Thoracolumbar Spine Fractures and Dislocations

The patient’s history, physical examination, and imaging
studies may all be useful for understanding the mechanism of injury for
a specific thoracolumbar fracture. In these situations, the spinal
column may be subjected to either a single or more often a combination
of forces including axial, flexion, extension, shear, and rotational
moments; the injury pattern is largely determined by the overall
direction and magnitude of these vectors. This information is not only
obligatory for elucidating the pathogenesis of these injuries but it
may also be critical for assessing their stability and directing
subsequent treatment. Unfortunately, even with a detailed history and
supporting imaging studies, the mechanism of injury is often impossible
to determine precisely and is often debated among spinal care providers.

The application of axial loads to the flexed spine may
generate compression fractures in which there is disruption of the
anterior vertebra with sparing of the middle and posterior portions of
the body (Fig. 43-1). These end plate or
wedge-shaped fractures are usually considered to be stable, but any
evidence of damage to the posterior ligamentous tension band may be
indicative of a more serious injury. With more significant axial
forces, the fracture line may extend posteriorly through the entire
body, which is characteristic of a burst injury (Fig. 43-2).
By definition, these fractures are more unstable than compression
injuries and frequently bring about compression of the neural elements
secondary to the retropulsion of bony fragments into the spinal canal.
Although in the past burst fractures were attributed primarily to axial
loading of the spine, the authors of a recent in vitro biomechanical
study reported that at least some degree of extension was also required
to reproduce the interpedicular widening and canal compromise that are
regularly observed in conjunction with these injuries.⁸⁴

A flexion moment may yield a variety of fracture
patterns depending on the position of the rotational axis and the
anatomic structures through which it passes (i.e., bone or disc space) (Fig. 43-3).
When the center of rotation is located within the spinal column in
close proximity to the posterior longitudinal ligament, the vertebra
will undergo compression as the posterior elements are distracted.
However, if this point is situated more anteriorly as is the case with
“seatbelt” fractures, the entire spine will fail in tension. In
contrast, a hyperextension mechanism has been implicated in the
development of shear fracture-dislocations, also referred to as
“lumberjack” injuries, in which the body is distracted while the
posterior spine is exposed to either compressive or tensile forces⁴² (Fig. 43-4).

A thoracolumbar fracture should be suspected in the
setting of polytrauma until proved otherwise, especially when the
patient may be distracted by injuries to other organ systems; in one
series, approximately 24% of all thoracolumbar fractures were initially
missed in these cases.¹² The
emergent care of these individuals should commence immediately in the
field where appropriate measures should be initiated to immobilize the
entire spinal column and minimize the risk of further injury as they
are extricated and transported to a hospital facility. Because these
fractures are routinely associated with other life-threatening
injuries, the basic treatment principles set forth in the Advanced
Trauma Life Support (ATLS) protocol should be meticulously adhered to
to ensure that the airway, breathing, and circulation are adequately
maintained. Resuscitation with supplemental

oxygen,
intravenous fluids, or cardiac pressors may also be warranted in
certain clinical scenarios, although it is important to differentiate
patients with hypovolemia who are tachycardic and hypotensive from
those with neurogenic shock where sympathetic dysfunction leads to both
decreased blood pressure and a paradoxical bradycardia.¹³⁹
Strict spinal precautions including logrolling techniques and the use
of a backboard for all transfers must be followed until the existence
of any unstable injuries have been clearly ruled out by clinical
examination or imaging modalities, especially in trauma victims who
remain unconscious.

Individuals with documented spinal cord injuries may be
reasonable candidates for the administration of corticosteroids. The
initial mechanical contusion to the neural elements has been found to
precipitate a complex biochemical cascade, which ultimately leads to
tissue edema, microvascular changes with ischemic damage, and the
production of inflammatory factors.³⁹ The primary objective of steroids and other pharmacologic

