The cranium is a rigid structure. The major intracranial contents are the brain (to include the neuroglial elements and interstitial fluid), blood (arterial and venous), and cerebrospinal fluid (Table 19.1). When a new intracranial mass is introduced, a compensatory change in volume must occur through a reciprocal decrease in venous blood or cerebrospinal fluid (CSF) to keep the total intracranial volume constant. This is the Monro-Kellie doctrine (Fig. 19.1), which has been confirmed by many experimental and clinical observations. Only in children, with open fontanelles and whose sutures have not yet fused, can the cranium itself expand to physically accommodate extra volume.

TABLE 19.1 Intracranial Contents and Their Respective Volumes

Component	Volume	Percentage of Total Volume
Brain (70%) and interstitial fluid (10%)	1400 mL	80%
Blood	150 mL	10%
Cerebrospinal fluid	150 mL	10%
TOTAL	1700 mL	100%

FIGURE 19.1 Monro-Kellie doctrine holds that total intracranial volume remains constant. A, Physiological state with normal intracranial pressure (ICP). The major intracranial components are brain (80%), arterial and venous blood (10%), and cerebrospinal fluid (CSF) (10%). The cranium is a rigid container, the intracranial volume is constant, and the normal intracranial contents are shown with ICP within physiological range (10 to 15 mm Hg). B, Intracranial mass with compensation (normal ICP). This patient has an intracranial mass (space-occupying lesion) of moderate size. Because intracranial volume is constant, the increasing volume caused by the mass is compensated by a decrease in the intracranial content. Venous volume decreases through egress of venous blood from the intracranial cavity into the jugular veins. CSF volume decreases because of egress of CSF through the foramen magnum into the spinal canal. The brain itself is nearly incompressible, and thus no significant change in its volume occurs; neither is there change in arterial volume. Intracranial volume is constant and there is no net rise in ICP (pressure buffering). C, Intracranial mass with decompensation and elevated ICP. The intracranial mass is much larger, beyond the pressure-buffering capacity of venous blood and CSF; there is a net rise in ICP.

Compliance (dV/dP) is the change in volume observed for a given change in pressure. This represents the accommodative potential of the intracranial space. In clinical practice, however, what is actually measured is elastance (dP/dV). Elastance is the inverse of compliance and is the change in pressure observed for a given change in volume. It represents the resistance to outward expansion of an intracranial mass. The elastance curve (not compliance) is what is plotted in Figure 19.2. Although technically a misnomer, the term compliance is actually entrenched in the literature to describe the aforementioned phenomenon instead of using the proper term elastance. Because of this we, too, will conform to the traditional nomenclature for the rest of this chapter.

FIGURE 19.2 Pressure-volume curve. If an intracranial balloon is expanded slowly in a laboratory animal and the intracranial pressure (ICP) and the balloon’s volume are plotted, an exponential curve results. In the initial phase there is virtually no increase in ICP because of the compensatory decrease in cerebrospinal fluid (CSF) and venous volumes; the compliance is high (elastance low). With further expansion of the balloon, the ICP begins to rise; compliance is low (elastance high). In the terminal stages, when the compensatory mechanisms are exhausted, there is a steep rise in pressure (no compliance, highest elastance).

In accordance with the Monro-Kellie doctrine, several innate homeostatic mechanisms exist in an attempt to maintain intracranial pressure within the physiological range—generally considered less than 20 mm Hg in the acute inpatient setting. First, the venous system collapses easily and squeezes venous blood out through the jugular veins or through the emissary and scalp veins. Likewise, CSF can be displaced from the ventricular system through the foramina of Luschka and Magendie into the spinal subarachnoid space.¹ When these compensatory mechanisms have been exhausted, small changes in volume produce precipitous increases in pressure. This effect can be demonstrated experimentally by inserting a Foley catheter into the epidural space of a rat and gradually inflating a balloon with increasing volumes. The curve produced by plotting intracranial pressure against volume is the so-called compliance curve in Figure 19.2. The innate homeostatic pressure-buffering mechanism offered by displacement of CSF and venous blood keeps this curve flat until a “critical volume” is reached. After this critical volume, small volumetric changes result in precipitous increases in pressure, and intracranial hypertension naturally ensues. Brain parenchyma and arterial blood do not participate, to any significant extent, in the innate intracranial pressure-buffering mechanisms.

