Chapter 5 Head Trauma

Traumatic brain injury (TBI) has an incidence of more than 500,000 cases per year, and is the leading cause of disability and death in children and young adults in the United States. The annual cost of TBI, including direct costs and lost income, is estimated to be above $25 billion. It has a peak incidence in 15- to 24-year-olds, with males being injured two to three times more frequently than females. The simplest form of TBI is concussion, which may be defined as a transient impairment of neurologic function associated with or without loss of consciousness occurring at the time of injury.

Imaging of head trauma has a primary role both in diagnosing the extent of the traumatic injury and in expediently determining the appropriate therapy. The most efficient method of triage for acute trauma remains computed tomography (CT). It is fast and usually very accurate at detecting acute hemorrhage (high density on unenhanced scan) (Fig. 5-1). CT is excellent for assessing facial and skull fractures. Neurosurgeons are interested in knowing the precise source of the patient’s clinical problems with respect to the trauma. Their most overriding concern is whether there is a treatable lesion. Such lesions could be an epidural hematoma, a large subdural hematoma, actively bleeding parenchymal hemorrhage, brain herniation, or a significantly depressed skull fracture. CT does have some pitfalls of which the radiologist should be aware, particularly during the 4-hour board examination. Not all hemorrhage is high density. Isodense to low density hemorrhages are seen in patients who are severely anemic or in those with disseminated intravascular coagulopathy (DIC) or in the hyperacute actively bleeding stage (Fig. 5-2). Fluid-fluid levels and sedimentation imply active bleeding, which may be prolonged in patients on anticoagulants. CT does not easily detect extracerebral intracranial hemorrhages of the infratemporal region, subfrontal region, or posterior fossa. It also is less sensitive than magnetic resonance (MR) imaging in detecting diffuse axonal injury and vascular injury.

Figure 5-1 Computed tomography of acute traumatic intraparenchymal hematoma. Multiple large intraparenchymal hematomas are identified as high-density mass lesions associated with edema in the frontal region. Subfalcine herniation is present (arrows). Also, note hemorrhagic cavity with blood level, which can be observed in traumatic intraparenchymal hemorrhage (open arrows).

Figure 5-2 Computed tomography of isodense subdural hematoma. Note that the density of the subdural hematoma (arrowheads) is similar to that of cortex.

When trauma is evaluated with CT, review the images with brain, subdural, and bone settings at the workstation or demand films with those windows. The wide window setting aids in separating high-density blood from the high density of bone and is particularly useful in acute subdural and epidural hematomas, which can be thin and difficult to differentiate from the calvarium. Bone windowing is essential in the search for fractures, as is the combination of coronal and axial images with very thin sections, particularly of the skull base, temporal bone, orbit, and face (Fig. 5-3).

Figure 5-3 Comminuted fracture of right temporal bone (white arrow).

Patients with TBI have been classified into mild, moderate, and severe grades by the Head Injury Interdisciplinary Special Interest Group of the American Congress of Rehabilitation Medicine based on (1) a period of loss of consciousness, (2) loss of memory for events immediately before or after the trauma, (3) change in mental status, and (4) focal neurologic deficits. The Glasgow Coma Scale (GCS) was devised to provide a uniform approach to the clinical assessment of patients with acute head trauma and is the most frequently used means of grading the severity of injury within 48 hours (Table 5-1).

Table 5-1 Glascow Coma Score Grading

One issue with the GCS is the marked heterogeneity of location and type of brain lesions in moderate to severe TBI patients; thus, individuals with the same GCS could have markedly different outcomes depending on the nature of the causative lesion (subdural versus epidural hematoma versus intraparenchymal hemorrhage versus diffuse axonal injury [DAI]).

Approximately two thirds of patients with head trauma in the United States are classified as having minor head injury, and less than 10% of these cases have positive results on CT and less than 1% require neurosurgery. The following findings in patients with head trauma were associated with a positive CT: headache, vomiting, age over 60 years, drug or alcohol intoxication, deficits in short-term memory, physical evidence of trauma above the clavicles, coagulopathy, and seizure. Of course, the patient would not be seen in the emergency department (ED) after minor head injury if he or she were normal.

