Chapter 111. Management of Cerebral Edema and Elevated Intracranial Pressure

Most forms of acute head injury increase the volume of the intracranial contents. Cerebral edema, hemorrhage, acute hydrocephalus, or rapidly growing tumors all result in intracranial hypertension (see Chapter 104 and Table 104-5). The relationship between the volume increase and the rise in intracranial pressure is not linear. Small volumes can be accommodated either by stretch, if the cranial sutures are still open, or by displacement of cerebrospinal fluid (CSF) into the spinal canal (this is why in most forms of intracranial hypertension, with the notable exception of hydrocephalus, the ventricles appear small). To a lesser extent, the blood contained at any time in the cerebral vessels, especially the cerebral veins, diminishes. As the volume increase produced by the injury becomes larger, however, the pressure increases more rapidly, a circumstance that can be described as a volume-dependent decrease in the compliance of the cranium. Any additional volume causes a disproportionate elevation in intracranial pressure.

The most immediate and dangerous consequence of increased intracranial pressure is a reduction in cerebral blood flow. Cerebral herniation (see Chapter 104) tends to be a later development. In most organs, blood flow is directly proportional to the difference between arterial and venous pressure (the perfusion pressure) and is inversely proportional to the resistance of the organ’s vascular bed. Some organs, such as the heart and the brain, regulate their vascular tone (and thus their vascular resistance) to maintain constant blood flow over a wide range of perfusion pressures appropriate for age (autoregulation). Below or above this autoregulation range, blood flow changes in proportion to perfusion pressure. (Other organs, such as the intestine and the skin, have their flow regulated by neural and humoral inputs in response to a variety of physiological circumstances, sometimes without change in blood pressure.)

The cranial enclosure introduces a new variable for consideration. The outside surface of the cerebral blood vessels is exposed to intracranial pressure. If this pressure exceeds venous pressure, then the effective perfusion pressure is the difference between arterial and intracranial pressure, the venous pressure becoming irrelevant in defining blood flow (just as the height of a waterfall is irrelevant in determining the flow of the river). Intracranial hypertension can reduce cerebral perfusion pressure below the autoregulation range of the cerebral blood vessels, resulting in cerebral ischemia.

Cerebrospinal fluid production and subsequent absorption is, like cerebral blood flow, quite dynamic. The choroid plexus accounts for at least 70% of the brain’s production of cerebrospinal fluid, and the transependymal movement of fluid from the brain parenchyma to the ventricular system accounts for the rest. The average volume of cerebrospinal fluid in children ages 4 to 13 years is 90 mL, and the rate of formation is approximately 500 mL per day, resulting in an hourly turnover of about 14% of the total volume.

The rate of production remains fairly constant and declines only slightly with increased intracranial pressure, but the rate of absorption increases linearly as the pressure exceeds approximately 15 mm Hg to 40 mm Hg to where the rate of absorption is triple the rate of production.2

The brain has specialized systems to regulate traffic of water and solutes in and out of the blood vessels. Brain endothelial cells, for instance, differ from those in other organs by their lack of fenestrations and the presence of an extensive patchwork of tight junctions. In addition, the capillary basement membranes, the pericytes embedded within this membrane, and aquaporin-4 expressing astrocytic end-feet3 form a sheath around every blood vessel, creating a very selective membrane known as the blood-brain barrier (Fig. 111-1). This barrier provides an ideal system for selective transfer of nutrients and metabolic products while protecting the brain cells from exposure to potentially injurious circulating molecules and from excessive accumulation of water (cerebral edema).

Based on the operative mechanisms, cerebral edema is classified as vasogenic, cytotoxic, or interstitial. Vasogenic edema is caused by increased capillary permeability and disruption of the blood-brain barrier following injury. It is seen with brain tumors, abscesses, hemorrhage, and trauma. Cytotoxic edema results from alterations in the membrane functions of neurons, glia, and endothelial cells such as those that occur after hypoxic-ischemic insults, infection, and trauma. Interstitial edema occurs when the usual flow of transependymal fluid to the ventricular system is altered or impeded.2 This happens when cerebrospinal fluid absorption is blocked or when production is increased beyond the brain’s ability to reabsorb it. In most clinical situations, cerebral edema results from a combination of more than one of these mechanisms.

