Chapter 28. Traumatic Brain Injury and Spinal Cord Injury

Traumatic brain injury (TBI) and traumatic spinal cord injury (TSI) represent pathologies that result from a diverse spectrum of primary insults to the central nervous system (CNS). Nearly 2 million cases of neurotrauma, including both TBI and TSI, occur annually in the United States making it an important public health issue.1–3 In addition to the long-term physical disabilities and the psychosocial impairments seen in neurotrauma survivors, the economic burden of TBI and TSI is significant. The cost of TBI in the United States is estimated to be somewhere between $40 and $200 billion.3–6 For TSI, it is estimated that the lifetime total cost directly attributable to spinal cord injury in a 25-year-old patient may exceed $3 million.2

Advancements in our understanding of the pathophysiology of CNS injury post-trauma have led to improvements in the critical care of patients with TBI and TSI. Indeed, the development of standardized guidelines for aggressive medical and surgical management of these patients has been credited for helping to improve outcomes.7,8 For severely brain- and/or spinal cord–injured patients, it is crucial that a multidisciplinary approach be taken from the outset. The foremost principle guiding the management of TBI and TSI patients is to minimize the secondary neural injury that inevitably follows a primary CNS insult. Appropriate and timely emergency stabilization, critical care management, and surgical interventions are essential for delaying the progression of secondary CNS injury. Toward this end, the clinician treating TBI and TSI patients must be able to assess, monitor, and treat the multitude of physiologic derangements that result from and also facilitate CNS injury. In this chapter, we review the epidemiology, pathophysiology, and critical care management of TBI and TSI patients. As neurosurgical intervention—whether at the bedside or in the operating room—is generally necessary for TBI and TSI patients, the surgical indications for pathology encountered in the emergency and critical care setting will also be reviewed.

Traumatic Brain Injury

The Centers for Disease Control and Prevention (CDC) estimates that 1.7 million people sustain TBI annually.1 Of this total, approximately 52,000 die, 275,000 are hospitalized, and 1.365 million are treated and released from an emergency department. However, the actual number of TBI cases is uncertain as many patients either receive care in the field or do not seek medical attention at all. The causes of TBI in all age groups combined are: falls in 35.2%, motor vehicle accident in 17.3%, being struck in 16.5%, assault in 10%, and other or unknown causes in 21%. Among all age groups, males have higher incidences of TBI and on average present with TBI about 1.4 times more frequently than females. The CDC identified three age groups—children aged 0–4 years, adolescents aged 15–19 years, and adults aged 65 years and older—as most likely to sustain TBI. Children aged 0–4 years had the highest rate of TBI-related emergency department visits (1,256 per 100,000 population) while hospitalization (339 per 100,000 population) and death (57 per 100,000 population) rates were highest among those 75 years and older. TBI-related death rates have substantially declined in the past 30 years and can mostly be attributed to primary prevention. Motor vehicle–related TBI deaths declined 22% and firearm-related TBI deaths declined 14% between 1989 and 1998.9 Recent data indicate more modest declines or relative stability in TBI mortality trends.1,10

Traumatic Spinal Cord Injury

The National Spinal Cord Injury Statistical Center (NSCISC) collects and dispenses the most comprehensive epidemiologic data on spinal cord injury in the United States. They estimate the annual incidence of spinal cord injury at 40 cases per million population, representing about 12,000 new cases each year in the United States.2,11 About 80% of these injuries occur in males. The most common causes of spinal cord injury include motor vehicle accident in 41.3%, fall in 27.3%, and violence in 15%. Cervical spinal cord injuries are the most common comprising over 50% of lesions within the NSCISC database, followed by thoracic, lumbar, and sacral lesions. Among all levels of injury, cervical lesions confer the worst prognosis with ventilator dependency having a strong negative association with survival. Death in spinal cord–injured patients most commonly results from respiratory infections and septicemia associated with urinary infections and decubitus ulcers.

