Cerebral Edema and Elevated Intracranial Pressure

Matthew A. Koenig, MD, FNCS p. 1588-1602 December 2018, Vol.24, No.6 doi: 10.1212/CON.0000000000000665
REVIEW ARTICLES
BROWSE ARTICLES
Continuum Audio Article as PDF
KEY POINTS

Corticosteroids are ineffective for the treatment of cytotoxic edema and are contraindicated in the treatment of patients with severe traumatic brain injury.

Frequent administration of hypertonic saline may cause hyperchloremic metabolic acidosis, which is associated with higher mortality in neurocritical care. Buffering hypertonic solutions with acetate may lower the chance of developing hyperchloremic metabolic acidosis.

Serum osmolality should be monitored in patients treated with frequent doses of mannitol. Mannitol should be held when serum osmolality exceeds 320 mOsm/kg to 340 mOsm/kg or when the osmolar gap exceeds 20 mOsm/kg.

Osmotic agents can be used as a temporizing measure to treat mass effect from cytotoxic edema related to stroke and intracerebral hemorrhage, but evidence for efficacy is weak. These lesions may require surgical decompression.

In states of poor intracranial compliance due to global cerebral edema, small changes in intracranial volume related to flat head position, hyperemia, hypercarbia, fever, or pain may result in exaggerated increases in intracranial pressure.

Spontaneous oscillations in intracranial pressure called Lundberg A and B waves may cause self-limited increases in intracranial pressure that last several minutes. These oscillations are an indicator of poor intracranial compliance, but they typically resolve spontaneously without treatment.

Augmentation of cerebral perfusion pressure with systemic vasopressors may lower intracranial pressure by causing reflex cerebral vasoconstriction, thereby lowering the intracranial volume of blood. Excessive cerebral perfusion, however, may contribute to vasogenic edema in regions with a disrupted blood-brain barrier.

Current Brain Trauma Foundation guidelines recommend maintaining an intracranial pressure of 22 mm Hg or less and a cerebral perfusion pressure of at least 50 mm Hg to 60 mm Hg in patients with severe traumatic brain injury.

A recent clinical trial of an intracranial pressure–based treatment protocol for severe traumatic brain injury compared to a treatment protocol based on clinical examination and imaging without intracranial pressure monitoring showed no difference in outcomes between the two groups.

Pentobarbital infusion for intracranial pressure reduction can result in severe medication side effects such as propylene glycol toxicity, immunosuppression, impaired gastrointestinal motility, and distributive shock.

A recent clinical trial of early induced hypothermia as a neuroprotective strategy in severe traumatic brain injury showed reduction of intracranial pressure–directed interventions, but neurologic outcomes were worse in patients treated with hypothermia.

Induced hypothermia continues to be a useful third-line intervention for refractory intracranial pressure elevation, but an effective hypothermia protocol that includes multimodality treatment of shivering is required.

A recent clinical trial demonstrated that decompressive craniectomy for refractory intracranial pressure elevation in severe traumatic brain injury improves survival and reduces the chance of severe disability, but more patients survived in a vegetative state compared to medical management alone.

With transtentorial herniation, the pupil dilation is ipsilateral to the mass lesion, but hemiplegia may be contralateral because of direct corticospinal tract involvement or ipsilateral because of compression of the contralateral cerebral peduncle against the tentorial edge (Kernohan notch phenomenon).

Serial bedside quantitative pupillometry may detect reduction in the pupillary constriction velocity hours prior to frank clinical signs of transtentorial herniation, which may afford time for preemptive treatment.

Although patients can survive and recover from transtentorial herniation in some cases, the sequelae of herniation can include Duret brainstem hemorrhage, ipsilateral anterior cerebral artery stroke, and contralateral posterior cerebral artery stroke.

Posterior fossa mass lesions can cause cerebellar tonsillar herniation or upward herniation of the cerebellum through the tentorial incisura, which may not be accompanied by intracranial pressure elevation or pupillary changes.

Article Level Metrics

PURPOSE OF REVIEW: This article reviews the management of cerebral edema, elevated intracranial pressure (ICP), and cerebral herniation syndromes in neurocritical care.

RECENT FINDINGS: While corticosteroids may be effective in reducing vasogenic edema around brain tumors, they are contraindicated in traumatic cerebral edema. Mannitol and hypertonic saline use should be tailored to patient characteristics including intravascular volume status. In patients with traumatic brain injury who are comatose, elevated ICP should be managed with an algorithmic, multitiered treatment protocol to maintain an ICP of 22 mm Hg or less. Third-line ICP treatments include anesthetic agents, induced hypothermia, and decompressive craniectomy. Recent clinical trials have demonstrated that induced hypothermia and decompressive craniectomy are ineffective as early neuroprotective strategies and should be reserved for third-line management of refractory ICP elevation in severe traumatic brain injury. Monitoring for cerebral herniation should include bedside pupillometry in supratentorial space-occupying lesions and recognition of upward herniation in patients with posterior fossa lesions.

