Extract
Management of thoracolumbar and sacral spinal fractures is one of the most
controversial areas in modern spinal surgery. Early fusion with
instrumentation is a generally accepted treatment method for patients with
clearly unstable injuries and a complete neurological deficit; it results in
more rapid mobilization, fewer complications due to prolonged recumbency, and
lower medical costs. The optimal treatment for patients with mild-to-moderate
deformity, an incomplete neurological deficit, and residual spinal canal
compromise remains largely unknown. A review of the literature revealed a wide
range of conflicting results and recommendations, and the vast majority of the
clinical studies can be criticized on the basis of retrospective design,
heterogeneous patient populations and treatment strategies, limited follow-up,
and poorly defined outcome measures.
Management of thoracolumbar and sacral spinal fractures is one of the most
controversial areas in modern spinal surgery. Early fusion with
instrumentation is a generally accepted treatment method for patients with
clearly unstable injuries and a complete neurological deficit; it results in
more rapid mobilization, fewer complications due to prolonged recumbency, and
lower medical costs. The optimal treatment for patients with mild-to-moderate
deformity, an incomplete neurological deficit, and residual spinal canal
compromise remains largely unknown. A review of the literature revealed a wide
range of conflicting results and recommendations, and the vast majority of the
clinical studies can be criticized on the basis of retrospective design,
heterogeneous patient populations and treatment strategies, limited follow-up,
and poorly defined outcome measures.
Mechanical failure of the spinal column following high-energy trauma
frequently occurs at the thoracolumbar junction as a result of its
transitional anatomy and biomechanical environment. The most common fracture
patterns at the thoracolumbar junction include compression fractures, burst
fractures, flexion-distraction injuries, and fracture-dislocations. These
injuries can be classified with use of either the anatomical three-column
model of spinal stability described by
Denis1
(Fig. 1) or the mechanistic
classification system of Ferguson and
Allen2. In the
anatomical three-column model described by Denis, the anterior column includes
the anterior longitudinal ligament and the anterior half of the vertebral body
and the anulus fibrosus, the middle column includes the posterior half of the
vertebral body and the anulus fibrosus along with the posterior longitudinal
ligament, and the posterior column includes the bone and ligamentous
structures posterior to the posterior longitudinal ligament.
With the three-column model, thoracolumbar fractures are differentiated on
the basis of the pattern of injury to the middle column. Compression fractures
involve failure of the anterior column in compression without injury to the
middle column, whereas burst fractures result in compression failure of both
the anterior and the middle column. Injury to the middle column is considered
to be a potentially unstable fracture pattern in this classification scheme.
Failure in distraction is characteristic of Chance fractures and Chance
variants, whereas any translation or rotation through the middle column
indicates a high degree of instability, characteristic of a rotational burst
injury or a fracture-dislocation.
Compression fractures at the thoracolumbar junction result from an axial
loading force acting on a flexed spine. The anterior spinal column fails in
compression, while the middle column remains intact. The posterior column may
remain intact or fail in tension, depending on the energy level of the injury.
The integrity of the posterior ligamentous structures making up the posterior
spinal column is the primary determinant of spinal stability in this fracture
pattern.
The vast majority of spinal compression fractures occur between T11 and L2.
The thoracolumbar junction appears to be uniquely vulnerable as a result of
its transitional anatomy and biomechanical environment. The rib cage and its
attaching radiate and costotransverse ligaments stabilize the thoracic spine
and provide increased resistance to bending moments in the sagittal and
coronal planes as well as in axial
rotation3,4.
This level of protection and relative stiffness is in sharp contrast to the
more flexible and relatively unprotected subjacent lumbar spine. The frontal
or coronal orientation of the thoracic spinal facet joints further constrains
mobility in the flexion-extension plane compared with the increased motion in
flexion-extension allowed by the sagittally oriented lumbar facet joints. The
natural kyphosis of the thoracic spine and lordosis of the lumbar spine also
serve to absorb and dissipate axial loads, whereas the straighter
thoracolumbar junction provides less shock-absorbing capacity or
potential.
Motor-vehicle accidents and falls from a height are responsible for most
thoracolumbar compression fractures in young and middle-aged
adults5,6.
Sports and recreational activities are also the cause of a large number of
these injuries, particularly in children and
adolescents7,8.
In the elderly population, osteoporotic compression fractures resulting from
low-energy injuries are more
common9.
A thoracolumbar compression fracture can be readily diagnosed on plain
radiographs and with computed tomography. The initial evaluation, however,
must rule out a potentially more unstable burst fracture. On lateral plain
radiographs, a typical compression fracture is seen as a wedge-shaped vertebra
with loss of the height of the anterior body and preservation of the height of
the posterior body (Fig. 2).
Measurements should be made at the level of injury and compared with those at
the more cephalad and caudad levels. Any posterior cortical disruption seen on
a lateral plain radiograph or interpedicular widening seen on an
anteroposterior radiograph suggests a possible burst fracture.
The posterior vertebral angle can be measured by drawing lines parallel to
the vertebral end plates and posterior vertebral cortex of the fractured
vertebra. Some investigators have reported that this angle is useful for
distinguishing compression from burst fractures and have suggested that any
measurement of >100° indicates a potentially unstable burst
fracture10.
However, it has been shown that relying solely on plain radiographs for
evaluation of these injuries can lead to a diagnostic error rate of up to
25%11. Often the
posterior vertebral angle is difficult to visualize and therefore to measure
on plain radiographs. When a burst fracture is suspected, computed tomography
is the imaging modality of choice for demonstrating the integrity of the
posterior vertebral column.
Denis described four different types of compression fractures
(Fig.
3)1. A
type-A fracture involves failure of both the superior and the inferior
vertebral end plate, type B involves failure of the superior vertebral end
plate, type C involves failure of the inferior vertebral end plate, and type D
involves failure of the central vertebral body with less involvement of the
vertebral end plates. Type B has been the most commonly noted compression
fracture pattern in large observational studies.