Acute Intracranial Hypertention

Symptoms and Signs

The symptoms and “sine qua non” of acute increased intracranial pressure are altered levels of consciousness with progressive neurological decline and decreasing wakefulness, usually followed by obtundation. The obtundation is usually due to one of two causes: (1) compression of the suprameduallary reticular activating system or (2) bithalamic or bicortical damage. This is typically manifested by a Glasgow Coma Scale (GCS) score less than 8, irregular breathing patterns, posturing, and possibly pupillary inequality.

Cerebral Blood Flow

Normal cerebral blood flow averages 55 to 60 mL/100 g brain tissue per minute. In the gray matter the blood flow is 75 mL/100 g brain tissue per minute, whereas in the white matter it is around 45 mL/100 g brain tissue per minute. This flow is sufficient to meet the metabolic needs of the brain. The brain usually begins to show signs of ischemia at 20 mL/100 g/minute and permanent damage usually results when the blood flow drops below 10 mL/100 g/minute in most models. The most significant factor that determines cerebral blood flow at any given time is the cerebral perfusion pressure (CPP). The CPP is the effective blood pressure gradient across the brain. CPP is the difference between the incoming mean arterial pressure (MAP) and the opposing intracranial pressure (ICP): CPP = MAP − ICP. The mean arterial pressure can be calculated in two ways: (1) the diastolic pressure plus one third of the pulse pressure (DP + 1⁄3PP) or (2) two thirds of the diastolic pressure plus one third of the systolic pressure (2⁄3DP + 1⁄3SP). The ICP has to be measured directly, which can be done with various devices and will be discussed later in this chapter. With increased ICP there is an obvious tendency for the cerebral perfusion pressure to decrease.

Three major factors regulate cerebral blood flow under physiological conditions: CPP, concentration of arterial CO₂, and arterial PO₂. The ability to maintain constant blood flow to the brain over a wide range of CPPs (50-150 mm Hg) is called cerebral autoregulation. When the CPP is low, the cerebral arterioles dilate to allow adequate flow at the decreased pressure. Conversely, an increase in CPP causes the arterioles to constrict and maintain the flow within physiological range (Fig. 19.3). Decreases in CO₂ tension in the blood (i.e., hyperventilation) causes diffuse vasoconstriction.² Vasoconstriction decreases both cerebral blood flow and cerebral blood volume in the brain. The risks and benefits of inducing hyperventilation are discussed under the treatment section at the end of this chapter. Lastly, severe hypoxia causes cerebrovascular dilatation. This effect only becomes apparent when the oxygen tension in the blood drops to 50 mm Hg and becomes maximal around 20 mm Hg.

FIGURE 19.3 Three major factors regulate cerebral blood flow (CBF) under physiological conditions: cerebral perfusion pressure (CPP), concentration of arterial CO₂, and arterial PO₂. This graph can be read in regard to CPP to CBF, or PaO₂ and PaCO₂ to CBF. The ability to maintain constant blood flow to the brain over a wide range of CPPs (50-150 mm Hg) is called cerebral autoregulation. Patients with hypertension have their autoregulatory curve shifted to the right. Trauma or stroke patients can have a loss of the normal sigmoidal shape to the autoregulatory curve. Additionally, decreases in CO₂ tension in the blood causes diffuse vasoconstriction. This decreases both cerebral blood flow and thus cerebral blood volume in the brain. Severe hypoxia causes cerebrovascular dilatation. This only becomes apparent when the oxygen tension in the blood drops to 50 mm Hg, and becomes maximal around 20 mm Hg.

Under certain pathological conditions cerebral blood flow cannot always be autoregulated.³ When the CPP exceeds 150 mm Hg, such as in hypertensive crisis, the autoregulatory system fails. In this case there would be a passive increase in blood flow proportionate to the increase in systemic pressure, causing an exudation of fluid from the vascular system with resultant vasogenic edema.⁴ Additionally, certain toxins such as carbon dioxide can cause diffuse cerebrovascular dilatation and inhibit proper autoregulation.⁵ Lastly, during the first 4 to 5 days of head trauma, many patients can experience a disruption in cerebral autoregulation; this is often observed in the pediatric population and may result in hyperemia.⁶

Disruption of cerebral blood flow autoregulation in the setting of trauma could potentially lead to severe alterations in CPP when a patient’s brain is acutely injured.⁷ When disrupted, cerebral blood flow would be directly proportional to CPP under a much larger range than normal. This is likely one of the contributing reasons why even one episode of hypotension in the patient with acute head injury may lead to significantly worse outcomes.⁸ Thus, many institutions have adopted protocols to maintain a minimal CPP in the acute setting—usually around 50 to 60 mm Hg—and perhaps a maximum threshold of CPP if the intracranial hypertension can be linked directly to hyperemia.