strategies is to limit the extent of any secondary neurologic insults
by inhibiting the release of neurotoxic molecules, preserving membrane
integrity, scavenging free radicals, and correcting electrolyte
imbalances.¹⁴
According to a prospective, randomized multicenter trial known as the
second National Acute Spinal Cord Injury Study (NASCIS II), a steroid
regimen consisting of a 30 mg/kg bolus of methylprednisolone that is
continued at a rate of 5.4 mg/kg/hr resulted in superior long-term
neurologic outcomes.²² The ensuing
NASCIS III trial demonstrated that patients who received steroids
within 3 hours of the time of their injuries may necessitate only 24
hours of therapy, while those whose infusions were started between 3
and 8 hours should be treated for 48 hours.²³
Despite the potential complications of high-dose steroids and the
growing controversy surrounding the validity of these findings, these
guidelines are still widely accepted as the “standard of care,” which
in this litigious environment may compel practitioners to follow these
recommendations even though their efficacy has yet to be definitively
established.⁴⁴,⁶²,⁶⁹
Other preparations that have been investigated as neuroprotective
agents for spinal cord injuries include lipid peroxidase inhibitors,
calcium channel blockers, glycolipids, and opiate receptor antagonists.¹⁴

Given the amount of energy that is typically required to
disrupt the integrity of the spinal column and the close proximity of
numerous vital thoracic and abdominal viscera, it is not surprising
that more than 50% of individuals with thoracolumbar fractures will be
diagnosed with a variety of nonspinal injuries (involvement of one
other organ system, 30%; two systems, 20%; three or more systems, 5%).¹¹⁸
For instance, up to 45% of patients with “seatbelt” fractures will also
sustain some type of intraabdominal injury such as a laceration of the
spleen or liver.¹¹ Moreover, the incidence of noncontiguous spinal fractures in this population has been shown to be approximately 17% to 20%.⁵,¹²⁸
Many victims of motor vehicle accidents or falls from a significant
height may also present with head injuries or fractures of the
extremities.

It is essential that a thorough history be obtained from
the patient and any other individuals who may have witnessed the trauma
because the specific details of the event may alert the treating
physicians to the possibility that a thoracolumbar fracture may be
present as well as other spinal or appendicular injuries. Relevant
information includes the speed of the vehicle and the use of restraints
during automobile collisions as well as the distance and point of
impact of those who have fallen. In addition to axial spinal pain and
decreased range of motion, patients with thoracolumbar injuries may
describe a spectrum of neurologic symptoms depending on the status of
the spinal cord and other neural elements.

As part of the physical examination, the subject should
be carefully logrolled so that the posterior spine may be visually
inspected for ecchymoses, abrasions, lacerations, or swelling; it is
also important to note any “seatbelt” contusions across the anterior
abdomen, which are frequently observed with flexion-distraction
injuries.²⁷,⁵⁵,⁶⁰
Palpation of this region may elicit focal tenderness at the fracture
site and reveal an obvious stepoff between the spinous processes,
crepitus, soft tissue defects, or other signs of malalignment.

A comprehensive neurologic evaluation is completed
before moving the patient to establish the baseline level of function
and serial assessments must be repeated over time to ensure that any
further deterioration does not escape detection. The presence of any
neurologic injury is verified by performing motor, sensory, and reflex
testing, although this may not be feasible in those who are
unresponsive or otherwise unable to cooperate with the examination. In
adults the transition between the central and peripheral nervous
systems usually occurs

behind
the L1 vertebral body so that individuals with fractures localizing to
the thoracolumbar junction may exhibit a variety of abnormalities based
upon the anatomic structures that are affected. An isolated
radiculopathy may manifest as a dermatomal pattern of altered
sensation, myotomal weakness, or hyporeflexia, while injuries to the
spinal cord, conus medullaris, or cauda equina may give rise to any
number of deficits ranging from diffuse sensory and motor changes in
the lower extremities to dense paraplegia and sphincter incontinence. A
complete spinal cord injury must also be distinguished from spinal
shock, which represents a transitory block of neurologic impulses that
ordinarily lasts no longer than 48 hours. The bulbocavernosus reflex is
elicited by stimulating the glans penis or clitoris, which should bring
about an involuntary contraction of the anus; the long-term prognosis
of a patient cannot be reliably ascertained until this reflex arc
returns, which signifies that the episode of spinal shock has resolved.
Furthermore, the sparing of sacral sensation and maintenance of rectal
tone are also signs of an incomplete injury because they affirm that at
least some neural pathways are still intact. Spinal cord injuries may
be classified using either the Frankel system or the more elaborate
American Spinal Injury Association (ASIA) scoring method in which
muscle strength is graded from 0 to 5 and sensation to both pin prick
and light touch is recorded throughout the entire spine⁴⁹,⁹² (Fig. 43-5).

