motor neuropathy, nephropathy, abdominal colic

C. late spinal instability: especially after surgery

28.12. Penetrating trauma to the neck

Most often, injuries to the soft-tissues of the neck fall into the purvey of general/trauma surgeons and/or vascular surgeons. However, depending on local practice patterns, neurosurgeons may participate in care of these injuries, or they may get involved by virtue of associated spinal injuries. Also, see Gunshot wounds to the spinepage 998.

Trauma surgeons have traditionally divided penetrating injuries of the neck into 3 zones284, and although definitions vary, the following is a general scheme285:

Zone I: inferiorly from the head of the clavicle to include the thoracic outlet

Zone II: from the clavicle to the angle of the mandible

Zone III: from the angle of the mandible to the base of the skull

The mortality rate for penetrating injury to the neck is ≈ 15%, with most early deaths due either to asphyxiation from airway compromise, or exsanguination externally or into the chest or upper airways. Late death is usually due to cerebral ischemia or complications from spinal cord injury.

Vascular injuries: Venous injuries occur in ≈ 18% of penetrating neck wounds, and arterial injuries in ≈ 12%. Of the cervical arteries, the common carotid is most usually involved, followed by the ICA, the ECA, and then the vertebral artery. Outcome probably correlates most closely with neurologic condition on admission, regardless of treatment.

Vertebral artery (VA): the majority of injuries are penetrating. Due to the proximity of other vessels, the spinal cord and nerve roots, injuries are rarely isolated to the VA. 72% of documented VA injuries had no related physical findings on exam286.


Neurologic examination: global deficits may be due to shock or hypoxemia due to asphyxiation. Cerebral neurologic deficits are usually due to vascular injury with cerebral ischemia. Local findings may be related to cranial nerve injury. Unilateral UE deficits may be due to nerve root or brachial plexus involvement. Median or ulnar nerve dysfunction can occur from compression by a pseudoaneurysm of the proximal axillary artery. Spinal cord involvement may present with complete injury, or with an incomplete spinal cord injury syndrome (see page 948). Shock due to spinal cord injury is usually accompanied by bradycardia (see page 930), as opposed to the tachycardia seen with hypovolemic shock.

Cervical spine x-rays: assesses trajectory of injury and integrity C-spine.

Angiography: indicated in most cases if the patient is stable (especially for zone I or III injuries, and for zone II patients with no other indication for exploration, or for patients with penetration of the posterior triangle or wounds near the transverse processes where the VA may be injured). Patients actively hemorrhaging need to be taken to the OR without pre-op angiography. Angiographic abnormalities include:

1. extravasation of blood

A. expanding hematoma into soft tissues: may compromise airway

B. pseudoaneurysm

C. AV fistula

D. bleeding into airways

E. external bleeding

2. intimal dissection, with

A. occlusion, or

B. luminal narrowing (including possible “string sign”)

3. occlusion by soft tissue or bone


Airway: stable patients without airway compromise should not have “prophylactic” intubation to protect the airway. Immediate intubation is indicated for hemodynamically unstable patients or for airway compromise. Options:

1. endotracheal: preferred

2. cricothyroidotomy: if endotracheal intubation cannot be performed (e.g. due to tracheal deviation or patient agitation) or if there is evidence of cervical spine injury and manipulation of the neck is contraindicated, then cricothyroidotomy is performed with placement of a #6 or 7 cuffed endotracheal tube (followed by a standard tracheostomy in the OR once the patient is stabilized)

3. awake nasotracheal: may be considered in the setting of possible spinal injury

Exploration: surgical exploration has been advocated for all wounds that pierce the platysma and enter the anterior triangles of the neck287, however, 40-60% of these explorations will be negative. Although a selective approach may be based on angiography, false negatives have resulted in some authors recommending exploration of all zone II injuries288.

Carotid artery: choices are primary repair, interposition grafting, or ligation. Patients in coma or those with severe strokes caused by vascular occlusion of the carotid artery are poor surgical candidates for vascular reconstruction due to a high mortality rate ≥ 40%285, however the outcome with ligation is worse. Repair of injuries is recommended in patients with no or only minor neurologic deficit. ICA ligation is recommended for bleeding that cannot be controlled and was used for extravasation of dye at the base of the skull in 1 patient289.

Vertebral artery: injuries are more often managed by ligation than by direct repair290, especially when bleeding occurs during exploration. Less urgent conditions (e.g. AV fistula) requires knowledge of the patency of the contralateral VA and the ability to fill the ipsilateral PICA from retrograde flow through the BA before ligation can considered (arteriographic anomalies contraindicate ligation in 15% of cases). Proximal occlusion may be accomplished with an anterior approach after the sternocleidomastoid is detached from the sternum. The VA is the normally the first branch of the subclavian artery. Alternatively, endovascular techniques may be used, e.g. detachable balloons for proximal occlusion, or thrombogenic coils for pseudoaneurysms. Distal interruption may also be required, and this necessitates surgical exposure and ligation. Optimal management of a thrombosed injured VA in a foramen transversarium is unknown, and may require arterial bypass if ligation is not a viable option.

28.13. Delayed deterioration following spinal cord injuries

Etiologies include:

1. posttraumatic syringomyelia: see page 513. Latency to symptoms: 3 mos-34 yrs

2. subacute progressive ascending myelopathy (SPAM): rare. Median time of occurrence: 13 days post injury (range: 4-86 days)291. Signal changes extending to ≥ 4 levels above the original injury

3. unrecognized spinal instability292: mean delay in diagnosis was 20 days

4. tethered spinal cord: may be due to scar tissue at site of injury

5. delayed spinal epidural hematoma (SEH): most symptomatic SEH occur within 72 hours of surgery, however longer delays have been reported293

6. apoptosis of neurons, oligodendrogliocytes, and astrocytes294: initiated during the acute phase, deterioration occurs during the chronic phase of SCI (months to years after SCI)

7. glial scar formation: mass effect as well as release of factors that may damage surviving neurons295 (p 43-5)

28.14. Chronic management issues with spinal cord injuries

Most of the following topics are treated elsewhere in this manual, but are pertinent to spinal cord injured (SCI) patients, and reference to the specific section is made.

1. autonomic hyperreflexia: see below

2. ectopic bone, includes para-articular heterotopic ossification: ossification of some joints that occurs in 15-20% of paralyzed patients

3. osteoporosis and pathologic fracture: see page 992

4. spasticity: see page 536

5. syringomyelia: see page 510

6. deep vein thrombosis: see below and also page 42

7. shoulder-hand syndrome: possibly sympathetically maintained


In attempting to wean high level SCI patients from a ventilator, it may be helpful to change tube feedings to Pulmonaid® which lowers the CO2 load.

Patients with cervical SCIs are more prone to pneumonia due to the fact that most of the effort in a normal cough originates in the abdominal muscles which are paralyzed.


image Key concepts:

• exaggerated autonomic response to normally innocuous stimuli

• in spinal cord injury, occurs only in patients with lesions above ≈ T6

• patients complain of pounding headache, flushing and diaphoresis above lesion

• can be life threatening, requires rapid control of hypertension and a search for an elimination of offending stimuli

AKA autonomic dysreflexia. Autonomic hyperreflexia296297 (AH) is an exaggerated autonomic response (sympathetic usually dominates) secondary to stimuli that would only be mildly noxious under normal circumstances. It occurs in ≈ 30% of quadriplegic and high paraplegic patients (reported range is as high as 66-85%), but does not occur in patients with lesions below T6 (only patients with lesions above the origin of the splanchnic outflow are prone to develop AH, and the origin is usually T6 or below). It is rare in first 12-16 weeks post-injury.

During attacks, norepinephrine (NE) (but not epinephrine) is released. Hypersensitivity to NE may be partially due to subnormal resting levels of catecholamines. Homeostatic responses include vasodilatation (above the level of the injury) and bradycardia (however, sympathetic stimulation may also cause tachycardia).