An appreciation for the pathophysiologic mechanisms at work after TBI and TSI is important for the development and implementation of effective clinical therapeutic strategies. Although typically treated separately, the pathophysiologies of both TBI and TSI have a number of similarities. Most notably, the injury due to both TBI and TSI can be understood in terms of primary and secondary insults to neural tissue. Primary injury denotes the initial mechanical damage secondary to energy transmission during impact while secondary injury results from the destructive tissue-intrinsic and body systemic response to primary injury. The critical care of TBI and TSI patients is directed at minimizing secondary injury. Detailed accounts of the molecular and cellular mechanisms of TBI and TSI have been given.12–15 Here we present a concise review of the pathophysiology of TBI and TSI with emphasis on delineating pathologic processes that are routinely targeted clinically.

Blood Flow

Both TBI and TSI are associated with focal and/or global hypoperfusion to the brain and spinal cord, respectively.13,14,16,17 Hypoperfusion can result from a number of mechanisms including microvascular or macrovascular damage, vasospasm, neurogenic/spinal shock, loss of autoregulation, or mechanical tissue disruption. Decreased blood flow to neural tissue results in ischemia and ultimately infarction as cellular metabolic demands exhaust available resources. For these reasons, it is unsurprising that reductions in cerebral blood flow (CBF) and hypotension portend a worse prognosis for TBI patients.18–23 The evidence suggests this may be true in TSI patients as well; however, further studies are warranted.24,25

Paradoxically, focally increased blood flow resulting in hyperemia may also result from acute injury to neural tissue. Hyperemia is similarly as deleterious to the injured brain and spinal cord as is hypoperfusion. Mechanistically, both processes result in a mismatch between blood flow and cellular metabolism. Additionally, by facilitating oxidative damage to cells, promoting tissue edema, and increasing intracranial pressure (ICP), hyperemia further promotes secondary injury in the acute setting.12–14

Metabolism

Metabolic dysfunction results after TBI and TSI due to impaired delivery and/or utilization of oxygen and glucose within the injured brain and spinal cord.12–14 As neural cells depend on the production of high-energy molecules from aerobic metabolism to meet their energy requirements, even modest reductions in oxygen and glucose are poorly tolerated.12,26 The ionic fluxes associated with primary cellular injury result in the initiation of energy-dependent processes such as membrane transport in an attempt to restore homeostasis. As energy stores become depleted, especially within the ischemic penumbra, cell death occurs.12

Inflammation

The robust inflammatory reaction seen within the damaged brain and spinal cord is a major component of both negative secondary injury processes and positive, reparative processes.12–14,26 At the site of injury, leukocyte recruitment and concurrent expression of inflammatory mediators such as TNF-α, interleukins, and complement molecules promote vascular permeability, edema, and progressive tissue damage. Inhibition of this cytotoxic inflammatory milieu is a major target for the development of neuroprotective therapies.

Excitotoxicity

TBI and TSI are associated with excessive release of excitatory neurotransmitters such as glutamate in response to hypoxia.12–14,26 This in turn results in major ionic flux involving sodium, potassium, and calcium across cell membranes. The accumulation of intracellular calcium in particular is associated with many toxic processes such as lipase and peroxidase activation as well as free radical generation.

TBI can be caused by a number of mechanisms that involve deforming brain tissue at various strain rates. This may range from the relatively low strain rates imparted during simple collision to the higher rates associated with ballistic and blast mechanisms. Injuries can be divided into two main categories—open (including penetrating) and closed head injuries. Open head injuries involve violation of the skull and can result from foreign body (i.e., bullet, arrow, knife) entry or direct blunt force (i.e., blow from baseball bat). Closed head injury is most often caused by direct blunt force; however, primary blast– or shock wave–induced injuries are becoming increasingly prevalent in the military theater.

