1. health and fitness
2. disease

The Spine Journal 4 (2004) 230–240
Review Article
Gunshot wounds to the spine
Christopher M. Bono, MDa,*, Robert F. Heary, MDb
a
Department of Orthopaedic Surgery, Boston Medical Center, Boston University School of Medicine, Dowling 2 North,
One Boston Medical Center Place, Boston, MA 02118-2393, USA
b
Spine Center of New Jersey, Department of Neurological Surgery, New Jersey Medical School,
University of Medicine and Dentistry of New Jersey, 90 Bergen Street, Suite 7400, Newark, NJ 07103, USA
Received 6 November 2002; accepted 13 June 2003
Abstract
BACKGROUND CONTEXT: The incidence of violent crimes has risen over the past decade.
With it, gunshot injuries have become increasingly more common in the civilian population. Among
the most devastating injuries are gunshot wounds to the spine.
PURPOSE: The purpose of this article is to provide a thorough review of the pathomechanics,
diagnosis and treatment of gunshot wounds to the spine.
STUDY DESIGN/SETTING: Literature review article.
CONCLUSIONS: Treatment of gunshot spine fractures differs from other mechanisms. Fractures
are usually inherently stable and rarely require stabilization. In neurologically intact patients, there are
few indications for surgery. Evidence of acute lead intoxication, an intracanal copper bullet or new
onset neurologic deficit can justify operative decompression and/or bullet removal. Overzealous
laminectomy can destabilize the spine and lead to late postoperative deformity. For complete and
incomplete neural deficits at the cervical and thoracic levels, operative decompression is of little
benefit and can lead to higher complication rates than nonsurgically managed patients. With gunshots
to the T12 to L5 levels, better motor recovery has been reported after intracanal bullet removal
versus nonoperative treatment. The use of steroids for gunshot paralysis has not improved the
neurologic outcome and has resulted in a greater frequency of nonspinal complications. Although
numerous recommendations exist, 7 to 14 days of broad-spectrum antibiosis has lead to the lowest
rates of infection after transcolonic gunshots to the spine. 쑖 2004 Elsevier Inc. All rights reserved.
Keywords:
Gunshot wounds; Gunshot; Spine; Fractures; Vertebral column; Spinal cord injury; Ballistics
Introduction
With the rise of violence throughout the world, the incidence of gunshot injuries continues to increase. Among the
most troublesome are gunshot wounds to the spine, which
account for 13% to 17% of all spinal cord injuries each year
[1–3]. Although most common in the thoracic region, they
are perhaps most devastating in the cervical spine, where
the potential for severe functional impairment is greatest
[4,5]. With a predilection for weekend occurrence, spinal
gunshots are most frequently sustained by young, male
FDA device/drug status: not applicable.
Nothing of value received from a commercial entity related to this
research.
* Corresponding author. Department of Orthopaedic Surgery, Boston
Medical Center, Boston University School of Medicine, Dowling 2 North,
One Boston Medical Center Place, Boston, MA 02118-2393, USA. Tel.:
(617) 414-6281; fax: (617) 414-6292.
E-mail address: [email protected] (C.M. Bono)
1529-9430/04/$ – see front matter 쑖 2004 Elsevier Inc. All rights reserved.
doi:10.1016/S1529-9430(03)00178-5
minorities between the ages of 15 and 34 years [1,6,7].
Spinal cord damage after gunshot wound is more likely to
result in complete injury compared with blunt trauma. Like
spinal cord injury in general, incomplete injuries have much
better prognoses [1].
Pertinent ballistics
According to the formula KE⫽1/2 mv2 (where KE⫽
kinetic energy, m⫽mass and v⫽velocity), the energy of
the bullet is influenced by both the mass and speed. As the
equation dictates, increases in velocity have exponential effects on the energy of the injury. Thus, by convention, muzzle
velocities of less than 1,000 to 2,000 feet/second are categorized as low energy, whereas speeds greater than 2,000
to 3,000 feet/second are considered high energy. Lowvelocity, low-energy firearms include pistols and handguns.
High-velocity, high-energy weapons include military assault
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
231
rifles, such as the AK-47 and M-16. In contrast, shotguns
are low-velocity weapons that incur high-energy wounds because of the combined mass of the numerous fired pellets.
In addition, shotgun wounds are further complicated by the
presence of large pieces of “wadding” that are a gross contaminant and require prompt surgical debridement [8].
In general, civilian gunshot wounds are considered low
energy. However, an increasing frequency of high-energy
injuries is being reported in the nonmilitary population [9].
Greater availability of military-type firearms, especially
among the criminal element, is a contributing factor. In
1995, assaults with high-energy weapons accounted for 16%
of homicides in New York City [10]. Close-range gunshots
lose less energy during transit and therefore transfer more
energy to the victim. Differentiation between high- and lowenergy injuries is crucial, because treatments are distinct.
In vivo experiments in monkeys have demonstrated the
local necrotic effect of copper on brain tissue [12]. Similar
studies have been performed on the spinal cord. Tindel
et al. [13] demonstrated spinal cord necrosis around copper
fragments implanted within the dura of rabbit spinal cords.
Whereas lead fragments caused less axonal necrosis than
copper, aluminum resulted in minimal pathologic changes.
Importantly, extradural fragments, regardless of the metal
type, did not cause appreciable pathologic changes within the
neural tissue. The authors extrapolated that these effects
would be similar to those in the human spinal cord. Although
these findings might support removal of intradural copper
bullet fragments [14], the clinical challenge is determining
the material of the bullet, which in many instances is not
known, as well as avoiding further neurologic injury during
explantation. The clinical benefits of bullet removal to avoid
the neurotoxic effects of copper remain unclear.
Bullet material and design
Evaluation
The bullet energy is not the sole determinant of the extent
of injury. Fragmentation of the missile can increase the
zone of tissue destruction. Hollow-point bullets explode on
impact, leading to multiple fragments that deviate from initial linear path. The phenomenon of yaw refers to the tumbling of a bullet along its longitudinal axis. Longer bullets
have a tendency for greater yaw, which can further increase
the zone of destruction. In its mid-path, one can imagine
the bullet turned 90 degrees to its long axis so that it is in
effect being pulled through the tissues sideways.