Stimulus sources causing episodes of autonomic hyperreflexia:

1. bladder: 76% (distension 73%, UTI 3%, bladder stones…)

2. colorectal: 19% (fecal impaction 12%, administering enema or suppository 4%)

3. decubitus ulcers/skin infection: 4%

4. DVT

5. miscellaneous: tight clothing or leg bag straps, procedures such as cystoscopy or debriding decubitus ulcers, case report of suprapubic tube


• paroxysmal HTN: 90%

• anxiety

• diaphoresis

• piloerection

• pounding H/A

• ocular findings:

image mydriasis

image blurring of vision

image lid retraction or lid lag

• erythema of face, neck and trunk: 25%

• pallor of skin below the lesion (due to vasoconstriction)

• pulse rate: tachycardia (38%) or mild elevation over baseline, bradycardia (10%)

• “splotches” over face and neck: 3%

• muscle fasciculations

• increased spasticity

• penile erection

• Horner’s syndrome

• triad seen in 85%: cephalgia (H/A), hyperhidrosis, cutaneous vasodilatation


In the appropriate setting (e.g. a quadriplegic patient with an acutely distended bladder), the symptoms are fairly diagnostic.

Many features are also common to pheochromocytoma. Studies of catecholamine levels have been inconsistent, however they can be mildly elevated in AH. The distinguishing feature of AH is the presence of hyperhidrosis and flushing of the face in the presence of pallor and vasoconstriction elsewhere on the body (which would be unusual for a pheochromocytoma).


1. immediately elevate HOB (to decrease ICP), check BP q 5 min

2. treatment of choice: identify and eliminate the offending stimulus

A. make sure bladder is empty (if catheterized check for kinks or sediment plugs). Caution: irrigating bladder may exacerbate AH (consider suprapubic aspiration)

B. check bowels (avoid rectal exam, may exacerbate). Palpate abdomen or check abdominal x-ray (AH from this usually resolves spontaneously without manual disimpaction)

C. check skin and toenails for ulceration or infection

D. remove tight apparel

3. HTN that is extreme or that does not respond quickly may require treatment to prevent seizures and/or cerebral hemorrhage/hypertensive encephalopathy. Caution must be used to prevent hypotension following the episode. Agents used include: sublingual nifedipine298 10 mg SL, IV phentolamine (alpha cholinergic blocker, see page 681) or nitroprusside (Nipride®) (see page 19)

4. consider diazepam (Valium®) 2-5 mg IVP (@ < 5 mg/min). Relieves spasm of skeletal and smooth muscle (including bladder sphincter). Is also anxiolytic


Good bowel/bladder and skin care are the best preventative measures.

Prophylaxis in patients with recurrent episodes

1. phenoxybenzamine (Dibenzyline®): an alpha blocker. Not helpful during the acute crisis. May not be as effective for alpha stimulation from sympathetic ganglia as with circulating catecholamines299. The patient may also develop hypotension after the sympathetic outflow subsides. Thus this is used only for resistant cases (note: will not affect sweating which is mediated by acetylcholine).

Rx Adult: wide range quoted in literature: average 20-30 mg PO BID

2. beta-blockers: may be necessary in addition to α-blockers to avoid possible hypotension from ß2 receptor stimulation (a theoretical concern)

3. phenazopyridine (Pyridium®): a topical anesthetic that is excreted in the urine. May decrease bladder wall irritation, however, the primary cause of irritation should be treated if possible.

Rx Adult: 200 mg PO TID after meals. SUPPLIED: 100, 200 mg tabs.

4. “radical measures” such as sympathectomy, pelvic or pudendal nerve section, cordectomy, or intrathecal alcohol injection have been advocated in the past, but are rarely necessary and may jeopardize reflex voiding

5. prophylactic treatment prior to procedures may employ use of anesthetics even in regions rendered anesthetic by the cord injury. Nifedipine 10 mg SL has also been used effective for AH during cystoscopy and prophylactically298

28.15. References

1. White A A, Panjabi M M: The problem of clinical instability in the human spine: A systematic approach. In Clinical biomechanics of the spine. J.B. Lippincott, Philadelphia, 2nd ed., 1990: pp 277-378.

2. Waters R L, Adkins R H, Yakura J, et al.: Profiles of spinal cord injury and recovery after gunshot injury. Clin Orthop 267: 14-21, 1991.

3. Atkinson P P, Atkinson J L D: Spinal shock. Mayo Clin Proc 71: 384-9, 1996.

4. Chesnut R M: Emergency management of spinal cord injury. In Neurotrauma, Narayan R K, Wilberger J E, and Povlishock J T, (eds.). McGraw-Hill, New York, 1996: pp 1121-38.

5. Hirsch S A, Hirsch P J, Hiramoto H, et al.: Whiplash syndrome: Fact or fiction? Orthop Clin North Am 19: 791-5, 1988.

6. Riley L H, Long D, Riley Jr. L H: The science of whiplash. Medicine (Baltimore) 74: 298-9, 1995.

7. Spitzer W O, Skovron M L, Salmi L R, et al.: Scientific monograph of the quebec task force on whiplash-associated disorders: Redefining “whiplash” and its management. Spine 20: 1S-73S, 1995.

8. Radanov B P, Sturzenegger M, Di Stefano G: Long-term outcome after whiplash injury. Medicine (Baltimore) 74: 281-97, 1995.

9. Hamilton M G, Myles S T: Pediatric spinal injury: Review of 61 deaths. J Neurosurg 77: 705-8, 1992.

10. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Management of pediatric cervical spine and spinal cord injuries. Neurosurgery 50 Supplement (3): S85-99, 2002.

11. Hamilton M G, Myles S T: Pediatric spinal injury: Review of 174 hospital admissions. J Neurosurg 77: 700-4, 1992.

12. Mandabach M, Ruge J R, Hahn Y S, et al.: Pediatric axis fractures: Early halo immobilization, management and outcome. Pediatric Neurosurgery 19: 225-32, 1993.

13. Garton H J L, Park P, Papadopoulos S M: Fracture dislocation of the neurocentral synchondroses of the axis. Case illustration. J Neurosurg (Spine 3) 96: 350, 2002.

14. Suss R A, Zimmerman R D, Leeds N E: Pseudospread of the atlas: False sign of Jefferson fracture in young children. AJR 140: 1079-82, 1983.

15. Bailey D K: The normal cervical spine in infants and children. Radiology 59: 712-9, 1952.

16. Townsend E H, Rowe M L: Mobility of the upper cervical spine in health and disease. Pediatrics 10: 567-74, 1952.

17. Jacobson G, Bleeker H H: Pseudosubluxation of the axis in children. Am J Roentgenol 82: 472-81, 1959.

18. Bachulis B L, B L W, Hynes G D, et al.: Clinical indications for cervical spine radiographs in the traumatized patient. Am J Surg 153: 473-8, 1987.

19. Harris M B, Waguespack A M, Kronlage S: ‘clearing’ cervical spine injuries in polytrauma patients: Is it really safe to remove the collar? Orthopedics 20: 903-7, 1997.

20. Podolsky S M, Baraff L J, Simon R R, et al.: Efficacy of cervical spine immobilization methods. J Trauma 23: 687-90, 1983.

21. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Clinical assessment after acute cervical spinal cord injury. Neurosurgery 50 Supplement (3): S21-9, 2002.

22. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Management of acute spinal cord injuries in an intensive care unit or other monitored setting. Neurosurgery 50 Supplement (3): S51-7, 2002.

23. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Blood pressure management after acute spinal cord injury. Neurosurgery 50 Supplement (3): S58-62, 2002.

24. Meguro K, Tator C H: Effect of multiple trauma on mortality and neurological recovery after spinal cord or cauda equina injury. Neurol Med Chir 28: 34-41, 1988.

25. Bracken M B, Shepard M J, Holford T R, et al.: Administration of methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in the treatment of acute spinal cord injury. JAMA 277: 1597-604, 1997.

26. Bracken M B, Members of the National Acute Spinal Cord Injury Study Group: Comments. Neurosurgery 50 Supplement (3): Guidelines for the management of acute cervical spine and spinal cord injuries: Sxiv-xx, 2002.

27. Hurlbert R J: Methylprednisolone for acute spinal cord injury: An inappropriate standard of care. J Neurosurg (Spine 1) 93: 1-7, 2000.

28. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Pharmacological therapy after acute cervical spinal cord injury. Neurosurgery 50 Supplement (3): S63-72, 2002.

