Injury, Int. J. Care Injured (2004) 35, 661—670
Damage control surgery for spine trauma Thomas Kossmanna,*, Larissa Treasea, Ilan Freedmana, Gregory Malhamb a
Department of Trauma Surgery, The Alfred Hospital/Monash University, Commercial Road, Melbourne, Vic. 3004, Australia b Department of Neurosurgery, The Alfred Hospital/Monash University, Commercial Road, Melbourne, Vic. 3004, Australia
KEYWORDS Damage control; Spine; Trauma; Multitrauma; Spinal injury
Summary The concept of ‘‘damage control’’ surgery was originally developed for massive abdominal trauma and also successfully applied to the management of lone bone injuries. More recently this has been extended to severely injured patients with spine injuries. This paper provided an overview of how damage control principles can be applied to multitrauma patients with spine injuries, to patients with isolated spine injuries and to spine injuries with and without neurology. The role of neuroimaging in acute spine trauma and controversies in the pharmaceutical approach to spine injuries are discussed. Additional prospective controlled trials are required to delineate the role and timing of damage control surgery in acute spine injury. With improved neuroimaging early spinal damage control surgery will be formally established in the management of spine trauma. ß 2004 Elsevier Ltd. All rights reserved.
Introduction The concept of ‘‘damage control’’ surgery was originally developed for treating hemodynamically unstable patients with massive abdominal trauma, and more recently has been adapted to the management of patients with musculoskeletal injuries including the spine. Application of damage control principles to the spine varies depending on the pattern of injury to the bony structures and extent of spinal cord involvement. In this context, the term ‘‘spine trauma’’ refers to musculoskeletal spine injuries and is differentiated from ‘‘spinal trauma’’ which pertains specifically to spinal cord injury. While the concept of maintaining spine stability following trauma has been established,47 no consensus exists as to whether immediate operative or conservative approaches are prefer*Corresponding author. Tel.: þ61-3-9276-3386; fax: þ61-3-9276-3804. E-mail address:
[email protected] (T. Kossmann).
able. Moreover, the timing, choice of surgical approach and extent of surgery in the acute management of spinal trauma is controversial.1,7,8,14,15,18,19,25,31,36—38,40,42,47,48,54 The orthopaedic trauma literature demonstrates that early stabilisation of long bone fractures reduces both morbidity and mortality in polytrauma patients.38 Various approaches and techniques for early fracture stabilisation of long bone fractures in trauma patients have been developed. Despite the benefits of early fracture stabilisation in long bones there are indications that premature extensive surgery on hemodynamically unstable patients may predispose to adverse outcomes.7,8,31 In this context, the role and timing of surgical intervention in the acute management of spine injuries in combination with spinal cord injury has not yet been defined. In our setting, the concept of damage control surgery entails a three-phase approach to the surgical management of the trauma patient. In the initial phase, the prime aim is to establish rapid
0020–1383/$ — see front matter ß 2004 Elsevier Ltd. All rights reserved. doi:10.1016/j.injury.2004.03.005
662
control of haemorrhage and to identify and to control major injuries. The next step is to stabilise major fracture(s), to reduce dislocated joint(s) and to decontaminate open wound(s) with appropriate surgical methods, which may or may not be definitive. Once normal physiology has been restored a final phase entails definitive injury repairs. Some situations may allow for definitive surgery on day one but the patient’s condition, available infrastructure and experience of the surgical team may delay definitive surgery until a later date.53 This paper provides a contemporary overview of how the current concept of damage control surgery has been extended to the management of traumatic injuries to the spine.