Bullets can be nonjacketed, partially jacketed or fully
jacketed. The “jacket” refers to a thin metallic layer covering
the surface of the bullet. Fully jacketed bullets exhibit little
deformation with firing and are designed for accuracy to hit
long-range targets. They are highly precise and incur clean
exit and entrance wounds. Partially or nonjacketed missiles
expand on tissue impact and, like hollow-point bullets, exponentially extend the circumference of tissue damage. Such
designs are intended for close-range targets.
Various materials have been used to produce bullets. Most
bullets are composed of a lead core, which may be combined
with a number of different metals to achieve a desired
hardness. Similarly, the jacketing material can be copper,
brass or nickel. These substances can have both systemic
and local toxicities on tissues. More common in synovial
joints, lead poisoning has been reported with bullets lodged in
the intervertebral disc [11]. Although cerebrospinal fluid
would intuitively be an effective solvent, systemic lead elution with intramedullary bullets is uncommon. In cases of
retained lead bullets, lead levels can be measured periodically. Notwithstanding other clinical factors, significant increases in lead levels and characteristic hematopoietic
changes on bone marrow biopsy can be an indication for
bullet removal. In actual clinical practice, bullets very rarely
need to be removed from the spine to treat or prevent
lead toxicity.
Acute examination
The initial evaluation of the gunshot victim must address
all life-threatening injuries. Maintenance of airway, breathing and circulation (ABCs) is paramount. The region of injury
should be considered. Gunshot wounds to the neck are
frequently complicated by airway injuries that can necessitate emergent intubation [15]. Action should not be delayed
for radiographic “clearance” of the cervical spine, because
gunshot fractures are usually inherently stable [5,16–18].
Carotid and vertebral artery perforations should be suspected
with pulsatile neck bleeding. Immediate vascular consultation to restore cerebral blood flow is important, as temporary
stents placed emergently can enable definitive repair at a
later time. Pharyngeal and esophageal wounds must also be
detected and evaluated, because they are often associated
with infections. In particular, large hypopharyngeal lacerations are associated with a higher rate of contamination
because of the pooling of secretions in this area. Endoscopic
techniques can be used effectively and expediently for surveillance in the acute setting. Open surgical exploration is
mandatory to repair any documented tears.
The thoracic spine is the most commonly affected with
gunshots [4]. The lungs, heart and great vessels are at risk
with such injuries. Careful chest auscultation can detect
asymmetric breath sounds indicating hemothorax or pneumothorax. Cardiac monitoring, including distal pulses, can
suggest heart perforation, aortic disruption or tamponade.
Until stabilization of ventilation and hemodynamics, the
evaluation of any spinal injury remains secondary. Similarly,
the examination of the abdomen is focused on suspected
vascular or viscus injuries. In particular, colonic perforations
that occur before the missile passing through the spine must
be recognized, because they are associated with a high rate
of spinal infection if not treated with an appropriate antibiotic
regimen [19–22]. In paraplegic or quadriplegic patients, the
232
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
abdominal examination may be less reliable because of loss
of visceral sensation. This may lower the threshold for obtaining advanced diagnostics, such as peritoneal lavage or
abdominal computerized tomography, to rule out injuries.
Although numerous organs exist in the pelvic area, sacral
gunshot wounds are most often complicated by hemorrhage
[23]. After posterior gunshot wounds, sterile packing of
the posterior bullet hole using methyl cellulose and bone wax
has been reported effective in facilitating tamponade [23].
Likewise, emergent angiographic embolization can diminish
or halt severe bleeding, with the risk of possible devascularization of the neural structures [23].
Evaluation of the spinal injury
After patient stabilization by the trauma team, focus is
turned toward the spinal injury. Witnesses, emergency medical technicians and law enforcement officers can offer useful
information about the direction of the injury, the type of
weapon or bullet and the proximity of the shot. Often the
alert and awake patient can account for the number of shots
fired. The cervical spine should be immobilized with a hard
cervical collar until initial radiographs are complete.
The patient must be stripped of all clothing while log
rolling the patient to protect the spine from further harm.
Entrance and exit wounds should be examined thoroughly.
Often, only an entrance hole can be found if the bullet is
lodged in tissue. The entrance hole may be differentiated
from the exit wound. As a bullet pierces the skin from
the outside, the margins of the wound are cleaner and more
defined. Upon exit, the bullet can take a more random path
(fragmentation) and has more of a “blown out” appearance.
In the abdomen, it is critical to determine if the bullet
traveled from front to back. Although deep probing is not
recommended, superficial inspection for imbedded pieces of
clothing or foreign materials is useful. After sterile wound
care and dressing, radiopaque markers (eg, electrocardiogram leads) are placed over all wounds. This facilitates
roentgenographic deduction of the gunshot path.
The physical examination should proceed as for other
types of spinal trauma. Each spinous process must be palpated for tenderness and crepitus. An in-depth neurologic
examination, including motor, sensory, reflexes and anal
sphincter tone is mandatory and must be documented precisely. In the intubated or unconscious patient, neurologic
testing is limited. However, if paralytic agents have not been
administered, deep tendon and bulbocavernosus reflexes
may still be elicited.
this path. Fragments can be located in soft tissue, intervertebral disc, bone or the spinal canal. Such parameters as vertebral body height, interpedicular distance and segmental
kyphosis can be measured. In the awake, neurologically
intact patient in whom spinal stability is in question, dynamic
active flexion-extension views may be obtained. This is best
delayed until 2 weeks after injury when pain and spasm
have sufficiently subsided.
The necessity of radiographically clearing the cervical
spine after gunshot wounds to the head and face has been
questioned in a number of recent articles. In a series of 53
consecutive patients sustaining gunshots to the cranium,
none had cervical spine fractures [24]. Similarly, Kennedy
et al. [17] found wounds limited to the calvaria had no
concomitant cervical spine trauma in 105 cases. However,
with gunshot wounds extending outside the calvaria, 10%
sustained a cervical spine fracture. Kihtir et al. [16] studied
the association of cervical injuries with gunshots to the face.