29. Bracken M B, Shepard M J, Collins W F, et al.: A randomized, controlled trial of methylprednisolone or naloxone in the treatment of acute spinal-cord injury. N Engl J Med 322: 1405-11, 1990.

30. Bracken M B, Shepard M J, Collins W F, et al.: Methylprednisolone or naloxone treatment after acute spinal cord injury: 1-year follow-up data. J Neurosurg 76: 23-31, 1992.

31. Heary R F, Vaccaro A R, Mesa J J, et al.: Steroids and gunshot wounds to the spine. Neurosurgery 41: 576-84, 1997.

32. Short D J, El Masry W S, Jones P W: High dose methylprednisolone in the management of acute spinal cord injury - a systematic review from a clinical perspective. Spinal Cord 38 (5): 273-86, 2000.

33. Qian T, Guo X, Levi A D, et al.: High-dose methylprednisolone may cause myopathy in acute spinal cord injury patients. Spinal Cord 43 (4): 199-203, 2005.

34. Resnick D K, Kaiser M G, Fehlings M, et al.: Hypothermia and human spinal cord injury: Position statement and evidence based recommendations from the AANS/CNS Joint Section on Disorders of the Spine and the AANS/CNS Joint Section on Trauma. 2007. URL http://www.spinesection.org/hypothermia.php. Date accessed: May 30, 2009.

35. Hamilton M G, Hull R D, Pineo G F: Venous thromboembolism in neurosurgery and neurology patients: A review. Neurosurgery 34: 280-96, 1994.

36. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Deep venous thrombosis and thromboembolism in patients with cervical spinal cord injuries. Neurosurgery 50 Supplement (3): S73-80, 2002.

37. Green D, Lee M Y, Ito V Y, et al.: Fixed- vs adjusted-dose heparin in the prophylaxis of thromboembolism in spinal cord injury. JAMA 260: 1255-8, 1988.

38. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Radiographic assessment of the cervical spine in asymptomatic trauma patients. Neurosurgery 50 Supplement (3): S30-5, 2002.

39. Como J J, Diaz J J, Dunham C M, et al.: Practice management guidelines for identification of cervical spine injuries following trauma: Update from the Eastern Association for the Surgery of Trauma practice management guidelines committee. J Trauma 67 (3): J Trauma: 651-9, 2009.

40. Hoffman J R, Mower W R, Wolfson A B, et al.: Validity of a set of clinical criteria to rule out injury to the cervical spine in patients with blunt trauma. National emergency x-radiography utilization study group. N Engl J Med343 (2): 94-9, 2000.

41. Stiell I G, Clement C M, McKnight R D, et al.: The Canadian C-spine rule versus the NEXUS low-risk criteria in patients with trauma. N Engl J Med 349 (26): N Engl J Med: 2510-8, 2003.

42. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Radiographic assessment of the cervical spine in symptomatic trauma patients. Neurosurgery 50 Supplement (3): S36-43, 2002.

43. Chendrasekhar A, Moorman D W, Timberlake G A: An evaluation of the effects of semirigid cervical collars in patients with severe closed head injury. Am Surg 64 (7): Am Surg: 604-6, 1998.

44. Stelfox H T, Velmahos G C, Gettings E, et al.: Computed tomography for early and safe discontinuation of cervical spine immobilization in obtunded multiply injured patients. J Trauma 63 (3): J Trauma: 630-6, 2007.

45. Hunt K, Hallworth S, Smith M: The effects of rigid collar placement on intracranial and cerebral perfusion pressures. Anaesthesia 56 (6): Anaesthesia: 511-3, 2001.

46. Mobbs R J, Stoodley M A, Fuller J: Effect of cervical hard collar on intracranial pressure after head injury. ANZ J Surg 72 (6): ANZ J Surg: 389-91, 2002.

47. Walter J, Doris P, Shaffer M: Clinical presentation of patients with acute cervical spine injury. Ann Emerg Med 13: 512-5, 1984.

48. Nichols C G, Young D H, Schiller W R: Evaluation of cervicothoracic junction injury. Ann Emerg Med 16: 640-2, 1987.

49. Clark W M, Gehweiler J A, Laib R: Twelve significant signs of cervical spine trauma. Skeletal Radiol 3: 201-5, 1979.

50. Shaffer M, Doris P: Limitation of the cross table lateral view in detecting cervical spine injuries: A retrospective analysis. Ann Emerg Med 10: 508-13, 1981.

51. MacDonald R L, Schwartz M L, Mirich D, et al.: Diagnosis of cervical spine injury in motor vehicle crash victims: How many x-rays are enough? J Trauma 30: 392-7, 1990.

52. Holliman C, Mayer J, Cook R, et al.: Is the AP radiograph of the cervical spine necessary in evaluation of trauma? Ann Emerg Med 19: 483-4, 1990 (abstract).

53. Harris J H: Radiographic evaluation of spinal trauma. Orthop Clin North Am 17: 75-86, 1986.

54. Tehranzedeh J, Bonk T, Ansari A, et al.: Efficacy of limited CT for non-visualized lower cervical spine in patients with blunt trauma. Skeletal Radiol 23: 349-52, 1994.

55. Miller M D, Gehweiler J A, Martinez S, et al.: Significant new observations on cervical spine trauma. AJR 130: 659-63, 1978.

56. Wales L, Knopp R, Morishima M: Recommendations for evaluation of the acutely injured spine: A clinical radiographic algorithm. Ann Emerg Med 9: 422-8, 1980.

57. Hollis P H, Malis L I, Zappulla R A: Neurological deterioration after lumbar puncture below complete spinal subarachnoid block. J Neurosurg 64: 253-6, 1986.

58. Horn E M, Lekovic G P, Feiz-Erfan I, et al.: Cervical magnetic resonance imaging abnormalities not predictive of cervical spine instability in traumatically injured patients. J Neurosurg Spine 1 (1): J Neurosurg Spine: 39-42, 2004.

59. Schuster R, Waxman K, Sanchez B, et al.: Magnetic resonance imaging is not needed to clear cervical spines in blunt trauma patients with normal computed tomographic results and no motor deficits. Arch Surg 140 (8): Arch Surg: 762-6, 2005.

60. Benzel E C, Hart B L, Ball P A, et al.: Magnetic resonance imaging for the evaluation of patients with occult cervical spine injury. J Neurosurg 85 (5): J Neurosurg: 824-9, 1996.

61. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Initial closed reduction of cervical spine fracture-dislocation injuries. Neurosurgery 50 Supplement (3): S44-50, 2002.

62. Botte M J, Byrne T P, Abrams R A, et al.: Halo skeletal fixation: Techniques of application and prevention of complications. J Am Acad Orthop Surg 4 (1): J Am Acad Orthop Surg: 44-53, 1996.

63. Robertson P A, Ryan M D: Neurological deterioration after reduction of cervical subluxation: Mechanical compression by disc material. J Bone Joint Surg 74B: 224-7, 1992.

64. Garfin S R, Botte M J, Triggs K J, et al.: Subdural abscess associated with halo-pin traction. J Bone Joint Surg 70A: 1338-40, 1988.

65. Dill S R, Cobbs C G, McDonald C K: Subdural empyema: Analysis of 32 cases and review. Clin Inf Dis 20: 372-86, 1995.

66. Schneider R C, Crosby E C, Russo R H, et al.: Traumatic spinal cord syndromes and their management. Clin Neurosurg 20: 424-92, 1972.

67. Wagner F C, Chehrazi B: Early decompression and neurological outcome in acute cervical spinal cord injuries. J Neurosurg 56: 699-705, 1982.

68. American Spinal Injury Association: International standards for neurological classification of spinal cord injury, revised 2000. 6th ed. American Spinal Injury Association, Chicago, IL, 2000.

69. Ditunno J F, Jr.: New spinal cord injury standards, 1992. Paraplegia 30 (2): 90-1, 1992.

70. Lucas J T, Ducker T B: Motor classification of spinal cord injuries with mobility, morbidity and recovery indices. Am Surg 45: 151-8, 1979.

71. Frankel H L, Hancock D O, Hyslop G, et al.: The value of postural reduction in the initial management of closed injuries of the spine with paraplegia and tetraplegia. Part I. Paraplegia 7: 179-92, 1969.

