Mechanism

Blunt injury is the most common mechanism and includes low- and high-velocity impacts. These injuries may be from a direct blow (e.g., a club) or a rapid deceleration force (e.g., falling or striking the windshield in a car crash). As already noted, motor vehicle collisions and falls cause most of the severe TBIs.

Gunshot wounds are the most prevalent penetrating head injuries (Fig. 50B.1), accounting for 35% of U.S. deaths from TBI in individuals younger than age 45. Self-inflicted injuries—for example, with power tools like nail guns (Fig. 50B.2)—are another cause of craniocerebral trauma (Testerman and Dacks, 2007), but gunshot wounds are the most lethal type of TBI, 90% of them resulting in death. They cause soft-tissue damage, often comminuted depressed skull fractures, and direct injury to the brain tissue from the missile. Beyond the laceration along the bullet path, further damage is done by shock waves, especially with high-velocity weapons. Depending on the type of penetrating injury, contamination may be a concern. However, bullet fragments are considered sterile owing to their exposure to heat from the firearm and are not routinely removed.

Fig. 50B.1 Picture of gunshot wound to the head, a form of penetrating injury. The bullet leaves a path of destruction marked by blood and bone and bullet fragments. Brain tissue around the bullet path is damaged as well by the propagating pressure wave. However, this is not visible on CT (axial computed tomography).

Fig. 50B.2 Another form of penetrating injury is shown on this anterior-posterior x-ray of the skull showing nails from a nail gun stuck in the skull and penetrating the right frontal lobe. This patient was neurologically intact on presentation. The difficulty is the bleeding that can occur deep in the brain when the nails are removed. This patient is also at high risk for developing pseudoaneurysm later on.

Vascular injury can occur with any type of TBI but is more common with penetrating trauma (25%–36% incidence) than with blunt injuries (<1%). Traumatic aneurysms, or pseudoaneurysms, can develop when all layers of the vessel wall rupture and surrounding cerebral tissue forms the aneurysmal wall. Such pseudoaneurysms can present as delayed subarachnoid hemorrhage (SAH) and can lead to death or severe disability. Therefore patients with penetrating injuries routinely undergo a cerebral angiogram to screen for traumatic pseudoaneurysm. A follow-up angiogram within the first months after a penetrating injury is recommended (du Trevou and van Dellen, 1992).

Blast is a rare mechanism of injury in civilian life, but it is common in combat. With ongoing militarily conflicts, blast injury has become more frequent in U.S. military personnel. American military members who returned from deployment to Iraq or Afghanistan self-reported a 12% to 20% incidence of mild TBI; explosions were recognized as a common injury mechanism (Wolf et al., 2009). Blast injuries can occur (1) as a direct result of supersonic waves of intense air (primary blast injury), (2) from being struck by objects set in motion by the blast (secondary blast injury), and (3) when a person is forcefully propelled by the blast (tertiary blast injury) (DePalma et al., 2005; Kocsis and Tessler, 2009). The brain is obviously vulnerable to secondary and tertiary blast injury, but whether primary blast forces directly injure the brain is controversial (Taber et al., 2006). The severity of blast exposure necessary to cause persistent symptoms is unclear (Hicks et al., 2010).

The severity of injury resulting from exposure to a blast can range from mild to fatal. Displacement, stretching, and shearing forces of the primary blast wave can affect the brain directly and lead to such sequelae as concussion, hemorrhage, severe edema, or diffuse axonal injury (DAI). Systemic acute air embolism from pulmonary disruption is believed to occlude the blood vessels of the brain or spinal cord and lead to stroke (Guy et al., 2000; Wolf et al., 2009). Reports from war zones suggest that brain swelling occurs within hours after a blast injury, and that the risk of mortality can be decreased by early decompressive craniectomy. Others have found that cerebral vasospasm occurred more often after a blast injury than with blunt TBI and led to a worse outcome than blunt severe TBI. Blast exposure on the mild end of the spectrum led to somatic, behavioral, psychological, and cognitive symptoms similar to postconcussion syndrome. There is well-recognized overlap between posttraumatic stress disorder and blast injury, and misdiagnosis can occur in both directions.

Injury Severity

There are different methods to stratify TBI by severity, and all are arbitrary. The most commonly used is the Glasgow Coma Scale (GCS), which was first published in 1974 by Graham Teasdale and Bryan K. Jennett, two neurosurgeons at the University of Glasgow (Teasdale and Jennett, 1974). The GCS assesses level of consciousness. It consist of three subscores: eye opening (maximal score: 4), verbal response (maximal score: 5), and motor response (maximal score: 6). Despite some shortcomings, it has excellent interrater agreement (weighted kappa >0.75) for verbal and total scores and intermediate agreement (weighted kappa 0.4-0.75) for motor and eye scores (Holdgate et al., 2006). Today it has been widely adopted by emergency medicine services and physicians to communicate the severity of injury in a consistent way. However, it is not designed to detect focal neurological deficits and is not a replacement for a thorough neurological examination.