The initial assessment of neurotrauma patients should begin with the familiar ABCs: evaluate the airway, confirm breathing with effective ventilation, and assess the circulatory status. Cervical hard collar placement and body immobilization on a rigid backboard is warranted for all trauma patients.27 Clinically deteriorating patients and those with a Glasgow Coma Scale (GCS) score of 8 or less should be intubated as they are unable to adequately protect their airway. Precautions such as logrolling and inline stabilization during intubation are prudent until spinal stability is verified.27 Cardiac, hemodynamic, respiratory, and pulse oximetry monitoring is necessary for all patients with moderate and severe TBI (see below) as well as in all TSI patients.16,17,28 Hypoxemia (SaO2 <90%) and hypotension (systolic blood pressure <90 mm Hg) should be avoided in both TBI and TSI patients. TBI outcomes have clearly been shown to be worse if hypoxia or hypotension is present, and the available evidence suggests that the same holds true for TSI.16,17,24,28 Intubation, volume resuscitation with intravenous fluids or transfusion, and the use of vasopressor medications may be necessary to achieve these initial goals.

Visual inspection of the general physical condition of the patient should be carried out. Evidence of basal skull fracture (periorbital or postauricular ecchymoses, cerebrospinal fluid [CSF] rhinorrhea/otorrhea), facial fracture, or spine deformity should be noted. At a minimum, the initial neurologic examination should encompass assessment of the following: (1) level of consciousness with a determination of the GCS (Table 28-1), (2) cranial nerve (CN) function with particular attention to the size, symmetry, and reactivity of pupils, and (3) gross motor and sensory examination of the extremities. In cases of suspected spinal cord injury, a more detailed motor exam and determination of a specific sensory level is appropriate. Additionally, a digital rectal exam checking for voluntary anal sphincter contraction should be performed and an American Spinal Injury Association (ASIA) impairment scale grade (see below) should be assigned.29

While in the acute period, all patients with TBI should be neurologically examined on a regular basis—at least every hour for the first 24 hours (hyperacute period) and then less often as clinically indicated. In general, cerebral edema is greatest from 48 to 96 hours following injury. Thereafter, edema resolution ensues with most patients showing some measure of clinical improvement.

Clinical Severity of Brain Injury

Patients with a GCS of 8 or less are considered to have severe TBI and will require advanced medical care in the prehospital setting as well as neurocritical care management. Severe TBI is associated with significant neurologic injury often with structural lesions revealed by neuroimaging (e.g., head computed tomography [CT] scan revealing skull fracture, intracranial hemorrhage, and early diffuse cerebral edema). After initial resuscitation and stabilization in the field, patients with severe TBI should be evacuated to the nearest Level 1 trauma center with neurosurgical capability. Patients presenting with moderate TBI (GCS 9–12) should also be treated in the critical care setting while those with mild TBI (GCS 13–15) may or may not require hospital admission.30–33

It should be noted that in addition to the GCS parameter, mild TBI is formally defined as the presence of loss of consciousness for 30 minutes or less, post-traumatic amnesia not greater than 24 hours, any alteration in mental status at the time of injury, or focal neurologic deficit.34 In clinical practice, concussion and mild TBI are often used interchangeably; however, the terms are subtly distinct if only because they were independently defined by different expert panels. The American Academy of Neurology (AAN) defines concussion as a trauma-induced alteration in mental status with confusion and amnesia being the hallmarks.35 As the GCS does not provide enough detail to give a useful clinical picture in cases of mild TBI or concussion, a variety of scales have been developed (Table 28-2).35,36

A great deal of clinicopathologic heterogeneity exists in head injury classifications and so physician judgment taking into account the entire clinical picture as well as neuroimaging findings is paramount.37 While the classifications and grading scales developed by expert consensus are clinically useful, their use is typically not supported by class I evidence.32

Spinal Cord Injury Syndromes

Spinal cord injuries may be categorized as either complete or incomplete (Table 28-3). Both categories of acute spinal cord injury require patients to be monitored and treated in a critical care setting.17

A complete spinal cord injury (ASIA A) results in loss of all motor and sensory function at or just caudal to the cord lesion level. Acutely, spinal shock with flaccid paralysis, areflexia, and autonomic dysfunction is seen below the lesion level.38,39 For cervical lesions, this may result in bradyarrhythmias, conduction block, and hypotension due to disrupted sympathetic outflow. Although this clinical picture represents a spinal cord transection syndrome, an actual anatomic transection is rare.38

In contrast to complete injury, incomplete spinal cord injuries (ASIA B–D) are associated with varying levels of motor and sensory preservation. Several incomplete spinal cord injury syndromes have been described and include the following:

1. Central cord syndrome: This syndrome presents with upper greater than lower extremity weakness, urinary dysfunction, and varying degrees of sensory disturbance. It is thought to result from a cervical hyperextension injury in the setting of cervical spondylosis.