Open head injury involves the intracranial contents communicating through the skull and scalp. In closed head injury, there is no communication between the intracranial contents and the extracranial environment. An organized approach to the categorization of head trauma is essential in examining patients undergoing diagnostic imaging. Accurate classification enables prediction of prognosis and institution of appropriate therapeutic management.

MRI is more sensitive than CT in detecting all the different stages of hemorrhage (i.e., acute, subacute, and chronic—although, within the first 4 hours, for hyperacute hemorrhage, CT may be more sensitive than low field MR). However, detecting acute subarachnoid hemorrhage is difficult if fluid-attenuated inversion recovery (FLAIR) imaging is not performed with the MR examination (Fig. 5-4). The direct multiplanar capability and absence of beam hardening artifact of MR enable excellent visualization of the inferior frontal and temporal lobes and the posterior fossa (Fig. 5-5). MR is more sensitive than CT in demonstrating DAI (see discussion of DAI under Primary Injury in this chapter), isodense subdural hematomas, minor deposits of hemorrhage with susceptibility-weighted scanning (particularly at higher field), and coexistent ischemia. CT/CT angiography (CTA) and MR/MR angiography (MRA) are both excellent means for detecting vascular dissection, but give the nod to CT for in–emergency department (ED) convenience.

Figure 5-4 Subarachnoid hemorrhage on FLAIR. A, FLAIR scan shows intraventricular blood as well as high signal in the sulci of the posterior sylvian fissure and inferior frontal sulci bilaterally. B, Here is the computed tomography scan. Blood there.

Figure 5-5 Sagittal T1-weighted image demonstrating inferior frontal and inferior temporal lobe subdural hematoma extending along the tentorium posteriorly (white arrows).

Skull films have little role in the triage or care of the patient with significant head trauma. They are useful only to ascertain whether linear skull fractures are present. CT is excellent at demonstrating depressed or comminuted skull fractures; however, fractures parallel to the plane of the scan slice, such as those involving the vertex of the skull, may be missed. Always assess the scout view from the CT for evidence of fracture or high cervical injury.

At present, angiography is indicated to define the anatomy of traumatic fistulas, particularly before therapeutic endovascular obliteration, and, in some cases, for diagnosis of vascular dissection and pseudoaneurysm. Another indication for angiography is the source diagnosing and treating of an expanding hematoma in the neck.

CLASSIFICATION OF INJURY

Traumatic brain damage may be separated into two major categories: (1) primary injury, and its associated primary complications, directly related to immediate impact damage; and (2) secondary complications (not present at the time of injury) resulting from the primary injury over time. Why all the fuss? Well, you obviously cannot prevent the primary injury. Importantly, secondary lesions are preventable or treatable.

Primary Injury

Primary injury is the consequence of brain damage occurring at impact. It may result in contusions of the brain and DAI. Contusions may be defined as bruises that are associated with petechial hemorrhages and are the most common primary intra-axial lesion. DAI is produced by stress-induced movements of the head that severely disrupt axons either completely or incompletely. Primary complications may progress over time. These include mass lesions, such as extraparenchymal or intraparenchymal hemorrhage, and brain swelling. Primary injury can be divided further into focal or diffuse injury. Focal injuries represent contusions and other mass lesions, such as epidural or subdural hematomas, which may generate shift of the midline brain structures and herniation with compression of and damage to the brain stem. Diffuse injury, in its fulminant form, shows widespread disruption of axons throughout the brain, resulting in permanent coma (persistent vegetative state) or death.