A GCS score of 15 to 13 is considered to indicate mild TBI, and a score of 8 or below is a universally accepted definition of coma and severe head injury. Scores between 13 and 8 are classified as moderate. It is true that moderate and severe TBI are rare; mild TBI is eight times more common. Moderate and severe TBI are seen equally often. One can imagine that the GCS score shows significant ceiling effect in patients with mild TBI. Therefore, more sensitive measures exist for this purpose. The two most often used are the Cantu system (Cantu, 1996) and the American Academy of Neurology system (1997).

Mild TBI is extremely common. In North America, between 1.5 and 2 million patients visit an emergency department each year for head trauma, with 70% to 90% of cases being mild TBI. Because this number does not include the many people who choose not to seek medical attention, the true prevalence of mild TBI is difficult to ascertain.

Despite its high incidence, no objective test to diagnose mild TBI currently exists. Only one of ten head CT scans has a positive finding in mild TBI patients. Therefore, special factors need to be considered when determining the need of a CT scan (Box 50B.1). Diagnosis is hampered by the lack of obvious injuries on computed tomography (CT) or conventional magnetic resonance imaging (MRI) and usually is based solely on clinical symptoms. Preliminary studies are investigating brain injury biomarkers to diagnose mild TBI (Dash et al., 2010; Hergenroeder et al., 2008; Kochanek et al., 2008). Although the initial progress has been promising, these markers are not yet used in the clinical setting (Springborg et al., 2009; Unden and Romner, 2009). Others have shown that an integrated approach with magnetoencephalography and diffusion tensor imaging (DTI) is more sensitive than conventional CT and MRI in detecting subtle neuronal injury in mild TBI and postconcussion syndrome (Huang et al., 2009).

A concussion is defined as injury to the brain caused by a hard blow or violent shaking, producing a sudden and temporary impairment of brain function such as a brief loss of consciousness or disturbance of vision and equilibrium. Concussions are equivalent to a mild TBI with negative CT findings. Their annual incidence in the United States is 128 per 100,000 population (Ropper and Gorson, 2007). The loss of consciousness is believed to result from rotational forces exerted on the upper midbrain and thalamus, impairing the function of the reticular neurons. Headache, nausea, dizziness, irritability, and impaired ability to concentrate can persist for days after the event. Persistence of these symptoms for weeks is called postconcussion syndrome and can last from 1 month to a year.

Patients with severe TBI (GCS score ≤ 8) often have injuries in at least one other organ system (Saul and Ducker, 1982). There is also a 5% incidence of associated spine fractures. About one-fourth of patients with a severe TBI undergo neurosurgical intervention. Early diagnosis of TBI improves outcome by reducing secondary injury, which can develop subsequent to the impact and includes edema, hypoxia, hypotension, ischemia, and elevated intracranial pressure (ICP) (Chesnut et al., 1993a; Chesnut et al., 1993b).

Morphology

Traumatic brain injury can also be classified by the underlying morphology and can be subdivided as injuries to the skull itself and intracranial injuries, which can be focal or diffuse (see Table 50B.1).

Skull Fractures

There are two types of skull fractures. Linear fractures traverse the full thickness of the skull from the outer to inner table, are usually fairly straight, and involve no bone displacement. The common method of injury is blunt force trauma in which the energy from the blow is transferred over a wide surface area of the skull. Fractures can be located in the calvarium or in the skull base. Skull base fractures are associated with cranial nerve (CN) palsy. Depending on their location, CN VII (temporal bone) and CN VI (clivus) are at risk; they are also prone to causing cerebrospinal fluid (CSF) leaks (Fig. 50B.3). In depressed skull fractures, the method of injury is similar to linear skull fractures, but in this case the inner and outer table are more shattered, and pieces are displaced into the brain (Fig. 50B.4). They may or may not lacerate the dura. The presence of skull fractures consistently has been associated with a higher incidence of intracranial lesions, neurological deficits, and poorer outcome (Bullock et al., 2006a). Hence, cranial fractures are important indicators of intracranial injuries, and CT scanning should be performed to rule out serious intracranial pathology in all patients with a known or suspected cranial fracture. A large study of patients with a TBI found that 71% of 850 patients with a cranial fracture had an intracranial lesion (e.g., contusion, hematoma), compared with only 46% of 533 patients without a cranial fracture (Macpherson et al., 1990).

Fig. 50B.3 Basilar skull fractures through the anterior skull base typically cause tears in the adjacent dura and cerebrospinal fluid rhinorrhea. Computed tomography scans through the base of the skull may not show the fracture itself, but fluid in the sphenoid sinus (arrows) or the other paranasal sinuses is often seen. (Axial computed tomography scan, bone window.)

Fig. 50B.4 Depressed skull fracture (axial computed tomography).

Intracranial Injuries

Contusions occur when the brain moves within the skull enough to collide with the skull, causing bruising of the brain parenchyma (hemorrhage and edema). The most common locations are the frontal and temporal poles, because these areas most often strike the bony surfaces of the skull base. Often, a second injury (contrecoup) is seen diagonally across the site of the direct impact (coup injury) when the accelerated brain strikes the skull on the opposite site. In contrast to concussion, contusions are visible on a head CT scan. Edema surrounding a contusion has lower signal intensity than brain on CT scan (Fig. 50B.5). Hemorrhages have higher CT signal intensity and are commonly seen as multiple bright areas of variable size. Progression of contusions is common, with delayed hemorrhage occurring over the first 24 hours in 25% of cases. However, neurosurgical intervention is rare.