Following formulation of the diagnosis and treatment, plain radiographs are
optimal for assessing injury severity and following the evolution of any
deformity over time. Both the percentage loss of vertebral height and the
kyphosis angle should be recorded initially and at each follow-up evaluation.
The percentage loss of vertebral height is determined by dividing the anterior
vertebral height by the posterior vertebral height and subtracting the derived
value from one. Studies have shown that the most accurate and reproducible
measurement of local kyphosis is the Cobb angle, which is the angle subtended
by lines drawn perpendicular to the superior end plate of the vertebral level
cephalad to the fracture and the inferior end plate of the vertebral body
caudad to it12.
These radiographic landmarks provide the highest interobserver and
intraobserver reproducibility and have the advantage of allowing simultaneous
assessment of the progressive angulation through the adjacent disc spaces.
Although the integrity of the middle column is widely considered to be the
primary determinant of spinal stability, compression fractures may be unstable
in the presence of major disruption of the posterior ligamentous structures
making up the posterior
column13. Injury to
the posterior interspinous and supraspinous ligaments may allow progressive
vertebral wedging to occur. With increasing kyphosis, a greater moment arm is
created, further exposing the anterior spinal column to axial loads and
potentially leading to more anterior vertebral body collapse and kyphosis.
The deformity associated with progressive kyphosis may eventually result in
substantial functional impairment and, in extreme cases, neurological
impairment. The association between thoracolumbar kyphosis and back pain,
however, remains unclear. Although many studies have failed to identify a
direct relationship, a majority of investigators believe that marked kyphosis
is a cause of back pain. When chronic back pain occurs, it tends to be
localized to the apex of the kyphotic segment or more caudad, possibly
reflecting compensatory hyperlordosis of the lower lumbar spine. Occasionally,
delayed progression of the kyphotic deformity and worsening back pain develop
in association with increased radiodensity of the apical vertebral body on
plain radiographs. This phenomenon has been attributed to posttraumatic
osteonecrosis, also known as Kümmell disease, and may be
associated with substantial back pain and, in the setting of canal occlusion,
a neurological
deficit14,15.
Thoracolumbar burst fractures are usually caused by a substantial axial
loading force that results in compression failure of the anterior and middle
spinal columns. As is the case for compression fractures, falls from a height
and motor-vehicle accidents are responsible for the majority of these
injuries16-18.
The sudden application of a supraphysiological axial load results in vertebral
end-plate failure as adjacent disc tissue is driven into the vertebral body.
An earlier theory that axial loading creates a sudden increase in internal
vertebral body pressure that results in a burst fracture appears to be
incorrect19. Like
compression fractures, burst fractures have a predilection for the
thoracolumbar (T11-L2) spinal segments.
The system that is most commonly utilized for classification of
thoracolumbar burst fractures is that described by Denis
(Fig.
4)1.
According to this system, a type-A fracture involves failure of both the
superior and the inferior end plate, type B involves failure of the superior
end plate only, type C involves failure of the inferior end plate only, type D
results in an axial loading and rotational injury, and type E results in an
axial loading and lateral flexion injury. Type B is the most frequent fracture
pattern, followed by type A. The other types are relatively rare.
The vast majority of burst fractures are associated with some degree of
canal compromise, typically as a result of retropulsion of an osseous fragment
or fragments from the superior end plate. The initial radiographic evaluation
should include assessment of loss of vertebral body height and the kyphosis
angle on lateral radiographs as well as widening of the interpedicular
distance on anteroposterior radiographs (Figs.
5-A,
5-B, and
5-C). Computed tomography
imaging is useful for demonstrating the extent of canal compromise. Magnetic
resonance imaging is recommended for patients with a neurological deficit, to
identify possible spinal cord or cauda equina injury, hemorrhage, or epidural
hematoma.
Once the fracture and any associated neurological injuries have been
characterized, spinal stability must be evaluated. The often quoted definition
of clinical instability by White and Panjabi is "the loss of the ability
of the spine under physiologic loads to maintain its pattern of displacement
so that there is no initial or additional neurological deficit, no major
deformity, and no incapacitating
pain."4
Unfortunately, a useful translation of this definition into practical clinical
guidelines has been elusive. Several different systems for determining
clinical spinal stability, including complicated point systems and checklists,
have been proposed, but all have proven too cumbersome for popular
use20-22.
Determinants of burst fracture instability common to these systems include a
progressive neurological deficit, progressive kyphosis, radiographic evidence
of substantial posterior column injury, and >50% loss of vertebral body
height in association with substantial kyphosis. Although many different
absolute numbers have been proposed for defining the percentage loss of height
that signifies an unstable fracture, 50% was popularized following a clinical
study suggesting that, when patients with this percentage of height loss are
treated nonoperatively, late progression of kyphosis and spinal stenosis tend
to develop22.
Flexion-distraction, or so-called Chance, fractures occur as a result of a
primary anterior force vector acting along an axis of rotation located
anterior to the middle column. The posterior and middle columns fail in
tension, whereas the anterior column may fail in either tension or compression
depending on the location of the axis of rotation (at or anterior to the
anterior spinal column). This injury pattern is typically observed following
high-speed motor-vehicle accidents in which the lap belt was used without the
shoulder belt. It is also commonly seen in young children who are too small
for the seat belt. An extremely high rate of intra-abdominal injury (45%) has
been observed in association with this injury
pattern23. Because
hollow viscus injury may not appear on initial computed tomography imaging,
there should be a high index of suspicion when this fracture pattern is seen
in order to exclude the possibility of intraabdominal pathology, especially in
the acute postinjury period. Overall, the risk of substantial neurological
injury in association with a flexion-distraction injury is 10% to
15%24.
When a patient has a flexion-distraction injury and a neurological injury,
magnetic resonance imaging is useful for identifying ongoing neural element
compression and to rule out an epidural hematoma. A computed tomography scan
with sagittal reconstructions is recommended to confirm the fracture pattern
and to ascertain that there is no comminution of the middle spinal column.