Various ways to evaluate a patient’s autoregulatory abilities have been developed to include computed tomography (CT) perfusion scans and transcranial Doppler (TCD) studies. Interpreting these studies, however, is a matter of current debate. Although still somewhat in its infancy, knowing the status of a patient’s autoregulatory capacity in traumatic brain injury is beginning to shape individualized care in the intensive care unit (ICU).

Monitoring Intracranial Pressure

The most significant factor determining morbidity and mortality risk in patients with acute cranial neurosurgical disorders continues to be increased ICP. Continuous ICP monitoring is very useful for assessing intracranial dynamics in patients with suspected intracranial hypertension. There are no clinical indicators that can be used in the early stage of rising ICP to forestall a further rise. Classical clinical indicators described in the literature (cranial nerve abnormalities, posturing, etc.) occur in the end stage and they are not sensitive enough to show subtle changes in pressure. The two most common indications for ICP monitoring are closed head injury and spontaneous subarachnoid hemorrhage.

Indications for Monitoring Intracranial Pressure

The advent of continuous ICP monitoring has improved outcomes in traumatic brain injury ^9,¹⁰ by reducing the effects of secondary injury occurring after the initial brain insult. In the modern intensive care unit ICP monitoring in the adult patient with severe head injury is a recommended practice guideline based on available clinical evidence. Indications of ICP monitoring following head injury include a GCS score between 3 and 8 and an abnormal CT scan. Unfortunately, no class I evidence is available regarding the use of ICP monitors in pediatric head injury patients. Nevertheless, strong evidence supports the association between elevated ICP and poor outcomes,¹¹ and aggressive treatment of elevated ICP is associated with the best clinical outcomes. ICP monitoring provides valuable data that can inform medical management.

Techniques of Monitoring Intracranial Pressure

Intracranial pressure is best measured directly and continuously from the cranial cavity. Although lumbar puncture can indirectly indicate ICP, it is neither safe nor accurate in the setting of intracranial hypertension as it might precipitate tonsillar/transtentorial herniation by increasing the pressure gradient between the cranial and spinal subarachnoid compartments.

There are two commonly used pressure-monitoring systems in contemporary neurosurgical practice. An intraventricular catheter connected to a manometer and a drainage system is the standard against which all other systems are compared.¹²^–¹⁴ The ventricular catheter ideally should be tunneled under the skin and brought out through a separate stab wound, well away from the ventricular entry site, to minimize the risk of infection. Infection is the most significant complication of intraventricular pressure monitoring. With the use of an electronic transducer, the waveform can also be monitored (Fig. 19.4). The major advantage of the method is that the ventricular catheter is used not only to measure the pressure, but also as a therapeutic modality allowing intermittent or continuous drainage of CSF when the pressure exceeds physiological limits.

FIGURE 19.4 Intracranial pressure (ICP) monitoring system using ventricular catheter, pressure transducer, manometer, and drainage bag.

A second method is the use of the fiberoptic transducer-tipped catheter system. The transducer-tipped catheter can be placed within the brain parenchyma or in the subdural space, depending on the surgeon’s choice and the clinical situation.¹⁵^–¹⁷ The pressure monitor gives both digital readout and a waveform. The advantages of this system are that the zero point does not have to be reset with changes in head position because the pressure-sensing transducer is within the cranial cavity, and it is not susceptible to blockage by debris because it is not a fluid-coupled system. Also, insertion of a fiberoptic cable into the brain parenchyma is not affected by ventricular size or shift, which can make insertion of a ventricular catheter in the setting of intracranial hypertension quite trying. Lastly, the infection rates with fiberoptic probes are significantly lower than with ventriculostomy catheters, reported as less than 1% in one of the the largest reported series.¹⁸ The disadvantages of this system, however, are (1) higher cost and (2) baseline drift, which means there is evidence that with prolonged fiberoptic monitoring the indicated pressures may become less accurate in an unpredictable fashion.^19,²⁰

Intracranial Pressure Waveforms

The waveform of normal ICP typically shows three components: the percussion wave (P1), the tidal wave (P2), and finally the dicrotic notch (P3) (Fig. 19.5A). Under physiological conditions P1 is the tallest and is thought to represent the pressure at peak cardiac systole transmitted throughout the choroid plexus. P2 follows P1 and its resultant smaller peak is thought to represent the filling of the intracranial arteries with systolic blood and results in a rebound increase in ICP from this increase in intracranial volume. Last is the dicrotic wave.