Fig. 50B.5 Temporal lobe contusions (black arrow) must be followed closely because even a slight enlargement can cause uncal herniation (white arrow), often without an increase in the intracranial pressure. (Axial computed tomography scan.)

An epidural hematoma (EDH) is bleeding between the dura and the skull. The peak incidence of EDH occurs in the second decade, and the mean age of TBI victims with EDH is between 20 and 30 years of age. Traffic-related accidents, falls, and assaults account for 53% of EDH in adults; falls account for half of EDH in children. Most (70%–90%) EDH are associated with a skull fracture and bleeding from a lacerated artery or indirect bone bleeding. Arterial bleeding was identified as the source of EDH in 36% of adults in a systematic review. Rupture of the middle meningeal vein, diploic veins, or venous sinuses also can cause EDH. The most common location of EDH is temporal; because the temporal bone is thinner than the rest of the skull, it breaks more easily, often resulting in a laceration of the middle meningeal artery (Bullock et al., 2006b) (Fig. 50B.6).

Fig. 50B.6 A, Right frontoparietal epidural hematoma, typical lenticular-shaped hyperdense lesion in the epidural space, with mass effect and midline shift. Also note absence of sulci, consistent with brain in children. B, Postoperative result. Air is visible under the bone flap.

Subdural hematoma (SDH) develops between the brain parenchyma and the dura (Fig. 50B.7). It occurs when the brain moves within the skull enough to tear a surface vein, also called a bridging vein, that runs from the brain surface to the dural venous sinus. The most common locations are the frontal and parietal convexities. On a CT scan, an acute SDH typically appears as a bright (hyperdense), extraaxial, crescent-shaped homogeneous collection of fluid that conforms to the cerebral surface. Unlike an EDH, its spread is not limited by suture lines; it can spread over the whole convexity, but it almost never crosses the midline. An SDH can be acute or chronic. Herein, acute SDH is defined as an SDH diagnosed within 14 days of TBI. Chronic SDH contain blood older than 14 days or blood of different ages. This accounts for their appearance. They may be of mixed density or isodense to gray matter. In these cases, it can be identified by its mass effects, including sulcal effacement, inward buckling of the gray/white interface, and presence of midline shift.

Fig. 50B.7 Left-sided acute subdural hematoma (SDH). Note concave shape along the convexity, in contrast to convex shape seen in epidural hematoma. This patient was comatose on arrival in the emergency room and required emergency surgery for left-sided craniotomy and evacuation of SDH. Also seen is significant subfalcine herniation (red arrows) (axial computed tomography).

In studies conducted after the introduction of CT, the incidence of SDH was estimated to be 11% of all patients with TBI (Servadei et al., 2000). The most common injuries involving traumatic SDH are motor vehicle collisions in the younger population and falls in those older than 75 years. Between 37% and 80% of patients with acute SDH present with an initial GCS score of 8 or less, and dilated pupils are seen preoperatively in 50% of these patients. Mortality among patients who arrive at the hospital in a coma and undergo surgical evacuation is between 57% and 68%.

Traumatic subarachnoid hemorrhage (tSAH), seen with mild and severe TBI alike, refers to microhemorrhages into the subarachnoid space from crushed or ruptured smaller vessels. CT scans show hyperdense linear areas in superficial brain sulci (Fig. 50B.8). In contrast to aneurysmal SAH, the blood is superficial in the cortex and not present in the basal cisterns. In some instances, for example, if blood is found in the sylvian fissure, a vascular study (CT angiography, four-vessel digital subtraction angiography) is needed to rule out an aneurysm rupture. Most of the time, traumatic SAH indicates a more severe underlying brain injury and is seen in conjunction with DAI, SDH, or EDH. It has been shown that the presence of traumatic SAH is an independent factor for poor outcome. This finding could reflect a more extensive underlying injury that CT scan or conventional MRI cannot detect, or it could be attributable to a less recognized phenomenon: posttraumatic vasospasm. The incidence of posttraumatic vasospasm in patients with a mean GCS score of 7 (range 3-15) is similar to that following aneurysmal SAH (see Chapter 51C). Posttraumatic vasospasm is often overlooked, but one study found that hemodynamically significant vasospasm has been found in 44% of patients with tSAH (Oertel et al., 2005). Patients were at highest risk of developing hemodynamically significant vasospasm on day 3. Younger patients and patients with GCS scores of 8 or less were more likely to develop posttraumatic vasospasm.

Fig. 50B.8 Axial computed tomography scan of 35-year-old trauma patient with traumatic subarachnoid hemorrhage in the left frontal gyrus.

Diffuse axonal injury is very common severe head injuries and is associated with significant morbidity. It is characterized clinically by rapid progression to coma in the absence of specific focal lesions. Adam classified DAI as mild, moderate, or severe (Adams et al., 1989). The duration of coma after a TBI correlates with the severity of DAI. Patients who survive the most severe form rapidly lapse into coma and remain unconscious, vegetative, or severely disabled.