This distinction is important when planning surgical treatment, as the
compression forces that would be necessary to reduce a flexion-distraction
injury may be contraindicated if the middle column is disrupted in order to
prevent middle column retropulsion. The magnetic resonance images and computed
tomography scans with sagittal reconstruction are also valuable for
delineating the path of injury (through bone or soft tissue), as this
determines the likelihood of the injury healing in a brace
(Fig. 6). The selection and
timing of treatment depend in part on the presence of associated
intraabdominal injuries as well as on the neurological profile.
Fracture-dislocations are the result of a violent complex shearing force
and can occur anywhere along the thoracolumbar spine. Most series, however,
have demonstrated a predilection for the thoracolumbar junction. These
injuries are extremely unstable and, by definition, involve disruption of all
three spinal columns. The rate of complete neurological injury is the highest
with this injury subtype.
Fracture-dislocations are usually readily recognized on plain radiographs.
Any horizontal translation or rotation through the injury level should raise
the suspicion that a fracture-dislocation is present
(Figs. 7-A and 7-B). Computed
tomography scanning is useful for planning surgical treatment as it identifies
incompetent osseous structures and aids in the selection of the implant size.
In the less common case of an incomplete neurological injury, magnetic
resonance imaging should be performed to characterize ongoing neural element
compression and the nature of injury of the spinal cord and cauda equina.
Advances in prehospital care have substantially improved the short and
long-term outcomes for patients with a thoracolumbar spine fracture.
Extrication and transportation of trauma patients on a backboard and adherence
to Advanced Trauma Life Support (ATLS) protocols for resuscitation have been
credited for much of this improvement. A spine fracture should be suspected in
any patient who has sustained high-energy trauma, and the patient should be
treated accordingly. Examination of a patient with a possible spine fracture
should begin with visual inspection of the back. The presence and location of
lacerations, abrasions, ecchymoses, and swelling provide clues about the
mechanism of injury. Palpation of the spine for localized tenderness,
step-offs, spinous process gaps, and malalignment may provide evidence of
substantial injury and spinal instability.
The neurological examination should include assessment of spinal cord
function as well as assessment of nerve root and peripheral nerve integrity.
The spinal cord terminates most commonly at the inferior border of the L1
vertebral body in adults, although it may extend caudad to L2 in some people.
As a result, injuries to the thoracolumbar spine can present with a varied
clinical picture of neurological deficits arising from damage to the terminal
spinal cord, conus medullaris, cauda equina, and thoracolumbar nerve roots.
Radiculopathy is identified by a dermatomal pattern of paresthesias or sensory
alterations with or without myotomal weakness and hyporeflexia. A more diffuse
distribution of lower-extremity paresthesias, weakness, and reflex loss may
signify acute injury to the cauda equina, conus medullaris, or spinal cord.
When there is a potential spinal cord injury, the bulbocavernosus reflex
should be evaluated to assess for spinal shock. In the absence of this reflex,
loss of sensorimotor function may be temporarily due to spinal shock and may
not necessarily reflect a complete neurological injury. The presence of rectal
tone and perineal pinprick sensation in a patient with a substantial
neurological deficit is vital information. It indicates integrity of at least
some pathways within the spinal cord passing through the level of injury and
represents an incomplete cord injury, with a dramatically improved
prognosis.
All patients with a confirmed spinal cord injury should receive a high dose
of methylprednisolone intravenously, beginning with a bolus of 30 mg/kg over
one hour and continued at an infusion at a rate of 5.4 mg/kg/hr. The second
National Acute Spinal Cord Injury Study
(NASCIS)25, a
prospective, randomized multicenter trial, demonstrated that this intervention
had a substantial benefit in terms of ultimate neurological recovery. Results
from the third NASCIS study suggested that patients started on
methylprednisolone within three hours after the injury should receive the drug
for a total of twenty-four hours, whereas patients who begin receiving
methylprednisolone between three and eight hours after the injury should be
treated for a total of forty-eight
hours26. Recently,
investigators have questioned the validity of these pharmacological
investigations.
The two most commonly used systems for grading spinal cord injury are the
Frankel system (Table
I)5 and
the more comprehensive system developed by the American Spinal Injury
Association (ASIA) (Fig.
8)27.
Grading of the spinal cord injury plays a major role in determining treatment
and ultimate prognosis for many patients following thoracolumbar spinal
trauma.
Anteroposterior and lateral radiographs of the cervical, thoracic, and
lumbar spine are standard imaging studies following spinal trauma. Because of
the high prevalence of contiguous and noncontiguous associated spinal
fractures, comprehensive radiographic evaluation, including the entire
cervical, thoracic, lumbar, and sacral spine, is recommended for any patient
who has sustained a high-energy injury and in whom a spinal injury is
suspected. Specific injury mechanisms and fracture patterns should trigger a
targeted search for commonly associated nonspinal injuries. For example, the
association between burst fractures and calcaneal or tibial plateau fractures
following a fall from a height has been well established. Similarly, Chance
fractures or flexion-distraction Chance variants are strongly associated with
potentially life-threatening intraabdominal injuries.
Computed tomography scanning is generally the next step after plain
radiographic evaluation. Axial fine cuts and sagittal reconstructions are
helpful for determining fracture patterns and the degree of compromise of the
spinal canal. The canal at the injured segment should be measured in the
anteroposterior and transverse planes and compared with the levels cephalad
and caudad to it. One of us (A.R.V.) and colleagues found the most clinically
useful measurement to be the ratio of the sagittal to the transverse canal
diameter18. In that
study, a smaller midsagittal diameter and a greater transverse diameter (a
widened interpedicular distance suggests higher-energy injury) correlated with
an increased risk of neurological deficit.
In the absence of neurological injury, magnetic resonance imaging scans
usually are not required for thoracolumbar injuries in the acute setting. They
can occasionally be helpful for identifying a ligamentous lesion that is
suspected but not confirmed on plain radiographs and computed tomography scans
(Fig. 9). The presence of gas
in the posterior subcutaneous tissue or within the spinal elements suggests a
possible flexion-distraction injury (Figs.
