The management of complex spine fractures with incomplete spinal cord injury (SCI) in the severely injured polytrauma patients presents the spine surgeon with a difficult clinical scenario.¹ These patients often have severe comorbid conditions such as pulmonary contusions, metabolic acidosis, long bone fractures, or acute respiratory distress syndrome. Controversy exists as to the optimal treatment and timing of intervention for unstable spine fractures with neurologic involvement in the severely injured.^2–9 For example, on June 30, 2006, a 74-year-old man was the pilot of a helicopter that crashed after striking power lines. He suffered severe injuries including an open tibia/fibular fracture, complex pelvic fractures, abdominal compartment syndrome requiring emergent laparotomy, bilateral hemopneumothoraxes, traumatic subarachnoid hemorrhage, burst fracture of the L1 and L2 vertebrae, and T12 vertebral body fracture (Fig. 228-1). Neurologically, the patient had no rectal tone but was able move his feet. The patient presented to the ICU after his laparotomy in acute respiratory distress syndrome and was placed on a RotoRest bed (KCI, San Antonio, TX) for ventilatory support. The patient was eventually paralyzed, and an intracranial pressure monitor was placed for declining neurologic exam. The patient underwent tracheostomy and bilateral chest tube placement for pulmonary effusions. He remained in the ICU for nearly 4 weeks. He was not thought to be stable enough to tolerate a prolonged prone procedure for internal reduction and fixation of his thoracolumbar fractures. This scenario is all too common in modern level 1 trauma centers.

FIGURE 228-1 A 75-year-old pilot sustained several injuries including burst fractures of L1 and L2 and paraparesis after crashing his helicopter while filming a movie. Axial CT scans of L1 (A) and L2 (B) show the fractures with retropulsion of bone into the canal. The MRI (C) demonstrates the fractures with canal narrowing, posterior root herniation at L2, and large subcutaneous hematoma. A follow-up anteroposterior radiograph (D) 16 months postoperatively demonstrates instrumentation.

Early reports from Frankel et al.,⁹ Guttmann,¹⁰ and Bedbrook^11,12 were heavily weighted toward nonoperative treatment. In recent years, nonoperative treatment of intact thoracolumbar fractures has been shown to be associated with excellent results.^13–19 However, there have been few randomized studies comparing early versus late surgery in SCI. The study by Vaccaro et al.⁸ involved 34 patients with cervical spine fractures that were randomized to early (<72 hours) surgery and 28 that were randomized to late (>5 days). Unfortunately, the follow-up at 1 year was available on only 23 and 19 patients, respectively. The unanticipated finding by the authors was that there was no significant difference in outcome between the two groups. A retrospective review by Mirza et al.⁶ involved 43 patients with cervical spine injuries treated at two facilities. Their findings demonstrated improved outcome with early surgery (<72 hours) compared to delayed surgery performed 10 to 14 days after injury. No increased complications were noted in the early surgery group.

Similarly, conflicting results with timing of surgery have been encountered with thoracolumbar fractures. A retrospective study involving thoracolumbar fractures divided patients into three surgical groups: early surgery (<8 hours), more than 8 hours after surgery, and after 10 days.² There were 26, 50, and 12 patients, respectively, in the three groups with an average follow-up of 5.6 years. Using the Frankel score, this study revealed better outcomes in the early surgery group compared to late surgical groups. There was also better pain control with early rather than delayed surgery. Complication rates did not differ between the three groups. McLain and Benson⁵ conducted a nonrandomized prospective study in thoracolumbar fractures. Fourteen patients underwent urgent surgery within 24 hours, and 13 patients underwent delayed surgery between 24 and 72 hours following injury. At 2-year follow-up, no significant difference in neurologic outcome between the two groups was noted. There was also no difference between groups in rates of complications, mortality, or morbidity. Cengiz et al.,²⁰ in a prospective, nonrandomized study, compared early stabilization (<8 hours) in 12 patients with thoracolumbar fractures with late (>3 days) stabilization in 15 patients. A statistically significant improvement in motor recovery was noted with early surgery in partial SCI patients. Hospital stays were shorter with early surgery, and the rate of complications appeared higher in the delayed-surgery group.