Two different mechanisms are principally responsible for primary injury: (1) direct contact between the skull and an object (contact phenomena), and (2) inertial injury resulting from the differential accelerations between white and gray matter. The former is responsible for scalp laceration, skull fractures, intracerebral and extracerebral hematoma, and contusion. Contact phenomena generate stress waves, which create skull base fractures, contusions, and hemorrhages remote from the site of contact. Severe contact injury is associated with subdural hematoma and massive intraparenchymal hemorrhage (“burst” lobe). Inertial injury results in DAI and pure subdural hematoma after shearing of bridging veins.

Secondary Complications

Secondary complications are temporally removed from the original trauma and result in brain injury, including (1) raised intracranial pressure (ICP), leading to increased cerebral edema and hydrocephalus; (2) hypoxia; (3) infection; and (4) infarction secondary to or in addition to (5) brain herniation. The gross release of excitatory neurochemicals (glutamate, aspartate) can cause cellular swelling, unsolicited calcium and magnesium influx into the cells, generation of free radicals, cytokine release, activation of caspases (proteases that cleave proteins that induce apoptosis [transcription-dependent method of cell death]), mitochondrial dysfunction, breakdown of the cytoskeleton, and additional neuronal death. Many of these specific steps in the cascade have been targeted with pharmacologic treatments.

Recent information suggests that in the acute period (hours to days) cytotoxic swelling detected by diffusion-weighted imaging (DWI) is the most important factor responsible for increased intracranial volume, leading to raised ICP.

Factors noted on imaging may predict TBI outcome: (1) levels of reduction in N-acetyl aspartate (NAA); (2) compression or obliteration of the mesencephalic cistern (risk of elevated ICP is increased threefold if the cistern is obliterated); (3) subarachnoid hemorrhage (increases the possibility of death twofold); (4) acute subdural hematoma and DAI (associated with two thirds of deaths in head injury); (5) positive DWI (implies worse outcome and more severe damage); and (6) perfusion deficits indicating coexistent ischemia. Still, 10% to 42% of patients with severe head injury have normal-appearing CT scans; elevated ICP still develops in 10% to 15% of these patients.

EPIDURAL HEMATOMA

The potential space between the inner table of the skull and the dura is the epidural space. The most common cause of epidural hematoma is head trauma with skull fracture of the temporal bone (90% of cases) crossing the vascular territory of the middle meningeal artery or vein. In children, the greater elasticity of the skull permits meningeal vascular injury without coexistent fracture. Tears of the middle meningeal artery (60% to 90%) or venous structures (middle meningeal vein, venous sinus, or diploic veins; 10% to 40%) result in the extravasation of blood and acute epidural hematoma. After injury there may be a lucid interval (in 50% of patients) as the lesion expands before causing midline shift and deterioration (as opposed to DAI, where coma occurs immediately after the injury). Slower bleeding from the meningeal vein or dural sinus temporally delays symptoms. Chronic epidural hematoma has also been observed. Epidural hematomas are usually biconvex and are the result of the firm adherence of the dura to the inner table and its attachment to the sutures (Fig. 5-6). Most frequently, these hematomas are observed in the temporal parietal region. When fractures run through the middle meningeal artery and vein, occasionally a fistula may develop between the two. Pseudoaneurysm of the meningeal artery may also occur. Epidural hematomas of the posterior fossa result from tearing of a venous sinus and may be continuous with the supratentorial and infratentorial space. Venous epidural hematoma is most often noted in the pediatric population and carries a lower incidence of skull fracture.

Figure 5-6 Computed tomography of acute epidural hematoma. Note the convex shape of the epidural hemorrhage.

Computed tomography reveals a high density extra-axial mass acutely (Fig. 5-7). There may be some low density in the acute hemorrhagic mass, probably representing serum extruded from the clot. In these cases, perform bone window settings to examine the scout image for fractures and upper cervical spine subluxations not visible on axial sections. Chronic epidural hematomas reveal low density and peripheral enhancement, and they may be concave. Chronic epidural hematomas must be distinguished from other epidural masses such as infection, inflammation, and tumor, and from subdural lesions such as empyemas.