FIGURE 19.5 A, Intracranial pressure (ICP) waveform under physiological conditions. Three peaks are generally recognizable. The first, and usually tallest, is the percussion wave, followed by the tidal wave, the dicrotic notch, and the dicrotic wave. Note that the tidal and dicrotic waves have progressively lesser amplitudes than the percussion wave. B, Abnormal intracranial waveform with high ICP. The amplitude of the tidal wave exceeds the amplitude of the percussion wave.

When the ICP rises and compliance decreases, one will observe P2 becoming taller than the P1. This is because a similar volume in intracranial arterial blood is now resulting in a higher change in ICP (Fig. 19.5B). Such alterations in the morphology can give the treating physician further clues regarding not only the compliance but also the autoregulatory capacity of the brain.

When the ICP waveforms are registered over a period of time certain trends may become apparent ¹⁴ (Fig. 19.6). Plateau waves, or type A waves, are characterized by an abrupt elevation in ICP for 5 to 20 minutes followed by a rapid fall in the pressure to resting levels. The amplitude may reach as high as 50 to 100 mm Hg. Plateau waves may be clinically marked by a decreasing level of consciousness, restlessness, increased tone in the extremities, and tonic-clonic movements. Plateau waves may represent transient surges in ICP secondary to increased cerebral blood volume likely related to CO₂ retention.

FIGURE 19.6 Examples of abnormal intracranial pressure (ICP) waveforms with trend recording. Plateau waves, or A waves, are characterized by abrupt elevation in ICP for 5 to 20 minutes followed by a rapid fall in pressure to the resting level. Plateau waves herald neurological worsening. B and C waves have questionable clinical significance.

B waves, which have a frequency of 0.5 to 2 waves per minute, are related to rhythmic variations in breathing. The C waves are rhythmic variations related to the Traub-Meyer-Hering waves of systemic blood pressure and have smaller amplitude, with a frequency of about 6 per minute. B and C waves have questionable clinical significance, although there is some evidence that B waves may precede A wave formation.

Types of Brain Herniation

The brain is supported by dural folds that prevent undue movements of the brain within the cranial cavity (Fig. 19.7). There are two major dural folds—the falx cerebri and the tentorium cerebelli. The falx cerebri is a sickle-shaped dural fold in the midline that separates the two cerebral hemispheres. The tentorium cerebelli is a tent-shaped dural fold that separates the occipital lobes from the posterior fossa structures. The tentorial incisura, or hiatus, is the dural opening that surrounds the rostral brainstem. Care must be taken during ICP management so as to not to exacerbate the pressure differences between compartments and to inadvertently encourage herniation.

FIGURE 19.7 Dural folds within the cranial cavity and associated herniation sites.

Tonsillar Herniation

Acute tonsillar herniation generally results from acute expansion of posterior fossa lesions. It may result from an ill-advised lumbar puncture in a patient with a mass lesion within the cranial cavity. The tonsils of the cerebellum herniate through the foramen magnum into the upper spinal canal, compressing the medulla (see Fig. 19.8). The manifestations of acute medullary compression are cardiorespiratory impairment, hypertension, high pulse pressure, Cheyne-Stokes respiration, neurogenic hyperventilation, and impaired consciousness. The patient may have a stiff neck or be in an opisthotonic position. Decorticate or decerebrate posturing may be present as well.

Upward Transtentorial Herniation

Acute upward cerebellar herniation occurs when the pressure in the infratentorial space exceeds the pressure in the supratentorial space. This can be iatrogenically caused from aggressive drainage of CSF from an external ventricular drain in the setting of a posterior fossa mass lesion. Upward transtentorial herniation does not have a clear-cut clinical syndrome, but results in obtundation, pupillary abnormalities, and posturing. Additionally, unilateral or bilateral superior cerebellar artery (SCA) infarctions could occur as the SCA becomes compressed against the bottom of the tentorium. As the SCA supplies the deep cerebellar nuclei, this would obviously portend a poor prognosis.