While a myriad of classification systems for
thoracolumbar fractures have been introduced over the past several
decades, given the inherent complexity of normal spinal anatomy and
biomechanics as well as the ongoing uncertainties surrounding the
factors that define “stability,” there is still no consensus among
trauma experts regarding the best method for categorizing these
injuries.⁴⁰,⁶⁸,⁷⁷,⁹¹,⁹⁵,¹⁰⁶
As with any heterogeneous collection of pathologic conditions, the
ideal method for classifying spinal injuries should be comprehensive
yet easy to apply so that it possesses sufficient reliability and
reproducibility. By integrating modern imaging technology and taking
advantage of recent advances in the understanding of the natural
history of thoracolumbar fractures, it should be possible to stratify
these injuries according to their severity, which may be helpful for
directing treatment and predicting the outcomes of these patients. An
effective classification algorithm also successfully minimizes any
confusion among different practitioners participating in the care of
these individuals and facilitates prospective clinical research in the
field of spinal trauma.

One of the earliest mechanistic classification schemes
was conceived by Denis, who attempted to elucidate the concept of
spinal stability by assigning osseous and soft tissue structures into
one of three columns: anterior (anterior half of vertebra/disc and
anterior longitudinal ligament), middle (posterior half of
vertebra/disc and posterior longitudinal ligament), and posterior
(posterior elements including the pedicles and facet joints and the
remaining ligaments)⁴⁰ (Fig. 43-10).
This system divides thoracolumbar injuries into four principal
categories —compression, burst, Chance, and fracture-dislocations—with
an additional 16 subgroups (Fig. 43-11). With
this paradigm, the greatest emphasis is placed on the middle column
such that any injury extending into this portion of the spine is
generally assumed to be unstable. However, because it does not take
into account the results of advanced imaging modalities or the status
of the PLC, the Denis algorithm may be overly simplistic and does not
appear to be beneficial for directing the management of these fractures.¹⁰⁷,¹⁵²

Unlike the Denis scheme, the AO (Magerl) system uses the
primary vector forces that are applied to the spinal column as the main
criterion for segregating thoracolumbar injuries.⁹¹
In this format, groups A, B, and C are composed of fractures generated
by compression, distraction, and torsional/rotational loads,
respectively (Fig. 43-12). The AO algorithm features multiple

levels of organization, which are not only designed to specify the
location, morphology, and direction of displacement for each fracture
but also allow for a distinction to be made between bony and soft
tissue injuries. The interobserver and intraobserver reliabilities of
the Denis and AO systems have been reported by a number of independent
investigators. In one study, the most basic subcategory of the Magerl
scheme (i.e., A, B, or C) exhibited only a fair amount of agreement (κ
= 0.33) with an interobserver reliability of 67%.¹⁸ In their comparisons of these classification systems, Oner et al.¹⁰⁷ and Wood et al.¹⁵³
both concluded that the Denis scheme was more reproducible than the AO
method. However, from these analyses it is clear that neither algorithm
is without its flaws. The AO system is certainly more inclusive but is
not as practical to implement because its complicated alphanumeric
scoring protocol, which undoubtedly reduces its overall reliability.
Conversely, the more straightforward Denis paradigm is associated with
improved interobserver agreement, but it may be too simplistic so that
unusual fracture configurations may be inadvertently excluded.

Other classification systems have also been adopted
because of their potential to guide treatment and provide prognostic
information about these injuries. After reviewing the radiographs and
CT scans of 100 thoracolumbar fractures, McAfee et al. separated these
injuries into 6 discrete groups: wedgecompression, stable and unstable
burst, Chance, flexion-distraction, and translational.⁹⁴
With its emphasis on the mechanism by which the middle column failed,
this scheme was able to determine which type of instrumentation (i.e.,
distraction or compression) was most suitable for each fracture.
McCormack et al.⁹⁵ later devised the
“load sharing classification,” which uses a grading system to assess
vertebral body comminution, spread of the bony fragments, and
posttraumatic kyphosis as a means of establishing which injuries may be
appropriately managed with immobilization alone or short-segment
transpedicular constructs limited to only the levels immediately above
and below the fracture site (Fig. 43-13). By
identifying cases that were complicated by implant breakage, the
authors suggested that a point total greater than 6 justified a
concomitant anterior arthrodesis with a strut graft. The load sharing
classification algorithm has since been validated by both in vitro
biomechanical experiments and other clinical series.⁶,³⁴,¹⁰⁹,¹⁴²

