Incidence and Classification

Currently, the annual incidence of spinal cord injury (SCI) in the United States is estimated to be 40 per 1 million people, or 12,000 new cases each year^5,6 The majority of SCIs are the result of motor vehicle accidents and falls; however, violence (primarily gunshot wounds) results in 15% of injuries. Prior to 1980, violence accounted for 13.3% of SCIs, and this rate peaked in the 1990s at 24.8% before declining to 15.1% in 2005.⁶ In less developed countries or those with strict gun laws, nonmissile PSIs are proportionally more common.⁷ According to the National Spinal Cord Injury Database, the mean age of PSI is 29.7 years, with a 4:1 male predominance.⁸

As with any SCI, PSIs should be described by both anatomic level and degree of neurologic impairment. Anatomically, 20% of injuries are reported in the cervical spine, 50% in the thoracic spine, and 30% in the lumbar spine, which partially reflects the relative lengths of the spine segments.^8–13 Approximately 40% of patients are shot in the back, while 19% are shot anteriorly.¹⁴ Management and prognosis of injuries is highly dependent on associated injuries, which vary based on trajectory. Associated injuries are reported in 25% of civilian missile (handgun) injuries and 67% of military missile (high-velocity rifle) injuries.² Injuries to the neck, chest, and abdomen take precedence over spine injury during initial management if vascular or visceral injury threatens patient survival.

It must be determined whether the injury is complete or incomplete and whether it involves the cauda equina. If incomplete or involving the cauda equina, motor, sensory, and sphincter function below the level of injury must be recorded. The American Spinal Injury Association’s International Standards for Neurologic Classification of SCI provides the best assessment of neurologic impairment.¹⁵ Patients injured by missiles are more likely to suffer a complete SCI than those suffering from nonmissile injuries. Of patients suffering from missile injuries as reported in the literature, 49% to 83% had complete injuries, 12% to 43% had incomplete injuries, and 17% to 20% had cauda equina injuries.^{10,14,16–18} Of patients suffering from nonmissile injuries, 21% to 33% had complete injuries and 67% to 79% had incomplete injuries.^10–12 Data comparison is limited, because some authors did not consistently categorize injuries to the cauda equina separately.

Clinical Evaluation

In any PSI, initial evaluation must follow advanced trauma life support (ATLS) guidelines; thus, neurologic examination can occur only after the patient is hemodynamically stable and airway control has been secured. Until the patient is clinically and radiographically cleared, care should be taken to avoid unnecessary movement of the spine. Thus, depending on the situation, a cervical collar, backboard, log-roll, and awake fiberoptic intubation should be considered to avoid iatrogenic neurologic deterioration. Spinal instability is a rare complication of gunshot wounds and is even less common with stab wounds.¹⁹ Some centers advocate not immobilizing the patient in the prehospital setting, given that immobilization may complicate or delay care of medically urgent associated injuries.^5,20–22

Following the primary survey, an accurate neurologic examination should be performed, along with subsequent serial evaluations. Confounding factors such as intubation, associated injuries, alcohol or drug impairment, and traumatic brain injury may make examination difficult. In addition, by merely observing in the clinical examination, entry and exit points may be misleading, as bullets and fragments may ricochet or cavitation may cause more damage than initially anticipated.

Appropriate imaging studies are needed to fully understand the path of the penetrating object and the resultant anatomic damage, even in asymptomatic patients.²³ Plain radiographs may be helpful in identifying metallic fragments, even if an exit wound is clearly seen (Fig. 180-1). Computed tomography (CT) is useful to assess bone damage and canal compromise (both by fracture and by weapon fragments) and to give an initial evaluation of alignment. CT myelogram may also be useful in determining canal and foraminal compromise, or root avulsion injuries. Unfortunately, metal artifact from bullet/knife fragments may make these tests difficult to interpret; yet these same fragments would likely preclude magnetic resonance imaging (MRI). MRI may be useful to assess ligamentous and spinal cord damage, as well as hemorrhage in all compartments (extradural, intradural, and intramedullary) if it is verified that there are no metal fragments. Further tests to determine spinal stability may be needed if clinical or radiographic evidence suggests this may be a problem. Injuries with a “side-to-side” trajectory may be more likely to cause instability.²⁴ Clinical judgment regarding additional tests, including thoracic, abdominal, and vascular imaging studies, should be exercised based on the clinical picture and the suspected trajectory of the penetrating object. In cases of suspected peripheral nerve injury, electromyelography and nerve conduction studies are seldom useful in the acute setting.