Treatment

With almost no exception, all patients with acute elevated ICP can be managed initially by the same conservative steps. First, the importance of elevating the head of the bed to 30 to 45 degrees cannot be overemphasized. This posture promotes venous drainage to enhance the innate compensatory mechanisms of the CNS. The neck should be kept straight and not allowed to kink over to any side. Second, normothermia should be achieved because an elevated brain temperature can increase metabolism, which can not only elevate ICPs but also promote ischemia. Lastly, sedation and analgesia are invaluable tools in the management of acute ICP. Propofol is the most commonly used sedating agent not only because of its “quick on-quick off” effect but also because of its ability to lower cerebral metabolism. After these initial steps have been taken, further ICP therapy requires enhanced knowledge of physiology as described in the following section.

The Blood-Brain Barrier

In order to understand the optimal way to treat intracranial hypertension, knowledge of the components of the blood-brain barrier is paramount. Not all substances that are carried in the blood can reach the neural tissue—a barrier composed of astrocytic foot process wrapping around a capillary endothelium composed of tight junctions ⁴ (Fig. 19.9). These endothelial tight junctions are the barrier to the passive movement of many substances in order to protect the sensitive neural tissue from toxic materials. There are two mechanisms by which materials may be transported naturally across the endothelial cells. First, lipid-soluble substances can usually penetrate all capillary endothelial cell membranes in a passive manner.²¹ Second, amino acids and sugars are transported across the capillary endothelium by specific carrier-mediated mechanisms.²² The measurement of how easily something diffuses across the blood-brain barrier is called the reflection coefficient. A coefficient of 0 is complete passive diffusion; 1 indicates no passive diffusion.^23,²⁴

FIGURE 19.9 Differences between somatic and brain capillaries. In the somatic capillary, fenestrations between endothelial cells allow free flow of plasma components into the tissues. In addition, there is bulky flow of plasma components across endothelial cells via pinocytotic vesicles. In the brain capillary, the endothelial cells are attached to each other by tight junctions. There are no intervening fenestrae. Certain selected plasma components cross the endothelial membrane if they are lipid soluble; others, such as amino acids and sugars, are transported across the endothelial cells through a carrier-mediated mechanism. The large number of mitochondria in the brain endothelial cells generate energy for active transport.

Decompressive Hemicraniotomy

The most direct way to normalize raised ICP to the physiological range is to eliminate its cause. If the increased pressure is the result of a mass effect, such as a blood clot, prompt evacuation of the offending lesion will restore ICP to normal more effectively than any other measure. The goal of a decompressive craniotomy is to perform a robust craniotomy that extends from the frontal lobe to the occipital lobe and includes the temporal lobe so that uncal herniation can be alleviated. After the bone is removed, the dura is often flapped open and the mass lesion evacuated. The brain can then be covered with a dural substitute to permit expansion. The bone flap can be placed back or kept off depending on intraoperative findings and personal surgical preferences. If the bone flap is kept off, the dural substitute is then covered with the scalp and a subgaleal drain is usually placed (Fig. 19.10). The bone flap may be either placed in a subcutaneous abdominal pocket, or more commonly at the present, storaged in a –80°C freezer in a sterile fashion. A cranioplasty or cranial bone replacement should be performed at a later date if the patient survives the acute injury.

FIGURE 19.10 Decompressive hemicraniectomy. Note the extensive boney decompression from the frontal bone to the occiput and the remaining interhemispheric subdural hematoma. The brain parenchyma on the left can be seen protruding beyond the previous boney-boundary indicating how severe the intracranial hypertension was. The hyperdensity seen above the left frontal lobe is part of the Jackson-Pratt subgaleal drain.