Dividing the face into three zones, spinal fractures were noted
in 10% and 20% of mid-face (maxillary) and orbital injuries,
respectively, whereas surprisingly none were noted after
gunshots to the lower face (mandibular region). From these
data, it would appear that gunshots to the maxilla and orbits
carry the highest risk for concomitant cervical spine injury
and that radiographic imaging should be pursued more aggressively after such incidents. Clearing the cervical spine
after calvarial injuries is probably not necessary.
After plain films have determined the level of the bullet
and/or fracture, computed tomography (CT) should be
sought. Currently, this is the advanced modality of choice for
spinal gunshot trauma. CT images allow better localization
of the bullet within the vertebral segment and can more
clearly demonstrate foreign bodies in the spinal canal. Thin
slices (1 to 3 mm) are used to evaluate the integrity of the
bony elements. Of note, images are often obscured by artifact
when metallic bullet fragments are present (Fig. 1). This
can limit the amount of detail appreciable.
Imaging
Two orthogonal plain radiographic views of the spine can
help detect gunshot fractures and locate bullet fragments.
Using radiopaque bullet hole markers, the bullet path can
be deduced. Damage must be suspected to structures along
Fig. 1. Computed tomography images are highly useful in evaluating the
location of bullet and bone fragments. However, their utility can sometimes
limited by metallic artifact.
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
The use of magnetic resonance imaging (MRI) to evaluate
gunshot wounds to the spine is controversial. Bullet migration from the pull of the strong magnet can possibly lead
to further neurological or soft tissue damage. However, numerous reports of the use of MRI after gunshot wounds to
the spine have not supported this concern [25]. The advantages of MRI over CT include markedly less artifact, better
soft-tissue imaging and coronal, sagittal and axial visualization of the neural elements [26,27]. Although these advantages are attractive, the use of MRI should be on an individual
and patient-specific basis. For example, bullets lodged in a
vertebral body may have fewer propensities to migrate than
intracanal fragments. However, these contentions remain
unproven in a controlled scientific study. In the authors’
experience, the most common complaint from patients undergoing MRI scans is a complaint of a heat sensation in the
area of the bullet (particularly jacketed types), which can
lead to discomfort and early abortion of the study.
Treatment
Initial treatment of the gunshot victim is an integral part
of the trauma evaluation. As discussed above, maintenance of
ABCs is the primary goal. Management can be facilitated
with the use of a structured algorithm [4]. Bishop et al. [4]
examined the results of a comprehensive treatment scheme
for both blunt and penetrating chest and abdominal trauma.
Although the authors noted substantially higher mortality
for cases of blunt trauma in which treatment deviated from
the algorithm, this relationship was not demonstrated in
gunshot victims. Among the most common causes of death
after gunshot wounds were exasanguination and hypoxemia
from massive hemothorax.
Antibiosis
Tetanus prophylaxis must be considered in all instances
of spinal gunshot wounds. If there is any question regarding
the most recent immunization, tetanus prophylaxis should
be administered in the emergency room at the time of initial evaluation.
Although some advocate initial bullet wound cultures,
there is little support of this practice in the literature. Extrapolating the findings from the available literature for open long
bone fractures, the utility of wound culture is minimal [28].
Broad-spectrum antibiosis should be initiated as soon as
possible. Recommendations for the duration of antibiosis
vary. For gunshot wounds not complicated by viscus perforations, it is generally recommended to maintain 48 to 72
hours of prophylaxis. The length of antibiotic treatment is
dictated in part by the amount of soft tissue damage and
should be considered in regards to clinical evidence of infection both in spinal and extraspinal regions.
Viscus perforations carry a higher risk of infection
[19,21,22,29]. Rates are highest with colonic wounds that
occur before the bullet entering the spine. It is thought that the
stomach and small bowel are sterile, although case reports
233
of spinal infection have been documented. In an early series of
20 low-velocity transabdominal gunshot wounds to the
spine, Romanick et al. [22] realized an 88% rate of spinal
infection after colonic perforation versus no infections if the
bullet penetrated the stomach or small intestine. Although
the authors concluded that surgical debridement may be
beneficial for transcolonic spinal injuries, it is noted that
antibiotics were continued for only 48 to 96 hours. Spinal
infections included meningitis in two cases, abscess in three
cases and vertebral osteomyelitis in two cases. Culturespecific antibiotics and thorough drainage and debridement
were effective in controlling infection. Diverting colostomy
was a useful adjunct in two cases of spinal abscess and one
case of osteomyelitis. Shotgun wounds were excluded from
the study.
After colonic perforation, the lowest infection rates have
been documented with antibiotics continued for 7 to 14 days
after injury [19,21]. Roffi et al. [21] retrospectively studied
the outcome of 42 patients who sustained spinal gunshot
wounds with a perforated viscus. In all cases, the missile
passed through the viscus before the spine. Patients received
at least 6 days of broad-spectrum antibiosis. Various combinations of drugs were used, including cefoxitin, gentamicin,
clindamycin and penicillin. The use of a diverting colostomy
was not reported. Three spinal infections occurred, two paraspinal abscesses in transcolonic injuries and one case of
meningitis after gastric perforation. Initial prophylaxis was
continued for only 6 days in one case of transcolonic injury.
The authors advised continuation of prophylaxis for at least
7 days if the bowel or stomach were violated. Supportingly,
Kumar et al. [19] recently reported no spinal infections in
13 patients treated with at least 7 days of broad-spectrum
antibiotics after spinal gunshot wound that first perforated
the colon. Dehiscence of a primarily repaired colon was
associated with a high rate of intraabdominal abscess and
peritonitis but did not increase the rate of spinal infection.
From these data, it is the authors’ practice to maintain antibiotic prophylaxis for a minimum of 10 days for transabominal
gunshot wounds with a perforated viscus.