In an effort to overcome many of the inadequacies of these conventional classification systems, Vaccaro et al.¹³⁸
developed the Thoracolumbar Injury Severity Score (TLISS), an
innovative paradigm that focuses on three key parameters that reflect
the global stability of the disrupted spinal column: (a) mechanism of
injury as interpreted from imaging studies, (b) integrity of the PLC,
and (c) neurologic status. The designated point values for these
categories are combined to calculate a total score that may assist in
the clinical decision-making process for each of these fractures.
Despite its excellent reliability as demonstrated in previous
prospective investigations,⁶⁵,¹³⁴
the TLISS scheme was revised so that the mechanism category was
replaced with fracture morphology, which was thought to be a more
objective variable for surgeons to evaluate; once these changes were
instituted, this modified algorithm was renamed the Thoracolumbar
Injury Classification and Severity Score (TLICS)¹³⁶ (Table 43-1).
In a recent prospective, comparative study of these two systems, the
interrater reproducibility of the TLISS scheme was found to be superior
to that of the TLICS method, which implies that a retrospective
reconstruction of the pathomechanisms underlying a fracture may
actually be more informative than a description of its morphology.¹⁴⁶
At any rate, since mechanism and morphology are closely related and
each would be expected to contribute to spinal instability following an
injury, it is likely that both of these factors will need to be
addressed during the classification and subsequent treatment of
thoracolumbar fractures.

The optimal method for managing thoracolumbar fractures
is governed by both biomechanical and clinical considerations such as
the ability of the spine to withstand physiologic stresses as well as
the presence of any neurologic deficits, associated traumatic injuries,
or other relevant medical issues. The objectives of every therapeutic
strategy are the same regardless of whether a fracture is addressed
surgically or with nonoperative measures: maintain or restore spinal
stability, correct any deformities in the sagittal or coronal planes,
maximize neurologic recovery, improve pain, and allow for prompt
rehabilitation. In general, many patients with thoracolumbar injuries
who do not demonstrate any clinical or radiographic findings consistent
with neural compression or instability may be treated nonsurgically
with immobilization and early ambulation.

Although plaster jackets were used extensively in the
past to support the fractured spinal column, these techniques have
largely been supplanted by a number of different external braces. A
Jewett device and other hyperextension appliances are designed to
resist flexion but are less apt to control rotation or lateral bending.
A prefabricated or custom-fit “clamshell” thoracolumbosacral orthosis
(TLSO) reduces motion in multiple planes and may be better suited for
more unstable fractures.⁸⁵ Miller et al.¹⁰³
compared the effects of a simple corset, a Jewett brace, and a TLSO on
the mobility of the lumbar spine. While both the Jewett orthosis and
the TLSO limited motion in the lower lumbar spine, none of these braces
were able to completely eliminate all movement. Moreover, the L5-S1
segment was not sufficiently stabilized by any of these appliances and
some of these individuals actually exhibited hypermobility at this
level while in these devices. There are some data to suggest that the
supplementation of these braces with a leg cuff may impart even greater
stability to the spine.¹⁴⁰
Similarly, a cervical extension may also be beneficial for inhibiting
any pathologic motion that may occur with fractures of the upper
thoracic spine (i.e., above T5).

Given the overall stability of their injuries, the
majority of patients with compression fractures may experience bony
healing and symptomatic relief with approximately 12 weeks of
immobilization.¹³⁵ Nevertheless,
compression injuries that demonstrate greater than 50% height loss or
25 degrees of focal kyphosis, disruption of the PLC, or any other signs
of obvious instability on imaging studies should be followed closely
because these injuries are predisposed to further collapse and
progressive deformity despite the initiation of a suitable bracing
regimen. Folman and Gepstein⁴⁸ noted
in their retrospective review of 85 patients with traumatic
thoracolumbar wedge fractures that even though most of these injuries
were adequately managed with conservative measures alone, a large
proportion of these individuals continued to have chronic low back pain

whose intensity correlated with the amount of segmental kyphosis.