The indications for performing this operation are most often (1) control of increased ICP that is recalcitrant to medical therapy and (2) progressively worsening neurological examination with evidence of massive hemispheric brain swelling, contusion, or intracranial blood or mass lesion. Despite the widespread use of decompressive hemicraniotomy worldwide,²⁵^–²⁷ no level I data are available that show its superiority over aggressive medical treatment. In fact, a recently published randomized control trial evaluated patients with diffuse brain swelling (i.e., no mass effect) with refractory ICP.²⁸ This study found that even though decompressive bifrontal craniectomy reduced ICP more than medical management, it did not translate in better outcomes and indeed appeared to actually worsen them in some circumstances. There are some critiques of this trial, including the definition of “refractory ICP,” but nonetheless, it reinforces that the use of decompressive hemicraniectomy in diffuse brain swelling is an area of controversy. In addition, surgery is obviously fraught with comorbidity, and thus, nonsurgical options is reasonable as a first-line therapy when there is no obvious mass lesion, contusion or intracranial blood. However, we still employ decompressive craniectomy in trauma patients who have failed medical therapy and have contusions and intracranial blood with elevated ICP. Indicated as a first-line therapy when there is no obvious mass lesion. These options are discussed in the following sections.

Hyperventilation

Hyperventilation causes a fall in ICP by reducing intracranial blood flow and blood volume through vasoconstriction. Hyperventilation is generally initiated for hyperacute management of increased ICP. Sustained hyperventilation in acute intracranial hypertension is a mode of therapy that has not been found to be beneficial and is actually contraindicated. In the setting of increased ICP, prolonged hyperventilation will promote ischemia through decreased cerebral blood flow supply. If undertaken, hyperventilation should be moderate, reducing the carbon dioxide between 30 and 35 mm Hg.

Mannitol

Mannitol is the osmotic agent universally used to treat cerebral edema.²⁹^–³¹ Other osmotic agents such as urea and glycerol are no longer used in contemporary neurosurgical practice because of side effects such as hemolysis. Mannitol decreases ICPs in two distinct mechanisms. First, when mannitol is initially delivered it creates an osmotic gradient from all body tissues into the intravascular space. This transiently increases intravascular volume and improves the rheology of cerebral blood. This transient increase in cerebral blood flow is thought to trigger the intrinsic autoregulatory mechanism of the intracranial vasculature (if it is intact) to vasoconstrict and decrease intracranial arterial blood volume. This decrease in blood volume translates into decreases in ICP, and is the mechanism that gives mannitol almost immediate efficacy in the setting of intracranial hypertension.

Once mannitol begins to circulate through the intracranial vasculature, the cumulative effects of the osmotic gradient can draw fluid from the brain interstitial space into the vascular space. Key to this process is intact tight junctions. If the tight junctions are not intact, mannitol will slowly permeate the neural tissue, preventing the formation of an osmotic gradient.

Mannitol is generally given in boluses rather than as a continuous drip. The usual dose is 0.25 g to 1 g/kg at 4- to 6-hour intervals. Mannitol boluses are generally effective for 48 to 72 hours. Its routine/scheduled use beyond 72 hour can become ineffective because mannitol’s reflection coefficient is 0.9, meaning that even with an intact blood-brain barrier it will slowly leak into the interstitial fluid/brain parenchyma. Serum osmolality, kidney function, and electrolytes should be carefully monitored during mannitol therapy.

Hypertonic Saline

High-dose aliquots of hypertonic saline have recently emerged as a potential first-line therapy for elevated ICP. Thirty-milliliter boluses of 23.4% sodium chloride given through a central line over 15 minutes have been shown to be at least as efficacious as mannitol in multiple small prospective studies in lowering intracranial pressure.^32,³³ The mechanism by which 23.4% saline works is similar to that of mannitol, in that it creates an osmotic gradient to draw fluid from the interstitial space of the brain across the blood-brain barrier and into the intravascular space.

The reflection coefficient of hypertonic saline is 1.0, meaning that if a patient’s blood-brain barrier is intact it will not accumulate in the interstitial space/brain parenchyma. This makes hypertonic saline appealing for longer than 48- to 72-hour use. Also, unlike mannitol, hypertonic saline does not have a systemic diuretic effect and thus MAPs/CPPs are more stable during its use. Therefore, hypertonic saline is a potentially better alternative to treating a polytrauma patient’s elevated ICP who may have concomitant systemic hypovolemia and hypotension. A single dose of 23.4% hypertonic saline does not seem to alter systemic sodium concentrations. However, with repeated use, bolus therapy of 23.4% saline can causes electrolyte abnormalities and thus potentially cardiac conduction abnormalities and should be monitored.

Loop Diuretics

Furosemide has been classically used as an adjunct to mannitol because it seems to have a synergistic effect. Furosemide alone, however, cannot be relied upon to reduce ICP. Because its primary action is on the kidney, it is not dependent on the intact blood-brain barrier for its effect. It is also possibly thought to reduce CSF production.