10-A and 10-B). When a patient has a neurological deficit,
however, magnetic resonance imaging is recommended to identify any ongoing
spinal cord compression, evaluate cord anatomy, and rule out an epidural
hematoma.
The most important factors to consider when deciding on treatment for
patients with a thoracolumbar spine fracture are neurological status, spinal
stability, degree of deformity, and associated injuries. Most compression
fractures are relatively stable injuries conducive to nonoperative treatment
in a thoracolumbar orthosis for approximately twelve weeks. Care must be taken
to rule out a potentially more unstable burst fracture and to confirm the
integrity of the posterior column. Burst fractures are typically more unstable
than compression fractures and, by definition, involve an injury to the middle
column. There is often some degree of canal compromise, and the risk of
neurological injury is correspondingly greater.
The extent of collapse and kyphosis as well as the integrity of the
posterior column are key determinants of stability and should be considered
when deciding between surgical and nonoperative treatment. Progressive
neurological deterioration in the presence of substantial canal compromise is
an indication for surgical decompression and stabilization. The treatment of
flexion-distraction injuries largely depends on the predominant type of tissue
injury. Pure Chance fractures in which disruption occurred through the bone of
the vertebral body, pedicles, laminae, and spinous process often will heal
reliably if immobilized in a hyperextension orthosis, especially in an
immature patient. A soft-tissue Chance variant with an injury vector passing
through the disc space, facet capsules, and interspinous ligament will not
heal predictably in an adult and often requires operative stabilization.
Radiographic evidence of substantial translational or rotational malalignment
indicates a highly unstable shear injury or fracture-dislocation requiring
surgical stabilization.
Nonoperative Compared with Operative Treatment
The principal advantage of nonoperative treatment is the avoidance of
operative morbidity, including postoperative infection, iatrogenic
neurological injury, pseudarthrosis, failure of instrumentation, and
complications related to anesthesia. Moreover, when the type of treatment is
being decided, it must be kept in mind that many studies have failed to reveal
a substantial difference in functional outcome between operative and
nonoperative treatment of thoracolumbar spine fractures, regardless of the
presence of neurological
injury28-35.
Ultimately, the treatment goals are identical regardless of whether
surgical or nonoperative modalities are chosen. Preeminent considerations are
prevention and limitation of neurological injury as well as restoration of
spinal stability. Secondary issues include deformity correction, minimizing
motion loss, and facilitating rapid rehabilitation. The aim of treatment
should be to provide a biological and biomechanical environment conducive to
osseous and soft-tissue healing in order to recreate a stable pain-free spinal
column. In the end, these goals should be accomplished with the introduction
of as little additional risk or morbidity as possible.
Although most studies have not supported the concept that surgery leads to
greater neurological improvement than does nonoperative treatment, most
investigators have recommended surgery for patients with progressive
neurological loss or a major neurological deficit in the setting of
substantial canal compromise. In these situations, the goals of surgery are to
provide adequate anatomic decompression of the neural elements and rigid
stabilization of the injured segments until biological fusion occurs. The
appropriate timing of surgery following spinal cord injury is the subject of
considerable controversy with conflicting information from animal and clinical
studies. Although animal models have suggested that early decompression may
improve neurological recovery, the window of opportunity appears to be very
small, and the only prospective, randomized study of the timing of surgery for
spinal cord injury of which we are aware revealed no substantial difference in
outcome between early and late surgical decompression and
stabilization36.
As previously discussed, the concept of spinal instability has not been
well translated from the theoretical definition provided by White and
Panjabi4 into a more
usable practical guideline. Various models (classification systems) for
conceptualizing thoracolumbar spinal injury have been introduced. These
include Holdsworth's initial two-column
model37, Denis's
three-column model1,
and the "neutral zone" paradigm of Panjabi et
al.38. Although the
two-column model is one of the earliest models of spinal stability, there has
been a resurgence of support for it as a result of both biomechanical and
clinical studies suggesting that the importance of the middle column in terms
of long-term stability has been overvalued. Holdsworth's model defines only
two spinal columns—an anterior column consisting of the anterior
longitudinal ligament, vertebral body, intervertebral disc, anulus fibrosus,
and posterior longitudinal ligament and a posterior column consisting of the
facet joint complex, ligamentum flavum, and interspinous and supraspinous
ligaments. Cadaveric models of burst fractures have provided evidence
supporting Holdsworth's model by demonstrating that integrity of the posterior
spinal column provides the greatest resistance to progressive kyphosis and
that little additional stability is provided by the middle
column39. However,
further clouding the stability issue is a study suggesting that a posterior
ligament injury does not have a negative impact on the outcome of nonoperative
treatment of thoracolumbar fractures with bracing and early
mobilization40.
The neutral zone theory described by Panjabi et
al.38 emphasizes
the continuum between stability (with physiological motion), hypermobility
(associated with normal degenerative changes), and frank instability, with all
patterns possible following a traumatic episode. This concept has come to
dominate the thinking behind laboratory investigations of instability but has
yet to be applied in the clinical setting.
The relationship between posttraumatic kyphotic deformity and chronic back
pain is similarly ambiguous. Although most clinicians believe that kyphotic
deformity of the thoracolumbar junction portends a poor clinical outcome, few
studies have provided convincing evidence that moderate kyphosis is associated
with either pain or
disability41. On
the other hand, several studies have suggested that there is no direct
relationship between kyphosis and back pain or functional
impairment13,29,31,32,34,35.
Recently, the importance of so-called radiographically invisible
soft-tissue injury as a determinant of long-term pain and functional outcome
has reemerged as a possible explanation for the diversity of clinical results
following radiographically similar injury
patterns42,43.
Specifically, traumatic insult to the discovertebral complex and the posterior
facet joints has been blamed for the development of chronic axial back pain as
well as for progressive kyphosis in the absence of progressive vertebral
compression.