2. Brown-Séquard syndrome: Penetrating trauma resulting in spinal cord hemisection is the usual cause of this syndrome. The clinical findings of Brown-Séquard syndrome include: (1) ipsilateral motor paralysis and loss of posterior column function (proprioception and vibratory sense) below the lesion and (2) contralateral dissociated sensory loss with loss of pain and temperature but preserved light touch sensation.

3. Anterior cord syndrome: Cord infarction in the territory of the anterior spinal artery causes this syndrome. Traumatic etiologies such as disc herniation or retropulsion of a vertebral body fragment may result in occlusion of the anterior spinal artery causing paralysis below the lesion level and bilateral dissociated sensory loss (pain and temperature sensation) with sparing of position sense.

4. Conus medullaris syndrome: This syndrome affects the most caudal region of the spinal cord and presents with bladder and anal sphincter dysfunction, impotence, and saddle anesthesia. Lower extremity motor weakness may be limited. Conus medullaris syndrome must be distinguished from cauda equina syndrome, which affects lumbosacral nerve roots rather than the spinal cord. Pain and lower extremity weakness are more prominent in cauda equina syndrome.

5. Posterior cord syndrome: This syndrome is poorly defined and uncommonly seen in TSI. Lesions causing this syndrome damage the dorsal columns causing reduced proprioception, paresthesias, and dysesthetic (burning) pain. Alternate descriptions include additional corticospinal tract involvement producing weakness below the lesion.

Head Injury Imaging

A noncontrast head CT should be the initial imaging modality used to evaluate TBI patients. Although criteria have been developed to help identify TBI patients who require head CT,40,41 its liberal use appears justifiable in even mild cases to rule out pathology requiring immediate surgical intervention.32,42 The Marshall CT classification that segregates diffuse brain injury into several categories is often a helpful prognostic guide (Table 28-4).43 Conventional catheter cerebral and cervical angiography should be performed in cases of penetrating brain injury when breach of a major vessel is suspected. Brain magnetic resonance imaging (MRI) is generally not indicated in the acute setting.

Spinal Cord Injury Imaging

In patients with evidence of spinal cord injury, a CT scan of the entire spine should be performed to evaluate known spinal lesions and to rule out additional, noncontiguous vertebral injuries.24 An MRI of the known or suspected area of spinal cord injury should also be performed. MRI is superior to CT for the detection of traumatic intervertebral disc herniation, spinal ligament disruption, and epidural hematoma (EDH).

Cervical Spine Clearance

Cervical spine injury is found in association with TBI in 2–6% of patients. Thus, cervical spine immobilization with a hard collar should be continued in all trauma patients until either clinical or radiographic clearance can be accomplished.44,45 Initial spine imaging in both obtunded and awake symptomatic patients should include cervical spine x-rays in three views (anteroposterior [AP], lateral, odontoid) with clear visualization from the craniocervical junction to the C7–T1 junction. Any x-ray abnormalities should be followed up with cervical spine CT. If cervical radiographs and CT are unrevealing, either dynamic flexion–extension x-ray (fluoroscopy for obtunded patients) or cervical spine MRI within 48 hours of injury is required for clearance.

The initial management of TBI and TSI patients aims to both stabilize and identify pathology requiring emergent surgical intervention. A working knowledge of the surgical indications for the most commonly seen neurosurgical pathology in this population is therefore essential for any physician caring for TBI and TSI patients. Through the combined efforts of the Brain Trauma Foundation and the Congress of Neurological Surgeons, evidence-based recommendations were codified in the Guidelines for the Surgical Management of Traumatic Brain Injury.46 The specific intracranial traumatic lesions and the indications for surgery are outlined below. As the indications for emergency surgery in TSI patients are not well defined, only a brief discussion will be given. It is important to keep in mind that guidelines only provide recommendations for patients meeting specific criteria. Thus, clinical judgment must also guide management in each unique scenario.

