Subdural Hematoma

Classically, a subdural hematoma (SDH) (Figure 1) has been said to develop after tearing of a bridging vein, that is, a vein passing directly from the cortex to the overlying dura. The mechanical forces of the trauma can cause tearing of these veins. More recent evidence indicates that at least some of these hematomas actually form from splitting of inner and outer layers of the dura, that is, they may actually be “intradural hematomas.” Finally, some SDHs are caused by direct bleeding into the subdural space from parenchymal contusions or hematomas or from injured cortical arteries or veins.

Figure 1 Large acute subdural hematoma (SDH) with midline shift. Subarachnoid hemorrhage is evident contralateral to the SDH.

Epidural Hematoma

Epidural hematomas (EDHs) classically arise after a blow to the side of the head results in a fracture of the thin temporal bone immediately overlying the middle meningeal artery. The patient may briefly lose consciousness after the initial impact, but he or she quickly awakens; thus, the brain injury was mild. Unfortunately, the fracturing of the skull lacerated the middle meningeal artery. Continued bleeding from this source produces an enlarging EDH, the presence of which may be signaled by such symptoms as severe and worsening headache, vomiting, and decreasing level of consciousness. The period between awakening from the initial concussion and subsequent lapsing into a coma has historically been described as a “lucid interval.” Importantly, loss of consciousness does not always occur after the skull is fractured, and many patients with large EDHs are awake until they begin to lapse into a terminal coma. It must also be mentioned that many EDHs are not associated with meningeal arterial bleeding. In these cases, perhaps the source of the hematoma is oozing from the overlying edges of fractured bone.

Referring again to the classic scenarios, patients with EDHs are said to fare better than patients with similarly sized SDHs. Why should this be so? The answer is that a “pure” EDH essentially represents a skull fracture with no direct parenchymal injury to the brain. On the other hand, the rotational forces that are said to be an important cause of SDHs via tearing of bridging veins may also cause widespread axonal injury, as discussed later. Thus, SDH is said to be associated with a greater burden of parenchymal injury, which explains the worse outcomes. Of course, this explanation refers only to the extreme ends of the spectrum of the pathophysiology of EDHs and SDHs. Many patients with EDHs will do poorly, while SDH patients often recover well from their injuries. Nevertheless, this explanation is a useful way to conceptualize the interactions between mass lesions and diffuse injury.

Subarachnoid Hemorrhage

The most common post-traumatic intracerebral hemorrhage is not a mass lesion, but rather diffuse subarachnoid hemorrhage (SAH) (see Figure 1). Several retrospective series report that SAH after TBI is independently associated with worse outcomes, but the mechanism that might explain such an association is unclear. In the acute setting, SAH does not seem to have much effect on patient management, which is driven instead by more immediately pressing concerns.

Parenchymal Lesions

Contusions occur commonly after TBI, especially at the base of the frontal lobes and at the anterior edges of the temporal lobes. The brain in these regions is said to continue moving over or into the skull base after the head suddenly stops moving after a violent blow or rapid rotational movement. Fortunately, most such contusions remain small and surgically insignificant. Emergency surgery may be required, however, for larger lesions or for smaller ones that subsequently enlarge.

Unlike contusions, in which extravasated blood mixes freely with brain tissue, parenchymal hematomas consist of solid blood clots within the brain itself. They occur less commonly than contusions. They tend to be more variable in their distribution.

Ischemia

Diffuse injury may be of several types. Ischemia is a common form of diffuse injury. In some cases, mass effect from a traumatic hematoma may cause elevated intracranial pressure (ICP) and local compression of underlying tissue that can lead, respectively, to global and local reduction of cerebral blood flow (CBF). However, post-traumatic cerebral ischemia may also occur when no mass lesion is present, especially very early after injury. Although the CT scans in these cases may be relatively unimpressive, these patients may be quite vulnerable to the effects of such secondary insults as hypotension or hypoxia. Over the subsequent hours and days, CBF usually increases, but the damage from early ischemia may have already been done long before the increase in CBF (Figure 2). Some centers have used xenon CT or perfusion CT to measure CBF and identify ischemia immediately after injury.

Figure 2 Infarction of an entire cerebral hemisphere is evident on the left side of this computed tomography scan.

Diffuse Axonal Injury

Another common type of diffuse injury is diffuse axonal injury (DAI). Although the axonal disconnection that characterizes DAI is commonly thought to occur at the time of injury, such immediate loss of axonal continuity probably occurs only when an injury produces severe cerebral parenchymal disruption. Instead, it appears more likely that the rotational and mechanical forces that are operant during the traumatic event produce a focal impairment of axoplasmic flow, which, in turn, culminates in axonal disconnection several hours after injury.² This slight delay creates hope that a therapeutic window may exist for the administration of a yet-to-be-developed treatment that would prevent loss of axonal integrity and function.