DAI is a result of shearing, stretching, and/or angular forces pulling on axons and small vessels. Impaired axonal transport leads to focal axonal swelling and may result in axonal disconnection. The most common locations are the corticomedullary (gray matter/white matter) junction (particularly in the frontal and temporal areas), internal capsule, deep gray matter, upper brainstem, and corpus callosum.

MRI is more sensitive than CT in detecting DAI. Gradient echo MRI is most sensitive to areas of hemorrhage, and fluid-attenuated inversion recovery images are best for visualizing nonhemorrhagic lesions. Diffusion tensor imaging is a new imaging technology more sensitive to DAI than conventional MRI. With DTI, the connectivity and integrity of neuronal tracts can be investigated. DTI is helpful when imaging tissue with an internal structure such as neural axons in white matter of the brain. Water molecules will then diffuse more rapidly in the direction aligned with the internal structure, and more slowly as it moves perpendicular to the preferred direction. With this information, injury to the white matter (diffusion anisotropy) can be detected and neural tracts followed through the brain (Fig. 50B.9). DTI measurements correlate with injury severity (acute GCS score) and outcome (Rankin Score at discharge) (Li and Feng, 2009; Sugiyama et al., 2007).

Fig. 50B.9 Two examples of patients with diffuse axonal injury. Diffusion tensor imaging (DTI) fiber tractography around the corpus callosum in a midsagittal plain. In the pictures on the left there were fewer fibers extending to the frontal, parietal, and occipital white matter, and they were even absent in some areas (arrows) in comparison with the normal volunteer image.

(From Sugiyama, K., Kondo, T., Higano, S., et al., 2007. Diffusion tensor imaging fiber tractography for evaluating diffuse axonal injury. Brain Inj 21, 413-419.)

Evaluation

Early intervention is essential in successful TBI care; as in any emergency situation, the CABs (circulation, airway, breathing) apply. Another priority of prehospital care is to optimize perfusion and oxygenation to prevent hypoxia and hypertension. Once in the emergency room, further measures can be taken if necessary to achieve adequate cardiopulmonary function. Next, a neurological examination is performed to triage patients accordingly. The GCS score is widely used to convey the severity of TBI, but a low GCS score upon arrival in the emergency room may be the result of sedation in the field. Only after the patient is weaned from sedation can injury severity be accurately assessed. Trauma patients with altered mental status, pupillary asymmetry, and flexion or extension posturing are at high risk for SDH or EDH compressing the brain and brainstem and must be evaluated with a CT scan of the head. CT is central to acute TBI diagnosis. There are two externally validated rules for when to require a CT scan after mild TBI: the Canadian CT Head Rule (CCTHR) (Stiell et al., 2001) and the New Orleans Criteria (NOC) (Haydel et al., 2000). These rules have 100% sensitivity for neurosurgical lesions and 83% to 98% sensitivity for nonoperative lesions. The CCTHR has a greater specificity and hence has led to fewer CT scans than the NOC (Levine, 2010). Neither addresses the presence of coagulopathy and anticoagulation, which others deem important risk factors for intracranial hemorrhage (Cohen et al., 2006; Ivascu et al., 2008). These criteria were updated in 2007 by the World Health Organization Taskforce on mild TBI and the Neurotraumatology Committee of the World Federation of Neurosurgical Societies (Fabbri et al., 2004) and in 2008 by the Centers for Disease Control and Prevention and the American College of Emergency Physicians. All in all, research findings agree on several indications for urgent CT scanning of the head after a minor TBI (see Table 50B.1).

Depending on CT findings, patients will undergo emergency surgery or nonoperative management of TBI. Neurocritical care and any therapeutic intervention, operative or nonoperative, have one goal: prevention of secondary injury, which unlike an irreversible primary injury can be limited.

Management of Diffuse Injuries

Patients with complicated mild TBI (positive CT findings) and moderate TBI are admitted to the intensive care unit (ICU) and followed with serial neurological examinations. Although intracranial complications are uncommon after a mild TBI, in 6% to 12% of patients, CT shows abnormalities (Ibanez et al., 2004; Smits et al., 2005) in the form of a contusion, tSAH, SDH, or EDH. Another CT scan obtained within 12 to 24 hours is currently the standard of care in many hospitals. However, studies of 1 million emergency department visits for mild TBI found that neurosurgical intervention was required in only 0.13% to 0.3% of these patients (Fabbri et al., 2004; Haydel et al., 2000; Jeret et al., 1993; Miller et al., 1996; Nagy et al., 1999). Over the past several years, evidence has suggested that the routine use of follow-up CT scans in patients with complicated mild TBI may not be necessary (Fabbri et al., 2008; Huynh et al., 2006; Nagy et al., 1999; Sifri et al., 2006; Velmahos et al., 2006). Most of these cases are not operative, and the patient can be discharged home after a short observation period (Thiruppathy and Muthukumar, 2004). The only exception is an EDH, which can present with a lucid interval after a brief loss of consciousness and then progress to coma and death (Levine, 2010).

Concussions are graded according to the severity of symptoms: presence or absence of loss of consciousness and time (< or > 15 minutes) for resolution of symptoms (Table 50B.2). Data to guide treatment of concussion are lacking. Clinical experts suggest a benefit from the use of mild analgesics for headache, avoidance of narcotics, and use of meclizine or promethazine and vestibular exercise for dizziness. Brief seizures may occur immediately after the initial impact and easily can be mistaken as a sign of a severe head injury.