72. Epstein N, Epstein J A, Benjamin V, et al.: Traumatic myelopathy in patients with cervical spinal stenosis without fracture or dislocation: Methods of diagnosis, management, and prognosis. Spine 5: 489-96, 1980.

73. Schneider R C, Cherry G, Pantek H: The syndrome of acute central cervical spinal cord injury. J Neurosurg 11: 546-77, 1954.

74. Merriam W F, Taylor T K F, Ruff S J, et al.: A reappraisal of acute traumatic central cord syndrome. J Bone Joint Surg 68B: 708-13, 1986.

75. Levi L, Wolf A, Mirvis S, et al.: The significance of dorsal migration of the cord after extensive cervical laminectomy for patients with traumatic central cord syndrome. J Spinal Disord 8: 289-95, 1995.

76. Chen T Y, Lee S T, Lui T N, et al.: Efficacy of surgical treatment in traumatic central cord syndrome. Surg Neurol 48: 435-40, 1997.

77. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Management of acute central spinal cord injuries. Neurosurgery 50 Supplement (3): S166-72, 2002.

78. Massaro F, Lanotte M, Faccani G: Acute traumatic central cord syndrome. Acta Neurol (Napoli) 15: 97-105, 1993.

79. Rothman R H, Simeone F A, (eds.): The spine. 3rd ed., W.B. Saunders, Philadelphia, 1992.

80. Fox J L, Wener L, Drennan D C, et al.: Central spinal cord injury: Magnetic resonance imaging confirmation and operative considerations. Neurosurgery 22: 340-7, 1988.

81. Ducker T B: Comment on Fox J L, et al.: Central spinal cord injury: Magnetic resonance imaging confirmation and operative considerations. Neurosurgery 22: 346-7, 1988.

82. Bose B, Northrup B E, Osterholm J L, et al.: Reanalysis of central cervical cord injury management. Neurosurgery 15: 367-72, 1984.

83. Penrod L E, Hegde S K, Ditunno J F: Age effect on prognosis for functional recovery in acute, traumatic central cord syndrome. Arch Phys Med Rehabil 71: 963-8, 1990.

84. Schneider R C: The syndrome of acute anterior spinal cord injury. J Neurosurg 12: 95-122, 1955.

85. Brown-Sequard C E: De la transmission des impressions sensitives par la moelle epiniere. C R Soc Biol 1 (192-4): C R Soc Biol, 1849.

86. Roth E J, Park T, Pang T, et al.: Traumatic cervical Brown-Sequard and Brown-Sequard plus syndromes: The spectrum of presentations and outcomes. Paraplegia 29: 582-9, 1991.

87. Rumana C S, Baskin D S: Brown-Sequard syndrome produced by cervical disc herniation: Case report and literature review. Surg Neurol 45: 359-61, 1996.

88. Kobayashi N, Asamoto S, Doi H, et al.: Brown-Sequard syndrome produced by cervical disc herniation: Report of two cases and review of the literature. Spine J 3 (6): Spine J: 530-3, 2003.

89. Kim J T, Bong H J, Chung D S, et al.: Cervical disc herniation producing acute Brown-Sequard syndrome. J Korean Neurosurg Soc 45 (5): J Korean Neurosurg Soc: 312-4, 2009.

90. Gehweiler J A, Clark W M, Schaaf R E, et al.: Cervical spine trauma: The common combined conditions. Radiology 130: 77, 1979.

91. Powers B, Miller M D, Kramer R S, et al.: Traumatic anterior atlanto-occipital dislocation. Neurosurgery 4: 12-7, 1979.

92. Alker G J, Leslie E V: High cervical spine and craniocervical junction injuries in fatal traffic accidents: A radiological study. Orthop Clin North Am 9: 1003-10, 1978.

93. Bucholz R W, Burkhead W Z, Graham W, et al.: Occult cervical spine injuries in fatal traffic accidents. J Trauma 19: 768-71, 1979.

94. Traynelis V C, Marano G D, Dunker R O, et al.: Traumatic atlanto-occipital dislocation. Case report. J Neurosurg 65: 863-70, 1986.

95. Harris J H, Jr., Carson G C, Wagner L K, et al.: Radiologic diagnosis of traumatic occipitovertebral dissociation: 2. Comparison of three methods of detecting occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 162 (4): 887-92, 1994.

96. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Diagnosis and management of traumatic atlanto-occipital dislocation injuries. Neurosurgery 50 Supplement (3): S105-13, 2002.

97. Przybylski G J, Clyde B L, Fitz C R: Craniocervical junction subarachnoid hemorrhage associated with atlanto-occipital dislocation. Spine 21 (15): 1761-8, 1996.

98. Dziurzynski K, Anderson P A, Bean D B, et al.: A blinded assessment of radiographic criteria for atlanto-occipital dislocation. Spine 30 (12): 1427-32, 2005.

99. Horn E M, Feiz-Erfan I, Lekovic G P, et al.: Survivors of occipitoatlantal dislocation injuries: Imaging and clinical correlates. J Neurosurg Spine 6 (2): J Neurosurg Spine: 113-20, 2007.

100. Harris J H, Carson G C, Wagner L K: Radiologic diagnosis of traumatic occipitovertebral dissociation: 1. Normal occipitovertebral relationships on lateral radiographs of supine subjects. AJR Am J Roentgenol 162 (4): 881-6, 1994.

101. Rojas C A, Bertozzi J C, Martinez C R, et al.: Reassessment of the craniocervical junction: Normal values on CT. AJNR Am J Neuroradiol 28 (9): 1819-23, 2007.

102. Bertozzi J C, Rojas C A, Martinez C R: Evaluation of the pediatric craniocervical junction on MDCT. AJR Am J Roentgenol 192 (1): AJR Am J Roentgenol: 26-31, 2009.

103. Werne S: Studies in spontaneous atlas dislocation. Acta Orthop Scand Suppl 23: 1-150, 1957.

104. Pang D, Nemzek W R, Zovickian J: Atlanto-occipital dislocation: Part 1--normal occipital condyle-C1 interval in 89 children. Neurosurgery 61 (3): Neurosurgery: 514-21; discussion 521, 2007.

105. Kaufman R A, Carroll C D, Buncher C R: Atlantooccipital junction: Standards for measurement in normal children. AJNR Am J Neuroradiol 8 (6): 995-9, 1987.

106. Pang D, Nemzek W R, Zovickian J: Atlanto-occipital dislocation--part 2: The clinical use of (occipital) condyle-C1 interval, comparison with other diagnostic methods, and the manifestation, management, and outcome of atlanto-occipital dislocation in children. Neurosurgery 61 (5): Neurosurgery: 995-1015; discussion 1015, 2007.

107. Dublin A B, Marks W M, Weinstock D, et al.: Traumatic dislocation of the atlanto-occipital articulation (AOA) with short-term survival. With a radiographic method of measuring the AOA. J Neurosurg52 (4): 541-6, 1980.

108. Lee C, Woodring J H, Goldstein S J, et al.: Evaluation of traumatic atlantooccipital dislocations. AJNR Am J Neuroradiol 8 (1): 19-26, 1987.

109. Wholey M H, Bruwer A J, Baker H L: The lateral roentgenogram of the neck; with comments on the atlanto-odontoid-basion relationship. Radiology 71: 350-6, 1958.

110. Maserati M B, Stephens B, Zohny Z, et al.: Occipital condyle fractures: Clinical decision rule and surgical management. Journal of Neurosurgery: Spine 11 (4): Journal of Neurosurgery: Spine: 388-395, 2009.

111. Bell C L: Surgical observations. Middlesex Hosp J 4: 469, 1817.

112. Jacoby C G: Fracture of the occipital condyle. AJR Am J Roentgenol 132 (3): 500, 1979 (letter).

113. Anderson P A, Montesano P X: Morphology and treatment of occipital condyle fractures. Spine 13 (7): 731-6, 1988.

114. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Occipital condyle fractures. Neurosurgery 50 Supplement (3): S114-9, 2002.

115. Sonntag V K H, Dickman C A: Treatment of upper cervical spine injuries. In Spinal trauma: Current evaluation and management, Rea G L and Miller C A, (eds.). Neurosurgical topics. Committee A P. American Association of Neurological Surgeons, 1993: pp 25-74.