The purpose of treating intracranial hypertension is to avoid further brain injury caused by elevated intracranial pressure. Injury can be from focal or global brain ischemia due to decreased blood flow. The ultimate and most severe form of ischemia takes place during a herniation event that compresses the arterial supply to the brain as discussed in Chapter 104. There are limited therapies that are proven to be beneficial over the long term, and there are fewer yet that have data supporting improved outcomes in children.

The application of these therapies is often the result of generalizing the results obtained in patients treated for head trauma, sometimes with questionable reasoning. Intracranial hypertension in patients with traumatic brain injury (TBI) can be due to multiple causes: space-occupying lesions from extra-axial bleeding; hydrocephalus from edema or mass lesions around the third or fourth ventricles; vasogenic edema from vascular injury; alterations in the intracranial vault from a depressed skull fracture; and cytotoxic brain edema from ischemia, which can occur as a primary or secondary mode of injury. A consensus statement on managing traumatic brain injury and intracranial hypertension specifically in children was published in 2003.4

Traumatic brain injury–induced intracranial hypertension is defined as an intracranial pressure greater than 20 mm Hg. This threshold is based on numerous published studies, but none have been validated in a prospective fashion. Although there may be a lower threshold for younger children or those with open sutures and fontanel, there are no data that support this. However, it should be noted that brain ischemia and herniation can occur at pressures lower than 20 mm Hg, depending on the location of edema and the presence of mass lesions. Factors that are known to exacerbate cerebral edema such as hyperthermia, hypoxia, hypotension, and hypercarbia should be avoided.

Sedation and neuromuscular blockade are often used to treat elevated intracranial pressure (ICP) caused by increases in cerebral metabolism due to noxious stimuli and pain.4,5 Although commonly used in patients with traumatic brain injury and intracranial hypertension, there are no studies in children that demonstrate an improvement in outcome from using sedation or neuromuscular blockade. Draining cerebrospinal fluid via an external ventricular drain to reduce total intracranial contents is an effective means of lowering intracranial pressure, although its long-term benefits have not been sufficiently evaluated.4,6

For decades, hyperosmolar therapy with either mannitol or hypertonic saline has been a mainstay of treatment for intracranial hypertension. Mannitol is an osmotic diuretic based on a six-carbon sugar that was first used for decreasing elevated ICP in the 1950s.7 Mannitol’s effect on intracranial hypertension is probably related to several different functional properties.8 It has historically been considered clinically effective due to its osmotic properties, which promote movement of extravascular fluid into the capillaries and then out of the intracranial vault. This “brain dehydration” effect of mannitol may certainly be a factor in its ICP-lowering properties, although changes in cerebral blood flow and volume may be more important. Mannitol alters blood viscosity, and therefore it may also alter blood flow by its rheological effects on vascular resistance.9

Treatment with hypertonic saline solutions for intracranial hypertension was first described by Weed and McKibben in 1919; these were the first osmotic agents studied for this purpose.4 Like mannitol, the mechanisms of action of hypertonic saline on intracranial hypertension are probably multifactorial. Hypertonic saline acts to lower ICP and, because it is a volume expander, acts to increase cardiac output and increase mean arterial blood pressure. In addition to its effects on ICP and cerebral perfusion pressure, hypertonic saline appears to have immunomodulatory and neurochemical effects.10 Hypertonic saline does have the benefit over mannitol of eliminating the rebound hypovolemia and possible hypotension seen with osmotic and nonosmotic diuretics.

Hyperventilation has also been a cornerstone of treatment of intracranial hypertension in traumatic brain injury based on data implicating hyperemia as the cause of intracranial hypertension in pediatric brain trauma.11 More recent studies have challenged the notion that hyperventilation is beneficial for traumatic brain injury, as there are concerns that it may cause regional brain ischemia and decreased brain oxygenation through the effects of hypocapnia on vascular tone.12 Over time, hypocapnia-induced changes in CSF pH are buffered by bicarbonate transport, so some patients may develop rebound intracranial hypertension when arterial PCO2 is allowed to normalize.13,14 The developing consensus is that the use of hyperventilation should be reserved for life-threatening neurological deteriorations such as herniation, when it may be the most rapidly applicable therapy. In addition, there may be a role for hyperventilation as a second-tier therapy for refractory ICH.

