THORACOLUMBAR FRACTURES: EVALUATION, CLASSIFICATION, AND INITIAL MANAGEMENT

In trauma management, the first priority is to preserve
the patient’s life. In some cases, the threat to life is evident (e.g.,
from hemorrhage, visceral trauma), but in others it is not.

Unstable thoracolumbar fractures are usually high-energy injuries. Anywhere from 40% to 80% result from

motor vehicle accidents, involving drivers and passengers of automobiles, riders of motorcycles, and pedestrians (1,7,17,18,20,23,25,29).
Other causes of spine fractures include falls from height, penetrating
trauma, and crush injuries, such as those sustained by a worker caught
beneath a collapsing structure. In these kinds of injuries, polytrauma
is common (Fig. 141.1). In our experience,
patients with unstable thoracolumbar fractures suffer an average of two
other major injuries in addition to their spinal fracture; some
patients may present with as many as six associated injuries (20).

Once the potentially life-threatening injuries have been addressed or ruled out, the next priority is to stabilize and

protect the patient’s spine so that a more formal evaluation and workup
can be carried out without injuring the spinal cord. This is
particularly important in the polytrauma patient who may be
unconscious, may require anesthesia and surgical care, and must be
moved repeatedly to manage other life-threatening injuries. Plain
radiographs of the cervical spine are mandatory before intubating the
patient, and if injury is seen or suspected, a fiberoptic nasotrachial
intubation is the safest.

Transfer of the patient is safest on a spine board or
slide board, and it should always be done with sufficient personnel to
make the transfer smoothly and without struggling. When log-rolling the
patient, the team must coordinate efforts to see that the shoulders and
pelvis move together as a unit. If the patient is hemodynamically
stable and does not require emergency procedures, he may be transferred
to a firm mattress and maintained under strict spinal precautions until
the workup is completed. Precautions include strict supine positioning,
log-rolling side to side every 2 hours for skin care, and periodic
reexamination of neurologic status. Head-injured and combative patients
may need to be sedated and intubated to avoid self-inflicted spinal
cord injury.

With the patient hemodynamically and mechanically
stabilized, return to the spinal injury assessment. Obtain a complete
history, paying close attention to reports of transient paresthesias,
acute back or neck pain, or temporary weakness or paralysis at the time
of injury. Record the location and radiation of pain symptoms, as well
as any radicular symptoms. Any history of previous injury, fracture, or
pain symptoms should be noted. A global examination of motor and
sensory function should rapidly focus on any areas of deficit. If the
patient cannot cooperate with the exam, carefully observe and note
spontaneous movements and withdrawal responses. Carry out a rectal exam
to assess rectal tone, voluntary rectal control, and the bulbocavanosus
reflex. If the patient is neurologically normal, log-roll him to one
side so that the spine can be palpated for step-offs, tenderness, or
kyphosis. Note the condition of the skin over the symptomatic area. If
a neurologic deficit exists, obtain radiographs of the symptomatic
level before moving the patient.

If a spinal cord injury is identified, start the patient on high-dose steroids to attempt to facilitate recovery (4).
Steroids have been shown to improve spinal cord recovery relative to
placebo and naloxone therapy, and they are thought to combat abnormal
biochemical processes brought on by thromboxanes and prostaglandins
released at the site of injury. Steroids must be given within 8 hours
of injury to have any beneficial effect. Patients treated with
high-dose steroids may be exposed to an increased infection rate and
risk of gastrointestinal hemorrhage.

A formal physical examination and history may not be
possible until the patient has been stabilized hemodynamically and has
recovered from initial resuscitation. When the patient is alert and
cooperative, a formal motor, sensory, and reflex examination should be
repeated, and a detailed history of the accident obtained.