Table 50B.2 Concussion Grading

AAN, American Association of Neurology; LOC, Loss of consciousness; PTA, posttraumatic amnesia.

Because 10% of all head injuries are sports related, return to play after concussion is an important issue. The American Academy of Neurology (1997), the Canadian Academy of Sports Medicine (2000), and several international symposia (McCrory et al., 2009) have developed guidelines for return to play after TBI, based mostly on expert opinion. Because no consensus has been reached regarding the most appropriate set of guidelines, the return-to-play decision remains in the realm of clinical judgment on an individualized basis. All guidelines agree that a symptomatic player should never return to play because of the potentially increased vulnerability of the brain to a second injury. This sensitivity may outlast the symptoms, which tend to resolve within 1 week. Evidence suggests that after a concussion, neuronal metabolic equilibrium and oxygen consumption are abnormal for several months. This suggests an initial and rapid phase of functional recovery driven by compensatory mechanisms, followed by a prolonged neuronal recovery period during which subtle deficits in brain functioning are present but not apparent to standard clinical assessment tools (Ellemberg et al., 2009). Depending on the grade of the concussion, the athlete may resume play within 1 week or 1 month (Mayers, 2008) (Table 50B.3). An objective measure of postural stability increases the sensitivity to ongoing neurological impairment in the return-to-play decision-making process and minimizes the consequences of mitigating factors such as practice effects and motivation. During recovery in the first few days after a concussion, it is important to emphasize to the athlete that physical and cognitive rest are required.

Table 50B.3 Guidelines for the Management of Sports-Related Concussion*

* These guidelines reflect consensus opinion, are not evidence-based, and are under revision. Adapted from the American Academy of Neurology guidelines.

† Testing includes orientation, repetition of digit strings, recall of word list at 0 and 5 minutes, recall of recent game events, recall of current events, pupillary symmetry, finger-to-nose and tandem gait tests, Romberg test, and provocative testing for symptoms with a 4-yd (3.5-m) sprint, 5 push-ups, and 5 knee bends.

Management of nonoperative severe TBI is directed by the evidence-based Guidelines for the Management of Severe Traumatic Brain Injury. The most recent (third) edition was published in 2007 by the Brain Trauma Foundation (Povlishock and Bullock, 2007) (Table 50B.4). With severe TBI, ICP-guided and/or CCP-guided therapy are the fundamental principles underlying all treatment. Patients with a GCS score of 8 or below qualify for ICP monitoring. The gold standard remains the external ventricular drain (EVD), which allows both ICP measurement and CSF drainage to reduce ICP. Other more expensive options are fiberoptic or microstrain gauge devices, which are placed in the brain parenchyma. These devices cannot be recalibrated during monitoring and have a negligible drift. Subarachnoid, subdural, and epidural monitors measure ICP via a fluid-coupled or pneumatic mechanism and are less accurate than an EVD.

An escalating protocol should be followed to control ICP (Box 50B.2), using sedation, hyperosmolar therapy in the form of mannitol and hypertonic saline, and hyperventilation. The sequence in which these treatment options are chosen varies from hospital to hospital. Currently, at my institution, patients with increased ICP initially receive a 250-mL hypertonic 3% saline bolus and continuous 3% saline infusion if needed. Mannitol is reserved for an emergency situation such as an acutely dilated pupil and imminent brain herniation. Serum sodium can be elevated safely to 155 mEq/L and a serum osmolality of 320 mOsm/kg. Beyond those levels, renal failure can develop. If hypertonic saline administration is stopped suddenly, rebound intracranial hypertension is a risk but can be avoided by tapering the hypertonic saline solution over 48 hours.

Box 50B.2 Escalating Protocol for Treatment of Intracranial Pressure Above 20 mm Hg for 10 Minutes

1. Verify intracranial pressure (ICP).

Check if external ventricular drain (EVD) is still present.

Check if waveform EVD is present and adequate.

Check if EVD ICP correlates with intraparenchymal monitor if present.

2. Check for 30-degree head elevation.

3. Loosen cervical collar if in place.

4. Open EVD for ICP > 20 mm Hg for 10 minutes, then close and transduce ICP.

Repeat as needed.

5. Treat temperature > 37.5°C with 650 mg Tylenol, once.

6. Sedation:

Titrate propofol to a Ramsay score of 4.

Do not exceed 5 mg/kg/h for more than 24 hours.

Check potassium, triglycerides, creatine kinase, and urinalysis for myoglobinuria q 8 h for 24 h.

If maximum dose of propofol is reached and ICP >20 mm Hg:

Start fentanyl drip to 0.8 µg/kg/h.

Apply bispectral index (BIS) monitor.

Titrate fentanyl drip to a BIS of 30 or to maximum 5 µg/kg/h.

7. Hyperosmolar therapy:

3% Hypertonic saline bolus of 250 mL

Before administering 3% hypertonic saline bolus, check if Na < 130 mEq/L.

Mannitol 0.25 to 1 mg/kg bolus once in emergency

Check sodium and serum osmolality q 4 h × 2 after every dose.