Of course, axonal pathology cannot be visualized on a CT scan. However, DAI frequently occurs in conjunction with small scattered parenchymal hemorrhages commonly described as “shear injuries,” probably because they have been postulated to occur as tissues at different depths of the brain undergo rotation at different velocities. The interface between these regions is said to undergo shearing, resulting in the small punctuate hemorrhages. Regardless of whether this explanation is true, the presence of scattered small hemorrhages may serve as a clue that DAI may be present.

Cellular and Molecular Factors

At the cellular level, the abnormalities caused by TBI are numerous and complex. Release of glutamate and other excitatory neurotransmitters may lead to excessive neuronal depolarization and intracellular calcium influx, with activation of proteases and other processes that lead to cell death. Inadequate blood flow can cause a conversion from aerobic to anaerobic metabolism. The lactic acid that is produced lowers local tissue pH, and the consequent acidosis contributes to tissue injury and death. Trauma-induced apoptosis may promote further cell death. These and other biochemical and cellular processes take place against the backdrop of an individual patient’s genetic makeup; the presence of specific alleles for various genes may make an individual more or less susceptible to the damaging effects of various pathophysiologic processes.

Clinical Examination

Ideally, the severity of TBI is determined and classified according to a patient’s neurologic examination. The size and appearance of a mass lesion as seen on imaging studies are not as important as the effect that the lesion may be having on a patient’s neurologic function and level of alertness.

The single most important question in the evaluation of a potentially head-injured patient is whether he or she obeys simple one-step commands. A simple definition of coma is that a person will not do such things as hold up two fingers or stick out the tongue when asked to do so. Failure to obey commands is widely used as an indicator of the presence of severe TBI. Other simple but important observations are the type of movement exhibited by the patient (localization of noxious stimuli, withdrawal, flexion, extension, etc.); whether the right and left sides are symmetrical; the type of speech; the presence or absence of eye opening; and pupillary size, reactivity to light, and bilateral symmetry.

Because the nerve fibers that mediate pupillary constriction lie on the surface of the third cranial nerve, compression of this nerve by herniating brain tissue that is being displaced by a large mass lesion may cause inactivation of the pupilloconstricting fibers. The resulting pupil appears large and unable to constrict in response to bright light. This physical finding in a comatose TBI patient suggests that an immediate CT scan is needed to identify a large acute hematoma. However, fixed and dilated pupils may also be caused by brainstem ischemia or by direct ocular trauma.

Many scales have been developed for the assessment of consciousness or neurologic status after injury, but by far the most widely used is the Glasgow Coma Scale (GCS)³ (Table 1). In conjunction with such information as the status of pupillary reactivity and the tempo or rate of change of a patient’s neurologic condition, the GCS is an extremely useful tool for assessing a patient’s baseline condition and subsequent progress.

Airway

In terms of control of the airway, prehospital endotracheal intubation should be considered in patients with severe TBI or with other problems that may impair movement of air or protection of the airway from aspiration. Recent retrospective reports in the trauma literature associate worse outcomes with prehospital intubation of severe TBI patients. One possible explanation of these findings may be the difficulty of performing successful endotracheal intubation in the prehospital setting, especially if prehospital providers do so only infrequently.

Breathing

In terms of breathing, standard recommendations advocate the lowest FiO₂ capable of maintaining adequate oxygenation. Although the minimum acceptable PaO₂ based on the oxygenhemoglobin dissociation curve is 60 mm Hg, most practitioners target a minimum of 80-100 mm Hg in TBI patients in order to create a bit of a cushion.

Hyperventilation is no longer recommended as a prophylactic measure to prevent intracranial hypertension.⁵ Hyperventilation is known to cause constriction of the cerebral vasculature, and the resultant decrease in cerebral blood volume can acutely lower ICP. However, the constriction of cerebral arteries may cause CBF to drop to critical levels. Also, within 24 hours of initiation of hyperventilation, the cerebral arteries probably dilate back to their original diameter.⁶ Subsequent attempts to allow the PaCO₂ to increase can cause the arteries to dilate even further, possibly raising ICP.

Most clinicians aim for the low-normal range of PaCO₂, targeting a value of approximately 35 mm Hg. Keeping the PaCO₂ toward the lower end of the normal range may optimize the ability of the cerebral vasculature to autoregulate.