The history should focus on three issues: mechanism of
injury; presence or absence of neurologic symptoms, and past history of
spinal trauma, surgery, or symptoms. In high-energy injuries, it is
often difficult to determine exactly what forces acted on the spine to
produce fracture, but knowledge of the injury mechanism can help
identify associated injuries and provide clues to the level of
instability to be expected. A lap-belted patient in a motor vehicle
accident may present with a straightforward flexion–distraction injury,
for instance, whereas a patient ejected from the vehicle or from a
motorcycle frequently will present with a more complex fracture pattern
consistent with the combination of torsional and axial loading forces
experienced when striking the ground. If the forces involved in the
fracture were rather low, an underlying pathologic process must be
considered. If the forces involved were very high, and multiple
injuries were sustained, the risk of prolonged recumbency to the
patient’s life must be considered in timing a surgical procedure.

If the patient can recall the event, it is important to
elicit any history of transient paresthesias or paralysis from the time
of injury. Even if the patient’s symptoms resolved quickly, they
suggest that some level of root or cord injury occurred at the time of
the injury; therefore, assume that the spinal fracture is unstable. If
the patient cannot give the details of the event, careful scrutiny of
field notes can provide important clues to whether the patient had
abnormal findings at the accident site. These notes are often gross
evaluations only, however, and a patient with a severe cauda equina
injury can still “move all four extremities.”

A history of previous trauma, surgery, or symptoms is
important to understanding the current injury. Lower extremity weakness
due to old injury or spinal disease can confuse the diagnostic picture
after an acute injury, and a preexisting deformity—compression
fracture, spondylolisthesis, or kyphosis—can be difficult to
differentiate from a new injury. Furthermore, the surgical approach to
the multiply operated back will be more demanding and may require
different instrumentation than for an ordinary fracture.

The physical examination for the spinal-injured patient centers around a careful, complete neurologic assessment.

Having examined the musculoskeletal system in the emergency department,
carefully reexamine the extremities for tenderness and pain, and
examine the back again to determine the level of discomfort and the
presence of step-offs or gaps between the spinous processes, and to
assess the skin over the area of injury.

Document a complete motor and sensory examination. Test
each motor group for the lumbar and sacral plexuses independently and
compare to the contralateral group (Fig. 141.2). Motor strength is recorded on six-point scale:

When extremity injuries are present, make an educated
assessment as to whether the patient is clinically weak or limited in
effort by pain. Also determine whether the pattern of weakness is
consistent with a cord lesion, a root lesion, or a peripheral nerve
injury.

The sensory examination begins at the chest wall and
seeks a level of anesthesia, root by root, down to the sacrum. Patients
with thoracic cord injuries will have an anesthetic level at or just
below their fracture. If the anesthetic level and the recognized
fracture do not coincide, obtain an MRI to determine the actual cause
of the cord impairment. Sensation in the lower extremities follows a
dermatomal pattern; test each dermatome for light-touch and pin-prick
sensation (Fig. 141.3).

Check reflexes at both the knees and the ankles. The
bulbocavernosus reflex, an involuntary contraction of the rectal
sphincter, can be triggered either by gently squeezing the glans penis
or glans clitoris or by gently tugging on the Foley catheter. If this
reflex is absent, the patient either is in spinal shock or has suffered
an injury to the caudal segments of the conus medullaris. Hyperactive
reflexes suggest disinhibition due to a cord-level injury. Absent
reflexes in an isolated distribution suggest an incomplete injury or
root lesion. Complete absence of reflexes may be due to either spinal
cord shock or a complete cauda equina injury. Spinal cord shock occurs
at the time of injury and may persist for 72 hours. While the patient
is in spinal cord shock, the neurologic examination remains
unreliable—an incomplete injury may appear complete due to the
overriding effects of cord shock. Once shock resolves and caudal
reflexes return, the examination provides clear prognostic information:
Incomplete injuries have potential for improvement, complete injuries
have almost none. The bulbocavernosus reflex is the most reliable level
for testing reflex return because it tests the most caudal segment of
the spinal cord.

The rectal examination deserves special comment. The
most caudal motor and sensory unit in the body is the rectum, the
function of which is crucial to independent social activity. Carry out
an independent examination of rectal tone, sensation, and reflex
activity. Do not rely on emergency department records if there is any
concern of neurologic injury. Explain the purpose of the examination to
the patient, who may be anxious over repeated exams. Document resting
tone, voluntary contraction, perianal sensation, and bulbocavernosus
reflex.