Start 3% hypertonic saline drip at 30 mL/h if ≥ three 3% hypertonic saline boluses within 6 h.

Check sodium and serum osmolality q 2 h while on drip.

8. Hyperventilation:

Do not hyperventilate in the first 24 hours (goal Paco₂ of 35-40 mm Hg).

If PbtO₂ is <20 mm Hg evaluate and treat brain hypoxia.

If CBF <10 mL/min/100 g white matter or <67 mL/min/100 g gray matter, evaluate and treat CBF.

If PbtO₂ and CBF are optimized, hyperventilate to 33-35 mm Hg.

9. Radiology:

If refractory—ICP >20 mm Hg despite above intervention—obtain portable head CT without contrast.

10. Consider surgery.

11. Induce pentobarbital coma:

!Only for diffuse, nonoperative injuries

Order continuous electroencephalography (EEG) if not already in place.

Have norepinephrine drip ready at the bedside for mean arterial pressure (MAP) <90 mm Hg, CPP (cerebral perfusion pressure) <60 mm Hg

Pentobarbital bolus/loading: 10 mg/kg, then 5 mg/kg hourly until burst suppression.

Pentobarbital maintenance dose: 1 mg/kg/h, titrate to burst suppression.

The TBI Guidelines recommend striving for a CPP above 60 mm Hg, a change from the goal of 70 mm Hg in the second edition published in 2000. Although maintaining CPP this high is helpful in patients with posttraumatic vasospasm, it can be harmful in patients with cerebral hyperemia. Multimodality monitoring can be helpful to identify this subpopulation with hyperemia. Vasoactive pharmacologic agents must be used with caution. They may have effects on cerebral vasomotor tone, causing vasoconstriction in the brain, and should be used only if the patient’s neurological status worsens and CPP cannot be maintained with intravenous fluids alone.

Hyperventilation has the potential to reduce ICP via reflex vasoconstriction in the presence of hypocapnia. The vasoconstriction leads to decreases in CBF, overall cerebral fluid volume, and therefore lower ICP. The potential dangers associated with inducing hypocapnia are related to the resultant vasoconstriction. Excessive hypocapnia may lead to ischemia secondary to insufficient CBF. Therefore, it is not recommended in the first 24 hours after TBI or prophylactically. Individual autoregulation of CBF with respect to hypocapnia varies widely and is difficult to predict. The measurement of brain tissue oxygenation can help guide hyperventilation.

Paralysis is less frequently used to lower ICP because of a higher incidence of ventilator-associated pneumonia and longer hospital stays.

ICP that remains elevated despite maximal medical efforts is an indication for decompressive craniectomy or bilateral frontal craniectomy. Depending on the underlying pathology, one side of the skull or the frontal bone bilaterally (Kjellberg procedure) is removed to give the swollen brain more room. To date, the effectiveness of decompressive craniectomy is controversial. Findings from non-randomized trials and controlled trials with historical controls, however, suggest that decompressive craniectomy may be a useful option for intractable ICP. The results of a trial published in 2011 did not show improved functional outcome after decompressive hemicraniectomy (Cooper et al., 2011). However, the results of this trial need to be interpreted with caution because of concerning design flaws. Nonetheless, decompressive craniectomy is not without risk. One study showed a 35% rate of complications that included subdural effusion, hydrocephalus, and infection. The incidence of posttraumatic hydrocephalus is higher in patients after decompressive craniectomy. This maybe due to disturbance of CSF dynamics after decompressive craniectomy. Bilateral decompressive craniectomy is sometimes performed but is associated with an unfavorable outcome in 46% of cases (Bao et al., 2010). Some patients suffer delayed neurological deterioration, described as “syndrome of the trephined.” It is thought that that brain function is impaired by the atmospheric pressure.

Hypothermia is an alternative therapy that should be reserved for patients with persistent intracranial hypertension refractory to other medical or surgical interventions. A recent meta-analysis of all prospective TBI clinical trials found that therapeutic moderate hypothermia (32°C–34°C [89.6°C–93.2°F]) resulted in less mortality (RR 0.51; 95% CI, 0.33, 0.79) and increased likelihood of a good outcome (RR 1.91; 95% CI, 1.28, 2.85) compared with normothermia during acute care, but the risk of pneumonia increased (RR 2.37; 95% CI, 1.37, 4.10) (Peterson et al., 2008). Problems associated with hypothermia include increased bleeding risk, arrhythmias, and increased susceptibility to infection and sepsis.

The Guidelines for the Management of Severe Traumatic Brain Injury recommend using high-dose barbiturates to control elevated ICP only after maximum standard medical and surgical treatment failed. High-dose barbiturate therapy has been shown to be effective in lowering ICP but never to improve outcome. High-dose barbiturate therapy, such as pentobarbital, has the potential benefit of suppressing cerebral metabolism, thus decreasing oxygen demand. Barbiturates have the added benefit of neuroprotection through mechanisms such as inhibition of free-radical lipid peroxidation and neuronal membrane disruption. However, clearance of barbiturates from the body varies among patients, and titration of therapy based on serum levels is difficult. Instead, continuous electroencephalographic monitoring can be used to titrate barbiturate infusions to achieve burst suppression. Barbiturate therapy also has major disadvantages. It is associated with serious cardiac side effects, potentially leading to both myocardial depression and hypotension and thus offsetting any ICP-lowering effect it may have. The ability to perform a valid neurological examination is also lost when barbiturates are used to control ICP. Additionally, barbiturate therapy may result in immunosuppression, leading to sepsis.