Figure 5-7 Epidural hematoma. A, Computed tomography scan after head trauma shows a patient with an actively bleeding epidural hematoma with the anterior border limited by the coronal suture (arrow). Disregard the coexistent arachnoid cyst (A) further anteriorly. B, The epidural hematoma crosses the midline at the vertex and was due to a sagittal sinus tear after trauma.

On MR, the extra-axial hemorrhage has different appearances depending on the interval between the traumatic event and imaging. If imaged less than 12 hours after trauma, acute hemorrhage demonstrates low intensity on T2-weighted images (T2WI) and isointensity on T1-weighted images (T1WI), whereas if the hemorrhage is imaged a few days after the incident (subacute hematoma), it is high intensity on T1WI (see Fig. 5-5). Intensity is not usually an issue because the biconvex morphology of the extra-axial mass in the setting of acute trauma makes the diagnosis obvious. One may visualize the medial dural margin as a hypointense rim on all pulse sequences. Linear fractures are poorly visualized on MR, but cortical veins and arteries are displaced inward as in any extra-axial lesion.

SUBDURAL HEMATOMA

Hemorrhage into the potential space between the pia-arachnoid and the dura is termed a subdural hematoma. Acute subdural hematomas result from significant head injury and are caused by the shearing of bridging veins (Fig. 5-8). This occurs because of rotational movement of the brain with respect to fixation of these veins at the adjacent venous sinus or dura. The subdural portion of the vein is not ensheathed with arachnoid trabeculae as is the subarachnoid portion and therefore is the weakest segment of the vein. Acute subdural hematomas also occur in the setting of the burst lobe (Fig. 5-9). Intracerebral clot is in direct continuity with the subdural hematoma. Penetrating injury can also result in acute subdural hematoma.

Figure 5-8 Computed tomography demonstrates bilateral acute subdural hematomas (open arrows). Lesions are concave medially.

Figure 5-9 Burst temporal lobe with associated acute subdural hematoma. Coronal proton density-weighted image demonstrates evidence of acute subdural hematoma (white arrow) associated with the burst left temporal lobe (T). A subacute subdural hematoma (arrowheads) is identified over the right convexity. There is also evidence of contusion of the right temporal lobe.

Subdural hematomas are seen in approximately 30% of patients with severe closed head trauma. They may be classified as simple (without associated brain parenchymal injury) and complicated (with parenchymal injury). In the latter the subdural component of the lesion is relatively insignificant compared with the parenchymal injury. Because of this, acute subdural hematoma is associated with a 35% to 50% mortality rate, and, of those who survive, most have functional limitations. Simple acute subdural hematomas have a better prognosis than complicated ones. The worst prognosis in acute subdural hematoma occurs in head trauma secondary to motorcycle injury or falls, in postoperative patients with ICP greater than 45 mm Hg, in the presence of severe mass effect, in rapid accumulation of the subdural hematoma, in delay (>4 hours) in evacuation of significant lesions, and in patients older than age 65 years (impaired regenerative capacity). Control of ICP (secondary injury) is very important.

Subdural hematomas in the elderly, at discovery, are larger because of the generalized loss of brain parenchyma allowing greater unimpeded growth before mass effect and symptoms. Subdural and epidural hematomas are contrasted in Table 5-2 (also compare Fig. 5-6 with Fig. 5-8).

Table 5-2 Subdural versus Epidural

The temporal nomenclature of the subdural hematoma is arbitrary. Lesions occurring at the time of the initial injury are considered acute, although symptoms may take up to a few days to become manifest. Those lesions becoming symptomatic between approximately 3 days to 3 weeks are subacute (Fig. 5-10), whereas those lesions that are diagnosed after 3 weeks are considered chronic. Subdural hematomas generally arise over the convexities but can also occur in the posterior fossa, middle cranial fossa, or along the tentorium (Fig. 5-11).

Figure 5-10 Right subacute subdural hematoma on T1-weighted image. Observe the extension to the floor of the right middle cranial fossa (arrows).