As a result of the foregoing clinical uncertainties, nonoperative treatment
must be considered the preferred method for thoracolumbar fractures in the
absence of neurological injury or clear evidence of instability. In
neurologically intact patients, determination of the long-term stability of a
fracture is paramount. The integrity of the posterior column is often the
decisive factor, and, if the results of the physical examination and the
preliminary radiographs are indeterminate, magnetic resonance imaging may be
useful. If stability remains in doubt, a trial of nonoperative bracing with
careful follow-up to identify any progressive kyphosis is a reasonable
option.
The presence of a substantial neurological deficit suggests spinal
instability sufficient to have resulted in injury to the neural elements at
the time of the traumatic episode. Some clinicians regard this as sufficient
evidence of instability to warrant surgical stabilization, although this
opinion remains controversial. Computed tomography or magnetic resonance
imaging findings of ongoing canal compromise in the presence of a neurological
deficit, however, provide a much stronger incentive for surgical decompression
and stabilization. There is general consensus that progressive kyphosis and
progressive neurological loss with evidence of canal compromise due to bone,
soft tissue, or an epidural hematoma are strong indications for urgent
surgical intervention.
Anterior and posterior as well as combined anteroposterior surgical
approaches have specific advantages and disadvantages. Many compression and
burst fractures are amenable to reconstruction through an anterior
thoracoabdominal, a retroperitoneal, or possibly an endoscopic approach,
depending on the level of the
injury44,45.
Anterior-only reconstruction appears to be adequate treatment for most
compression and burst fractures limited to the anterior and middle
columns—i.e., without substantial injury to the posterior column (Figs.
11-A,
11-B, and 11-C). A
posterior-only fusion with instrumentation from two or three levels cephalad
to the injury to two levels caudad to it is an acceptable alternative in
patients who do not require decompression, but it involves fusing a more
extensive portion of the spine. A so-called posterior short-segment fusion
with instrumentation and utilization of pedicle screw fixation from one level
cephalad to the injury to one level caudad to it has fallen out of favor
because of a relatively high rate of proximal screw pullout and loss of
correction in the setting of substantial anterior column compromise. In the
presence of substantial anterior collapse and posterior column disruption, a
combined anteroposterior fusion often provides the greatest degree of
iatrogenic spinal stability. Because of the nature of instability,
flexion-distraction injuries and fracture-dislocations are best approached
from the back initially, with posterior stabilization, followed by an anterior
decompression and reconstruction if needed.
In the setting of a complete neurological injury, surgery is recommended to
provide immediate spinal stability, obviating the need for a cumbersome
orthosis and facilitating more rapid rehabilitation. In this scenario, surgery
is typically performed through a posterior approach with the goal of restoring
sagittal and coronal alignment and stability. Limiting the number of lower
lumbar levels fused optimizes wheelchair mobility.
One of the most common clinical scenarios is a compression or burst
fracture in a neurologically intact patient with an initial local kyphosis of
25° to 35°. Treatment of such patients should be highly
individualized. Several factors should be considered when deciding on a
treatment plan. The patient's age, general health, occupation and lifestyle,
and body habitus are integral to obtaining a satisfactory outcome. Some
investigators have recommended more aggressive surgical treatment for younger
patients, with the rationale that fractures in strong healthy bone suggest a
higher-energy injury and greater potential for instability. Nonoperative
treatment is usually the optimal initial choice in borderline cases, given the
ill-defined advantages of surgery in these patients. However, close followup
is mandatory, and evidence of progressive deformity or new-onset neurological
deterioration should be addressed by considering surgery.
Issues in Surgical Treatment
Neural Decompression
Spinal canal decompression may appear intuitively desirable in the setting
of neural compression and a neurological deficit, but most clinical studies
have not demonstrated objective benefits of this approach in well-matched
patient populations. Although the extent of canal compromise at the time of
the initial evaluation has been associated in some studies with the severity
of the neurological deficit, the timing of surgical decompression has not been
well correlated with the degree of neurological
recovery16,46.
Moreover, following either nonoperative treatment or surgical decompression,
the amount of remaining canal compromise typically decreases by >50%
through a process of gradual
remodeling47,48.
However, despite continued controversy, most investigators agree that
progressive neurological deterioration in the setting of substantial canal
compromise is an indication for surgical intervention. Many surgeons also
recommend surgical decompression for a patient with a stable incomplete
neurological deficit when there is evidence of ongoing compression of the
neural elements. In the vast majority of these patients, the site of the
neural compression is located anteriorly, which dictates the site of surgical
exposure. A posterior surgical approach may be used in certain cases to
indirectly decompress the spinal canal through distraction instrumentation and
ligamentotaxis. A laminectomy alone further disrupts the posterior supporting
spinal elements, contributing to additional instability at the fracture site,
and may result in lower rates of neurological recovery compared with those
following anterior decompression and fusion or posterior decompression and
stabilization49. In
rare cases, a costotransversectomy or transpedicular approach may be employed
to gain access to compressing anterior bone fragments through the posterior
approach46.
However, limited visualization and the occasional need to manipulate or
retract the neural elements make these techniques suboptimal for most patients
with a thoracolumbar traumatic injury.
The ligamentotaxis effect of a posterior indirect decompression is based on
the integrity of Sharpey fibers or anular ligament attachments to the
displaced fracture fragments. Application of a distraction force by means of
posterior spinal instrumentation results in an anteriorly directed force on
the displaced osseous fragments. Careful attention must be paid to sagittal
alignment, as posterior distraction also tends to aggravate any preexisting
kyphosis. The efficacy of indirect decompression is greater when surgery is
performed within three days after the traumatic
event50,51.
This technique may be less effective in the setting of canal compromise of
>67%, which is associated with a higher prevalence of anular ligament
disruption52. One
study suggested that motor recovery and return of bowel and bladder function
after incomplete neurological injury is more reliable following an anterior
decompression53,
but several other studies have failed to reveal any substantial difference in
ultimate functional outcome between anterior direct and posterior indirect
decompression
methods54,55.
Stabilization
The advantages of early surgical stabilization of unstable fractures have
been well established in terms of improved fracture reduction, preservation of
neurological function, early mobilization, and fewer complications associated
with prolonged bed
rest56. Early
fusion techniques incorporating Harrington hook and rod constructs were
suboptimal because of the required length of the fusion (five or six motion
segments) and less reliable fixation in the middle and lower lumbar
regions57,58.