Hyperventilation should be reserved for acute deterioration accompanied by signs of a mass lesion, such as raised ICP with asymmetrical pupils or asymmetrical motor exam. In such cases, the assumption is that the patient will need emergency surgery to treat the lesion. Hyperventilation and other measures, like administration of mannitol, are intended only to buy a few minutes to obtain an emergency CT scan prior to going to surgery. If the scan reveals that no surgical lesion is present, attempts should be made to manage the elevated ICP without hyperventilation by using some of the steps explained in the next chapter.

Circulation

The “C” in ABCs stands for circulation. For brain-injured patients, this can be thought of as management of blood pressure and intravenous fluids.

In years past, common practice was to dehydrate patients “to prevent the brain from swelling.” In the 1990s, the pendulum swung the other way as TBI patients were aggressively managed with intravenous fluids and pressors in order to artificially elevate their blood pressure. However, subsequent studies showed that patients treated with aggressive elevation of blood pressure through fluids and pressors did not have improved neurologic outcomes and, in fact, had a five-fold higher likelihood of developing acute respiratory distress syndrome.⁷

Current management strategies call for maintenance of a normal blood pressure, with aggressive elevation of blood pressure reserved for those patients in whom clinical or physiologic monitoring suggests a need for such therapy, such as to treat cerebral hypoxia or to reverse neurologic deterioration.

Direct monitoring of brain tissue oxygen tension (PbtO₂) is now possible via small intraparenchymal catheters. Although some clinicians treat low PbtO₂ by increasing the FiO₂, we prefer to treat low brain tissue PO₂ by raising blood pressure in order to optimize perfusion of the affected tissue. Ischemic thresholds of brain tissue are difficult to identify with precision. A PbtO₂ below 15-20 mm Hg is generally regarded as low, whereas values below 8-10 mm Hg may suggest that further evaluation and/or intervention might be appropriate.

Computed Tomography Scanning

After initial assessment and stabilization of a TBI patient, the next order of business is the procurement of imaging studies. The most important radiologic study for the evaluation of acute TBI is a CT scan. This imaging modality is excellent for revealing acute hemorrhage, cerebral edema, and mass effect, which are the features of greatest interest during the initial assessment. Bone settings can also detect calvarial and skull base fractures. Another advantage is that CT scanning is a quick procedure that is widely available, at least in the United States. To overcome difficulties with prehospital intubation and the frequent use of sedation in these patients, Marshall et al.⁸ devised a TBI classification scheme based on CT scan findings (Table 2).

Table 2 Marshall CT Classification Scheme

CT, Computed tomography.

Magnetic Resonance Imaging

Magnetic resonance imaging (MRI) has limited application in the acute setting for several reasons. Obtaining a study requires a fair amount of time, access to the patient is limited during the study, and ferromagnetic monitoring and resuscitative equipment must be removed before the patient can enter the magnet. After patients have been stabilized, MRI may reveal the presence of subtle parenchymal injuries that may shed light on the unexplained failure of some patients to improve.

Angiography

Angiography continues to have an ambiguous role in the initial evaluation of the brain-injured patient. Prior to the development of CT scanning in the 1970s, neurosurgeons were limited to pneumoencephalography or angiography for detecting midline shift or other types of mass effect. Skull radiography could also indicate displacement of the calcified pineal gland.

Since the advent of CT scanning, cerebral angiography has usually been reserved for cases in which a high index of suspicion for intracranial vascular injury exists, such as pseudoaneurysm formation or arterial dissection. However, the optimal treatment of any such findings is unclear. Most acutely brain-injured patients should not receive heparin or warfarin, yet interventional radiologic procedures often require that the patient be anticoagulated. Similarly, critically low cerebral blood flow is present in a sizeable percentage of severe TBI patients. The wisdom of occluding a major artery in a patient who may already be ischemic is questionable.

Continuing advances in CT angiography may soon make this technology comparable to conventional angiography for detecting vascular injury. Thus, angiography will probably continue its historical trend of declining use in the acute assessment of TBI patients.

Glasgow Coma Scale

Many different schemes have been proposed for the grading and classification of brain injury. As discussed previously, the best known is the GCS. Introduction of this scale reinforced the need for an accurate neurologic examination as part of the assessment and classification of brain-injured patients. This point is as fundamental today as when the GCS was initially described. Because this scale made possible a more objective assessment of patients, interobserver and intercenter variability could be reduced, thus enabling the creation of multicenter and even multinational studies.