Barbiturates can be effective agents in lowering intracranial pressure in patients with ICH. It is likely that barbiturates act therapeutically by lowering cerebral oxygen consumption. Decreased oxygen demand will lead to decreased cerebral blood flow in those areas of the brain that can couple supply and demand, thus leading to a decrease in ICP.15 In addition to their ICP-lowering effects, barbiturates have direct neuroprotective actions, including inhibition of free radical–mediated lipid peroxidation.4 However, use of high-dose barbiturates is also associated with systemic complications such as myocardial depression and immune suppression. Thiopental and pentobarbital are both used for control of intracranial hypertension, but presumably other barbiturates may be as effective. The benefit-to-adverse-effect ratio decreases as the patients reach burst suppression on EEG, and there is little further reduction in cerebral metabolism achieved with higher doses.16 Although the adult traumatic head injury guidelines suggest barbiturate therapy in treating refractory intracranial hypertension, there is not as much support in the pediatric literature.

Hypothermia has both neuroprotective and ICP-lowering effects. It has been well established that hypothermia lowers cerebral metabolism and will thus reduce cerebral oxygen demand.17 The resultant decreased cerebral blood flow and volume contribute to the ICP-lowering effect of therapeutic hypothermia. In addition, it has been proposed that hypothermia reduces inflammatory injury by other mechanisms, but these have not been sufficiently investigated in humans. Hypothermia does have adverse effects that may counteract the beneficial effects of neuroprotection. Hypotension, cardiac arrhythmias, and immune suppression are three of the more significant drawbacks. The incidence of adverse effects in children compared to the adult incidence is not clear. Initial phase II clinical trials of therapeutic hypothermia in adult traumatic brain injury appeared quite promising; however, when a large multicenter study was done, there was no beneficial effect on outcome.18-20 Any beneficial effect of hypothermia may have been negated by the complications seen with its use. Children who do not have as much preexisting organ dysfunction may be more resistant to these serious adverse effects, and there are enough data supporting its use that it is now being researched in a large, multicenter, NIH-sponsored, randomized study.21 In adults, it is clear that hypothermia offers cerebral protection in the event of hypoxic brain injury secondary to cardiac arrest.22 In newborns with hypoxic-ischemic encephalopathy, there is evidence that early hypothermia may improve cognitive outcome.23

Corticosteroids have been used to treat intracranial hypertension because of their putative benefits of decreasing edema, limiting free radical production and lipid peroxidation, decreasing vasoactive and chemoattractive factors, restoring vascular permeability, and decreasing inflammatory response.24 The positive effects of glucocorticoids in edema surrounding brain tumorspromoted their use in other disease states.25 There has been widespread use of corticosteroids in traumatic brain injury, with as many as 64% of trauma centers using them routinely as recently as 10 years ago.26 There have been multiple studies of steroid use in adult head injury studies with outcome data that demonstrate no beneficial effect on outcome.27-30 The CRASH (corticosteroid randomization after significant head injury) trial included over 10,000 patients and was stopped early as mortality was increased in the steroid-treatment group.31 There are insufficient pediatric data to make guideline or standard recommendations for using steroids for treating intracranial hypertension, though based on the adult data, their use is limited primarily to the presence of newly diagnosed brain tumors.

An emerging treatment of adults and children with traumatic brain injury is decompressive craniectomy. Small case-controlled and randomized studies have confirmed the ICP-relieving effect of decompressive craniectomy and have shown a trend toward improved outcomes in these patients. There are insufficient data for standards or guidelines to be established for using this technique, but it should be considered in cases of severe traumatic brain injury with diffuse swelling, intracranial hypertension refractory to medical management, refractory intracranial hypertension in abusive traumatic brain injury, and those with indicators of “recoverable” brain injury.4 As with all incompletely studied therapies for severe traumatic brain injury and intracranial hypertension, it is possible that decompressive craniectomy will result in decreased mortality by increasing the incidence of vegetative and severely disabled survivors. There does, however, appear to be a selective benefit in the pediatric population, although timing, patient selection, and operative technique are three areas that need further study.32