Epidural Hematoma

An acute EDH with a volume greater than 30 mL should be evacuated regardless of the patient's GCS score. In patients with GCS less than 9, pupillary abnormalities, or other focal neurologic deficits, a craniotomy for evacuation should occur as soon as possible. An EDH with volume less than 30 mL, thickness less than 15 mm, and midline shift less than 5 mm in a patient with GCS greater than 8 and no focal deficits may be managed nonoperatively with serial head CT and close observation.47

Subdural Hematoma

An acute subdural hematoma (SDH) with thickness greater than 10 mm or midline shift greater than 5 mm should be evacuated regardless of the patient's GCS score. If the patient is comatose (GCS <9) and has pupillary abnormality or ICP >20 mm Hg or had a clinical decline by 2 or more GCS points, hematoma evacuation as soon as possible is also indicated.48

Parenchymal Lesions

Traumatic parenchymal lesions include both focal and nonfocal lesions. The focal lesions occur at the site of impact (coup) or opposite the site of impact (contrecoup) and include intracerebral hematoma (ICH), contusion, and infarction. Nonfocal lesions include diffuse injury typically resulting in hemispheric or global cerebral edema. Patients with traumatic parenchymal mass lesions causing neurologic deterioration, refractory intracranial hypertension, or evidence of mass effect on CT should be treated surgically. Similarly, any lesion greater than 50 mL should be evacuated. In patients with GCS 6–8, a lesion greater than 20 mL should be evacuated if it is frontal or temporal in location and causing more than 5 mm of midline shift and/or cisternal compression.49 Contusions commonly affecting the orbitofrontal and anterior temporal lobes must be observed with particular vigilance. Delayed hematomas that can manifest or “blossom” within hours to days may require urgent craniectomy (Figure 28-1).

Posterior Fossa Mass Lesions

A traumatic posterior fossa mass lesion should be evacuated by suboccipital craniectomy if there is radiographic evidence of mass effect or if neurologic dysfunction is referable to the lesion. As neurologic decline can be precipitous in patients with these lesions, surgery should be performed as soon as possible.50

Depressed Cranial Fractures

Closed (simple) nondepressed, typically linear, cranial fractures are not surgical lesions unless associated with an intracranial mass. On the other hand, depressed cranial fractures may be managed either surgically or nonsurgically depending on the particular case. Patients with open (compound) cranial fractures depressed greater than the thickness of the cranium should undergo early elevation of the bone fragments and debridement of the wound. Open cranial fractures with depression less than 1 cm and no dural penetration, significant intracranial hematoma, frontal sinus involvement, gross deformity, wound infection, pneumocephalus, or gross wound contamination may be treated nonoperatively. Although not rigorously supported in the literature, closed depressed cranial fractures are often surgically treated if the extent of depression is greater than the thickness of the adjacent calvarium to effect better cosmesis and lower rates of post-traumatic seizure (PTS) and neurologic deficit. Nonoperative management, however, is a treatment option in these cases.51

Penetrating Brain Injury

Penetrating brain injury involves both missile and nonmissile trauma to the brain. No strict guidelines dictate when surgical debridement, hematoma evacuation, and/or removal of protruding foreign body are warranted. In one study of gunshot wounds to the head, it was suggested that all patients with GCS 9–15 should have aggressive surgical therapy; patients with GCS 6–8 should have surgical therapy if no transventricular, multilobar, or dominant hemisphere injury is present; and patients with GCS 3–5 should have surgical therapy only if a large extra-axial hematoma is present.52

Spinal Decompression and Stabilization

There are no well-defined indications for emergency decompression and stabilization in TSI patients.53,54 While animal studies suggest that early decompression is beneficial, the available human studies do not consistently indicate improved neurologic outcomes. A recent systematic review concluded that urgent decompression in patients experiencing neurologic deterioration, with bilateral locked facets in the setting of incomplete tetraplegia, or cervical cord injury may be appropriate.55 However, further studies are warranted to clarify the role of emergent surgical intervention in TSI patients.