Plain radiographs should include anteroposterior (AP)
and lateral views of the thoracic spine and/or the lumbosacral spine,
depending on the symptomatic level (16). On
occasion, standard thoracic films will cut off T12–L1, and lumbosacral
films will start at L-1, giving an inadequate view of the most
frequently injured level. If fracture of the thoracolumbar junction is
suspected, repeat AP and lateral radiographs, centered at the T-12
level. In stable fractures (compression fractures, mild burst
fractures, and mild flexion–distraction injuries), plain radiographs
are sufficient to allow definitive treatment, and no further diagnostic
studies are needed. In unstable spine fractures, however, additional
imaging studies are often indicated.

Unstable fractures (severe burst fractures, fracture–
dislocations, significant flexion–distraction injuries, and any
fracture with a neurologic deficit) require further study to assess the
extent of bony disruption, spinal cord impingement, canal compromise,
and/or cord injury. A CT scan provides the most definitive information
on bony characteristics, such as fracture pattern and comminution (8,11,15).
The axial cuts of the CT scan can completely miss flexion–distraction
injuries, however. An MRI is superior for soft-tissue details such as
cord injury, cord compression, disc herniation, and ligamentous
disruption (28). The MRI has the added benefit of scanning the entire thoracolumbar spine, and it can pick up noncontiguous

fractures, cord injury, and epidural hematoma at levels other than that
of the primary fracture. Longitudinal MRI cuts show the soft-tissue
disruption and bony separation of flexion–distraction injuries well.

Myelography, the gold standard for assessing neural
compromise just a decade ago, is now replaced by the MRI in all but a
few cases. When an MRI is contraindicated (e.g., intraocular fragments,
cardiac valves), CT myelography is an appropriate but more invasive
alternative.

There are two absolute indications for ordering MRI or myelography in the acutely injured patient:

Plain tomography is sometimes useful for evaluating the
cervicothoracic junction when a CT scan cannot be obtained immediately.
Flexion–extension studies or nuclear medicine scans have little role in
acute trauma. There is no role for electrodiagnostic testing in the
acute management of spine trauma patients.

Once the diagnostic evaluation is complete, the fracture
can usually be classified according to one of a number of schemes.
Holdsworth first characterized spinal fractures according to a
two-column—anterior and posterior—model of the spinal column (13,14).
This model has since given way to the three-column model of Denis,
which considers the vertebral body, anulus, and posterior longitudinal
ligament to be the middle column, a discrete unit separate from the
anterior and posterior stabilizers (5,6,22) (Fig. 141.4).

The first consideration is whether to classify a specific

injury as stable or unstable. Unstable injuries include all those with any of the following:

Although all stable injuries may be treated
nonoperatively, not all unstable injuries need to be treated
operatively. A simple algorithm for treatment is indicated in Fig. 141.5.

After assessing the level of instability, classify the
fracture according to fracture type and severity. The Denis fracture
classification (Table 141.1) depends on
information about the fracture pattern, the mechanism of injury, and
the deforming forces that caused the fracture. The differences between
severe burst fractures and rotational fracture–dislocations, and severe
seat-belt injuries and flexion–distraction fracture–dislocations are
subtle, and of limited importance: These severe injuries are all
clearly unstable, and all require operative treatment.

Successful fracture treatment begins with a careful and
comprehensive initial evaluation. The key to success is, as always, to
look at the whole patient—never allowing a single, dramatic injury to
distract attention from more subtle, and potentially more dangerous,
injuries. Once the patient is hemodynamically stable, and the fracture
is recognized and classified, prepare a treatment plan based on the
fracture pattern, the severity of injury, and the patient’s overall
condition. The options for nonoperative and operative treatment are
extensive, and the correct choice for any patient must be determined by
weighing all the above considerations, as well as the surgeon’s
experience, against the potential risks of treatment.