Burst injuries are produced when the spine is subjected
to more substantial axial forces and are frequently observed with
higher energy trauma. Because they characteristically involve the
middle and posterior aspects of the vertebral body, burst fractures are
by definition more unstable than compression injuries and any
retropulsed fragments may lead to significant canal stenosis with
ensuing neurologic decline. In addition to espousing the same
radiographic criteria that have been used to determine which
compression fractures are most appropriate for nonoperative treatment,
other authors have reported that burst fractures with less than 50%
canal compromise may also be addressed without surgery¹³⁰;
unfortunately, these guidelines have not been properly validated by the
existing literature. A period of immobilization lasting up to 3 months
may be acceptable for patients with stable burst fractures who have a
normal neurologic examination and an intact PLC as well as those with
unstable injuries who are not able to tolerate an operative procedure.⁶⁷,⁹⁸,¹¹³
Individuals who present with complete spinal cord injuries secondary to
burst fractures may also be candidates for conservative care. Even
without surgery, the displaced fragments of burst injuries have been
shown to undergo spontaneous remodeling, which may increase the patency
of the spinal canal and perhaps relieve any impingement of the neural
elements.³⁸ Any burst fracture that
is being braced should be regularly assessed by repeating standing
radiographs in the orthosis to ensure that there is no further collapse
of the vertebral body or increase in segmental kyphosis over time.⁵⁰

Multiple investigators have confirmed that favorable
long-term outcomes may be achieved with the closed management of
thoracolumbar burst fractures in the absence of any instability or
neurologic abnormalities.⁶,⁹,¹⁰⁴,¹⁰⁵,¹⁴⁴ Chow et al.³¹
also recorded excellent results with hyperextension casting of 24 burst
injuries that were categorized as being unstable according to the Denis
classification system, but only 10 of these fractures actually
demonstrated interspinous widening or any other radiographic findings
indicative of PLC disruption.

Because of their gross instability affecting the
anterior and posterior spinal elements, flexion-distraction injuries
and fracture-dislocations are rarely managed without surgery with the
exception of certain purely osseous Chance fractures.⁵⁸,⁸⁸
In certain circumstances, patients with these injuries who have a
normal neurologic examination and less than 15 degrees of kyphosis may
be braced in hyperextension as long as a satisfactory reduction is able
to be maintained for 3 months or more until the bony fragments have
consolidated.¹⁰ Conversely, reliable
healing of soft tissue flexion-distraction injuries and virtually all
fracture-dislocations is unlikely to be attained if internal fixation
is not used to reestablish the normal alignment and biomechanical
integrity of the spine.

Even though a number of thoracolumbar fracture patterns
may be treated with a rigid orthosis and prompt mobilization, a large
subset of these injuries may be more effectively addressed with
surgery. Operative intervention is intended to convey immediate
stability to the spine, allow for the correction of deformities, and
optimize neurologic improvement by directly or indirectly relieving any
residual impingement of the neural elements; for these reasons,
surgical techniques may have the potential to enhance clinical
outcomes, facilitate rehabilitation, and eschew many of the adverse
consequences of nonoperative therapies.¹⁹,¹⁴⁹

The operative decision-making process is dictated by
several different factors such as the morphology of the fracture, the
status of the posterior ligamentous complex, the neurologic status of
the patient, and any other traumatic injuries or medical comorbidities.
Even so, it is widely believed that individuals with unstable fractures
who present with worsening kyphosis or intersegmental translation,
incomplete neurologic deficits in the setting of persistent compression
of the spinal cord, or radiographic evidence of damage to the posterior
ligamentous structures will presumably benefit from surgery.⁹⁸
Operative strategies may also be preferable for patients who cannot
tolerate external immobilization because of their body habitus or major
injuries involving the extremities or other organ systems.

A variety of surgical procedures have been introduced in
an attempt to stabilize and decompress the injured thoracolumbar spine,
all of which are performed either through a posterior, anterior, or
circumferential approach. The ideal method for treating these fractures
is still a matter of some debate, and in most cases the operative plan
is influenced by both clinical and radiographic concerns including the
neurologic examination, other critical traumatic injuries or pertinent
medical conditions, amount of sagittal plane deformity, encroachment of
the thecal sac on axial images, and any signs of spinal instability.