Figure 5-11 Infratemporal subdural hematoma. A, This small subacute hematoma along the base of the temporal lobe would be invisible on computed tomography because of beam hardening artifact, partial volume averaging, and reduced contrast sensitivity. Large right-frontal acute epidural hematoma is also present. B, Different patient with a similar finding. The sagittal T1-weighted image allows detection of the small subdural hematoma along the floor of the middle cranial fossa (arrows) and the extension over the tentorium and along the occipital lobe. The subdural hematoma turns the corner and extends up to the parietal region.

Imaging Characteristics of Subdural Hematomas

Acute subdural hematomas on CT are high-density crescentic extracerebral intracranial masses. The thickness of the hematoma can range from pencil-thin to large and can at times be convex inwardly. Rarely, acute subdural hematomas have been reported to be isodense or low density on CT. This may be correlated with significant anemia (hemoglobin level 8 to 10 g/dL), DIC, ongoing bleeding, or tears in the arachnoid membrane, leading to dilution of the red blood cells with cerebrospinal fluid (CSF).

Subacute subdural hematomas, on the other hand, are usually isodense to low density; however, mass effect is common in significant subacute hematomas. The lesion usually is concave but occasionally may be convex when chronic, with inward displacement of the convexity veins, which course along the inner surface of the arachnoid. Although this shape may simulate an epidural hematoma, subdural hematomas are not confined by the cranial sutures and the chronic collections are low density. Intermediate window settings on CT are useful in separating thin subdural hematomas from the bony calvarium. This is obviously not an issue with MR. Subacute and chronic subdural lesions enhance (but who gives contrast in this setting?). This is caused by the vascularization of the subdural membranes (usually being thick outside and thin inside), formed between 1 to 3 weeks after injury, which enhance. These vessels are not associated with tight junctions. They leak contrast and may easily be torn, resulting in various blood products and growth of the subdural hematoma. Repeated episodic bleeding results in fibrous septations and compartments within the hematoma. Subacute and chronic subdural hematomas may also display layering with dependent cells and cellular debris and an acellular supernatant (Fig. 5-12). The dependent portion is of higher density than the supernatant. Hematomas may also occur between the leaves of the tentorium. Tentorial subdural hematomas fade out laterally and stop abruptly at the medial margin of the tentorium (Fig. 5-13).

Figure 5-12 A, T2-weighted image (T2WI) of layered left subdural hematoma. High intensity and lower intensity gradients inferiorly can be identified in this layered subdural hematoma. (You are correct; there is also a subdural hematoma on the right.) B, This poor chap shows a blood-fluid layer over the right frontal convexities on the emergency department computed tomography scan. The clinicians rushed him to the magnetic resonance scanner, where axial T2WI (C), FLAIR (D), and T1WI (E) show the layering of the serum anteriorly (dark on T1WI, bright on T2WI) and the deoxyhemoglobin posteriorly (isointense on T1WI, dark on T2WI). F, The coronal fat-suppressed T1WI also showed an interhemispheric subacute hematoma (bright on T1WI).

Figure 5-13 Computed tomography of tentorial subdural hematoma. Observe that the high density on the left (inferior) image (A) is lateral to the medial high density on the right (superior) image (B). That is why it is called a TENTorium. The hemorrhage is contained within the leaves of the tentorium.

Chronic subdural hematomas are usually low density (Fig. 5-14). High and low density levels observed in these lesions may be caused by rebleeding into the chronic subdural collection. There may be no history or an insignificant history of head trauma. Other contributing conditions include coagulopathy, alcoholism, increased age, epilepsy, and surgery for ventricular shunts. More than 75% of chronic subdural hematomas occur in patients over age 50 years. Chronic subdural hematomas occur in infants as a result of birth injury, vitamin K deficiency, coagulopathy, or child abuse. In fact, the presence of multiple-aged subdural hematomas should raise the spectre of nonaccidental trauma (see discussion of child abuse in this chapter). Rarely, chronic subdural hematomas may ossify or contain fat.