With incomplete SCI, neurologic improvement has been reported with both operative^21–25 and nonoperative^{13–19,26,27} techniques, although no consistent improvement has been reported with either modality. Neurologic deficit is more likely due to contusion of the spinal cord at the time of injury and not from persistent bony compression of neural structures. The complications of surgical intervention are by no means negligible and include infection, hardware failure, the need for subsequent surgery, and deep vein thrombosis. These complications in general have exceeded those encountered in recumbency.^{17,22,23,25,27,28}

Owing to the severe pulmonary injury sustained by the patient discussed previously, spine surgery was delayed. The patient was immobilized throughout his intubation period and, once extubated, was fitted with a thoracolumbar spine orthosis and transferred to a rehabilitation facility. On follow-up, the patient remained paraparetic and incontinent of urine with poor rectal tone; he also complained of radicular pain bilaterally. Nearly 3 months following injury, when he had failed to regain the ability to walk, the patient was evaluated by neurosurgery and underwent L2-3 laminectomy, repair of the dural rent, and reduction of the herniated roots, with T12-L4 pedicle screw and rod fixation. Eighteen months postoperatively, the patient had improvement in his pain, had regained bladder but not bowel control, and was ambulatory with ankle braces and the assistance of one person.

The management of complex spine fractures, be they cervical, thoracic, or lumbar, in polytrauma patients remains a challenging arena for the spine surgeon.^1,7 The choice of treatment should be individualized to the patient’s medical condition and ability to undergo often lengthy prone spine procedures in the face of multiple comorbidities. Though class I evidence in favor of early surgery for spine fractures is sparse,²⁹ our policy has been to proceed with decompression and stabilization of spine fractures, particularly with deficits, as soon as the patient’s condition allows. Recumbency for thoracolumbar fractures and bracing or traction for cervical fractures can be used as a bridge to surgical stabilization when it is deemed safe and appropriate.

Acute spinal cord injury (SCI) has momentous repercussions in our society. Every year, more than 10,000 Americans are victims of SCI and become physically disabled and psychologically affected.¹ Without consideration for loss of productivity and income, the estimated cost of care of complete and incomplete SCI patients in the United States is estimated at $28,000 and $17,000 per patient per year,² respectively. Thus, the optimization of SCI treatment may imply a greater number of SCI patients with timely neurologic recovery, potential return to work, and reinstatement of wages.

Patients with incomplete SCI generally have a higher functional recovery rate than do patients with complete SCI.^3–5 Current neurosurgical and orthopedic practices therefore favor a more aggressive approach and early surgical intervention in this subset of SCI patients.^3,6–8 However, because of the lack of well-designed prospective randomized control trials, there remains a controversy in determining the optimal timing of surgery for SCI patients^3,7–9 Given the emotional driving forces surrounding the treatment of SCI patients, we strongly recommend early surgical intervention, as it comes with minimal risk but high hope of recovery.

Recent laboratory studies have established that SCI is a dynamic process with temporal evolution of pathophysiologic processes.^10–13 Thus, there are primary and secondary injuries that contribute to the progression of acute SCI. The primary injury corresponds to the immediate effect of trauma with the mechanical spinal cord compression and immediate cellular and axonal damage. The secondary injury refers to the effect of the inflammatory cascades, ischemia, and oxidative damage on the spinal cord parenchymae.¹⁴

The initiation of emergency surgical intervention is to prevent or limit secondary injury. However, these effects most likely take place in the first few hours following trauma. Therefore, the critical factor that remains unclear is what time interval defines the surgical window during which secondary injury is still reversible and whether surgery can be performed in the time frame necessary despite the numerous logistical challenges in the trauma population.