SUMMARY: Although elevated ICP, cerebral edema, and cerebral herniation are interrelated, treatments should be based on the distinct pathophysiologic process. Focal lesions resulting in brain compression are primarily managed with surgical decompression, whereas global or multifocal brain injury requires a treatment protocol that includes medical and surgical interventions.

Address correspondence to Dr Matthew A. Koenig, The Queen’s Health Systems, 1301 Punchbowl St, Neuroscience Institute QET5, Honolulu, HI 96816, mkoenig@queens.org.

RELATIONSHIP DISCLOSURE: Dr Koenig has received research/grant support as principal investigator of a study for the Hawaii Department of Health Neurotrauma Special Fund and receives publishing royalties from Rutgers University Press.

UNLABELED USE OF PRODUCTS/INVESTIGATIONAL USE DISCLOSURE: Dr Koenig reports no disclosure.

CONTINUUM AUDIO INTERVIEW AVAILABLE ONLINE

INTRODUCTION

Elevated intracranial pressure (ICP), cerebral edema, and cerebral herniation syndromes are distinct but overlapping processes in neurocritical care. Management of elevated ICP and cerebral edema is heavily dependent on the underlying mechanism and clinical context. In patients with cerebral edema, determination of whether the patient has vasogenic edema, cytotoxic edema, or hydrostatic edema is a critical first step in identifying the most effective management strategy. Determining whether elevated ICP is caused by global elevation in intracranial volume or focal injury that results in displacement of the brain is also crucial in choosing appropriate treatments. This article focuses on the etiology and treatment of cerebral edema in the neurocritical care setting, current concepts in the treatment of intracranial hypertension, and cerebral herniation syndromes.

CEREBRAL EDEMA

Cerebral edema results from the pathologic accumulation of excess water within the brain parenchyma. Vasogenic edema results from increased permeability of the blood-brain barrier with extravasation of proteins, electrolytes, and water into the parenchymal extracellular compartment. Common etiologies of vasogenic edema include intraaxial and extraaxial brain tumors and cerebral abscess. Vasogenic edema disproportionately affects subcortical white matter with relative sparing of the cerebral cortex and subcortical gray matter. Cytotoxic edema is caused by disruption of cell membranes within the brain parenchyma, resulting in water shifts from the extracellular to the intracellular compartment. The most common cause of cytotoxic edema is ischemic stroke. Less common etiologies include hepatic encephalopathy and Reye syndrome. Traumatic brain injury (TBI) and intracerebral hemorrhage (ICH) result in a combination of cytotoxic and vasogenic edema. Cytotoxic edema affects both gray matter and white matter structures, resulting in loss of cortical-subcortical distinction on imaging studies. Hydrostatic cerebral edema results from transependymal displacement of CSF from the ventricular compartment into the brain parenchyma, typically due to obstructive hydrocephalus. Cerebral edema contributes to an increase in intracranial volume. Global cerebral edema primarily results in a global rise in ICP, while focal cerebral edema can result in cerebral herniation syndromes with or without ICP elevation.

Treatment of Cerebral Edema

Treatment strategies for cerebral edema are heavily contingent on the underlying etiology and type of cerebral edema. The mainstay of treatment of hydrostatic edema due to obstructive hydrocephalus is CSF diversion, typically by placement of an external ventricular drain (EVD). Depending on the underlying etiology, vasogenic cerebral edema is treated with corticosteroids, osmotic agents, and surgical decompression. Treatment options for cytotoxic edema are much more limited. While osmotic agents have been used as a temporizing measure, the evidence for efficacy is poor, and surgical decompression may be considered in the appropriate clinical context.

CORTICOSTEROIDS

Dexamethasone has been a mainstay of treatment for peritumoral vasogenic edema for both intraaxial and extraaxial brain tumors since the 1960s. The role of corticosteroids in the treatment of vasogenic edema from cerebral abscess is less clear. Despite widespread use, few clinical trials have been conducted to determine the efficacy, optimal dose, and appropriate duration of corticosteroids for vasogenic edema. Dexamethasone is typically started at a dose of 4 mg every 6 hours, and subsequent dose adjustments are based on the clinical course. For brain tumors that are amenable to surgical resection and radiation, dexamethasone may be tapered off over several weeks. For patients with untreatable brain tumors, the dose of dexamethasone may need to be increased over time as a palliative measure. In the acute treatment of peritumoral edema, dexamethasone use should be limited to patients who have significant symptoms attributable to cerebral edema rather than focal neurologic involvement from the tumor itself. These symptoms include severe headache and depressed mental status from displacement of the brain. Caution should be applied in patients with newly discovered brain tumors awaiting tissue diagnosis when central nervous system lymphoma is a consideration; although uncommon, early initiation of corticosteroids can result in nondiagnostic biopsy specimens due to tumor necrosis. Outside of central nervous system lymphoma treatment, early initiation of corticosteroids has not been proven to alter the clinical course of brain tumors.