Given the severity of these injuries, it should come as
no surprise that thoracolumbar fractures may give rise to a wide range
of complications. Although patients with obvious neurologic
deterioration are usually treated with surgery in the acute setting, it
is not uncommon for individuals who have sustained spinal trauma to
experience increasing impairment secondary to other pathologic
conditions such as a synringomyelia. Whether surgical stabilization is
performed, the inherent instability of many thoracolumbar injuries also
predisposes many of these patients to the development of spinal
malalignment in any plane. In most cases, fractures that demonstrate
kyphosis greater than 30 degrees, height loss greater than 50%,
translation greater than 2.5 mm, canal compromise greater than 50%, or
signs of posterior ligamentous disruption on MRI studies may be
particularly susceptible to progressive deformity. Because these
fractures are frequently associated with injuries to the spinal cord
and abdominal viscera, there is also a fairly high incidence of
gastrointestinal abnormalities including ileus, gastroesophagel reflux

disease,
and constipation, all of which must be managed with appropriate
interventions. Individuals with thoracolumbar fractures, especially
those with concomitant spinal cord injuries, are clearly at risk for
thromboembolic events; in one investigation, the rate of symptomatic
deep vein thrombosis or pulmonary embolism was reported to be
approximately 2%, with the incidence of asymptomatic clots estimated to
be even higher.¹¹⁰
Consequently, some form of prophylaxis is indicated for these patients
such as chemical anticoagulation, intermittent pneumatic compression
devices, or vena cava filters. Despite the appearance of a healed
fracture, a large proportion of this population will continue to
complain of intractable pain, which may be refractory to both
conservative and operative measures. Finally, these types of injuries
regularly necessitate prolonged hospitalizations, which may lead to
problems related to pneumonia, pressure ulceration of the skin, or
malnutrition.

While the surgical techniques that are used to
decompress and stabilize thoracolumbar fractures are considered to be
reasonably safe, these procedures may still subject the patient to a
number of potential complications. Iatrogenic neurologic insults have
been shown to occur in 1% of posterior operations secondary to the
erroneous placement of pedicle screws.⁷³
Malpositioned implants may also bring about serious injuries to any
number of visceral or vascular structures, which are often
life-threatening, especially when they are not diagnosed and addressed
expediently (Fig. 43-31). Up to 10% of
instrumented fusions for thoracolumbar fractures may become infected,
all which require either long-term antibiotic therapy directed toward
the causative organism or possibly open irrigation and débridement if
indicated.¹¹² Symptomatic durotomies
should be repaired primarily if possible, but individuals with
unrelenting leakage of cerebrospinal fluid from fistulas are routinely
managed with protracted bed rest or even insertion of a lumbar
subarachnoid drain to decrease intradural pressure enough so that a
watertight closure may be obtained. Any remaining intersegmental
instability at the arthrodesis site may result in the formation of a
pseudarthrosis with hardware failure, recurrent deformity, and
unrelenting pain (Fig. 43-32). Unfortunately,

these surgical approaches are also not without their morbidity; that
patients undergoing anterior or posterior procedures may experience
substantial blood loss from a corpectomy or the epidural veins,
compromised pulmonary function, or poor wound healing.

Dai et al.³⁵
retrospectively reviewed a large series of 147 thoracolumbar fractures
that occurred in conjunction with multiple other injuries. According to
their analysis, the group that had been immobilized was significantly
more likely to contract pneumonia and required longer hospital stays
than the subjects who were managed with operative fixation. Shen et al.¹²¹
prospectively evaluated 80 neurologically intact patients with
single-level burst injuries at the thoracolumbar junction who were
either treated with a course of bracing or short-segment instrumented
posterior arthrodesis. In this series, surgery brought about greater
initial correction of kyphosis and earlier pain relief but at 24 months
there were no significant differences between these two cohorts. Wood
et al.¹⁵² also performed a
randomized, prospective study comparing surgical and conservative
treatments for stable burst fractures without any accompanying
neurologic impairment. At an average follow-up of 44 months, operative
stabilization of these injuries did not appear to provide any
meaningful advantages to these individuals; in fact, those who were
placed in a body cast or orthosis actually reported less disability,
greater pain relief, and fewer complications. However, another recent
prospective, randomized, multicenter trial supported the use of
short-segment posterior stabilization for these types of fractures by
showing that these procedures gave rise to decreased kyphosis, better
functional scores, and a faster return to work relative to
immobilization in a Jewett device.¹²³