No discrete segregation exists between the emergency care, the surgical care, and the critical care of neurotrauma patients. The intensive care management of TBI and TSI patients begins in the field with first responders and continues until the patient is stable for intensive care unit (ICU) discharge. At all phases of care, the goal is to prevent additional primary injuries and to minimize the extent of secondary injury to the CNS. The recommendations set forth in the Brain Trauma Foundation's Guidelines for the Management of Severe Traumatic Brain Injury7 and in the American Association of Neurological Surgeons/Congress of Neurological Surgeons Joint Section on Disorders of the Spine and Peripheral Nerves' Guidelines for the Management of Acute Cervical Spine and Spinal Cord Injuries56 are invaluable resources detailing medical management. Additional guidelines have been released in an attempt to optimize prehospital, combat-related, and neurosurgical care. These include the Guidelines for Prehospital Management of Traumatic Brain Injury,57 the Guidelines for the Field Management of Combat-Related Head Trauma,58 and the previously mentioned Guidelines for the Surgical Management of Traumatic Brain Injury.46

Blood Pressure and Oxygenation

As already emphasized in the section “Initial Assessment and Classification,” both hypotension (systolic blood pressure <90 mm Hg) and hypoxia (SaO2 <90% or PaO2 <60 mm Hg) should be avoided.28 As an additional measure to improve spinal cord perfusion in TSI patients, it is recommended that a mean arterial pressure (MAP) of 85–90 mm Hg be maintained for 7 days after the injury.16 At a minimum arterial and central venous pressure (CVP), line placement is usually done in neurotrauma patients. Swan–Ganz catheter placement or use of arterial catheter device monitors is also helpful for hemodynamic optimization.

Intracranial Pressure Monitoring

A number of devices are available for ICP monitoring. Intraventricular devices provide the most accurate and reliable measurements of ICP and also allow for therapeutic drainage of CSF. However, in the setting of global cerebral edema with collapse of the lateral ventricles, ventriculostomy may not be possible. Parenchymal ICP monitors are also accurate and useful when ventriculostomy is not possible but their inability to effect CSF diversion makes them less ideal. Subarachnoid, subdural, and epidural ICP monitors are the least favored devices.59

ICP monitoring is often initiated in all salvageable severe TBI (GCS 3–8) patients and in select patients with GCS >8 whose neurologic exam cannot be followed. The guidelines recommend ICP monitoring in salvageable severe TBI patients with: (1) an abnormal head CT or (2) a normal head CT and two or more of the following—age over 40 years, motor posturing, and systolic blood pressure less than 90 mm Hg.60

Cerebral Perfusion Pressure

Cerebral perfusion pressure (CPP) is defined as the difference between the MAP and the ICP. CPP maintenance is an important clinical parameter that can be used to prevent the reductions in CBF that are associated with poor outcome. While the literature points to no clear optimal CPP target, multiple studies suggest maintenance within the range of 50–70 mm Hg with a goal of 60 mm Hg for the severe TBI population.61 Addressing CPP begins with placing an arterial line and providing fluid resuscitation to euvolemia. In order to increase the osmolar gradient between the systemic vasculature and the brain, hyperosmolar intravenous solutions should be used. Normal saline—which is hyperosmolar relative to blood—is commonly used. Another option is hypertonic saline, which is a useful adjunctive therapy in managing cerebral edema and intracranial hypertension (see below).

Controlling Intracranial Pressure

Consensus recommendations based on class II evidence suggest that treatment for ICP should be initiated for values above 20 mm Hg.62 Nonetheless, it is important to remember that herniation may occur at pressures below this value and so the entire clinical picture must be assessed. The methods below detail a fairly stepwise approach to the treatment of intracranial hypertension in neurotrauma patients. At all stages during the escalation of ICP-lowering therapy, a high suspicion for enlarging mass lesion must be entertained with a low threshold for obtaining a head CT.