Corticosteroids have not demonstrated efficacy in the treatment of cytotoxic cerebral edema, and routine use of dexamethasone was not recommended in the latest American Heart Association (AHA) guideline for the management of cerebral and cerebellar infarction with swelling. Similarly, routine use of corticosteroids is not recommended for cerebral edema related to spontaneous ICH. For patients with TBI, the CRASH (Corticosteroid Randomisation After Significant Head Injury) trial was conducted to compare mortality rates of patients treated with 48 hours of methylprednisolone compared to placebo. This clinical trial showed significantly increased mortality among patients with severe TBI who were treated with corticosteroids. The 2017 Brain Trauma Foundation guidelines state, “In patients with severe TBI, high-dose methylprednisolone was associated with increased mortality and is contraindicated.”

OSMOTIC AGENTS

The mainstays of osmotic therapy in the treatment of cerebral edema in neurocritical care are mannitol and hypertonic saline. Hypertonic saline can be administered in several different concentrated solutions depending on institutional practices, ranging from 2% to 23.4%; 3% saline is commonly administered in 250 mL to 500 mL boluses either on an as-needed or standing basis (eg, 250 mL every 6 hours), while 23.4% is typically administered as a 30 mL bolus over 10 to 15 minutes. Faster administration of hypertonic saline may cause hypotension. A central line is required for administration of hypertonic saline more concentrated than 3% to avoid peripheral vascular injury. Hypertonic saline can also be administered as a continuous infusion, particularly in patients who have cerebral edema in the setting of serum hyponatremia. Most neurointensivists advise against increasing the serum sodium level above 160 mmol/L, and the safety and efficacy of iatrogenic hypernatremia beyond this value is not well studied. Frequent administration of hypertonic saline may result in hyperchloremic metabolic acidosis, which has been associated with higher mortality in patients with ICH. Buffering hypertonic saline with acetate may reduce the risk of metabolic acidosis in this circumstance.

Mannitol is a potent osmotic diuretic most commonly delivered as a 20% concentrated solution in a dose range of 0.5 g/kg IV to 2 g/kg IV either as a scheduled or as-needed bolus. Administration of mannitol requires use of an in-line filter because mannitol may crystallize, especially at lower storage temperatures. For this reason, mannitol administration is not recommended through temperature exchange catheters. Continuous infusion of mannitol is not recommended because mannitol may permeate across a disrupted blood-brain barrier and cause rebound cerebral edema. Most neurointensivists recommend routine monitoring of serum osmolality in patients being treated with frequent doses of mannitol with avoidance of routinely increasing the serum osmolality to more than 320 mOsm/kg or an osmolar gap of more than 20 mOsm/kg. The osmolar gap is the difference between the measured and calculated osmolality, where osmolality is calculated as:

However, observational studies have shown a low incidence of deleterious effects, including acute kidney injury, in patients with inadvertent elevation in osmolality to 340 mOsm/kg.

The choice of osmotic agent for a particular patient should be tailored to volume status, serum sodium concentration, and other patient-specific factors. As a rule of thumb, hypertonic saline is preferred in patients who would benefit from volume expansion (eg, patients with the combination of hypovolemic shock and cerebral edema), while mannitol is preferred in patients who would benefit from the diuretic effect.

The evidence for effectiveness of osmotic therapy in patients with cytotoxic edema from ischemic stroke is relatively weak. Because cytotoxic edema is caused by disruption of cell membrane and blood-brain barrier integrity, osmotic agents can permeate into infarcted brain tissue rather than remaining within the intravascular compartment. Osmotic agents may accumulate over time, resulting in rebound cerebral edema. Nevertheless, mannitol and hypertonic saline are often used as a temporizing measure in both ischemic stroke and ICH with mass effect as a bridge to more definitive surgical management. For patients near the period of peak edema, scheduled doses of mannitol or hypertonic saline alone may provide enough time for spontaneous resolution of cytotoxic edema. In the AHA guideline for cerebral and cerebellar infarction with swelling, there was Class IIa, Level C evidence that osmotic therapy is reasonable for patients with stroke with clinical deterioration due to cerebral edema. Routine use of osmotic agents in patients with stroke or ICH without clinical deterioration due to cerebral edema is not indicated.