Hemodynamic management can be quite challenging after a severe TBI. Hypo-osmolar solutions should be avoided because they lead to an increase in ICP, and hyperosmolar treatments of elevated ICP such as mannitol can actually lower CPP. Normal saline is commonly used to achieve euvolemia. Hypertonic saline bolus with a 3% sodium solution is optimal for patients who are volume depleted, need to be resuscitated, and have elevated ICP. The advantage of hypertonic saline is that it stays intravascular longer than normal saline and has an osmolar mechanism of action similar to that of mannitol, without the diuretic effects. Other theoretical benefits of hypertonic saline include improvements in vasoregulation, cardiac output, immune modulation, and plasma volume expansion.

Multimodality monitoring is more frequently used now, at least in centers specializing in TBI. Besides ICP, other factors monitored in patients with severe TBI include partial brain-tissue oxygen tension (PbtO₂), CBF, microdialysis variables, bispectral index, and cerebral oximetry. ICP measurements alone are only surrogate measures of CBF, brain oxygenation, and cerebral metabolism. A combination of ICP, CBF, and PbtO₂ can lead to a better understanding of brain pathophysiology after TBI.

The ability to limit secondary brain injury after a severe TBI relies on ensuring the delivery of an adequate supply of oxygen and metabolic substrate to the brain and to detect it as early as possible. Once the ICP rises, it may be too late to intervene, so measuring PbtO₂ directly should detect compromise in oxygen delivery earlier than relying on ICP alone. Several studies have investigated the relationship between outcome and focal PbtO₂, which can be measured with the Licox monitor (Stiefel et al., 2004; Narotam et al., 2006). They found that likelihood of death increased with duration of time PbtO₂ was less than 15 mm Hg, prompting the question, Does oxygen-directed therapy improve outcome?, which has not been shown so far. The effect of hyperoxemia after TBI are also unclear. While mild hyperoxemia is beneficial in TBI patients, a study of 3420 patients found worse outcomes with extreme hyperoxemia (O₂ 487 mm Hg). The detrimental effects of hyperoxemia have been demonstrated in patients with ischemic brain injury not due to TBI. Whether the same pathophysiological process of post-reperfusion hyperoxemia is harmful in TBI remains to be determined (Davis et al., 2009).

Transcranial cerebral oximetry is a noninvasive technique for monitoring changes in cerebral oxygenation. The method relies on measuring absorption of light at multiple wavelengths with near-infrared spectroscopy (NIRS), and this allows the transmission of photons through skin, bone, brain, and CSF. The values obtained with NIRS primarily represent the oxygenation status of the chromophores of the venous compartment (75%) or the cerebral vascular bed (arterial 20%, capillary 5%). Changes in transcranial cerebral oximetry result from changes in the balance between the cerebral oxygen supply and oxygen consumption. The NIRS light emitter and a receiver are usually placed at the lateral forehead. Scalp edema and forehead abrasion can interfere with cerebral oximetry.

Knowing about PbtO₂ and CBF can help the treating physician to prevent secondary injury more effectively. Along with PbtO₂ measurements, regional CBF can be measured with thermal diffusion probes. The latest edition of the TBI Guidelines has commented on the usefulness of these catheters for the first time.

High ICP can be secondary to malignant brain edema due to decreased cerebral perfusion or to hyperemia; treatment of high ICP in these cases would be very different. Treatment of high ICP due to malignant brain edema would entail hyperosmolar therapy to decrease the edema and increase CBF to maintain brain perfusion. Treatment of high ICP due to hyperemia would entail lowering CCP with aggressive hyperventilation.

The metabolic state of the brain can also be assessed with cerebral microdialysis, which monitors energy metabolites over extended periods but only in a small focal volume of the brain. The threshold for abnormal metabolism is evidenced by a lactate/pyruvate ratio of 40. Microdialysis in patients with severe TBI showed that tight systemic glucose control is associated with reduced cerebral extracellular glucose availability and increased prevalence of brain energy crisis (microdialysis glucose <0.7 mmol/L, with a lactate/pyruvate ratio > 40), which in turn correlates with increased mortality, so intensive insulin therapy may impair cerebral glucose metabolism after severe brain injury (Oddo et al., 2008). The combined use of microdialysis variables and PbtO₂ in addition to ICP, mean arterial pressure (MAP), and CPP was found to have the best predictive accuracy for outcome after severe TBI (Low, 2009).

One conundrum that remains unsolved, even with multimodality monitoring, is the value of continuous information versus global information about the condition of the brain. While PbtO₂ and CBF probes provide continuous information about these variables, they reflect only a very small portion of the brain. Positron emission tomography (PET) and Xenon-enhanced CT can provide global information about the metabolic status of the brain and blood flow, but only at a given time.