116. Fielding J W, Hawkins R J: Atlanto-axial rotatory fixation. (fixed rotatory subluxation of the atlanto-axial joint). J Bone Joint Surg 59A: 37-44, 1977.

117. Lourie H, Stewart W A: Spontaneous atlantoaxial dislocation: A complication of rheumatic disease. N Engl J Med 265: 677-81, 1961.

118. Wetzel F T, La Rocca H: Grisel’s syndrome. Clin Orthop (240): 141-52, 1989.

119. Fielding J W, Stillwell W T, Chynn K Y, et al.: Use of computed tomography for the diagnosis of atlanto-axial rotatory fixation. J Bone Joint Surg 60A: 1102-4, 1978.

120. Schneider R C, Schemm G W: Vertebral artery insufficiency in acute and chronic spinal trauma. With special reference to the syndrome of acute central cervical spinal cord injury. J Neurosurg 18: 348-60, 1961.

121. Banna M: Spinal fractures and dislocations. In Clinical radiology of the spine and the spinal cord. Aspen Systems Corporation, Rockville, Maryland, 1985: pp 102-59.

122. Govender S, Kumar K P: Staged reduction and stabilisation in chronic atlantoaxial rotatory fixation. J Bone Joint Surg Br 84 (5): J Bone Joint Surg Br: 727-31, 2002.

123. Dickman C A, Greene K A, Sonntag V K: Injuries involving the transverse atlantal ligament: Classification and treatment guidelines based upon experience with 39 injuries. Neurosurgery 38 (1): 44-50, 1996.

124. Bohrer S P, Klein M D, Martin W: “V” shaped pre-dens space. Skeletal Radiol 14: 111-6, 1985.

125. Dickman C A, Mamourian A, Sonntag V K, et al.: Magnetic resonance imaging of the transverse atlantal ligament for the evaluation of atlantoaxial instability. J Neurosurg 75 (2): 221-7, 1991.

126. Hadley M N, Dickman C A, Browner C M, et al.: Acute traumatic atlas fractures: Management and long-term outcome. Neurosurgery 23: 31-5, 1988.

127. Landells C D, Van Peteghem P K: Fractures of the atlas: Classification, treatment and morbidity. Spine 13 (5): 450-2, 1988.

128. Jefferson G: Fractures of the atlas vertebra: Report of four cases, and a review of those previously recorded. Br J Surg 7: 407-22, 1920.

129. Papadopoulos S M: Biomechanics of occipito-atlanto-axial trauma. In Spinal trauma: Current evaluation and management, Rea G L and Miller C A, (eds.). Neurosurgical topics. Committee A P. American Association of Neurological Surgeons, 1993: pp 17-23.

130. Alker G J, Oh Y S, Leslie E V, et al.: Postmortem radiology of head and neck injuries in fatal traffic accidents. Radiology 114: 611-7, 1975.

131. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Isolated fractures of the atlas in adults. Neurosurgery 50 Supplement (3): S120-4, 2002.

132. Levine A M, Edwards C C: Fractures of the atlas. J Bone Joint Surg Am 73 (5): 680-91, 1991.

133. Spence K F, Decker S, Sell K W: Bursting atlantal fracture associated with rupture of the transverse ligament. J Bone Joint Surg 52A: 543-9, 1970.

134. Schneider R C, Livingston K E, Cave A J E, et al.: ‘hangman’s fracture’ of the cervical spine. J Neurosurg 22: 141-54, 1965.

135. Wood-Jones F: The ideal lesion produced by judicial hanging. Lancet 1: 53, 1913.

136. Effendi B, Roy D, Cornish B, et al.: Fractures of the ring of the axis: A classification based on the analysis of 131 cases. J Bone Joint Surg 63B: 319-27, 1981.

137. Levine A M, Edwards C C: The management of traumatic spondylolisthesis of the axis. J Bone Joint Surg 67A: 217-26, 1985.

138. Levine A M: Traumatic spondylolisthesis of the axis: “hangman’s fracture”. In The cervical spine, The Cervical Spine Research Society Editorial Committee, (ed.). Lippincott-Raven, Philadelphia, 3rd ed., 1998: pp 429-48.

139. Starr J K, Eismont F J: Atypical hangman’s fractures. Spine 18 (14): 1954-7, 1993.

140. Francis W R, Fielding J W, Hawkins R J, et al.: Traumatic spondylolisthesis of the axis. J Bone Joint Surg 63B: 313-8, 1981.

141. Greene K A, Dickman C A, Marciano F F, et al.: Acute axis fractures. Analysis of management and outcome in 340 consecutive cases. Spine 22: 1843-52, 1997.

142. Burke J T, Harris J H, Jr.: Acute injuries of the axis vertebra. Skeletal Radiol 18 (5): 335-46, 1989.

143. The Cervical Spine Research Society Editorial Committee, (ed.) The cervical spine. 2nd ed., J.B. Lippincott, Philadelphia, 1989.

144. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Isolated fractures of the axis in adults. Neurosurgery 50 Supplement (3): S125-39, 2002.

145. Tuite G F, Papadopoulos S M, Sonntag V K H: Caspar plate fixation for the treatment of complex hangman’s fractures. Neurosurgery 30: 761-5, 1992.

146. Coric D, Wilson J A, Kelly D L: Treatment of traumatic spondylolisthesis of the axis with nonrigid immobilization: A review of 64 cases. J Neurosurg 85: 550-4, 1996.

147. Sonntag V K H, Hadley M N: Nonoperative management of cervical spine injuries. Clin Neurosurg 34: 630-49, 1988.

148. Youmans J R, (ed.) Neurological surgery. 2nd ed., W. B. Saunders, Philadelphia, 1982.

149. Hadley M N: Comment on Tuite G F, et al.: Caspar plate fixation for the treatment of complex hangman’s fractures. Neurosurgery 30: 761-5, 1992.

150. Husby J, Sorensen K H: Fracture of the odontoid process of the axis. Acta Orthop Scand 45: 182-92, 1974.

151. Crockard H A, Heilman A E, Stevens J M: Progressive myelopathy secondary to odontoid fractures: Clinical, radiological, and surgical features. J Neurosurg 78: 579-86, 1993.

152. Przybylski G J: Management of odontoid fractures. Contemp Neurosurg 20 (18): 1-6, 1998.

153. Anderson L D, D’Alonzo R T: Fractures of the odontoid process of the axis. J Bone Joint Surg 56A: 1663-74, 1974.

154. Hadley M N, Browner C M, Liu S S, et al.: New sub-type of acute odontoid fractures (type IIA). Neurosurgery 22: 67-71, 1988.

155. Scott E W, Haid R W, Peace D: Type I fractures of the odontoid process: Implications for atlanto-occipital instability: Case report. J Neurosurg 72: 488-92, 1990.

156. Naim-ur-Rahman, Jamjoom Z A, Jamjoom A B: Ruptured transverse ligament: An injury that is often forgotten. Br J Neurosurg 14: 375-7, 2000.

157. Hadley M N, Dickman C A, Browner C M, et al.: Acute axis fractures: A review of 229 cases. J Neurosurg 71: 642-7, 1989.

158. Apuzzo M L J, Heiden J S, Weiss M H, et al.: Acute fractures of the odontoid process. An analysis of 45 cases. J Neurosurg 48: 85-91, 1978.

159. Ekong C E U, Schwartz M L, Tator C H, et al.: Odontoid fracture: Management with early mobilization using the halo device. Neurosurgery 9: 631-7, 1981.

160. Dunn M E, Seljeskog E L: Experience in the management of odontoid process injuries: An analysis of 128 cases. Neurosurgery 18: 306-10, 1986.

161. Lennarson P J, Mostafavi H, Traynelis V C, et al.: Management of type II dens fractures: A case-control study. Spine 25 (10): 1234-7, 2000.

162. Bohler J: Anterior stabilization for acute fractures and non-unions of the dens. J Bone Joint Surg 64: 18-28, 1982.

163. Polin R S, Szabo T, Bogaev C A, et al.: Nonoperative management of types II and III odontoid fractures: The philadelphia collar versus the halo vest. Neurosurgery 38: 450-7, 1996.