Ventricular Drainage

Drainage of CSF from the ventricular system is a simple, effective, and quick method of decreasing ICP. Ventriculostomy is supremely effective in two unique patient populations: acute subarachnoid hemorrhage and spontaneous/traumatic intraventricular hemorrhage. Providing an alternative pathway for CSF drainage in these patient populations can dramatically reduce ICP and result in a remarkably quick improvement in clinical examination.

Ventriculostomy drainage can also be used in the setting of trauma, not only to monitor pressure continuously but also to drain the ventricular to provide therapeutic effects.

Barbiturate Coma and Paralytics

Induction of coma with short-acting barbiturates or inducing chemical paralysis is a last and controversial resort in the medical management of raised ICP when all other conservative measures fail. Barbiturates decrease the ICP by several mechanisms. They inhibit the release of fatty acid perioxidation products by scavenging free radicals from the mitochondrial respiratory chain. They also inhibit cerebral metabolism and reduce cerebral blood flow. The most commonly used drug is thiopental, which is given in a loading dose of 3 to 10 mg/kg over a 10-minute period and a maintenance dose of 1 to 2 mg/kg/hour. The serum level should be maintained at 3 to 4 mg/L. Patients in barbiturate coma require intensive monitoring of hemodynamic function, ICP, and blood gases. Vasopressors may have to be administered if hypotension results. Barbiturate therapy is usually withdrawn when the ICP normalizes and there is a good intracranial compliance; no clinical indicators are available to dictate termination of therapy.

Subacute to Chronic Intracranial Hypertension

Causes of subacute to chronic intracranial hypertension are mass lesions (tumors, chronic subdural hematomas, etc.) that not only increase intracranial volume but also can induce interstitial edema. The most common symptom of subacute/chronic elevations is headache that is generalized in nature and is worse in the morning—presumably because of the increase in CO₂ tension (from a physiological decrease in respiratory rate) and increased venous pressure from laying supine. The pain-sensitive structures within the cranial cavity are the dura and the blood vessels. Focal mass lesions that distort the dura and stretch the vessels tend to cause headache more frequently than diffuse generalized increases in ICP without focal mass effect, such as a pseudotumor cerebri or hydrocephalus. The headache, if present, may be associated with vomiting. The vomiting associated with increased ICP is usually not associated with nausea. Papilledema is another cardinal sign of increased ICP.³⁴

A sixth cranial nerve/lateral rectus palsy is another classic sign of a subacute to chronic presentation of increased ICP. The long intracranial course of the abducens nerve makes it susceptible to injury from elevated ICP. The abducens nerve exits the ventral brainstem at the pontomedullary junction and travels in the subarachnoid space before ascending up the clivus to the petrous apex. Once at the petrous apex it travels beneath the petroclinoid ligament of Gruber before piercing the dura at Dorello’s canal and entering the cavernous sinus. It travels through the cavernous sinus to the superior orbital fissure before entering the orbit. Elevated ICP is postulated to compress the abducens nerve at the petroclinoid ligament.

Cerebral Edema

Cerebral edema is a prominent cause of subacute to chronic intracranial hypertension. It may be defined as a state of increased brain volume as a result of an increase in water content. There are three types of cerebral edema (Table 19.2).

TABLE 19.2 Types and Characteristics of Cerebral Edema

Vasogenic edema is the most common form of brain edema encountered in clinical practice (Fig. 19.11). It results from increased permeability of capillaries of the blood-brain barrier. The tight junctions between the endothelial cells become incompetent, allowing plasma filtrate to escape into the intercellular space³⁵^–³⁷ (Fig. 19.12). The phenomenon of contrast enhancement in CT and magnetic resonance imaging (MRI) scans is, in part, because of the breakdown of the blood-brain barrier and resultant vasogenic edema. Vasogenic edema is most commonly seen with tumor, abscess, and chronic reaction to a blood clot. The edema is more marked in white matter than in gray matter.

FIGURE 19.11 Computed tomography (CT) scan of a patient with vasogenic edema caused by metastatic tumor. Note that the edema involves predominantly the white matter.