The so-called rod long-fuse short technique was developed in an attempt to
reduce the number of motion segments requiring fusion with distraction
instrumentation. This method involved fusing only the two levels adjacent to
the injured vertebra while spanning additional levels cephalad and caudad to
the instrumentation, which was eventually removed after a year. Unfortunately,
the outcomes were poor as a result of accelerated arthritis in the immobilized
but unfused segments, progressive kyphosis following rod removal, and an
increased prevalence of late back pain; thus, this technique has fallen out of
favor56,59-61.
The improved rigidity and versatility of modern pedicle screw-based systems
allow more reliable fixation while fusing fewer motion segments. Posterioronly
fusion techniques often incorporate only two levels cephalad to and two caudad
to the fracture level and, in appropriately selected cases, should provide
acceptable long-term sagittal alignment and clinical
results62. As
previously mentioned, short-segment posterior fusion with pedicle screw
instrumentation (from a level cephalad to the fracture site to a level caudad
to it) is typically associated with a high rate of early hardware failure and
late loss of sagittal plane
correction63-65.
The development of techniques and instrumentation systems for anterior spinal
reconstruction has allowed anterior short-segment fusion to emerge as a
preferable alternative to posterior short-segment fusion for selected
fractures with substantially greater stability under most loading
conditions66. The
exception is a flexion-distraction injury with an intact anterior
osteoligamentous hinge. In that case, posterior short-segment instrumentation
under compression provides excellent stabilization and a reliable fusion
(Figs. 12-A and
12-B). Fracture-dislocations
are highly unstable injuries requiring more extensive fusion procedures,
typically spanning three levels cephalad to the injury level to two or three
levels caudad to it.
The transitional anatomy of the thoracolumbar spine renders it uniquely
vulnerable to the high-energy deceleration trauma associated with falls and
motor-vehicle collisions. Although the predominant injury patterns and
overriding management principles have been well defined, specific treatment
recommendations remain the subject of considerable controversy. The primary
objectives of initial evaluation and diagnosis include characterization of the
injury and identification of any neurological deficit. Spinal stability
remains a poorly defined concept but is of paramount importance in terms of
determining appropriate treatment. Whether surgical or nonoperative treatment
is selected, the ultimate therapeutic goals are identical and consist of
preservation of neurological function and restoration of spinal stability
while incurring minimal additional morbidity and cost.
DenisF.
The three column spine and its significance in the classification of acute
thoracolumbar spinal injuries. Spine.1983;8:
817-31.8817
1983
[PubMed][CrossRef]
FergusonRL,
Allen BL Jr. A mechanistic classification of thoracolumbar spine
fractures. Clin Orthop.1984;189:
77-88.18977
1984
[PubMed]
AndriacchiTP,
Schultz A, Belytschko T, Galante J. A model for studies of mechanical
interactions between the human spine and rib cage. J
Biomech.1974;7:
497-507.7497
1974
[CrossRef]
WhiteAA 3rd,
Panjabi MM.Clinical biomechanics of the spine.
Philadelphia: Lippincott; 1990.
1990
FrankelHL,
Hancock DO, Hyslop G, Melzak J, Michaelis LS, Ungar GH, Vernon JA, Walsh
JJ. The value of postural reduction in the initial management of closed
injuries of the spine with paraplegia and tetraplegia.
Paraplegia.1969;7:
179-92.7179
1969
[PubMed][CrossRef]
DicksonJH,
Harrington PR, Erwin WD. Results of reduction and stabilization of the
severely fractured thoracic and lumbar spine. J Bone Joint Surg
Am.1978;60:
799-805.60799
1978
HubbardD.
Injuries of the spine in children and adolescents. Clin
Orthop.1974;100:
56-65.10056
1974
KeeneJS.
Thoracolumbar fractures in winter sports. Clin Orthop.1987;216:
39-49.21639
1987
[PubMed]
KimDH, Silber
JS, Albert TJ. Osteoporotic vertebral compression fractures.
Instr Course Lect.2003;52:
541-50.52541
2003
[PubMed]
McGroyBJ,
VanderWilde RS, Currier BL, Eismont FJ. Diagnosis of subtle thoracolumbar
burst fractures. A new radiographic sign. Spine.1993;19:
2282-5.192282
1993
[CrossRef]
BallockRT,
Mackersie R, Abitbol JJ, Cervilla V, Resnick D, Garfin SR. Can burst
fractures be predicted from plain radiographs? J Bone Joint Surg
Br.1992;74:
147-50.74147
1992
KukloTR, Polly
DW, Owens BD, Zeidman SM, Chang AS, Klemme WR. Measurement of thoracic and
lumbar fracture kyphosis: evaluation of intraobserver, interobserver, and
technique variability. Spine.2001;26:
61-6.2661
2001
[PubMed][CrossRef]
NicollEA.
Fractures of the dorso-lumbar spine. J Bone Joint Surg
Br.1949;31:
376-94.31376
1949
BenedekTG,
Nicholas JJ, Reece GJ. Kümmell's disease: a rare cause of
posttraumatic back pain. Arthritis Rheum.1980;23:
653.23653
1980
BrowerAC,
Downey EF Jr. Kummell disease: report of a case with serial radiographs.
Radiology.1981;141:
363-4.141363
1981
[PubMed]
KimNH, Lee HM,
Chun IM. Neurologic injury and recovery in patients with burst fracture of
the thoracolumbar spine. Spine.1999;24:
290-4.24290
1999
[PubMed][CrossRef]
KnopC, Fabian
HF, Bastian L, Blauth M. Late results of thoracolumbar fractures after
posterior instrumentation and transpedicular bone grafting.