In transabominal gunshots that perforate a viscus before
spinal injury, surgical debridement does not appear to be of
benefit (Fig. 2). A minimum of 7 days of broad-spectrum
antibiotics is recommended. Although intestinal diverting
procedures (ostomies) might help reduce the incidence of
intraabdominal infection, it has little influence on the development of spinal infection. Barring other indications, such
as neurologic deterioration or lead toxicity, bullet extraction
is not advocated to decrease this risk.
The role of antibiotic prophylaxis after esophageal and
upper airway perforation is less clear. Pooled secretions in
the hypopharynx are thought to increase the propensity for
infection with gunshots involving this structure. The decision to explore such wounds is usually based on the size of
the lesion, because small rents can effectively be treated
nonoperatively. To the authors’ knowledge, there are no
controlled studies of the effects of antibiotic prophylaxis
234
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
Fig. 2. (Left) Anteroposterior and (right) lateral radiographs of a patient who sustained a transabdominal gunshot fracture to the lumbar spine. During
exploratory laparatomy, a colonic perforation was primarily repaired. The spinal wound was not debrided or explored. The patient was maintained on 14
days of broad-spectrum antibiotics without development of spinal infection.
duration after upper airway injuries associated with gunshot
wounds to the spine. Notwithstanding evidence of frank
infection or meningitis, it is prudent to extend prophylaxis
for at least 48 to 72 hours. Delayed exploration for developing neck infections is recommended, although the role of
cervical gunshot wound debridement remains to be defined.
Specific injuries
Gunshot wound: cerebrospinal fluid leak
In the initial treatment of gunshot wound patients, the
bullet entry site should be treated with debridement of any
devitalized skin and superficial soft tissues. Cerebrospinal
fluid can be differentiated from other forms of clear drainage
by the presence of beta-2 transferrin if the diagnosis is in
question [30]. If a cerebrospinal fluid leak appears to be
present, then a lumbar subarachnoid drain should be placed.
In cases where a persistent cerebrospinal fluid leak or fistulae
is present through the bullet entry or exit sites at the level
of the skin, consideration of open surgery must be entertained
[21,22]. Because of the risk of meningitis resulting from a
persistent cerebrospinal fluid leak, the treatment would involve a laminectomy with repair of the dural violation either
primarily or with use of a dural graft [31,32]. In these relatively rare instances, placement of a temporary lumbar subarachnoid drain after the laminectomy may be beneficial to
supplement the dural repair.
Gunshot injury: no neurologic deficit
In addition to the principles of antibiosis described above,
the treatment of gunshots to the spine with a neurologically
intact patient is similar to that from blunt trauma. Especially in the neurologically intact patient, spinal stability
must be assessed. Although Denis’ three-column theory is
useful for blunt traumatic injuries, it is of limited utility for
gunshot wounds to the spine, In Denis’ original work, the
proposed mechanisms of injury implied an abrupt acceleration/deceleration of the body/spine in space [33]. With gunshots, the body/spine can be considered stationary and the
bullet is the directional force. This concept can be likened
to the magician who pulls a tablecloth from a table set with
glasses and plates. The bullet acts as the tablecloth. If pulled
very quickly, the glasses and plates (ie, spinal elements) stay
in place. In contrast to blunt trauma, two- or even threecolumn involvement is therefore less likely to result in
instability.
In the best-case scenario, a through-and-through bullet
wound will damage only those structures that lie directly in
its effective path. Low-energy gunshots have a narrower
circumference of damage than high-energy wounds. These
factors influence the amount of spinal instability after gunshot wounds to the spine.
In the awake, cooperative, neurologically intact patient,
dynamic lateral cervical radiographs can be obtained. Careful flexion and extension of the spine can demonstrate pathologic mobility of adjacent spinal segments. Again, this is
best delayed after 2 weeks of collar immobilization to allow
pain and spasm to subside. If stability must be determined
immediately, an MRI may be obtained (if not considered a
risk to neurologic status) to assess ligamentous integrity,
although the clinical significance of signal changes within
the soft tissues after gunshot wounds has not been well
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
characterized. If, after appropriate measures, cervical stability remains in question, surgical stabilization may be considered to prevent untoward neurologic injury, particularly
if an intracanal bullet or bone fragment is in close proximity to
the spinal cord or nerve roots.
In contrast to blunt traumatic injuries, which have more
defined radiographic criteria for instability, any degree of
abnormal angulation or translation may represent instability
after gunshot wounds to the cervical spine. In cases of instability, the affected segments can be stabilized with a variety
of instrumentation and fusion constructs. In the patient with
neurologic deficit, spinal stability is deduced from imaging
findings. Severe comminution of both anterior and posterior
elements with evidence of segmental deformity is indicative
of instability.
The majority of cases of instability after gunshot wounds
have been associated with overly aggressive decompression [34]. Fortunately, there are few indications for decompression in a neurologically intact patient. New onset or
progressive neurological deficit associated with an intramedullary bullet, bone fragment or expanding epidural hematoma is an indication for urgent decompression. Importantly,
neurologic progression is most accurately detected by means
of serial examinations performed by a single experienced
observer and documented in the chart so that follow-up
examinations can be compared with baseline. Less frequently cited indications are for the treatment of lead intoxication from the bullet or contact of a copper jacketed missile
with the neural elements [11,31,35]. Routine removal of
the bullet is not warranted for transcolonic injury and is
associated with a high rate of complications.
Gunshot injury: neurologic deficit
Spinal cord injuries are devastating to the patient, family
and society. In a study by Roye et al. [2] in 1988, the average
initial admission hospital costs for survivors of cervical
level injury can be greater than $50,000. High-level complete
injuries frequently leave patients ventilator dependent,
requiring highly specialized units for prolonged care. Gunshot wounds more frequently result in complete injuries
compared with blunt trauma, which may be paraplegic or
quadriplegic depending on the level. Approximately 59%
of penetrating cervical spine injuries result in complete
spinal cord injury, compared with 49% for blunt trauma [36].