At this time, the superiority of one operative procedure
over the others has yet to be borne out in the current literature. Two
centers independently proposed that anterior decompression of burst
fractures may generate more substantial improvements in neurologic
status compared to indirect reduction of these injuries through a
posterior approach.²⁴,⁹³
In another series of unstable burst fractures, an anterior corpectomy
and instrumented arthrodesis were found to be more efficacious than
posterior short-segment fixation for restoring normal sagittal
alignment.¹¹⁷ While Been and Bouma¹⁶
also observed better deformity correction with circumferential
surgeries than with posterior implants alone, the neurologic outcomes
of these cohorts were equivalent. However, Danisa et al.³⁶
identified no significant differences between patients with unstable
burst injuries addressed anteriorly, posteriorly, or with a combined
technique. Esses et al.⁴⁵ conducted
a prospective, randomized investigation in which subjects with acute
thoracolumbar burst fractures were alternately treated with posterior
distraction and pedicle screws or direct anterior decompression in
conjunction with a fusion construct composed of an autogenous
tricortical iliac crest strut graft and a screw-rod device. Although
the authors noted no obvious disparities between the clinical or
radiographic results of these two strategies, they emphasized that
anterior operations permitted a more thorough decompression of the
spinal canal. In a more recent prospective clinical trial, 38 patients
with stable burst injuries located at the thoracolumbar junction and no
attendant neurologic deficits were randomized so that they either
underwent a posterior instrumented arthrodesis or an anterior
reconstruction with an allograft strut and lateral fixation.¹⁵¹
At a minimum of 2-year follow-up, the fusion rates and functional
outcomes of these two cohorts were comparable but there was a trend
toward more frequent complications and a higher incidence of revision
surgeries with a posterior approach.

For patients with ongoing neurologic decline, it is
generally accepted that prompt operative intervention may yield
superior clinical outcomes.²⁸ In
animals, rapid decompression appears to correlate with enhanced
neurologic recovery but these benefits have not necessarily been
recorded in humans.²⁶ However, the
optimal time to perform surgical procedures in individuals who are
neurologically intact is still a controversial issue. The authors of
one retrospective study observed a trend toward greater return of
neurologic function among a group of thoracolumbar fractures that were
addressed surgically within 8 hours of the traumatic event compared to
injuries in which these operations were completed in a delayed fashion,
but otherwise there were no significant time-dependent effects
identified in this series.⁵¹ In similar investigation, Chipman et al.²⁹
concluded that even though early surgery did not provide any additional
improvements in terms of postoperative neurologic status, patients that
were treated less than 3 days after sustaining their fractures
exhibited significantly fewer complications and shorter
hospitalizations than others whose interventions occurred after this
point. Based on this limited data, patients with thoracolumbar
fractures who do not exhibit neurologic compromise requiring emergent
decompression may be carefully evaluated before any type of operation;
for instance, it may be prudent to postpone definitive surgical
stabilization until any hemodynamic instability has been corrected and
traumatic injuries to all other organ systems have been thoroughly
ruled out (e.g., cardiac, pulmonary).

Minimally invasive spinal surgery (MISS) has become
increasing popular in recent years because of its potential for
decreasing operative morbidity and preserving the stability of the
vertebral column, which many advocates claim may translate into
improved long-term functional outcomes. Nevertheless, many of these
purported advantages are still largely theoretical and have not been
supported by any randomized, prospective, controlled clinical trials
comparing MISS with open techniques. Nevertheless, the rationale for
MISS is undoubtedly intuitively pleasing to surgeons and patients
alike. While MISS may be inappropriate for many thoracolumbar
fractures, such as those that are associated with a substantial amount
of canal occlusion and an incomplete neurologic injury, that would be
best served with a wide decompression through an open approach, this
feasibility

of
MISS for this application has already been preliminarily established.
One center has reported their considerable experience with
endoscopically assisted anterior corpectomy, interbody reconstruction,
and instrumentation for thoracolumbar fractures.¹⁷,⁷⁸
While the initial results have been favorable, the authors concede that
this strategy is very difficult and entails a very steep learning curve.

Another MISS procedure that is currently under
investigation includes the use of vertebroplasty/kyphoplasty for
traumatic thoracolumbar injuries, either with¹,²,³¹,⁸⁰,⁸¹,¹⁴¹ or without supplemental fixation³⁷,¹²⁷ (Fig. 43-33).
One major concern regarding the injection of cement into a fractured
vertebral body with disruption of the posterior cortex is the
possibility that this material may extrude into the canal and produce
an iatrogenic neurologic deficit. Clearly, multiple well-designed
studies are warranted to confirm the safety and efficacy of this and
other surgical techniques for the treatment of the entire spectrum of
thoracolumbar spinal injuries.