FIGURE 19.12 Normal appearance of a cerebral capillary contrasted with the changes that occur in vasogenic or cytotoxic edema. Under normal conditions, the intercellular tight junctions are intact. In vasogenic edema, the tight junctions are not competent, allowing leakage of plasma into the interstitial space. In cytotoxic edema, there is a primary failure of adenosine triphosphate (ATP)-dependent sodium pump mechanism resulting in intracellular accumulation of sodium and secondarily water.

Cytotoxic edema most commonly results from hypoxia of the neural tissue. The hypoxia affects the adenosine triphosphate (ATP)-dependent sodium/potassium pump mechanism in the cell membrane, promoting an accumulation of intracellular sodium and the subsequent flow of water into the cell to maintain osmotic equilibrium. Thus, the edema is primarily intracellular and affects virtually all cells, including the endothelial cells, astrocytes, and neurons. Because of the swelling of these cells, the interstitial space is considerably narrowed (see Fig. 19.12). The two most common causes of cytotoxic edema are tissue hypoxia and water intoxication. The CT scan commonly shows only subtle initial changes or no changes at all, indicative of cytotoxic edema in the early phases of ischemic stroke.

Interstitial edema results from transudation of CSF in obstructive hydrocephalus. It is best observed on CT or MRI scan as periventricular low-density areas because of the retrograde transependymal flow of CSF into the interstitial space of the white matter (Fig. 19.13). Interstitial edema is most commonly observed in the frontal region. This finding generally indicates active hydrocephalus requiring surgical therapy.

FIGURE 19.13 Computed tomography (CT) scan of a patient with active hydrocephalus showing periventricular low density representing transependymal flow of cerebrovascular fluid (CSF) (interstitial edema).

Treatment

Craniotomy

As previously mentioned, the most direct way to normalize raised ICP to the physiological range is to eliminate its cause. If the increased pressure is the result of a mass effect, such as a tumor or chronic subdural hematoma, prompt evacuation of the offending lesion will restore ICP to normal more effectively than any other measure. Ventriculostomy may also be used to stabilize ICP.

Steroids

Dexamethasone is used for treating chronic increases in ICP related to vasogenic edema caused by primary or metastatic neoplasms. It is ineffective in the management of cytotoxic edema related to head trauma or cerebral infarction. Steroids are thought to stabilize the cell membrane and restore the normal permeability of endothelial cells. This stability is brought about by inhibition of lysosomal activity, suppression of polyunsaturated fatty acid production that causes edema, and decrease of free radical production. The usual loading dose is 10 mg, given intravenously, followed by 4 mg every 6 hours. When the appropriate therapeutic goal has been achieved, the dose should be slowly tapered over a period of 3 to 4 days. Patients can make a remarkable neurological improvement on steroids if vasogenic cerebral edema is the main cause of the intracranial hypertension. Side effects include a relative leukocytosis, elevated systemic glucose, mildly increased blood pressures, stomach ulcers, immunosuppression, increased appetite, acute psychiatric disease such as mania or irritability, and osteoporosis with long-term use. If steroids are used for a protracted period, a long taper should be used rather than abrupt discontinuation to avoid complications with potential adrenal suppression.

Novel Adjunctive Therapies

In the setting of acute head trauma, much focus has now turned to prevention of secondary injury. This approach includes not only ICP control but also monitoring and attempted manipulation at cerebral oxygenation.

Brain Tissue Oxygenation Monitoring

A polarographic cerebral oxygen monitor (Licox system) can be placed to the brain parenchyma itself to monitor brain tissue oxygenation levels (Fig. 19.14). Several retrospective studies correlated poor outcome with low cerebral oxygenation (jugular saturation) and low values for partial pressure of oxygen in brain tissue (PBTO₂ (<10 mm Hg for >2 hours).³⁸^–⁴¹ ⁴² This finding opens up the possibility that therapeutric interventions to manipulate the PBTO₂ value could potentially alter the traditionally poor outcome in traumatic brain injury, notably the 36% mortality rate.^43,⁴⁴ Indeed, several retrospective studies have suggested that active treatment of low PBTO₂ (defined as <20 mm Hg) does improve outcomes as measured on the Glasgow Outcomes Scale.^45,⁴⁶ However, a few other papers suggest it does not and simply adds to the cost of care and stay in ICU.⁴⁷ The precise role of PBTO₂ measurement in management of ICP is evolving, and utility remains to be determined. However, its popularity in centers that manage a large population of head-injured patients has certainly expanded over the past decade.