164. Darakchiev B J, Bulas R V, Dunsker S: Use of calcitonin for the treatment of an odontoid fracture: Case report. J Neurosurg (Spine 1) 93: 157-60, 2000.

165. Paridis G R, Janes J M: Posttraumatic atlanto-axial instability: The fate of the odontoid process fracture in 46 cases. J Trauma 13: 359-67, 1973.

166. Fielding J W, Hensinger R N, Hawkins R J: Os odontoideum. J Bone Joint Surg 62A: 376-83, 1980.

167. Ricciardi J E, Kaufer H, Louis D S: Acquired os odontoideum following acute ligament injury. J Bone Joint Surg 58A: 410-2, 1976.

168. Clements W D, Mezue W, Mathew B: Os odontoideum: Congenital or acquired? That’s not the question. Injury 26 (9): 640-2, 1995.

169. Spierings E L, Braakman R: The management of os odontoideum. Analysis of 37 cases. J Bone Joint Surg Br 64 (4): 422-8, 1982.

170. Menezes A H, Ryken T C: Craniovertebral abnormalities in down’s syndrome. Pediatr Neurosurg 18 (1): 24-33, 1992.

171. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Os odontoideum. Neurosurgery 50 Supplement (3): S148-55, 2002.

172. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Management of combination fractures of the atlas and axis in adults. Neurosurgery 50 Supplement (3): S140-7, 2002.

173. Bono C, Vaccaro A R, Fehlings M, et al.: Measurment techniques for lower cervical spine injuries: Consensus statement of the Spine Trauma Study Group. Spine 31: 603-9, 2006.

174. Vaccaro A R, Hulbert R J, Patel A A, et al.: The sub-axial cervical spine injury classification system: A novel approach to recognize the importance of morphology, neurology, and integrity of the disco-ligamentous complex. Spine 32 (21): 2365-74, 2007.

175. White A A, 3rd, Panjabi M M: Update on the evaluation of instability of the lower cervical spine. Instr Course Lect 36: 513-20, 1987.

176. Allen B L, Ferguson R L, Lehmann T R, et al.: A mechanistic classification of closed, indirect fractures and dislocations of the lower cervical spine. Spine 7: 1-27, 1982.

177. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Treatment of subaxial cervical spine injuries. Neurosurgery 50 Supplement (3): S156-65, 2002.

178. Hall R D M: Clay-shoveller’s fracture. J Bone Joint Surg 22: 63-75, 1940.

179. Gershon-Cohen J, Budin E, Glauser F: Whiplash fractures of cervicodorsal spinous processes. JAMA 155: 560-1, 1954.

180. Abitbol J-J, Kostuik J P: Flexion injuries to the lower cervical spine. In The cervical spine, The Cervical Spine Research Society Editorial Committee, (ed.). Lippincott-Raven, Philadelphia, 3rd ed., 1998: pp 457-64.

181. Fuentes J-M, Bloncourt J, Vlahovitch B, et al.: La tear drop fracture: Contribution à l’étude du mécanisme et des lésions ostéo-disco-ligamentaires. Nirochirurgie 29: 129-34, 1983 (French).

182. Schneider R C, Kahn E A, Arbor A: Chronic neurologic sequelae of acute trauma to the spine and spinal cord. The significance of acute flexion or teardrop cervical fracture-dislocation of the cervical spine. J Bone Joint Surg38A, 1956.

183. Torg J S, Vegso J J, Sennett B: The national football head and neck injury registry: 14-year report of cervical quadriplegia (1971-1984). Clin Sports Med 6: 61-72, 1987.

184. Harris J H, Edeiken-Monroe B, Kopaniky D R: A practical classification of acute cervical spine injuries. Orthop Clin North Am 17: 15-30, 1986.

185. Gehweiler J A, Osborne R L: The radiology of vertebral trauma. W. B. Saunders, Philadelphia, 1980.

186. Favero K J, Van-Peteghem P K: The quadrangular fragment fracture: Roentgenographic features and treatment protocol. Clin Orthop 239: 40-6, 1989.

187. Webb J K, Broughton R B K, McSweeney T, et al.: Hidden flexion injury of the cervical spine. J Bone Joint Surg 58B: 322-7, 1976.

188. Fazl M, LaFebvre J, Willinsky R A, et al.: Posttraumatic ligamentous disruption of the cervical spine, an easily overlooked diagnosis: Presentation of three cases. Neurosurgery 26: 674-7, 1990.

189. White A A, Johnson R M, Panjabi M M, et al.: Biomechanical analysis of clinical stability in the cervical spine. Clin Orthop 109: 85-96, 1975.

190. White A A, Southwick W O, Panjabi M M: Clinical instability in the lower cervical spine - a review of past and current concepts. Spine 1: 15-27, 1976.

191. Andreshak J L, Dekutoski M B: Management of uni-lateral facet dislocations: A review of the literature. Orthopedics 20: 917-26, 1997.

192. Payer M, Schmidt M H: Management of traumatic bilateral locked facets of the subaxial cervical spine. Contemp Neurosurg 27 (6): 1-4, 2005.

193. Rizzolo S J, Piazza M R, Cotler J M, et al.: Intervertebral disc injury complicating cervical spine trauma. Spine 16: S187-9, 1991.

194. Doran S E, Papadopoulos S M, Ducker T B, et al.: Magnetic resonance imaging documentation of coexistent traumatic locked facets of the cervical spine and disc herniation. J Neurosurg 79: 341-5, 1993.

195. Sonntag V K H: Management of bilateral locked facets of the cervical spine. Neurosurgery 8: 150-2, 1981.

196. Glasser J A, Whitehall R, Stamp W G, et al.: Complications associated with the halo vest. J Neurosurg 65: 76-9, 1986.

197. Sears W, Fazl M: Prediction of stability of cervical spine fracture managed in the halo vest and indications for surgical intervention. J Neurosurg 72: 426-32, 1990.

198. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Spinal cord injury without radiographic abnormality. Neurosurgery50 Supplement (3): S100-4, 2002.

199. Pang D, Wilberger J E: Spinal cord injury without radiographic abnormalities in children. J Neurosurg 57: 114-29, 1982.

200. Bondurant C P, Oró J J: Spinal cord injury without radiographic abnormality and Chiari malformation. J Neurosurg 79: 833-8, 1993.

201. Pollack I F, Pang D, Sclabassi R: Recurrent spinal cord injury without radiographic abnormalities in children. J Neurosurg 69: 177-82, 1988.

202. Madsen J R, Freiman T: Cervical spinal cord injury in children. Contemp Neurosurg 20 (26): 1-5, 1998.

203. Kotani Y, Abumi K, Ito M, et al.: Cervical spine injuries associated with lateral mass and facet joint fractures: New classification and surgical treatment with pedicle screw fixation. Eur Spine J 14: 69-77, 2005.

204. Roy-Camille R, Saillant G: Osteosynthese des fractures du rachis cervical. Actual Chir Orthop Hop R Poincarré Mason, Paris 8: 175-94, 1970.

205. Spector L R, Kim D H, Affonso J, et al.: Use of computed tomography to predict failure of nonoperative treatment of unilateral facet fractures of the cervical spine. Spine (Phila Pa 1976) 31 (24): Spine (Phila Pa 1976): 2827-35, 2006.

206. Aldrich E F, Crow W N, Weber P B, et al.: Use of MR imaging-compatible Halifax interlaminar clamps for posterior cervical fusion. J Neurosurg 74: 185-9, 1991.

207. Branch C L, Kelly D L, Davis C H, et al.: Fixation of fractures of the lower cervical spine using methylmethacrylate and wire: Technique and results in 99 patients. Neurosurgery 25: 503-13, 1989.

208. Cooper P R: Comment on Branch C L, et al.: Fixation of fractures of the lower cervical spine using methylmethacrylate and wire. Neurosurgery 25: 512-3, 1989 (comment).

209. McGuire R A: Cervical spine arthrodesis. In The cervical spine, The Cervical Spine Research Society Editorial Committee, (ed.). Lippincott-Raven, Philadelphia, 3rd ed., 1998: pp 499-508.

210. Vaccaro A, Ring D, Seuderi G, et al.: Vertebral artery location in relation to the vertebral body as determined by two-dimensional computed tomography evaluation. Spine 19: 2637, 1994.