The multitrauma patient with spine injuries The benefit of early stabilisation of long-bone fractures in reducing both the morbidity and mortality in polytrauma patients has been well demonstrated.7,8,31 Emphasis on minimising the adverse effects of major injury has led to the development of trauma systems and multidisciplinary trauma teams. Spinal precautions now pervade all major trauma clinical guidelines. Neurosurgical and spine services are increasingly incorporated within trauma system models. Safer early surgical management of the multitrauma patient with spine injuries has been established.38 Spinal cord injuries may result in devastating medical and social consequences for patients and their families and may necessitate indefinite dependence on support services.22 Considerable research efforts have focused on the pathophysiology of spinal cord injury and have reinforced Allen’s2 postulate that a primary mechanical spinal cord injury initiates a series of secondary injury events that exacerbate spinal cord damage. These processes include electrolyte fluxes, biochemical changes and vascular compromise which produce spinal cord ischaemia, impaired autoregulation, post-traumatic hypotension, vaspospasm, oedema, oxygen free radical formation, calcium ion entry, cytokine and catecholamine release which lead to cell apoptosis.18,52 Experimental animal models further suggest that the severity of these pathological changes and the degree of possible recovery are related to the duration of acute cord compression.13—15 As a consequence, early spine surgery even in multitrauma patients is focused on spinal cord decompression to avoid secondary spinal injury. This is accomplished through efforts to restore
T. Kossmann et al.
anatomical alignment, to decompress neural and vascular structures and to stabilise the vertebral column. In acute trauma spinal anatomy may be distorted, soft tissues are traumatised, the biomechanics of the vertebral column may be unstable or unpredictable and the spinal cord may be vascularly compromised. In this context, a two-stage management approach has been adopted in our institution. Relatively safe procedures such as posterior internal stabilisation of thoracic or lumbar fractures or the temporary external fixation of cervical spine fractures in halo thoracic braces or extension are performed. This enables the secondary phase to be scheduled at a later time with anterior reconstruction or complex stabilisation procedures as allowed by the patient’s physiological condition. Such a delay until definitive surgery allows for general recovery of the patient, facilitates safer anaesthesia and may also enable the use of hypotensive techniques to minimise intra-operative blood loss. Furthermore, adoption of a two-staged approach to surgery allows the definitive surgery to be electively scheduled at a time when an experienced spine surgeon is available. In the interval between the initial damage control surgery and second-stage definitive surgery, a high standard of general care encompassing adequate ventilatory support, pressure care, bowel and bladder management, thromboembolic prophylaxis and control of hospital acquired infections further minimises systemic damage to the patient. A recent prospective longitudinal study by McLain and Benson examined the outcome of early surgical stabilisation of spine fractures in polytrauma patients and demonstrated that this approach was safe and appropriate. Early surgery was also indicated in situations where chest trauma and pulmonary contusions would predictably result in pulmonary deterioration, allowed earlier patient mobilisation and rehabilitation and reduced the incidence of post-operative infection, pulmonary disease and thromboembolism. Furthermore, neurologic recovery in patients with persistent or progressive neurological deficits was better in patients treated within 24 h after injury than in those treated between 24 and 72 h post-injury.38
Open spine injuries Open spine injuries affecting the spinal canal or spinal cord result from penetrating trauma. Delayed or inappropriate treatment of these open injuries may lead to serious central nervous system infections. Penetrating injuries such as gunshot or stab
Damage control surgery for spine trauma
wounds may affect the integrity of the spine and spinal cord and also involve other organs or important structures. Primary aims in the management of penetrating spine injuries are the optimal management of the spine and spinal cord injury, and the management of associated vascular, visceral and musculoskeletal injuries, which may themselves be life-threatening. Penetrating cervical spine injuries are associated with a high incidence of vascular injury, while penetrating thoracic wounds are frequently associated with pulmonary and cardiac trauma.29 Lumbar spine knife or gunshot injuries may be accompanied by damage to the abdominal viscera, genitourinary system and major vascular structures29 and a bullet that transverses the colon and lodges in the lumbar spine may be complicated by infection. Urgent surgical exploration should be performed for repair of major anatomical structures such as vessels, organs, small or large bowel and the patient’s airway. In penetrating injuries to
663
the neck an expanding haematoma or haemopneumothorax must be assumed, specifically looked for and adequately treated.35 In approximately one third of gunshot injuries to the spine bullet fragments are retained within the canal.33,57 Plain radiographs of the involved region should be obtained in two planes3 to evaluate the projectile’s path, the degree of bony destruction, spine stability and for retained bullet fragments in the vertebrae or spinal canal (Fig. 1). Computed tomography (CT) provides supplementary information when spine instability or retained bullet fragments are suspected33 (Fig. 2). The role of Magnetic Resonance Imaging (MRI) in spine gunshot patients with possible retained metallic material is controversial20,32,51 and should be considered on an individual patient basis. Damage control surgery in the management of open spine injuries has been addressed mostly in case reports. Firoozmand and Velmahos21 reported a case of a young man who suffered a single gunshot
Figure 1 Single gunshot injury to the abdomen. Antero-posterior and lateral X-ray views indicate that fragments of the bullet are within the spinal canal. An emergency laparotomy had been performed to address the intra-abdominal injuries including the left urether. (i) Antero-posterior X-ray and (ii) lateral X-ray.
