The Role of Therapeutic Hypothermia After Traumatic Spinal Cord Injury—A Systematic Review

Background

Traumatic spinal cord injury (SCI) is a devastating neurologic entity characterized by a primary insult followed by a secondary pathologic cascade that propagates further injury. Hypothermia has an established clinical role in preventing SCI after cardiac arrest and thoracoabdominal aortic aneurysm repair, yet its emergence as a potential neuroprotectant after spinal cord trauma remains experimental. There are currently no pharmacologic interventions available to prevent secondary mechanisms of injury after spinal cord trauma.

Methods

Systematic review of literature.

Results

Experimental studies demonstrated that hypothermia diminishes secondary pathomechanisms, such as ischemia, oxidative stress, apoptosis, inflammation, and edema. Early onset and longer durations of hypothermia as well as concomitant steroids or neural stem cell engraftment combined with hypothermia appear to improve functional and histologic outcomes in animal models of spinal cord trauma. Recent clinical studies provide evidence that localized and systemic hypothermia may be applied safely and efficaciously in patients with severe acute SCI. Randomized clinical trials are needed to better evaluate optimal cooling parameters and the effectiveness of hypothermia after traumatic SCI.

Conclusion

Although variability exists in the literature, therapeutic hypothermia most likely confers neuroprotection after spinal cord trauma by diminishing the destructive secondary cascade. The available clinical data suggest that regional and systemic hypothermia is a relatively safe and feasible initial treatment modality for patients with acute SCI when combined with surgical decompression/stabilization with or without steroids. However, establishing a clinical role for therapeutic hypothermia after spinal cord trauma will invariably depend on future well-designed, multicentered, randomized, controlled clinical trial data.

Introduction

There are about 85,000 Canadians and 250,000 people in the United States living with the consequences of acute SCI1, 2 and approximately 12,000 new cases annually in the United States.³ A third of these patients leave the hospital with paraplegia or quadriplegia with only 1% discharged neurologically normal.3, 4 The consequences of SCI are severe and often fatal. The rate of mortality is between 48% and 79% with 4.4%–16% occurring before hospital discharge.5, 6

More than half of spinal cord injuries result from trauma. Motor vehicle accidents are the most common etiologic cause of SCI (40%–50%) followed by falls (20%), violence (14%), and recreational and work-related injuries.⁷ The annual incidence of SCI is 3- to 4-fold greater in men than women.¹ Two thirds of those affected by acute SCI are younger than 30 years of age; however, with an aging population, the prevalence and incidence of fall-related SCI among the elderly (≥65 year) is expected to increase.1, 8

The natural history of spontaneous SCI recovery is less than optimistic, but a variable degree of neurologic recovery is expected, peaking at 3–6 months. This variation can be a confounder when considering the therapeutic effects of various agents.³ Steeves et al.⁹ showed that among patients classified with complete cervical SCI, only 10% regain sensory function without motor recovery, and 10% regain some degree of motor function.

Acute traumatic SCI is a devastating neurologic event, characterized by an initial injury followed by a secondary pathologic cascade that further damages the spinal cord. Primary injury results from mechanical impact to the spine and concomitant compression, contusion, laceration, shear, and acute kinking of the spinal cord resulting from displacement of bone and intervertebral disk into the spinal canal.¹⁰ The severity of SCI can be classified according to the American Spinal Injury Association (ASIA) impairment scale, which is stratified into complete motor and sensory impairment (ASIA grade A), complete motor and incomplete sensory impairment (ASIA grade B), incomplete motor impairment with intact sensation (ASIA grades C and D), and normal neurologic status (ASIA grade E).¹⁰ Injury to the cervical spinal cord may cause paralysis of all 4 limbs (quadriplegia), whereas thoracic and lumbar injuries affect the lower limbs (paraplegia).¹⁰