In contrast to the predominance of cervical level injuries
after blunt trauma, the majority of spinal cord injuries after
gunshot wounds occur at the thoracic level [36]. Incomplete
injuries can present in a variety of manners, including
Brown-Sequard, central cord, anterior cord or in rare cases
cruciate hemiplegic syndromes [37].
Decompression/bullet removal
The first rule of medicine is primum non nocere, or “do
nor harm.” This must be kept in mind in management
of neurologic deficit after gunshot wounds to the spine (Fig.
235
3). It is our tendency as surgeons to be compelled to remove
bullets and decompress the spinal canal. However, the clinical benefit of these actions is inconsistent and is not without complications.
In both complete and incomplete spinal cord injuries, the
role of decompression has been studied. Stauffer et al. [34]
reviewed 185 cases of gunshot paralysis, half of which were
treated with laminectomy and half with observation only.
The authors documented no appreciable return of neurological function after both surgery and nonoperative management for complete lesions. With incomplete injuries, 71%
of decompressed spines and 77% of nonsurgically treated
spines demonstrated neural improvement. Incomplete lesions to the lumbar spine and thoracolumbar junction had
better neurologic recovery than more cranial levels regardless if surgery was performed. Although antibiotic regimen was not documented, four wound infections and six
spinal fistulae were reported in the operative group as well
as six cases of late spinal instability. There were no cases
of infection, cerebrospinal fluid fistula or spinal instability
in the nonsurgically treated patients. Both groups had a
high incidence of causalgia (19% and 15% for operative and
nonoperative groups, respectively). In support of a nonoperative approach, Robertson and Simpson [38] reported no neurologic improvements with lumbar laminectomy versus
nonsurgical treatment in 30 patients with gunshot wounds to
the cauda equina region. A high rate of postoperative complications was reported. In no case did the authors report an
intracanal bullet.
In a more recent prospective study of lesions associated
with intracanal bullets, Waters and Adkins [39] demonstrated statistically significant motor improvement after surgical decompression from the T12 to L4 levels compared
with nonoperatively treated spines (Fig. 4). At more cranial
sites in the thoracic and cervical regions, surgical removal and
decompression had no significant effect on neurologic outcome. No case of infection was reported in either group.
Surgical decompression is indicated for progressive neurologic deficits associated with compressive intracanal
blood, bullet or bone fragments. In these situations, the
surgeon is lead to conclude that neural damage is actively
advancing secondary to spinal cord compression and immediate operation is indicated. Intracanal bullets in the lumbar
spine should also be removed to improve the rate of motor recovery in cauda equina level injuries. The role of decompression and bullet removal in the cervical and thoracic levels
remains unclear, with the complications of surgery a foreboding detractor. Because of the possible potential benefits of
decompression, the authors sometimes elect to remove
cervical-level intracanal fragments, especially in incomplete spinal cord injuries. In the current authors’ opinion,
the possibility of even one or two levels of recovery justifies
the risk of cerebrospinal leak, fistula or infection in the
cervical region. This, however, would be difficult to justify in
the thoracic spine, where little functional return is sacrificed.
236
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
Fig. 3. Management of neurologic deficit after gunshot wounds to the spine.
Steroids
The use of corticosteroids in spinal-cord-injured patients
after gunshot wounds has recently been examined [40,41].
Levy et al. [41] retrospectively studied 252 cases of both
complete and incomplete gunshot spinal cord injuries. Administration of methylprednisolone according to National
Acute Spinal Cord Injury Study (NASCIS-II) protocol did
not significantly affect neurological prognosis. Similarly,
Heary et al. [40] demonstrated that administration of either
methylprednisolone or dexamethasone regimens did not significantly improve the neurologic recovery of patients with
either complete or incomplete injuries compared with those
receiving no steroids. Interestingly, the incidence of pancreatitis was statistically greatest in patients who received
methylprednisolone, while gastrointestinal complications
were highest in those who received dexamethasone. Although these data were not randomized prospective analyses,
they offer compelling evidence that emergent corticosteroid
infusion has no role in the treatment of spinal cord injury
after gunshot wound. In a recent editorial of the literature,
Fehlings [42] reinforced recommendations that steroids
should not be administered after penetrating spinal cord
injuries.
Surgical timing
Timing of decompressive surgery may be an important
consideration. Although there is some support for early
decompression to improve neurologic outcome after blunt
trauma, it has not been similarly supported for gunshot
wounds. In a retrospective review of 88 patients, Cybulski
et al. [43] found equivalent rates of neurologic recovery in
patients undergoing decompressive laminectomy for conus
or cauda equina level lesions within 72 hours (47.5% improved) versus those operated more than 72 hours after injury
(48.1% improved). Similar studies of thoracic or cervical
injuries have not been performed.
Although early surgery does not appear to improve neurologic recovery rates, surgical timing does appear to affect
the incidence of other complications. Higher rates of infection and arachnoiditis have been documented when surgery
was performed more than 2 weeks from injury [43]. In
contrast, it has been suggested that cerebrospinal-cutaneous
fistulae may be more common after early or immediate laminectomy [14,34].
For these reasons, some surgeons recommend delaying
bullet removal, if warranted, until 5 to 10 days after injury
[14]. However, it should be highlighted that others think
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
237
Fig. 4. Gunshot wound to the spine. Fig. 3. (Left) Plain radiographs suggestive of an intracanal missile, demonstrating a midline bullet at the T12 level
after a gunshot wound to the back in a patient with a nonprogressive, incomplete neurologic deficit. (Right) Computed tomography images clearly show
the bullet within the spinal canal. The patient was taken to the operating room for a delayed posterior decompression. The bullet was easily removed.
that if bullet removal is indicated that it should be performed
immediately. Further clinical studies are needed to compare
the relative risks and benefits of early versus late bullet
removal after gunshot wounds to the spine. Regardless of
the level of injury, most agree that a documented progression
of neurologic deficit is an indication for urgent bullet
removal.