1. General measures: The general measures used for decreasing ICP include ensuring neutral head position, elevating the head of bed (HOB), providing adequate sedation and analgesia, and avoiding fever. A neutral head position allows for unimpeded venous drainage out of the intracranial compartment, thus decreasing cerebral blood volume and ICP. Keeping the HOB elevated at 30–45° also enhances venous outflow. Agitation, discomfort, endotracheal tube–induced coughing, tensing the abdominal musculature, and elevated sympathetic tone all increase ICP. The use of adequate sedation and analgesia with agents such as propofol or midazolam is helpful with these issues.

   As fever is known to increase both the cerebral metabolic rate and ICP, maintenance of normothermia along with the prevention of shivering is important. Although TBI alone can induce a fever, this should be a diagnosis of exclusion. Febrile patients must be appropriately assessed for underlying infection with tests including complete blood count, chest x-ray, urinalysis, and cultures of blood and urine. Fever reduction can be achieved with acetaminophen, cooling blankets or vests, or intravascular devices. If a patient develops severe shivering, one should consider administering meperidine, buspirone, magnesium, or increasing sedation with or without pharmacologic paralysis.

2. CSF drainage: One of the major benefits to ICP monitoring with an intraventricular device is that it allows for CSF diversion. Unclamping the intraventricular catheter and draining 3–5 mL of CSF is one of the fastest and most effective methods of lowering ICP.

3. Hyperventilation: Hyperventilation reduces ICP by causing cerebral vasoconstriction. As CBF is also reduced by cerebral vasoconstriction, one must necessarily question the safety of hyperventilation in the setting of TBI. The only randomized controlled trial addressing this question found significantly worse outcomes at 6 months in a subgroup of patients hyperventilated to a Paco2 of 25 mm Hg as compared with the control group ventilated to a Paco2 of 35 mm Hg.63 Based on this along with additional class III data, prophylactic hyperventilation is not recommended in TBI patients.64 However, hyperventilation may be briefly used as a temporizing measure in an acute setting such as active or imminent herniation. The clinical response is very fast and lasts anywhere from a few minutes to hours. Efficacy subsequently wanes due to metabolic compensation within the CSF and brain tissue. When employed, efforts should be made to wean the patient from this type of therapy as soon as other interventions become effective.65

4. Osmotherapy: Mannitol is the most commonly used ICP-reducing osmotic agent in the setting of TBI. It is an osmotic diuretic that lowers ICP through both rheologic and osmotic effects. Mannitol administration immediately expands plasma volume reducing blood viscosity, increasing CBF, and increasing cerebral oxygen delivery. More delayed osmotic effects dehydrate the intracellular and interstitial brain compartments pulling water intravascularly. Bolus administration at doses of 0.25–1 g/kg is effective.66 In an effort to prevent renal toxicity, a cutoff serum osmolarity of 320 mOsm is often used. This value is somewhat arbitrary and so in the setting of ICP crisis a higher osmolarity may be tolerated. Avoiding severe hypernatremia (Na >160 mEq/L) is also prudent as this has been associated with increased mortality.67

   Hypertonic saline is also an osmotic agent that effectively lowers ICP. Hypertonic saline concentrations up to 23.4% are frequently used in clinical practice, and it may represent an important alternative to mannitol for controlling ICP.68 However, it is less well studied in the TBI population and recommendations regarding the use of hypertonic saline have been deferred.

5. Surgical decompression: Approximately 10–15% of severe TBI patients experience intracranial hypertension refractory to maximal medical therapy.69 At present, decompressive surgery is one of the last strategies save high-dose barbiturate therapy employed in the management of such patient. It remains to be seen whether the promising results obtained from retrospective cohort studies69 will be reflected in ongoing randomized controlled trials.