Secondary mechanisms of SCI start seconds after primary injury and continue for several weeks.⁸ The secondary pathologic cascade is strongly associated with vascular aberrations including vasospasm, thrombosis, microhemorrhage, impaired autoregulation of microcirculation, and ischemia. Systemic hypotension secondary to neurogenic shock exacerbates hypoperfusion of the spinal cord.5, 8 Electrolyte disturbances after SCI include loss of neuronal sodium gradient due to ischemic dysfunction of Na⁺/K⁺ ATPase and glutamate excitotoxic calcium overload. Oxidative stress promotes neuronal injury via production of free radicals such as reactive oxygen species and reactive nitrogen species and lipid peroxidation.⁵ Neuroinflammation mediated by the host immune response and molecules such as cytokines, chemokines, and complement promote the neurodegenerative processes within the injured spinal cord.¹⁰ Peripheral monocytes/macrophages and neutrophils and central microglia activated by nervous tissue injury produce proinflammatory cytokines and chemokines.11, 12 Chemokines such as CXCL molecules and complement anaphylatoxins such as C5a drive transmigration of leukocytes into the injured central nervous system (CNS).10, 13, 14 These infiltrates harm nervous tissue by releasing neurotoxic free radical molecules such as reactive oxygen species and reactive nitrogen species, proteases such as matrix metalloproteinases, and other inflammatory mediators.¹⁵ These pathophysiologic processes cause breakdown of the blood–spinal cord barrier, leading to the formation of edema and the extravasation of systemic toxins and proinflammatory mediators into injured spinal cord tissue and cerebrospinal fluid, promoting secondary neurotoxicity.10, 13, 16 Breakdown of the blood–spinal cord barrier disrupts the precisely regulated CNS milieu and promotes neuroinflammation and subsequent neurodegeneration.¹³ Myelin-related factors, such as Nogo-A, released by spinal cord damage and demyelination, inhibit adult CNS regeneration and create a cycle of damage and failure to repair.8, 10

The clearance of myelin debris is essential for the neuroregenerative process. Macrophages possess the most efficient capability to phagocytose myelin debris with minor contributions from resident microglia and oligodendrocytes.¹³ Limited macrophage access to clear myelin debris, inefficient microglial clearance of myelin, and restrictive astroglial scar formation impair neural regeneration. The self-perpetuating cycle of neuroinflammatory damage and failure to repair results in spinal cord edema, progressive neurologic deterioration, and expansion of the primary spinal cord lesion. There appear to be no pharmacologic agents or surgical approaches that prevent the secondary pathologic cascade after acute SCI. Therapeutic hypothermia has emerged as an experimental adjunct therapy for acute SCI capable of ameliorating secondary mechanisms of injury.

In 1940, the New York State Journal of Medicine published pioneering descriptions of therapeutic hypothermia by American neurosurgeon Dr. Temple Fay, wherein he described a patient with metastatic disease who underwent “refrigeration” (32.2°C) for 18 hours by surface cooling under anesthesia. The patient recovered without knowledge or discomfort of the procedure.¹⁷ Fay pioneered the early use of hypothermia to treat neurologic conditions, first administering hypothermia to patients with intractable pain and later to patients with intracranial processes, including traumatic brain injury, abscesses, and cerebritis.18, 19 By the 1950s hypothermia was revolutionizing cardiac surgery²⁰ and cerebral aneurysm repair²¹ by preserving neural function in patients undergoing these surgical procedures. Pontius et al.²² described a protective role for hypothermia in preventing paralysis during aortic occlusion experiments. Rosomoff23, 24, 25 built on these early concepts of therapeutic hypothermia and introduced hypothermia as an adjunct to neurosurgery for the treatment of intracerebral vascular lesions and brain injury. In the 1970s local cooling by applying cold saline directly to the exposed spinal cord during spinal decompression surgery was documented to promote functional recovery after cervical SCI in a series of case studies.26, 27 Small sample sizes and the confounding effects of toxic metabolite clearance by decompression and irrigation limited the interpretation of these early reports. Enthusiasm for the use of hypothermia as a neuroprotectant declined by the 1980s because of inconsistent benefits, availability of novel pharmacologic agents (steroids), and high rates of complications, such as pulmonary insufficiency, ventricular arrhythmia, bradycardia, coagulopathy, and infection.3, 28