Spine.2001;26:
88-99.2688
2001
[PubMed][CrossRef]
VaccaroAR,
Nachwalter RS, Klein GR, Sewards JM, Albert TJ, Garfin SR. The
significance of thoracolumbar spinal canal size in spinal cord injury
patients. Spine.2001;26:
371-6.26371
2001
[PubMed][CrossRef]
OchiaRS, Ching
RP. Internal pressure measurements during burst fracture formation in
human lumbar vertebrae. Spine.2002;27:
1160-7.271160
2002
[PubMed][CrossRef]
PanjabiMM,
Hausfeld JN, White AA 3rd. A biomechanical study of the ligamentous
stability of the thoracic spine in man. Acta Orthop
Scand.1981;52:
315-26.52315
1981
[CrossRef]
LouisR.
Spinal stability as defined by the three-column spine concept. Anat
Clin.1985;7:
33-42.733
1985
[CrossRef]
McAfeePC, Yuan
HA, Lasda NA. The unstable burst fracture. Spine.1982;7:
365-73.7365
1982
[PubMed][CrossRef]
AndersonPA,
Rivera FP, Maier RV, Drake C. The epidemiology of seatbelt--associated
injuries. J Trauma.1991;31:
60-7.3160
1991
[PubMed][CrossRef]
GumleyG,
Taylor TK, Ryan MD. Distraction fractures of the lumbar spine.
J Bone Joint Surg Br.1982;64:
520-5.64520
1982
[PubMed]
BrackenMB,
Shepard MJ, Collins WF, Holford TR, Young W, Baskin DS, Eisenberg HM, Flamm E,
Leo-Summers L, Maroon J, Marshall LF, Perot PL, Piepineier J, Sonntag VK,
Wagner FC, Willberger JE, Winn HR. A randomized, controlled trial of
methylprednisolone or naloxone in the treatment of acute spinal-cord injury.
Results of the Second National Acute Spinal Cord Injury Study. N
Engl J Med.1990;322:
1405-11.3221405
1990
[CrossRef]
BrackenMB,
Shepard MJ, Holford TR, Leo-Summers L, Aldrich EF, Fazl M, Fehlings M, Herr
DL, Hitchon PW, Marshall LF, Nockels RP, Pascale V, Perot PL Jr, Piepmeier J,
Sonntag VK, Wagner F, Wilberger JE, Winn HR, Young W. Administration of
methylprednisolone for 24 or 48 hours or tirilazad mesylate for 48 hours in
the treatment of acute spinal cord injury. Results of the Third National Acute
Spinal Cord Injury Randomized Controlled Trial. National Acute Spinal Cord
Injury Study. JAMA.1997;
277: 1597-604.2771597
1997
[PubMed][CrossRef]
FrankelHL,
Hancock DO, Hyslop G, Melzak J, Michaelis LS, Ungar GH, Vernon JD, Walsh
JJ. The value of postural reduction in the initial management of closed
injuries of the spine with paraplegia and tetraplegia. I.
Paraplegia.1969;7:
179-92.7179
1969
[PubMed][CrossRef]
BurkeDC,
Murray DD. The management of thoracic and thoraco-lumbar injuries of the
spine with neurological involvement. J Bone Joint Surg
Br.1976;58:
72-8.5872
1976
CantorJB,
Lebwohl NH, Garvey T, Eismont FJ. Nonoperative management of stable
thoracolumbar burst fractures with early ambulation and bracing.
Spine.1993;18:
971-6.18971
1993
[PubMed][CrossRef]
FredricksonBE,
Yuan HA, Bayley JC. The non-operative treatment of thoracolumbar injuries.
Semin Spine Surg.1990;2:
70-8.270
1990
KraemerWJ,
Schemitsch EH, Lever J, McBroom RJ, McKee MD, Waddell JP. Functional
outcome of thoracolumbar burst fractures without neurological deficit.
J Orthop Trauma.1996;10:
541-4.10541
1996
[PubMed][CrossRef]
MumfordJ,
Weinstein JN, Spratt KF, Goel VK. Thoracolumbar burst fractures. The
clinical efficacy and outcome of nonoperative management.
Spine.1993;18:
955-70.18955
1993
[PubMed][CrossRef]
RechtineGR
2nd, Cahill D, Chrin AM. Treatment of thoracolumbar trauma: comparison of
complications of operative versus nonoperative treatment. J Spinal
Disord.1999;12:
406-9.12406
1999
WeinsteinJN,
Collalto P, Lehmann TR. Long-term follow-up of nonoperatively treated
thoracolumbar spine fractures. J Orthop Trauma.1987;1:
152-9.1152
1987
[PubMed][CrossRef]
WeinsteinJN,
Collalto P, Lehmann TR. Thoracolumbar "burst" fractures
treated conservatively: a long-term follow-up. Spine.1988;13:
33-8.1333
1988
[PubMed][CrossRef]
VaccaroAR,
Daugherty RJ, Sheehan TP, Dante SJ, Cotler JM, Balderston RA, Herbison GJ,
Northrup BE. Neurologic outcome of early versus late surgery for cervical
spinal cord injury. Spine.1997;22:
2609-13.222609
1997
[PubMed][CrossRef]
HoldsworthF. Fractures, dislocations, and
fracture-dislocations of the spine. J Bone Joint Surg
Am.1970;52:
1534-51.521534
1970
PanjabiMM,
Oxland TR, Lin RM, McGowen TW. Thoracolumbar burst fracture. A
biomechanical investigation of its multidirectional flexibility.
Spine.1994;19:
578-85.19578
1994
[PubMed][CrossRef]
JamesKS,
Wenger KH, Schlegel JD, Dunn HK. Biomechanical evaluation of the stability
of thoracolumbar burst fractures. Spine.1994;19:
1731-40.191731
1994
[PubMed][CrossRef]
ChowGH, Nelson
BJ, Gebhard JS, Brugman JL, Brown CW, Donaldson DH. Functional outcome of
thoracolumbar burst fractures managed with hyperextension casting or bracing
and early mobilization. Spine.1996;21:
2170-5.212170
1996
[PubMed][CrossRef]
GertzbeinSD. Scoliosis Research Society. Multicenter spine
fracture study. Spine.1992;17:
528-40.17528
1992
[PubMed][CrossRef]
BaabOD.