Gunshots to the cervical spine
There is a paucity of information regarding cervical gunshot wounds. Kupcha et al. [5] reviewed the records of 28
patients. Laminectomy was performed in four patients and
anterior corpectomy in one patient, with no neurological
improvement compared with nondecompressed cases. Neck
exploration was undertaken for vascular damage in four
cases, expanding hematoma in two cases and airway difficulty in three cases. Long-term complications were primarily
thromboembolism, pulmonary congestion and urinary tract
infection. Posttraumatic syrinx developed in two patients.
Fistulae occurred in two operatively and two nonoperatively
treated patients. Despite a lack of description of the antibiotic
regimen, only one case of spinal infection (meningitis) was
reported. Extended antibiotic prophylaxis is prudent after
pharyngeal, hypopharyngeal or airway violations. Neck
wound exploration should be based on the severity of extraspinal pathology. The cervical spine should be protected
as best as possible during intubation and surgical exploration,
while not inhibiting life-saving procedures. This may be
effected by gentle in-line traction, either manually or using
cranial tongs. However, if endoscopic or direct surgical neck
exploration can be safely delayed without endangering the
patient’s life, it is preferable to determine cervical stability first.
Decompression or bullet removal at the cervical cord
level is probably not useful in improving static neural deficit
(Fig. 5), whereas it may be beneficial in cases of neurologic
progression. Because of the possible potential neurologic
benefit, intracanal bullet removal may be pursued in cases
of incomplete neural injury, provided the bullet appears to
be relatively accessible and can be removed without extensive dural dissection. When indicated, transoral decompression of upper cervical (C1–C2) lesions has resulted in low
rates of postoperative infection, whereas laminectomy is
useful for more caudal levels [44]. Ultimately, the optimal
approach is dependent on the location of the fragment.
Gunshots with herniated discs
Herniated disc after a gunshot wound to the spine is a
rare cause for neurologic compromise [45]. By lodging in
the intervertebral space, it is postulated that the pressure
in the nucleus pulposus is increased. With a defect in
the posterior or posterolateral annulus, the disc material
can be expelled into the canal or foramen to compress
the cord or root. Treatment recommendations are similar
to those for other acute disc herniations. Significant, acute
neurologic deficit can be ameliorated with disc excision.
Herniations associated with a cauda equina syndrome are a
surgical emergency. Bullet removal is neither necessary nor
indicated unless it can be easily removed without further
jeopardizing surrounding neural or bony structures.
238
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
Fig. 5. Lateral cervical radiograph of a patient who sustained a gunshot
wound to the neck with the bullet entering from the posterolateral side.
The patient had a complete neurologic deficit at the C4 level. There were
no associated pharyngeal, tracheal or vascular injuries. Computed tomographic images demonstrated an intracanal bullet. The patient was treated
nonsurgically.
High-energy gunshot wounds
The authors strongly emphasize that the above recommendations are relevant to low-energy gunshot wounds.
High-energy rifle and shotgun wounds are associated with
significantly more soft tissue devitalization and a larger zone
of injury. From experience with US combat troops in
Panama, Parsons et al. [46] advocated aggressive surgical
debridement after such injuries. Likewise, Splavski and
others [47] reported low complication rates (14%) in spinal
injuries treated with wide debridement and decompressive
laminectomy during the Croatian War in the early 1990s.
Although not controlled comparative studies, the authors
highlighted that decompression did not likely affect neurologic recovery but that surgery was indicated to remove
devitalized tissues to decrease the likelihood of secondary
infection and sepsis. With increasing prevalence of gang and
terrorist acts worldwide, use of high-energy weapons has
necessitated adaptation of similar surgical tactics for some
civilian injuries.
[49]. Bullets in close proximity to facet joints or the intervertebral disc may more likely lead to intoxication. The diagnosis is often missed and requires a high level of suspicion.
Peripheral blood lead levels can detect increases, and bone
marrow biopsy can confirm hematopoietic alterations. Treatment with chelating agents, such as dimercaprol and calcium
disodium edetate, is initiated immediately followed by carefully planned bullet removal.
Besides neurologic deficit, pain is probably the most
common long-term complication of gunshot wounds with
spinal cord injury. In particular, conus medullaris– and cauda
equina–level lesions are associated with a high rate of pain
[50]. The incidence of pain is not reduced with bullet removal
[39]. Pain is mediated by spontaneous discharges from deafferentated dorsal horn neurons and can be conservatively
managed with oral neuroleptic medication, such as amitriptyline (Elavil, Merck & Co., Inc., Whitehouse Station, NJ)
or gabapentin (Neurontin, Pfizer, New York, NY). In unresponsive cases, a vascularized omental pedicle graft can
be surgically transplanted to the affected cord level. Increasing the neurovascularity of the region as well as altering
the neurotransmitter milieu relieves pain. As a last resort,
the dorsal root entry zone procedure surgically removes the
nociceptive dorsal rootlets. Spaic et al. [50] have documented excellent results with this maneuver in six patients
with pain recalcitrant to omental transplant after thoracolumbar gunshot wounds.
Neuropathic (Charcot) arthropathy may rarely occur in
spinal-cord-injured patients after gunshot wounds. It is typified by pain associated with progressive spinal deformity
at or below the level of injury. Treatment is directed toward
correction and stabilization of the spine through anterior,
posterior or combined surgical approaches [51–53].
Other long-term sequelae can occur. Bullet migration can
cause neural deficit months to years after the initial injury.
Kuijlen et al. [54] documented neurogenic claudication 11
years after a gunshot wound to the abdomen. The bullet
migrated from the paraspinal muscles at the L3 level into
the spinal canal where it effectively disintegrated into multiple fragments causing a diffuse inflammatory reaction. The
patient responded favorably to decompressive laminectomy
and bullet removal. Conway et al. [55] reported a case of
cauda equina syndrome after a bullet previously lodged in
the intervertebral disc migrated into the spinal canal 9 years
after injury. Numerous other reports of bullet migration have
been documented within the spinal canal, within the dura,
and in a caudal to cranial direction [56–58]. Bullet migration
is not always associated with neurologic deficit.