Despite significant interest and research in the treatment of traumatic spinal cord injury (SCI), including advances in modern surgical and nonsurgical techniques to manage SCI, the prognosis for neurologic recovery in patients with SCI is still poor. In the United States alone, SCI affects 10,000 to 12,000 people annually.^1–3 Tragically, these injuries, which are the result of motor vehicle accidents, falls, and athletic injuries, disproportionately affect younger patients.⁴ SCI shortens the lives of those that it affects in proportion to the degree of injury.⁵ Economically, SCI contributes to the loss of many productive years by affected individuals and society at large with significant direct costs, estimated at $80,000 to $120,000 per patient annually for an estimated yearly direct liability of $4 billion in the United States.⁶

The role for and timing of surgical decompression and stabilization in the setting of traumatic SCI have been subject to much scrutiny and opinion in the medical literature for the last century. Despite multiple attempts to study the appropriate timing of surgical intervention, no overwhelming consensus has developed. Indeed, the variety of opinion in the literature has demonstrated that there is no clear correct answer to this dilemma. Despite our best efforts, the mechanism and natural history of SCI are still poorly understood, and research protocols to fully understand the effects of interventional therapies in human subjects are subject to ethical challenges that may make experimental determination of the correct treatment of such injuries virtually impossible. Therefore, however imperfect the situation may be, reliance on the current, largely retrospective, clinical literature and experiments done in animal models is the basis of modern treatment paradigms.

Indirect evidence provided by pharmacologic studies, postmortem analysis, and animal models have been the key to our basic understanding the mechanisms of SCI. As early as 1911, Allen, in a dog model for traumatic SCI, demonstrated that drainage of traumatic hematomyelia via myelotomy led to improved neurologic recovery in injured animals.⁷ By demonstrating postoperative neurologic recovery in the context of traumatic cord injury, this study suggested that secondary, reversible mechanisms of SCI exist beyond the injury associated with the initial neurologic insult. Since its publication, multiple mechanisms of secondary SCI have been proposed, including spinal cord ischemia, oxidative injury from free radical formation, catecholamine release, cytokine-mediated inflammatory responses, and calcium ion–mediated cell injury and death.^8–10 Indeed, the results of limited and often controversial pharmacologic studies aimed at targeting these various mechanisms have reported improved outcome in patients with incomplete SCI and have largely been the rationale for early interventions in affected patients.^11–13

Anatomic postmortem studies have also shed light onto SCI mechanisms. Analysis of spinal cords from deceased patients with incomplete and complete injuries has demonstrated greater structural continuity in some cases of complete injury.¹³ This finding has led some researchers to the conclusion that mechanisms beyond structural white matter disruption contribute to the overall disability of affected patients and suggests that even in the context of structural discontinuity, reversible mechanisms of injury might still exist.¹⁴

A host of animal studies have been used to delineate mechanisms of SCI. In these studies, two major factors have been identified that determine the reversibility of SCI: the magnitude of impact force at the time of injury and the persistence of compressive forces after injury.^15–17 While from a postinjury clinical treatment perspective, impact forces are unalterable, surgical and nonsurgical interventions can be utilized to treat persistent spinal cord compression. Multiple animal studies support the idea that early decompression of the spinal cord, within minutes to hours, may maximize neurologic outcomes.^15,18,19 These findings are the basis for the concept of rapid decompression of the spinal cord after trauma.

The experimental literature suggests reversibility and recovery of neurologic function with early pharmacologic and surgical intervention. Independent of controversial issues regarding the validity of the experimental pharmacologic data, the intellectual leap from experimental data to applied clinical intervention has been extensively debated in the literature. It is safe to say that despite a century’s passage since Allen’s studies, no absolute recommendation can be made as to the appropriate timing for surgery.

It is important to note that the existing clinical literature is so varied in definitions of what constitutes emergency or early intervention for SCI versus delayed intervention that we feel it necessary to define what constitutes emergent intervention in the context of this discussion. For the purposes of this discussion, surgery performed within 24 hours of injury constitutes emergency or early surgery. Delayed surgery, by definition, constitutes surgery performed after the 24-hour period. It should be noted that despite this rigid definition, delayed surgical intervention should be performed after medical optimization and stabilization regardless of how long a delay that might be. We advocate nonemergent intervention and discuss advantages, disadvantages, indications, and relative contraindications for nonemergent surgery.