FIGURE 180-1 Anteroposterior (A) and lateral plain (B) radiographs of the cervical spine in a 41-year-old man who was rendered immediately quadriplegic following a gunshot wound to the neck. A large-caliber bullet is present at C5-C6. No surgical intervention was undertaken.

Clinical Management

In the majority of patients suffering from PSI, conservative (nonoperative) treatment is indicated. Historically, surgical indications included decompression and débridement of neural tissue, removal of bony or metallic fragments, dural repair, correction of spinal instability, and relief of pain. Currently, absolute indications are probably limited to spinal instability, intractable CSF leak, and clear demonstration of a projectile, bone fragment, or hematoma that is resulting in ongoing deficit or pain. Once the patient has undergone a thorough evaluation, operative versus conservative management is decided case by case. The main factors influencing this decision are the extent of neural element compression and the patient’s potential for functional neurologic recovery.

Surgical decompression should be considered in cases of intradural or extradural neural compression from the missile, bone, disc, hematoma, gross contamination, or in-driven foreign bodies. Although PSI may result in spinal instability, this is rare. More relevant is iatrogenic instability resulting from a surgical procedure to decompress the neural elements. Such a situation must be addressed during preoperative planning. Patients with known or planned iatrogenic instability should be considered for stabilization or fusion. Removal of debris or washout of the wound to prevent infection is usually not indicated in missile injuries, because the explosive force of the firing weapon effectively “sterilizes” the bullet. The exception may be stab wounds or cases in which there is gross contamination.²⁵

The degree of neurologic compromise and potential for functional recovery is important in determining the aggressiveness of treatment. In patients who are neurologically well, local wound care alone may be appropriate. In patients with complete injuries, surgery is usually not performed unless they have spinal instability or a late complication. In patients with incomplete deficits, surgery in the setting of a compressive lesion may result in functional recovery and should be considered (Fig. 180-2).

FIGURE 180-2 Axial CT of the upper thoracic spine that demonstrates intracanalicular bone and bullet fragments. This 54-year-old man was running from his assailants when he was shot in the back. The bullet entered the interscapular region on the left side at the T5 level, resulting in paraparesis. After a T3 to T6 laminectomy with removal of bone and bullet fragments, he had partial, but remarkable, motor and sensory recovery.

The cauda equina is a special case, as root sparing and recovery is comparable to a peripheral nerve injury (Figs. 180-3 and 180-4). Also, almost all patients enjoy some recovery of nerve root function regardless of management.¹⁴ Thus, some practitioners recommend conservative treatment for these injuries,^18,26 while others favor surgical intervention.^13,16,27 The advantage of surgical intervention is the removal of blood products and necrotic debris, which may decrease the incidence of scarring, arachnoiditis, and chronic pain syndromes.

FIGURE 180-3 Axial (A) and sagittal reformatted (B) CT scans of the upper lumbar spine in a 24-year-old man who had a complete L3 paraplegia immediately following PSI. The scans demonstrate the intraspinal location of bone and bullet fragments. After a two-level lumbar laminectomy, removal of in-driven fragments, and primary dural closure, the patient regained a unilateral sensory level.

FIGURE 180-4 Anteroposterior (A) and later plain (B) radiographs of the lumbar spine of a 25-year-old man with right lower extremity weakness following a gunshot wound to the spine. Surgery was undertaken, but the patient did not regain function.

In some cases, radiographic studies may not explain the neurologic deficit. In high-velocity missile injury, such as that seen from a military rifle bullet, concussive effects can damage neural tissue without the missile penetrating the spinal canal.^28–31 In these cases, supportive and local wound care is appropriate. In addition, if a patient develops a progressive deficit without evidence of a new compressive lesion on MRI (e.g., a hematoma or movement of a debris fragment), surgery is not likely to result in improvement.^17,32

Involvement of surrounding structures may be important in surgical decision making. Bone and missile fragments in the foramen can result in radicular symptoms. Pain relief and functional recovery may occur if these fragments are removed.^33–35 In patients in whom the bowel has been perforated, infections can occur, and in addition to the repair of the bowel perforation, a brief course of antibiotics is recommended. Colonic perforations are at slightly greater risk for infection, thus more aggressive antibiotic coverage is appropriate.^36–39 Some authors have advocated débridement of the bone and bullet fragments in the spinal canal if the patient has an associated colonic injury, although we do not favor routine neurosurgical intervention in patients with transperitoneal or transintestinal PSI. In most cases of PSI, broad-spectrum antibiotics and tetanus prophylaxis are the sole indicated therapies.⁴⁰

Late complications from PSI may also warrant surgical intervention. A dural rent from PSI may cause CSF leak. In most cases, this resolves; however, persistent leaks (more than 4 days) may require exploration and closure. The goal of intervention is to prevent meningitis or prevent accumulation of fluid, as in the case of a leak into the pleural cavity. Infectious processes such as discitis, osteomyelitis, or abscess can result in late neural compression and subsequent deterioration and may warrant surgical intervention. In most cases of late complications, further imaging studies are warranted to provide the clinician with up-to-date information.