664
T. Kossmann et al.
Figure 2 CT-scans of the affected region demonstrate the presence of bullet fragments within the spinal canal.
wound to the neck. Early operative exploration revealed extensive osseous and soft tissue damage around C3 and C4 and spinal cord transection at C3. Bleeding vessels were ligated, the wound was packed and the patient was resuscitated in the intensive care unit. This was followed by angiography to embolise any further bleeding vessels and on the second post-operative day the patient returned to the operating room for further surgery and primary wound closure. Our experience in an open spine injury with initial complete tetraplegia has been reported by Platz et al.43 who adopted damage control principles in the management of a patient with penetrating knife neck stab wounds. Urgent surgical exploration of the injuries revealed a partial spinal cord transection combined with oesophageal penetration. A developing coagulopathy due to massive haemorrhage required temporary wound closure after control of the bleeding sites and subsequent patient transfer to the intensive care unit for ongoing resuscitation without repair of the oesophagus. Once hemodynamically stable, the patient was returned to the operating room (12 h later) and the oesophageal laceration was closed. Ongoing
increases in fluid drainage over the subsequent days required a second wound exploration. Cerebrospinal fluid leakage was identified and repaired with a dural patch. The patient ultimately made a partial neurologic recovery to incomplete tetraplegia. Aggressive early surgical treatment of penetrating spine injuries has great benefits in avoiding complications. Schaefer et al.46 demonstrated that the high incidence of vertebral osteomyelitis of the cervical spine after penetrating injury may be effectively eliminated with aggressive early debridement. To further decrease the incidence of infection Simpson et al. recommended intravenous cephalosporin antibiotics for high velocity gun shot wounds, addition of gentamycin for soft tissue or cavitating lesions, and penicillin for grossly contaminated wounds.49 Mandatory tetanus prophylaxis for all gun shot victims was also advocated.49 Kitchel recommended 2 weeks of broadspectrum parenteral antibiotics for patients where the bullet trajectory penetrated the alimentary tract and lodged in the spine or spinal canal.33 Kumar et al.34 subsequently demonstrated this to effectively help prevent late infection and osteomyelitis.
Damage control surgery for spine trauma
Heary et al. reviewed 254 consecutive patients with gunshot wounds to the spine and spinal cord injury and showed that administration of intravenous steroids to these particular patients did not confer neurological benefits. Higher rates of both infectious and non-infectious complications were observed in the steroid group and the authors concluded that patients with gunshot spinal cord injuries should not receive steroids.28 Routine laminectomy after penetrating cervical spine injuries is also not supported. Early cervical laminectomy may increase the risk of spinal column destabilisation and provoke neurologic deterioration in patients with anterior compressive lesions.29,35,40 Laminectomy for surgical decompression should be reserved for actual neurological deterioration or for debridement. Likewise, early bullet removal must be treated with circumspection and the decision to proceed should be assessed on a patient to patient basis. Waters and Adkins56 described an improvement in neurological function following surgical removal of bullets from the spinal canal below the T12 level but no improvement was demonstrated in gunshot wounds affecting the proximal thoracic spine (T1—T11) in either complete or incomplete injuries. Simpson reported that surgery for bullet removal does not improve neurological recovery and increases complications such as meningitis and wound infection.50 Neurological deterioration in patients with bullets left in situ has been described in case reports4,11,30 but the delay from injury to onset of complications has been reported up to 15 years.3 Patients with retained bullet fragments should be monitored for neurological deterioration and fragment removal should be considered if deterioration occurs.57
Controversies in the surgical approach to spine injuries with neurology No consensus of opinion exists regarding the impact of spine surgery on neurological outcome and disagreement persists regarding the optimal timing of surgery after traumatic injury to the spinal cord. The time between injury and surgery and the degree of spinal cord injury appear to be important prognostic factors1,18,19,40,52,54,55 but various authors have reported conflicting results.