Summary
Late sequelae and complications
Lead intoxication is a rare complication of spinal gunshot
wounds [11,35,48]. Elution from bullets bathed in cerebrospinal fluid is rare. Synovial fluid is a more effective solvent
Gunshot wounds to the spine continue to be a significant
clinical problem. Although prevention is the best solution,
effective treatment relies on understanding the principles of
wound ballistics, tissue injury and the role of diagnostic
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
modalities. Wounds from bullets passing through the colon
before the spine require at least 7 days of broad-spectrum
antibiotics. Bullet removal caudal to the T11 vertebra can
improve motor recovery. Regardless of the level of injury,
corticosteroid infusion has little effect on neurologic improvement. Nerve pain is a common long-term sequela and
can be responsive to conservative and surgical treatment.
Lead intoxication and bullet migration are infrequent causes
of late morbidity and require a high index of suspicion for
accurate diagnosis.
References
[1] Farmer J, Vaccaro A, Balderston R, Albert T, Cotler J. The changing
nature of admission to a spinal cord injury center: violence on the
rise. J Spinal Disord 1998;11:400–3.
[2] Roye W, Dunn E, Moody J. Cervical spinal cord injury—a public
catastrophe. J Trauma 1988;28:1260–4.
[3] Lin S, Vaccaro A, Reich S, Devine M, Cotler J. Low-velocity gunshot
wounds to the spine with an associated transperitoneal injury. J Spinal
Disord 1995;8:136–44.
[4] Bishop M, Shoemaker W, Avakian S, James E, Jackson G. Evaluation
of a comprehensive algorithm for blunt and penetrating thoracic and
abdominal trauma. Am Surg 1991;57:737–46.
[5] Kupcha P, An H, Cotler J. Gunshot wounds to the cervical spine.
Spine 1990;15:1058–63.
[6] Isiklar Z, Lindsey R. Gunshot wounds to the spine. Injury 1998;
29:SA7–SA12.
[7] Yoshida G, Garland D, Waters R. Gunshot wounds to the spine.
Orthop Clin North Am 1995;26:109–16.
[8] Simpson R, Venger B, Fischer D, Narayan R, Mattox K. Shotgun
injuries of the spine: neurosurgical management of five cases. Br J
Neurosurg 1988;2:321–6.
[9] Turgut M, Ozcan O, Gucay O, Saglam S. Civilian penetrating spinal
firearm injuries of the spine. Results of surgical treatment with
special attention to factors determining prognosis. Arch Orthop
Trauma Surg 1994;113:290–3.
[10] Zawitz M. Guns used in crime. Washington, DC: US Department of
Justice, Bureau of Justice Statistics Selected Findings, 1995:NCJ148201.
[11] Grogan D, Bucholz R. Acute lead intoxication from a bullet in an
intervertebral disc space. J Bone Joint Surg 1981;63A:1180–2.
[12] Cushid J, Kopeloff L. Epileptogenic effects of metal powder implants
in the motor cortex in monkeys. Int J Neuropsychiatr 1968;3:24–8.
[13] Tindel N, Marcillo A, Tay B, Bunge R, Eismont F. The effect of
surgically implanted bullet fragments on the spinal cord in a rabbit
model. J Bone Joint Surg 2001;83A:884–90.
[14] Eismont FJ. Gunshot wounds of the spine. In: Levine AM, Lampert R, Garfin SA, Eismont F, editors. Spine trauma. Philadelphia: WB
Saunders, 1998:525–43.
[15] Fetterman B, Shindo M, Stanley R, Armstrong W, Rice D. Management of traumatic hypopharyngeal injuries. Laryngoscope 1995;105:
8–13.
[16] Kihtir T, Ivatury R, Simon R, Nassoura Z, Leban S. Early management
of civilian gunshot wounds to the face. J Trauma 1993;35:569–75.
[17] Kennedy F, Gonzalez P, Beitler A, Sterling-Scott R, Fleming A. Incidence of cervical spine injury in patients with gunshot wounds to the
head. South Med J 1994;87:621–3.
[18] Cornwall EE, Chang DC, Bonar JP, et al. Thoracolumbar immobilization
for trauma patients with torso gunshot wounds: is it necessary? Arch
Surg 2001;136:324–7.
[19] Kumar A, Wood G, Whittle A. Low-velocity gunshot injuries of the
spine with abdominal viscus trauma. J Orthop Trauma 1998;12:514–7.
239
[20] Kihtir T, Ivatury R, Simon R, Stahl W. Management of transperitoneal
gunshot wounds of the spine. J Trauma 1991;31:1579–83.
[21] Roffi R, Waters R, Adkins R. Gunshot wounds to the spine associated
with a perforated viscus. Spine 1989;14:808–11.
[22] Romanick P, Smith T, Kopaniky D, Oldfield D. Infection about the
spine associated with low-velocity missile injury to the abdomen. J
Bone Joint Surg 1985;67:1195–201.
[23] Naude G, Bongard F. Gunshot injuries of the sacrum. J Trauma 1996;
40:656–9.
[24] Chong C, Ware D, Harris J. Is cervical spine imaging indicated in
gunshot wounds to the cranium? J Trauma 1998;44:501–2.
[25] Finitsis S, Falcone S, Green B. MR of the spine in the presence of
metallic bullet fragments: is the benefit worth the risk? Am J Neuroradiol 1999;20:354–6.
[26] Ebraheim N, Savolain E, Jackson W, Andreshak T, Rayport M. Magnetic
resonance imaging in the evaluation of a gunshot wound to the cervical
spine. J Orthop Trauma 1989;3:19–22.
[27] Bashir E, Cybulski G, Chaudhri K, Choudhury A. Magnetic resonance
imaging and computed tomography in the evaluation of penetrating
gunshot injury. Spine 1993;18:772–3.
[28] Gustilo R. Current concepts in the management of open fractures.
Inst Course Lect 1987;36:359–66.