6. Barbiturate therapy: TBI patients with intracranial hypertension refractory to maximal medical and surgical therapy may be treated with high-dose barbiturates.70 The beneficial effects of barbiturates are attributed to their ability to induce cerebral vasoconstriction, decrease cerebral metabolism and CBF, and act as a free radical scavenger. Given the plethora of complications associated with the use of high-dose barbiturates, one must consider this a “last ditch” effort as their prophylactic use has no benefit. If barbiturate coma is to be used, pentobarbital may be administered with an intravenous loading dose of 10 mg/kg over 30 minutes followed by an infusion at 1–3 mg/kg/h titrated to burst suppression on electroencephalogram (EEG). Barbiturates are myocardial depressants and thus aggressive cardiovascular management is necessary to maintain systemic blood pressure. ICP unresponsive to this therapy is an ominous sign and consideration for further neurosurgical decompression is warranted.

Hypothermia

Systemic hypothermia with target temperatures between 33°C and 35°C has been used in neurotrauma patients. Although there are known benefits in terms of ICP control to avoiding fever, the data are less clear regarding the neuroprotective effects of prophylactic hypothermia. No mortality benefit to prophylactic hypothermia has been demonstrated in TBI patients. However, benefit by other measures such as the Glasgow Outcome Score is supported by the available literature.71

In the TSI population as well, the evidence for therapeutic benefit from systemic hypothermia is not yet available and awaits the results of ongoing clinical trials.72 A committee from the Joint Section of the AANS and the CNS concluded that there is insufficient evidence to recommend either for or against the practice of local or systemic hypothermia in the treatment of acute spinal cord injury.73

Brain Oxygen Monitoring

Low brain oxygenation as measured globally by jugular venous oxygen saturation (SjvO2) or locally by brain tissue oxygen tension (PbtO2) is associated with poor outcomes.74 Desaturations of SjvO2 to values less than 50% have been shown to indicate various ischemia inducing pathologies such as elevated ICP, hypocarbia, arterial hypoxia, systemic hypotension, and cerebral vasospasm. Sheinberg et al showed a trend toward increasing death rates in TBI patients who had multiple SjvO2 desaturations.75 Similarly the length of time PbtO2 levels were less than 15 mm Hg is also associated with increased death rates.76 In a recent retrospective series, TBI patients managed with PbtO2-directed therapy had better outcomes than patients managed with standard ICP/CPP-directed therapy.77 Interestingly elevated SjvO2 levels (>75%) are associated with poor outcomes and may be an indication of hyperemia or infarction.78 Therefore, brain oxygen monitoring not only gives information on brain state but also provides another parameter that can be clinically optimized.

Steroids

1. Steroids in TBI: Corticosteroids have no role in the management of severe TBI. Class I data from large randomized controlled studies indicate that they neither improve outcome nor lower ICP in this setting.79,80 Indeed, the data suggest that the use of steroids is harmful in TBI patients.

2. Steroids in TSI: Findings from the second National Acute Spinal Cord Injury Study (NASCIS II) form much of the basis for using corticosteroids after TSI.81 This multicenter randomized controlled study concluded that the administration of methylprednisolone within the first 8 hours after spinal cord injury improves neurologic recovery. The dosage protocol used was an initial intravenous bolus of 30 mg/kg/h followed by infusion at 5.4 mg/kg/h for 23 hours. In a follow-up study—NASCIS III—improved motor recovery was seen if the methylprednisolone infusion continued for 47 hours in the subgroup of patients who began steroid therapy between 3 and 8 hours after injury.82 While this protocol is often implemented in clinical practice, it cannot be considered standard of care given the numerous criticisms leveled at the NASCIS studies. The current guidelines suggest methylprednisolone therapy as an option with the proviso that harmful side effects are more consistent than evidence of clinical benefit.83

Seizure Prophylaxis

Seizures may be seen in up to 25% of TBI patients within the first 7 days of injury and in up to 42% after this.84 The literature currently supports the use of seizure prophylaxis to decrease the occurrence of early (within 7 days of injury) PTS but not for late PTS. Therefore, continuation of anticonvulsant therapy for more than 1 week following TBI is not recommended.84 While phenytoin has been the historical drug of choice in the setting of TBI, use of levetiracetam or newer agents such as lacosamide may be similarly beneficial with a better side effect profile, no known significant drug–drug interactions, and no need for serum level monitoring.85–87