Cervical spine injuries with neurology Vaccaro compared ‘‘early’’ (5 days) spinal fixation in a randomised, prospective trial conducted in a Regional Spinal Cord Injury
665
Centre and enrolled 123 patients during a 37-month study period. Sixty-four patients with cervical spine injuries were admitted within 48 h of injury. For inclusion in this study, radiographically documented cord compression from anterior or posterior vertebral body elements, disc material or fracture subluxation or dislocation was required. Comparison of patients who underwent surgery within 72 h of injury to those who received surgery after 5 days demonstrated no statistical difference in neurological recovery.54 Wagner and Chehrazi55 reviewed 44 consecutive cases of cervical spinal cord injury over a 5-year period. Although patients with incomplete paraplegia made significantly greater neurologic recovery compared to those with complete spinal injuries, no significant differences were revealed in neurologic recovery between the incomplete paraplegia patients managed conservatively and those who underwent surgery within 48 h of their injuries. Mirza et al performed a retrospective study of 30 patients with cervical spine injuries who underwent surgery within 72 h post-injury or underwent immediate closed reduction followed by surgical stabilisation between days 10 and 14. Patients included in the study had acute traumatic cervical spinal injury between C2 and C7 and all had neurological deficits on admission. Some patients with complete spinal cord injury were included. Patients were regularly assessed using the American Spinal Injury Association Motor Index. Overall, patients who received surgery within 72 h demonstrated a significant improvement in score and Frankel grade compared to those who underwent delayed surgery. The authors did not attempt to differentiate between complete and incomplete spinal injuries as they believed that such differentiation in the acute period after injury may not be reliable.37 As they did not demonstrate any significant difference in either major or minor complication rates between the groups treated surgically within 72 h or at days 10—14 they suggested that surgical decompression and stabilisation within 72 h of the initial trauma should be considered in all acute cervical spine injuries with associated with neurological deficit. Similarly, Aebi et al.1 retrospectively reviewed 100 patients in whom a fracture, dislocation or fracture-dislocation of the cervical spine was treated surgically. Twenty eight complete tetraplegia, 40 incomplete paraplegia and 32 neurologicallynormal patients were included. Manual or surgical reduction was performed within 6 h of injury in 25% of the patients and 23% were operated on within the first 24 h. Overall, 31% of patients showed neurologic improvement. The greatest neurologic gains
666
were seen in the incomplete paraplegia group where 22 of 40 patients made a significant neurologic recovery, whereas only 2 of 28 complete tetraplegia patients improved to incomplete tetraplegia. Of the overall 31% patients who did improve, 75% were reduced within 6 h after injury whereas 85% of the 69 patients without neurologic recovery were reduced after 6 h. The authors concluded that prompt decompression of the spinal cord should be performed for spinal injuries with neurologic deficits and advised that patients with incomplete injuries derive the greatest benefit. Rizzolo et al.45 demonstrated that 40% of cervical spinal cord injury patients present with complete spinal cord injuries, 40% with incomplete injuries and 20% with either no cord or only root lesions. The authors argued that immediate cervical spine protection, pharmacologic treatment, correction of misalignment and spinal canal decompression optimises neurologic recovery in patients with cervical spine trauma. Multiple authors have similarly concluded that prompt reduction (