[29] Hales D, Duffy K, Dawson E, Delamarter R. Lumbar osteomyelitis
and epidural and paraspinous abscesses. Case report of an unusual
source of contamination from a gunshot wound to the abdomen. Spine
1991;16:380–3.
[30] Bateman N, Jones N. Rhinorrhoea feigning cerebrospinal fluid leak:
nine illustrative cases. J Laryngol Otol 2000;114:462–4.
[31] Wigle R. Treatment of asymptomatic gunshot injuries to the spine.
Am Surg 1989;55:591–5.
[32] Gentleman D, Harrington M. Penetrating injury of the spinal cord.
Injury 1984;16:7–8.
[33] Denis F. The three columns of the spine and its significance in the
classification of acute thoracolumbar spine injuries. Spine 1983;
8:817–31.
[34] Stauffer E, Wood R, Kelly E. Gunshot wounds of the spine: the effects
of laminectomy. J Bone Joint Surg 1979;61A:389–92.
[35] Linden M, Manton W, Stewart R, Thal E, Feit H. Lead poisoning from
retained bullets. Pathogenesis, diagnosis, and management. Ann Surg
1982;195:305–13.
[36] Young J, Burns P, Bowen A. Spinal cord injury statistics: experience
of the regional spinal cord injury systems. Phoenix: Good Samaritan
Medical Center, 1982:1–152.
[37] Ciappetta P, Salvati M, Raco A, Artico M. Cruciate hemiplegia: a
clinical syndrome, a neuroanatomical controversy. Report of two cases
and review of the literature. Surg Neurol 1990;34:43–7.
[38] Robertson D, Simpson R. Penetrating injuries restricted to the cauda
equina: a retrospective review. Neurosurgery 1992;31:265–9.
[39] Waters R, Adkins R. The effects of removal of bullet fragment retained
in the spinal canal. A collaborative study by the National Spinal
Cord Injury Model Systems. Spine 1991;16:934–9.
[40] Heary R, Vaccaro A, Mesa J, et al. Steroids and gunshot wounds to
the spine. Neurosurgery 1997;41:576–83.
[41] Levy M, Gans W, Wijesinghe H, SooHoo W, Adkins R, Stillerman C. Use
of methylprednisolone as an adjunct in the management of patients
with penetrating spinal cord injury: outcome analysis. Neurosurgery
1996;39:1141–8.
[42] Fehlings MG. Editorial: recommendations regarding the use of methylprednisolone in acute spinal cord injury: making sense out of the
controversy. Spine 2001;26:S56–7.
[43] Cybulski G, Stone J, Kant R. Outcome of laminectomy for civilian
gunshot injuries of the terminal spinal cord and cauda equina:. Neurosurgery 1989;24:392–7.
[44] Maniker A, Gropper M, Hunt C. Transoral gunshot wounds to the
atlanto-axial complex: report of five cases. J Trauma 1994;37:858–61.
[45] Robertson D, Simpson R, Narayan R. Lumbar disc herniation from a
gunshot wound to the spine. A report of two cases. Spine 1991;16:
994–5.
240
C.M. Bono and R.F. Heary / The Spine Journal 4 (2004) 230–240
[46] Parsons T, Lauderman W, Ethier D, et al. Spine injuries in combat
troops—Panama, 1989. Mil Med 1993;158:501–2.
[47] Splavski B, Vrankovic D, Saric G, Blagus G, Mursic B, Rukovanjski M. Early management of war missile spine and spinal cord injuries:
experience with 21 cases. Injury 1996;27:699–702.
[48] Cagin C, Diloy-Puray M, Westerman M. Bullets, lead poisoning, and
thyrotoxicosis. Ann Intern Med 1978;89:509–11.
[49] Machle W. Lead absorption from bullets lodged in tissues. JAMA
1940;115:1536–41.
[50] Spaic M, Petkovic S, Tadic R, Minic L. DREZ surgery on conus
medullaris (after failed implantation of vascular omental graft) for
treating chronic pain due to spine (gunshot) injuries. Acta Neurochir
1999;141:1309–12.
[51] Vaccaro AR, Silber JS. Post-traumatic spinal deformity. Spine 2001;
26:S111–8.
[52] Arnold PM, Baek PN, Stillerman CB, Rice SG, Mueller WM. Surgical
management of lumbar neuropathic spinal arthropathy (Charcot joint)
One Hundred
Forty Years
Ago in
Spine
In 1864, in London, Bernard Edward
Brodhurst, who has become known more for his contributions to early advancements in knee arthrotomy
[53]
[54]
[55]
[56]
[57]
[58]
after traumatic thoracic paraplegia: report of two cases. J Spinal Disord
1995;8:357–62.
Brown CW, Jones B, Donaldson DH, Akmajkian J, Brugman JL. Neuropathic (Charcot) arthropathy of the spine after traumatic spinal paraplegia. Spine 1992;17:S103–8.
Kuijlen J, Herpers M, Beuls E. Neurogenic claudication, a delayed
complication of a retained bullet. Spine 1997;22:910–4.
Conway J, Crofford T, Terry A, Protzman R. Cauda equina syndrome
occurring nine years after a gunshot injury to the spine. A case repo.
J Bone Joint Surg 1993;75:760–3.
Gupta S, Senger R. Wandering intraspinal bullet. Br J Neurosurg
1999;13:606–7.
Jeffery J, Borgstein R. Case report of a retained bullet in the lumbar
spinal canal with preservation of cauda equina function. Injury
1998;29:724–6.
Oktem I, Selcuklu A, Kurtsoy A, Kavuncu I, Pasaoglu A. Migration
of bullet in the spinal canal: a case report. Surg Neurol 1995;44:548–50.
and treatment of congenital dislocation of the hip, wrote a
monograph on scoliosis [1]. He proposed the idea that
curvature progresses inexorably until finally an ankylosis
between the vertebrae occurs and the curve progression
stops. The idea led Hibbs 50 years later to the concept
of surgically induced ankylosis by fusion to prevent
curve progression.
Reference
[1] Brodhurst BE. On lateral curvature of the spine. London: John
Churchill and Sons, 1864.