The potential for neurologic benefits using modest reductions in body temperature without the attending high rates of complications revived clinical interest in hypothermia as a neuroprotective therapeutic strategy in the late 1980s to early 1990s. Busto et al.²⁹ revealed that mild intraischemic variations in brain temperature (e.g., down to 34°C) ameliorated neuronal damage in a rat model of global cerebral ischemia, demonstrating that the extent of hypothermia required to improve neurologic outcome was milder than previously anticipated. Likewise, moderate hypothermia (≥30°C) rather than profound deep hypothermia (≤28°C) was found to have protective effects on neurologic outcome with lower rates of complications in patients with traumatic head injuries.30, 31, 32

Hypothermia has been studied as a neuroprotectant for decades in the context of cardiac arrest and thoracoabdominal aortic aneurysm (TAAA). The vertebral and cervical branches of the subclavian artery along with posterior intercostal arteries of the aorta contribute to the segmental medullary blood supply of the spinal cord. Damage to these vessels during TAAA repair results in spinal cord ischemia and sometimes paralysis. Interestingly, hypothermia lowers the rate of paralysis among patients who undergo TAAA repair. During the past 2 decades, surgeons at Massachusetts General Hospital have reported success with the use of epidural cooling for prevention of spinal cord ischemia complications after TAAA surgery.33, 34, 35 Studies also demonstrated neuroprotective benefit of hypothermia in out-of-hospital cardiac arrest survivors whose initial electrocardiogram demonstrated ventricular fibrillation.³⁶ To date, hypothermia is the only treatment shown to have neuroprotective effects against global ischemic injury associated with cardiac arrest³⁷ and newborn hypoxic ischemic encephalopathy.³⁸ The evidence is so convincing that The American Heart Association published recommendations that modest hypothermia (32–34°C) be instituted for 12–24 hours in out-of-hospital cardiac arrest.³⁶ Similarly, a Cochrane systematic meta-analysis reviewing data from 8 randomized control trials concluded that hypothermia benefits neonates with hypoxic ischemic encephalopathy.³⁹ Together, these studies indicate that hypothermia increases brain and spinal cord tolerance to ischemia.

Hypothermia is induced externally (local and surface cooling) or internally (intravascular infusion and endovascular heat-exchange catheter). Local cooling of the spinal cord can be achieved with the use of epidural thermal exchange devices or by irrigating the spinal cord with chilled saline. The human clinical trials of local spinal cord cooling were facilitated by the surgical practice of laminectomy and durotomy, which enables direct irrigation of the spinal cord.³ Local cooling was a popular method of inducing hypothermia SCI experiments in animals during the 1960s and 1970s.⁴ Promising results from these animal studies led to a few case series and case report studies in which human SCI was treated with local hypothermia.26, 27, 40, 41, 42, 43

Systemic hypothermia can be achieved by surface cooling by the use of ice packs, cooling blankets, or ice-cold gastric lavage, used in important clinical trials of hypothermia after acute traumatic head injury30, 44 and cardiac arrest.³⁷ Surface-cooling methods are problematic in obese patients, as they shiver and insulate warmth, especially in unanesthetized patients.³ Sedation and paralysis effectively decrease shivering. Meperidine is particularly effective in decreasing shivering, without the need for paralysis.28, 45 Surface cooling takes longer than intravascular cooling methods to achieve target core temperature.

More rapid systemic cooling can be achieved by intravascular infusion of cold saline⁴⁶ and endovascular heat exchange balloon catheter.¹² In healthy volunteers, infusion of 40 mL/kg ice-cold saline (4°C) reduced body temperature by 2.5°C over 30 minutes.⁴⁶ Infusion of 30 mL/kg ice-cold lactated Ringer's solution in comatose patients after cardiac arrest lowered core temperature by 1.6°C over 25 minutes.³⁷ Endovascular heat exchange balloon catheters, such as the Celsius Control System (Intercool Therapeutics, San Diego, CA, USA) and CoolGuard Icy Catheter (Alsius; Irvine, California, USA) function by cooling circulating blood. Endovascular heat exchange balloon catheters maintain target temperature better than surface cooling and do not alter the intravascular volume.3, 47 During the past decade, there has been an international effort among basic scientists and clinicians in the area of hypothermia to elucidate the therapeutic mechanisms underlying the putative benefits of hypothermia in acute SCI. The remainder of the paper will review the modern experimental and clinical evidence concerning hypothermia treatment of acute SCI.