Most patients with a severe TBI have associated comorbidities that also can lead to secondary brain injury. Hypoxia from adult respiratory distress syndrome or ventilator-associated pneumonia, and hyperthermia from infection or sepsis are the most common. Comatose patients are at high risk for deep vein thrombosis and pulmonary embolism. Sequential compression stockings should be used for the first 48 hours after TBI. After the first 48 hours, subcutaneous heparin or low-molecular-weight heparin can be used safely.

Hyperthermia following a severe TBI is common, potentiates secondary injury, and worsens neurological outcome. Induced normothermia (36.5°C–37.5°C) via an intravascular cooling catheter is effective prophylaxis to reduce fever burden (Puccio et al., 2009).

No evidence yet exists to guide the decision for blood transfusion (whether and when) after a severe TBI. Although intuitively it makes sense that increasing a patient’s hematocrit will lead to increased oxygen delivery to the brain, this benefit has not been clearly shown. Over recent years, evidence has emerged that blood transfusion may actually have a negative effect after TBI, even after controlling for anemia as a possible confounder (Carlson et al., 2006). Stored blood has decreased deformability and may in fact exacerbate ischemia.

Management of Focal Injuries

Patients with focal TBI and axial or extraaxial bleeding with mass effect are treated surgically. Sometimes patients with mass lesion without signs of mass effect and controlled ICP may be managed nonoperatively with intracranial monitoring and serial imaging. However, the development of mass effect may result in secondary brain injury, posing the risk of further neurological deterioration, herniation, and death. This condition necessitates emergent neurosurgical intervention. Traumatic SAH rarely causes any mass effect or warrants neurosurgical intervention. Patients with severe TBI and only traumatic SAH on CT scan are managed nonoperatively with intracranial monitoring and serial imaging.

Various surgical techniques have been advocated to evacuate intracranial hemorrhages. The most commonly used is craniotomy, with or without subtemporal decompressive craniectomy. Sometimes a large decompressive hemicraniectomy is performed at the same time if the brain appears swollen intraoperatively. Preoperative CT can help predict malignant brain swelling. More midline shift than one would expect from the extraaxial hematoma, effacement of basal cisterns and sulci, and loss of gray/white matter differentiation are warning signs.

Between 22% and 50% of patients with EDH are comatose on admission or immediately before surgery. Additionally, 47% of patients with an EDH have a lucid interval—that is, they awaken after initial unconsciousness and then deteriorate neurologically, which may obscure the development of EDH. A patient with an EDH can go from normal neurological status to coma within minutes. Most often, the patient becomes agitated and restless and may have an episode of emesis because of the increased ICP. This is generally followed by focal neurological signs such as hemiparesis, seizures, pupillary dilation, decrease in consciousness, and in the worst-case scenario, decortication. Because of the high association between skull fractures and EDH, screening x-rays of the skull can help identify the risk for EDH. Guidelines for surgical management of acute EDH list the following criteria for surgical evacuation of an EDH (Bullock et al., 2006b):

Regarding timing, emergent surgical evacuation is strongly recommended for comatose (GCS score of ≤8) patients with an acute EDH. Emergent craniotomy for evacuation also is indicated if an EDH thicker than 1 cm develops or leads to neurological symptoms. The mortality for patients in all age groups and GCS scores who require surgery for evacuation of EDH is approximately 10%, compared to 5% in the pediatric population.

The 2006 guidelines for surgical management of SDH give the following criteria for cases of SDH (Bullock et al., 2006c):

The GCS score, pupillary examination, comorbidities, CT findings, age, and ICP, influence the decision to evacuate an acute SDH. Neurological deterioration over time is common and also can influence the decision to operate. Although the time from neurological deterioration to surgery influences outcome, optimal timing of surgery is difficult to pinpoint, because most patients who undergo early surgery for SDH also have a more severe injury and worse outcome than patients undergoing delayed evacuation. The magnitude of preoperative midline shift also seems to correlate with unfavorable outcome (Nelson et al., 2010). Age is an independent predictor of outcome in patients with SDH; age older than 50 years is statistically correlated with poorer outcome. A study showed that at 3 months after SDH evacuation, patients between the ages of 18 and 30 years had 25% mortality, whereas those 50 years and older had 75% mortality (Hukkelhoven et al., 2003).

Isolated linear fractures are usually of little clinical significance unless they are near or traverse a suture or involve a venous sinus groove or vascular channel. In those cases, complications may include suture diastasis, venous sinus thrombosis, and EDH. Although rare, a growing linear skull fracture can develop in young children, especially if the parietal bone is fractured. CSF pulsation and pressure then prevent the bone from healing and cause enlargement of the fracture. A tear in the dura under the fracture is usually the underlying cause. Surgical obliteration of the CSF leak is needed so that the skull fracture can heal.

Patients with simple (closed) depressed cranial fractures may be treated nonoperatively if there is no clinical or radiographic evidence of dural penetration (pneumocephalus), significant intracranial hematoma, depression greater than 1 cm, frontal sinus involvement, cosmetic deformity, or gross wound contamination. Compound (open) depressed skull fractures are associated with infections and development of late epilepsy; they should undergo surgical débridement and elevation to decrease the possibility of infection. Any neurological deficit that can be associated with a skull fracture—open or closed—is an indication for surgery. The guidelines for surgical management of depressed skull fractures recommend routinely administering antibiotics.