The decision to perform surgical exploration of PSIs should also be carefully weighed against the possibility of difficult dural and tissue closure that may lead to persistent postoperative CSF leaks. In our experience, such situations may then lead to a series of complications, including the necessity for additional surgery or infection. Therefore, in most instances, unless a clear compressive lesion is identified (i.e., hematoma or bone or bullet fragment), nonsurgical management is the more prudent strategy for PSIs.

Nonsurgical Management

The majority of PSIs require nonsurgical management. In these cases, local wound care and supportive care are sufficient. The wound should be thoroughly cleaned and followed closely for signs of CSF leak or infection, especially if a dural injury has occurred. In addition, prophylactic antibiotics should be given for 24 to 48 hours once the wound has stopped draining. Skin, urinary, excretory, and pulmonary function should be monitored closely, as with any SCI. In addition, patients should begin a comprehensive SCI rehabilitation program as soon as possible. Steroid use in PSI is controversial. Although benefit has been seen in closed SCI, PSI patients were excluded from the National Spinal Cord Injury Study II.⁴¹ In retrospective studies, no benefit has been seen to steroid administration, and steroid administration is not recommended.^42,43

Surgical Treatment

The goal of surgical treatment is to decompress the neural elements, to repair dural tears to prevent CSF leak, and to stabilize the spine. However, surgery for decompression of the neural elements is seldom beneficial because the injury to the nervous system is most commonly due to the direct mechanical injury or the concussive effect of the penetrating element. Surgery for compressive hematoma is indicated but is an uncommon scenario. Likewise, removal of a projectile may be effective when it causes direct compression to neural elements. CSF leaks are seldom an issue. In our experience, surgery will likely increase the chances of a postoperative incisional CSF leak when compared to nonoperative cases. Therefore, surgical stabilization of an unstable spine is an absolute indication for surgery, but penetrating injuries to the spine seldom result in unstable injuries. Surgical planning should be individualized in each case (Fig. 180-5).

FIGURE 180-5 Management of penetrating SCI.

Aside from timing and surgical approach considerations, the surgeon must consider additional injuries and the extent of preoperative wound contamination. Wound contamination may affect antibiotic choice and the timing of instrumentation placement. Electrophysiologic monitoring can also be considered at the time of surgery for incomplete SCIs.

The timing of surgery remains controversial. While data supports early evacuation of hematomas in head injuries to reduce neural injury and optimize recovery,⁴⁴ this may not be directly applicable to PSI. Early surgery, defined as laminectomy performed within 3 days of injury, has not been clearly shown to improve neurologic function.⁴⁵ In addition, some authors note a high rate of CSF leaks and infection in cases operated on within 7 days of injury.⁴⁶ The former is likely due to poor wound closure, and the latter likely results from contamination at the time of injury. This has led some authors to suggest that surgery be delayed beyond 7 days. However, unlike the head injury study described previously, there are no data involving surgeries performed within hours of the injury.

In most cases, a posterior approach to the spine is used and a laminectomy is performed for decompression (Fig. 180-6). If the compressive fragment is located dorsally, and there is no dural incursion, the fragment may be removed once adequate bone has been resected. If intradural exploration is needed, a longitudinal durotomy can be made, or if a tear has occurred, this may be extended to provide adequate exposure of the intradural contents. In the rare circumstance that intramedullary exposure is needed, a longitudinal myelotomy may be performed, preferably midline to reduce injury to the posterior columns. The surgical durotomy should be closed in a watertight fashion, along with all accessible dural tears (Fig. 180-7). Patch grafting may be necessary.

FIGURE 180-6 A and B, Gunshot wound to the lumbosacral region, resulting in compressive lesions with complete cauda equina deficit. C, Transection of the cervical spinal cord with an associated intramedullary hematoma following PSI.