211. Bailes J E, Hadley M N, Quigley M R, et al.: Management of athletic injuries of the cervical spine and spinal cord. Neurosurgery 29: 491-7, 1991.

212. Torg J S, Corcoran T A, Thibault L F, et al.: Cervical cord neuropraxia: Classification, pathomechanics, morbidity, and management guidelines. J Neurosurg 87: 843-50, 1997.

213. Maroon J C: “burning hands” in football spinal cord injuries. JAMA 238: 2049-51, 1977.

214. Cantu R C, Mueller F O: Catastrophic spine injuries in football. J Spinal Disord 3: 227-31, 1990.

215. Torg J S, Ramsey-Emrhein J A: Management guidelines for participation in collision activities with congenital, developmental, or post-injury lesions involving the cervical spine. Clin Sports Med 16: 501-31, 1997.

216. Herkowitz H N, Rothman R H: Subacute instability of the cervical spine. Spine 9: 348-57, 1984.

217. Delfini R, Dorizzi A, Facchinetti G, et al.: Delayed post-traumatic cervical instability. Surg Neurol 51: 588-95, 1999.

218. Stein D M, Boswell S, Sliker C W, et al.: Blunt cerebrovascular injuries: Does treatment always matter? J Trauma 66 (1): J Trauma: 132-43; discussion 143-4, 2009.

219. Biffl W L, et al.: Western Trauma Association critical decisions in trauma: Diagnosis and management of blunt cerebral vascular injury. J Trauma In press: J Trauma, 2009.

220. Eastman A L, Chason D P, Perez C L, et al.: Computed tomographic angiography for the diagnosis of blunt cervical vascular injury: Is it ready for prime-time? J Trauma 60 (5): J Trauma: 925-9; discussion 929, 2006.

221. Miller P R, Fabian T C, Croce M A, et al.: Prospective screening for blunt cerebrovascular injuries: Analysis of diagnostic modalities and outcomes. Ann Surg 236 (3): Ann Surg: 386-93; discussion 393-5, 2002.

222. Biffl W L, Ray C E, Jr., Moore E E, et al.: Noninvasive diagnosis of blunt cerebrovascular injuries: A preliminary report. J Trauma 53 (5): J Trauma: 850-6, 2002.

223. Cogbill T H, Moore E E, Meissner M, et al.: The spectrum of blunt injury to the carotid artery: A multicenter perspective. J Trauma 37 (3): J Trauma: 473-9, 1994.

224. Mutze S, Rademacher G, Matthes G, et al.: Blunt cerebrovascular injury in patients with blunt multiple trauma: Diagnostic accuracy of duplex doppler us and early CT angiography. Radiology 237 (3): Radiology: 884-92, 2005.

225. Biffl W L, Moore E E, Offner P J, et al.: Blunt carotid arterial injuries: Implications of a new grading scale. J Trauma 47 (5): 845-53, 1999.

226. Biffl W L, Ray C E, Jr., Moore E E, et al.: Treatment-related outcomes from blunt cerebrovascular injuries: Importance of routine follow-up arteriography. Ann Surg 235 (5): Ann Surg: 699-706; discussion 706-7, 2002.

227. Edwards N M, Fabian T C, Claridge J A, et al.: Antithrombotic therapy and endovascular stents are effective treatment for blunt carotid injuries: Results from longterm followup. J Am Coll Surg 204 (5): J Am Coll Surg: 1007-13; discussion 1014-5, 2007.

228. Cothren C C, Moore E E, Ray C E, Jr., et al.: Carotid artery stents for blunt cerebrovascular injury: Risks exceed benefits. Arch Surg 140 (5): Arch Surg: 480-5; discussion 485-6, 2005.

229. Hermosillo A J, Spinler S A: Aspirin, clopidogrel, and warfarin: Is the combination appropriate and effective or inappropriate and too dangerous? Ann Pharmacother 42 (6): Ann Pharmacother: 790-805, 2008.

230. Anson J, Crowell R M: Cervicocranial arterial dissection. Neurosurgery 29: 89-96, 1991.

231. Biller J, Hingtgen W L, Adams H P, et al.: Cervicocephalic arterial dissections: A ten-year experience. Arch Neurol 43: 1234-8, 1986.

232. Biffl W L, Moore E E, Elliott J P, et al.: The devastating potential of blunt vertebral arterial injuries. Ann Surg 231 (5): 672-81, 2000.

233. Hart R G, Easton J D: Dissections of cervical and cerebral arteries. Neurol Clin North Am 1: 255-82, 1983.

234. Section on Disorders of the Spine and Peripheral Nerves of the American Association of Neurological Surgeons and the Congress of Neurological Surgeons: Management of vertebral artery injuries after nonpenetrating cervical trauma. Neurosurgery 50 Supplement (3): S173-8, 2002.

235. Berne J D, Norwood S H: Blunt vertebral artery injuries in the era of computed tomographic angiographic screening: Incidence and outcomes from 8292 patients. J Trauma: J Trauma, 2009.

236. Louw J A, Mafoyane N A, Small B, et al.: Occlusion of the vertebral artery in cervical spine dislocations. J Bone Joint Surg Br 72 (4): 679-81, 1990.

237. Willis B K, Greiner F, Orrison W W, et al.: The incidence of vertebral artery injury after midcervical spine fracture or subluxation. Neurosurgery 34 (3): 435-41; discussion 441-2, 1994.

238. Mas J L, Henin D, Bousser M G, et al.: Dissecting aneurysm of the vertebral artery and cervical manipulation: A case report with autopsy. Neurology 39: 512-5, 1989.

239. Caplan L R, Zarins C K, Hemmati M: Spontaneous dissection of the extracranial vertebral arteries. Stroke 16: 1030-8, 1985.

240. Smith W S, Johnston S C, Skalabrin E J, et al.: Spinal manipulative therapy is an independent risk factor for vertebral artery dissection. Neurology 60 (9): 1424-8, 2003.

241. Lee Y J, Ahn J Y, Han I B, et al.: Therapeutic endovascular treatments for traumatic vertebral artery injuries. J Trauma 62 (4): J Trauma: 886-91, 2007.

242. Denis F: The three column spine and its significance in the classification of acute thoracolumbar spinal injuries. Spine 8: 817-31, 1983.

243. Chedid M K, Green C: A review of the management of lumbar fractures with focus on surgical decision-making and techniques. Contemp Neurosurg 21 (11): 1-5, 1999.

244. Homnick A, Lavery R, Nicastro O, et al.: Isolated thoracolumbar transverse process fractures: Call physical therapy, not spine. J Trauma 63 (6): J Trauma: 1292-5, 2007.

245. Bradley L H, Paullus W C, Howe J, et al.: Isolated transverse process fractures: Spine service management not needed. J Trauma 65 (4): J Trauma: 832-6; discussion 836, 2008.

246. Hitchon P W, Jurf A A, Kernstine K, et al.: Management options in thoracolumbar fractures. Contemp Neurosurg 22 (21): 1-12, 2000.

247. Hitchon P W, Torner J C, Haddad S F, et al.: Management options in thoracolumbar burst fractures. Surg Neurol 49: 619-27, 1998.

248. Klerk L W L, Fontijne P J, Stijnen T, et al.: Spontaneous remodeling of the spinal canal after conservative management of thoracolumbar burst fractures. Spine 23: 1057-, 1998.

249. Dai L Y: Remodeling of the spinal canal after thoracolumbar burst fractures. Clin Orthop 382: 119-, 2001.

250. Court-Brown C M, Gertzbein S D: The management of burst fractures of the fifth lumbar vertebrae. Spine 12: 308-12, 1987.

251. Frederickson B E, Yuan H A, Miller H: Burst fractures of the fifth lumbar vertebrae: A report of four cases. J Bone Joint Surg 64A: 1088-94, 1982.

252. Vaccaro A R, Zieller S C, Hulbert R J, et al.: The thoracolumbar injury severity score: A proposed treatment algorithm. Journal of Spinal Disorders Tech 18: 209-, 2005.

253. Vaccaro A R, Lehman R A, Jr., Hurlbert R J, et al.: A new classification of thoracolumbar injuries: The importance of injury morphology, the integrity of the posterior ligamentous complex, and neurologic status. Spine 30 (20): Spine: 2325-33, 2005.