Fractures of the dorsal and lumbar spine. Clin Orthop.1966;49:
195-200.49195
1966
[PubMed]
YoungMH.
Long-term consequences of stable fractures of the thoracic and lumbar
vertebral bodies. J Bone Joint Surg Br.1973;55:
295-300.55295
1973
[PubMed]
MiyakoshiN,
Abe E, Shimada Y, Hongo M, Chiba M, Sato K. Anterior decompression with
single segment spinal interbody fusion for lumbar burst fracture.
Spine.1999;24:
67-73.2467
1999
[PubMed][CrossRef]
OkuyamaK, Abe
E, Chiba M, Ishikawa N, Sato K. Outcome of anterior decompression and
stabilization for thoracolumbar unstable burst fractures in the absence of
neurologic deficits. Spine.1996;21:
620-5.21620
1996
[PubMed][CrossRef]
GarfinSR,
Mowery CA, Guerra J Jr, Marshall LF. Confirmation of the posterolateral
technique to decompress and fuse thoracolumbar spine burst fractures.
Spine.1985;10:
218-23.10218
1985
[PubMed][CrossRef]
DaiLY.
Remodeling of the spinal canal after thoracolumbar burst fractures.
Clin Orthop.2001;382:
119-23.382119
2001
[PubMed][CrossRef]
WessbergP,
Wang Y, Irstam L, Nordwall A. The effect of surgery and remodelling on
spinal canal measurements after thoracolumbar burst fractures. Eur
Spine J.2001;10:
55-63.1055
2001
[CrossRef]
BohlmanHH,
Freehafer A, Dejak J. The results of treatment of acute injuries of the
upper thoracic spine with paralysis. J Bone Joint Surg
Am.1985;67:
360-9.67360
1985
WillenJ,
Lindahl S, Irstam L, Nordwall A. Unstable thoracolumbar fractures. A study
by CT and conventional roentgenology of the reduction effect of Harrington
instrumentation. Spine.1984;9:
214-9.9214
1984
[PubMed][CrossRef]
CrutcherJP Jr,
Anderson PA, King HA, Montesano PX. Indirect spinal canal decompression in
patients with thoracolumbar burst fractures treated by posterior distraction
rods. J Spinal Disord.1991;4:
39-48.439
1991
[PubMed]
GertzbeinSD,
Crowe PJ, Fazl M, Schwartz M, Rowed D. Canal clearance in burst fractures
using the AO internal fixator. Spine.1992;17:
558-60.17558
1992
[PubMed][CrossRef]
BradfordDS,
McBride CG. Surgical management of thoracolumbar spine fractures with
incomplete neurologic deficits. Clin Orthop.1987;218:
201-16.218201
1987
[PubMed]
GertzbeinSD,
Court-Brown CM, Marks P, Martin C, Fazl M, Schwartz M, Jacobs RR. The
neurological outcome following surgery for spinal fractures.
Spine.1988;13:
641-4.13641
1988
[PubMed]
EssesSI,
Botsford DJ, Kostuik JP. Evaluation of surgical treatment for burst
fractures. Spine.1990;15:
667-73.15667
1990
[PubMed][CrossRef]
JacobsRR,
Asher MA, Snider RK. Thoracolumbar spinal injuries. A comparative study of
recumbent and operative treatment in 100 patients.
Spine.1980;5:
463-77.5463
1980
[PubMed][CrossRef]
AnHS, Vaccaro
A, Cotler JM, Lin S. Low lumbar burst fractures. Comparison among body
cast, Harrington rod, Luque rod, and Steffee plate.
Spine.1991;16 (8
Suppl): S440-4.16S440
1991
[PubMed][CrossRef]
SassoRC,
Cotler HB. Posterior instrumentation and fusion for unstable fractures and
fracture-dislocations of the thoracic and lumbar spine. A comparative study of
three fixation devices in 70 patients. Spine.1993;18:
450-60.18450
1993
[PubMed]
KahanovitzN,
Arnoczky SP, Levine DB, Otis JP. The effects of internal fixation of the
articular cartilage of unfused canine facet joint cartilage.
Spine.1984;9:
268-72.9268
1984
[PubMed][CrossRef]
DekutoskiMB,
Conlan ES, Salciccioli GG. Spinal mobility and deformity after Harrington
rod stabilization and limited arthrodesis of thoracolumbar fractures.
J Bone Joint Surg Am.1993;75:
168-76.75168
1993
[PubMed]
ChenWJ, Niu
CC, Chen LH, Chen JY, Shih CH, Chu LY. Back pain after thoracolumbar
fracture treated with long instrumentation and short fusion. J
Spinal Disord.1995;8:
474-8.8474
1995
BeenHD, Bouma
GJ. Comparison of two types of surgery for thoraco-lumbar burst fractures:
combined anterior and posterior stabilisation vs. posterior instrumentation
only. Acta Neurochir (Wien).1999;141:
349-57.141349
1999
[PubMed][CrossRef]
SassoRC,
Cotler JM, Reuben JD. Posterior fixation of thoracic and lumbar spine
fractures using DC plates and pedicle screws. Spine.1991;16 (3 Suppl):
S134-9.16S134
1991
[PubMed][CrossRef]
CarlAL,
Tromanhauser SG, Roger DJ. Pedicle screw instrumentation for thoracolumbar
burst fractures and fracture-dislocations. Spine.1992;17 (8 Suppl):
S317-24.17S317
1992
[PubMed][CrossRef]
McLainRF,
Sparling E, Benson DR. Early failure of short-segment pedicle
instrumentation for thoracolumbar fractures. A preliminary report.
J Bone Joint Surg Am.1993;75:
162-7.75162
1993
[PubMed]
ShonoY, McAfee
PC, Cunningham BW. Experimental study of thoracolumbar burst fractures. A
radiographic and biomechanical analysis of anterior and posterior
instrumentation systems. Spine.1994;19:
1711-22.191711
1994
[PubMed][CrossRef]