FIGURE 180-7 A, The dorsal approach to the cervical region begins with subperiosteal dissection of paraspinous muscles from the laminae extending a couple of levels above and below the injured level. The normal laminae above and below are exposed first, and then exposure at the level of injury is accomplished with careful dissection. This is to prevent (1) inadvertent injury to herniated neural tissue and (2) bone fragments or surgical instruments from being accidentally pushed into the bony defect. B, After completing the laminectomy, the dura is opened and the cord is exposed at the level of injury. C, To adequately expose an intramedullary hematoma, it is often necessary to perform a dorsal midline myelotomy. The clot may then be evacuated. D, Using continuous sutures, a watertight dural closure is performed. A dural graft may be needed.

Anterolateral lesions may be approached by the anterior approach in the cervical spine and an anterior, anterolateral, posteroanterolateral, posterolateral, or lateral extracavitary approach in the thoracic and lumbar spine. Unfortunately, ventral dural closure is often quite difficult and may require a complex repair. In cases of dural tears, a lumbar drain should be considered to reduce CSF pressure and divert flow until the defect heals. All wounds should be closed in a multilayer, watertight fashion (Fig. 180-8).

FIGURE 180-8 A to D, The anterolateral approach to the thoracic spine first involves placing the patient in the strict lateral position. Starting in the midaxillary line, a skin incision is made and carried posteriorly, centered over the interspace that is one level above the level of interest. A standard thoracotomy is performed. After self-retaining retractors are placed, the lung is collapsed and packed off. After resecting the posterior 5 to 10 cm of the rib at the injured level, the head of the rib is drilled away. The pleura is the opened, and the spine is exposed.

A 5% to 10% complication rate has been reported following the surgical management of PSI.^33–35 Complications include infection, CSF fistulas, pseudomeningocele formation, and iatrogenic worsening of neurologic deficit. Spinal instability and delayed myelopathy can also occur. In three series, the complication rate of patients undergoing surgery was much higher, and 16% to 22% of patients developed postoperative meningitis, wound infections, CSF leak, pseudomeningocele, or spinal instability.^10,17,47 It may be possible to reduce this with appropriate patient selection, judicious use of antibiotics, and careful surgical technique. More recently, three studies noted a 0% mortality rate.^14,33,45

Postoperatively patients should be carefully followed and should begin a standard protocol for SCI. Extubation should occur only if the patient meets all weaning parameters, has an injury level below C5, and does not have head or chest injuries. As with all SCI patients, attention must be paid to pulmonary, urinary, excretory, and skin care. In addition, prophylactic minidose heparin can be considered to prevent deep venous thrombosis.⁴⁸ Rehabilitation efforts should begin as soon as possible.

Results

Regardless of treatment, prognosis correlates highly with severity of the initial deficit.^16,18,47 In patients with immediate, complete deficits following PSI, prognosis is generally poor. For that reason, they are not usually considered surgical candidates. Yashon et al. found no improvement in operated or nonoperated patients with complete deficits immediately after a single bullet wound to the spinal cord above the level of the conus (Fig. 180-9).¹⁸ In addition, there is no convincing evidence that surgery improves outcome in patients with incomplete deficits following PSI. Multiple retrospective, nonrandomized studies demonstrated similar rates of improvement between operated and nonoperated patients.^10,16–18 However, the possibility that surgery could reduce the risk of arachnoiditis due to the presence of necrotic tissue and blood or reduce the incidence of chronic pain has yet to be determined.

FIGURE 180-9 Sagittal (A) and axial (B) reconstruction of the thoracic spine of a man rendered quadriplegic following a gunshot wound. No surgical intervention was undertaken.

Cauda equina injuries appear to have a more favorable prognosis than other types of PSI, although the degree to which surgery benefits these patients has not been determined. In patients with both complete and incomplete penetrating injuries of the cauda equina, some authors report that 47% to 100% of patients improved following surgery. Even in patients with complete injuries, improvement was seen in 34% of patients following laminectomy and débridement.⁴⁵ Thus, it is recommended that patients with PSI limited to the cauda equina undergo surgery. However, it is important to document that the fragment or projectile is clearly causing the presenting deficit.

Conclusions

The role of surgical intervention in PSI has yet to be examined in a prospective, randomized trial; thus, surgeons must use retrospective studies, anecdotal evidence, and personal prejudices to determine management strategies for individual patients. Surgical intervention appears to benefit patients with cauda equina injuries, spinal instability, radiculopathies due to bullet or bone fragments, and incomplete lesions. In those with complete lesions, the benefits of surgery are less clear. The overall risk–benefit ratio has yet to be defined.