However, accurate determination of the GCS score is often difficult in patients who are intoxicated with alcohol or other drugs or in patients with injuries that cause such extensive periorbital edema that the eyes are swollen shut, making assessment of eye opening impossible. Other factors that complicate the determination of the GCS score reflect changes in prehospital and emergency department practices since the initial description of the GCS over a generation ago. Nowadays, many patients are endotracheally intubated in the prehospital setting, and paralytics and sedatives are often administered before an accurate and thorough neurologic assessment is performed. These problems have led to the creation of other assessment tools, such as those based on CT findings.⁸ Another way of getting around the problem of an inaccurate neurologic examination might be the use of serum markers to identify trauma patients who are likely to have serious central nervous system injury, analogous to the manner in which serum levels of cardiac enzymes are used in the diagnosis of acute myocardial injury.⁹ This is an area of rapidly growing interest.

Common practice classifies brain injury as mild, moderate, or severe based on the GCS score. Mild TBI is traditionally equated with a GCS score of 13-15, whereas moderate TBI refers to a GCS score of 9-12. Some authorities, however, consider a score of 13 to be more indicative of moderate injury. Severe TBI refers to patients with a GCS score of 8 or less.

It is worth emphasizing that this widely used scheme of classifying TBI is based on functional, not anatomic, criteria. This approach contrasts with that used in much of the general trauma literature, in which anatomic criteria are used as the primary means of classifying injuries.

Marshall Computed Tomography Scale

Marshall and colleagues developed a CT-based classification scheme using data from the Traumatic Coma Data Bank (see Table 2). This system, commonly referred to as the Marshall scale or the TCDB scale, classifies CT scans according to such factors as midline shift and compression of cerebrospinal fluid cisterns. Although this scale is useful in certain circumstances, it should not be considered a substitute for an accurate neurologic examination.

Abbreviated Injury Scale

The Abbreviated Injury Scale (Table 3) for the head assigns a score of 1 for minor scalp injuries such as abrasions, contusions, and lacerations.¹⁰ Longer and deeper lacerations receive a score of 2, whereas scalp injuries accompanied by significant blood loss or characterized by total scalp loss are scored as 3. Cranial nerve injuries are coded as 2.

Table 3 Abbreviated Injury Scale for Head Injury

LOC, Loss of consciousness.

Injuries to major cerebral vessels are generally coded as 3 or 4 for thrombosis or traumatic aneurysm formation, or as 4 or 5 for laceration.

Scores for fractures of the skull and skull base range from 2 for simple fractures of the vault, to 3 for skull base fractures orcomminuted vault fractures, to 4 for the most complex open fractures with exposed brain tissue or for significantly depressed closed fractures.

Scoring for brain parenchymal injuries ranges from 3 to 5. Small single or multiple contusions receive a score of 3, as does SAH, edema, or infarction directly related to trauma. Hematomas are scored as 4 or 5, depending on their size.

Massive destruction or crush injuries are scored as a 6.

The duration of loss of consciousness and presence of associated neurologic deficits may be used to score injury severity if such scores exceed those based on anatomical injuries. Such scores range from 2 to 5.

The American Association for the Surgery of Trauma organ injury scale does not address brain injuries.

Mild and Moderate Traumatic Brain Injury

This chapter and the following one focus on patients with severe TBI. Those with moderate brain injury, who are commonly described as patients with a GCS score of 9–12, often have significant intracranial pathology and yet still obey commands. Their management is similar to that of severe TBI patients in the sense that careful observation and prevention of secondary insults are of paramount importance. However, intracranial monitors may not be needed in many of these patients, and they may not require the same intensity of care for the same amount of time as severe TBI patients.

The vast majority of patients seeking medical attention after a brain injury have mild TBI. They are often diagnosed as having had a concussion. Importantly, the diagnosis of mild TBI (or concussion) does not require that a patient lose consciousness. Some data indicate that the majority of patients with mild TBI never lose consciousness, but they may complain of a variety of symptoms like headache, post-traumatic amnesia, difficulty concentrating, ringing in the ears, unsteadiness of gait, and so on.

Unlike the situation with severe TBI, the major concern with mild TBI is not whether a patient will live or die or whether a patient will become vegetative or severely disabled. Almost all these patients survive. Instead, the major morbidity relates to disturbances of memory, cognition, attention, emotional stability, and related areas. The lack of “hard” evidence of neurologic impairment leads many physicians to downplay the significance of these symptoms. However, these patients often go on to lose their jobs, drop out of school, divorce their spouses, or go through other major upheavals in their lives. Eventually, most of them recover, but such a process may take months. Reassurance of the patient and family may be all that is needed, and in fact, may be all that can be offered. Counseling and formal testing may be appropriate if objective documentation of injury is needed or if a physician suspects malingering or symptom magnification for secondary gain.