254. Bose B, Osterholm J L, Northrup B E, et al.: Management of lumbar translocation injuries: Case reports. Neurosurgery 17: 958-61, 1985.

255. Blumenkopf B, Daniels T: Intraoperative ultra-sonography (IOUS) in thoracolumbar fractures. J Spinal Disord 1: 86-93, 1988.

256. Consensus Development Conference: Prophylaxis and treatment of osteoporosis. Am J Med 90: 107-10, 1991.

257. Daniell H W: Osteoporosis of the slender smoker: Vertebral compression fracture and loss of metacarpal cortex in relation to postmenopausal cigarette smoking and lack of obesity. Arch Int Med 136: 298-304, 1976.

258. Shahinian V B, Kuo Y F, Freeman J L, et al.: Risk of fracture after androgen deprivation for prostate cancer. N Engl J Med 352 (2): 154-64, 2005.

259. Kanis J A, Melton J, Christiansen C, et al.: The diagnosis of osteoporosis. J Bone Miner Res 9: 1137-41, 1994.

260. Choice of drugs for postmenopausal osteoporosis. Med Letter 34: 101-2, 1992.

261. Riggs B L, Melton L J: The prevention and treatment of osteoporosis. N Engl J Med 327: 620-7, 1992.

262. Khosla S, Riggs B L: Treatment options for osteoporosis. Mayo Clin Proc 70: 978-82, 1995.

263. Drugs for prevention and treatment of postmenopausal osteoporosis. Med Letter 42: 97-100, 2000.

264. Rossouw J E, Anderson G L, Prentice R L, et al.: Risks and benefits of estrogen plus progestin in healthy postmenopausal women: Principal results from the Women’s Health Initiative randomized controlled trial. JAMA 288 (3): JAMA: 321-33, 2002.

265. Holmberg L, Anderson H, steering H, et al.: HABITS (hormonal replacement therapy after breast cancer--is it safe?), a randomised comparison: Trial stopped. Lancet 363 (9407): Lancet: 453-5, 2004.

266. Office of Dietary Supplement - National Institutes of Health: Dietary supplement fact sheet: Vitamin D. (December 8, 2009), 2009.

267. Once-a-week risedronate (Actonel). Med Letter 44 (1141): 87-8, 2002.

268. Raloxifene for postmenopausal osteoporosis. Med Letter 40: 29-30, 1998.

269. Bell N H: RANK ligand and the regulation of skeletal remodeling. J Clin Invest 111 (8): J Clin Invest: 1120-2, 2003.

270. McClung M R, Lewiecki E M, Cohen S B, et al.: Denosumab in postmenopausal women with low bone mineral density. N Engl J Med 354 (8): N Engl J Med: 821-31, 2006.

271. Kallmes D F, Comstock B A, Heagerty P J, et al.: A randomized trial of vertebroplasty for osteoporotic spinal fractures. N Engl J Med 361 (6): N Engl J Med: 569-79, 2009.

272. Buchbinder R, Osborne R H, Ebeling P R, et al.: A randomized trial of vertebroplasty for painful osteoporotic vertebral fractures. N Engl J Med 361 (6): N Engl J Med: 557-68, 2009.

273. Cotten A, Dewatre F, Cortet B, et al.: Percutaneous vertebroplasty for osteolytic metastases and myeloma: Effects of the percentage of lesion filling and the leakage of methyl methacrylate at clinical follow-up. Radiology 200: 525-30, 1996.

274. Fourney D R, Schomer D F, Nader R, et al.: Percutaneous vertebroplasty and kyphoplasty for painful vertebral body fractures in cancer patients. J Neurosurg 98: 21-30, 2003.

275. Choe D H, Marom E M, Ahrar K, et al.: Pulmonary embolism of polymethyl methacrylate during percutaneous vertebroplasty and kyphoplasty. Am J Roentgenol 183: 1097-102, 2004.

276. Bendok B R, Halpin R J, Rubin M N, et al.: Percutaneous vertebroplasty. Contemp Neurosurg 26 (8): 1-6, 2004.

277. Gibbons K J, Soloniuk D S, Razack N: Neurological injury and patterns of sacral fractures. J Neurosurg 72: 889-93, 1990.

278. Denis F, Davis S, Comfort T: Sacral fractures: An important problem. Retrospective analysis of 236 cases. Clin Orthop 227: 67-81, 1988.

279. Sabiston C P, Wing P C: Sacral fractures: Classification and neurologic implications. J Trauma 26: 1113-5, 1986.

280. Robertson D P, Simpson R K: Penetrating injuries restricted to the cauda equina: A retrospective review. Neurosurgery 31: 265-70, 1992.

281. Messer H D, Cereza P F: Copper jacketed bullets in the central nervous system. Neuroradiology 12: 121-9, 1976.

282. Benzel E C, Hadden T, Coleman J E: Civilian gunshot wounds to the spinal cord and cauda equina. Neurosurgery 20: 281-5, 1987.

283. Linden M A, Manton W I, Stewart R M, et al.: Lead poisoning from retained bullets. Pathogenesis, diagnosis, and management. Ann Surg 195: 305-13, 1982.

284. Monson D O, Saletta J D, Freeark R J: Carotid vertebral trauma. J Trauma 9: 987-9, 1969.

285. Perry M O: Injuries of the brachiocephalic vessels. In Vasc surg, Rutherford R B, (ed.). W.B. Saunders, Philadelphia, 4th ed., 1995, Vol. 1: pp 705-13.

286. Reid J D S, Weigelt J A: Forty-three cases of vertebral artery trauma. J Trauma 28: 1007-12, 1988.

287. Fogelman M J, Stewart R D: Penetrating wounds of the neck. Am J Surg 91: 581-96, 1956.

288. Meyer J P, Barrett J A, Schuler J J, et al.: Mandatory versus selective exploration for penetrating neck trauma. A prospective assessment. Arch Surg 122: 592-7, 1987.

289. Ledgerwood A M, Mullins R J, Lucas C E: Primary repair vs ligation for carotid artery injuries. Arch Surg 115: 488-93, 1980.

290. Meier D E, Brink B E, Fry W J: Vertebral artery trauma: Acute recognition and treatment. Arch Surg 116: 236-9, 1981.

291. Planner A C, Pretorius P M, Graham A, et al.: Sub-acute progressive ascending myelopathy following spinal cord injury: MRI appearances and clinical presentation. Spinal Cord 46: 140-4, 2008.

292. Levi A D, Hurlbert R J, Anderson P, et al.: Neurologic deterioration secondary to unrecognized spinal instability following trauma - a multicenter trial. Spine 41 (4): 451-8, 2006.

293. Parthiban C J K B, Majeed S A: Delayed spinal extradural hematoma following thoracic spine surgery and resulting in paraplegia: A case report. 2008. URL http://www.imedicalcasereports.com/content/2/1/141. Date accessed: June 9, 2008.

294. Liu X Z, Xu H M, Hu R, et al.: Neuronal and glial apoptosis after traumatic spinal cord injury. J Neurosci 17 (14): 5395-406, 1997.

295. Liverman C T, Altevogt B M, Joy J E, et al., (eds.): Spinal cord injury: Progress, promise and priorities. The National Academies Press, Washington, D.C., 2005: pp 360.

296. Erickson R P: Autonomic hyperreflexia: Pathophysiology and medical management. Arch Phys Med Rehabil 61: 431-40, 1980.

297. Kewalramani L S, Orth M S: Autonomic dysreflexia in traumatic myelopathy. Am J Phys Med 59: 1-21, 1980.

298. Dykstra D D, Sidi A A, Anderson L C: The effect of nifedipine on cyctoscopy-induced autonomic hyperreflexia in patients with high spinal cord injuries. J Urol 138: 1155-7, 1987.

299. Sizemore G W, Winternitz W W: Autonomic hyperreflexia - suppression with alpha-adrenergic blocking agents. N Engl J Med 282: 795, 1970.

300. Wardlaw D, Cummings S R, Van Meirhaeghe J, et al.: Efficacy and safety of balloon kyphoplasty compared with non-surgical care for vertebral compression fracture (FREE): A randomised controlled trial. Lancet 373 (9668): 1016-24, 2009.