INTRACRANIAL PRESSURE ELEVATION

The Monro-Kellie doctrine states that the intracranial compartment contains a fixed total volume determined by the rigid skull. The intracranial volume is determined by the relative volume of three primary compartments: blood, brain, and CSF. A transient increase in volume of one of these compartments results in a transient rise in ICP that is subsequently buffered by displacement of one of the other compartments. In normal physiology, CSF is the lowest pressure compartment and acts as the primary buffer for expanding space-occupying lesions. This concept is best evidenced by displacement of CSF from the subarachnoid space and intraventricular compartment with an enlarging brain mass. The relationship between ICP and intracranial volume is described by the property of compliance. In conditions leading to poor brain compliance, small changes in intracranial volume result in relatively large changes in ICP.

In normal physiology, as CSF is eluted from the choroid plexus into the ventricle, a transient and measurable rise in ICP occurs, which is subsequently buffered by displacement of CSF from the subarachnoid and ventricular compartments. This transient ICP elevation and buffering results in the characteristic ICP waveform, which is composed of the percussion wave (P1, cardiac systole), tidal wave (P2, brain parenchymal displacement restricted by the dura), and the dicrotic wave (P3, closure of the aortic valve).

As demonstrated in FIGURE 1-1, the CSF waveform can be used as a subjective indicator of intracranial compliance. In the setting of global cerebral edema, in which the CSF buffer is displaced from the intracranial compartment and the brain parenchyma abuts the rigid dura, further CSF production in the ventricle results in exaggeration of the P2 waveform relative to P1. This finding indicates poor intracranial compliance and suggests caution should be exercised to ensure that even relatively small changes in intracranial volume are avoided. In a state of poor compliance, physiologic changes that could be expected to increase intracranial volume include: flat head positioning, obstruction of venous return from the external jugular veins (restrictive tape or cervical collar, internal jugular vein catheter placement), hypercarbia, hyperemia, fever, pain and agitation, and increased intrathoracic or intraabdominal compartment pressure.

Slower oscillations in ICP also occur during normal and pathologic states. These Lundberg waves can be tracked by changing the time scale of physiologic ICP monitoring in order to evaluate trends over minutes to hours. Lundberg C waves can be seen in normal physiology and most likely represent interactions between the cardiac and respiratory cycles. These oscillations occur 4 to 8 times per minute and generally do not exceed 25 mm Hg. Lundberg B waves are an indicator of poor intracranial compliance. These oscillations occur 0.5 to 2 times per minute and generally do not exceed 30 mm Hg. Lundberg A waves are also called “plateau waves.” These waves are always pathologic and may be a harbinger of cerebral herniation. Lundberg A waves represent steep increases in ICP and may be as high as 40 mm Hg to 50 mm Hg and last for 5 to 10 minutes. Implicit in the discussion of Lundberg waves is that these spontaneous oscillations in ICP are self-limited and do not necessarily require urgent treatment. Although Lundberg A waves may be an indicator of impending cerebral herniation, each ICP plateau does not require treatment per se. Well-designed ICP treatment algorithms reserve interventions for sustained elevation of ICP that lasts more than 10 to 15 minutes. ICP treatment strategies that require an intervention every time the ICP exceeds 20 mm Hg will likely result in overtreatment of spontaneous ICP oscillations that would otherwise be self-limited.

Cerebral Perfusion Pressure

Another important principle of ICP management is the relationship between ICP and cerebral perfusion pressure (CPP). CPP is measured as the difference between the mean arterial pressure (MAP) and the ICP (in mm Hg), which determines the pressure gradient of cerebral perfusion as a global measure. The normal range of CPP in adults is 50 mm Hg to 70 mm Hg, but these values can be impacted by chronic hypertension, hydrocephalus, and other conditions. In the cerebral autoregulation graph depicted in FIGURE 1-2, cerebral blood flow is tightly maintained along a wide range of CPP through regulation of cerebral arterial vasoconstriction and vasodilation. This physiologic response creates a complex relationship between ICP, cerebral perfusion, and intravascular volume.

In states of cerebral hypoperfusion, the normal physiologic response is cerebral vasodilation. In patients with elevated ICP and poor intracranial compliance, cerebral vasodilation increases the volume of intracranial blood (most of which is in the venous compartment), further increasing ICP, which can paradoxically worsen cerebral ischemia by reducing CPP. In this situation, treatment with systemic vasopressors may be needed to augment the CPP to supraphysiologic values to allow subsequent cerebral vasoconstriction with reduction of intracranial blood volume and ICP. On the other hand, in brain regions with a disrupted blood-brain barrier, excessive CPP may further contribute to vasogenic edema, which will increase the parenchymal component of intracranial volume and further raise the ICP. Management of ICP in patients with regional hypoperfusion and cerebral edema, typical of patients with severe TBI, may require a trial-and-error process to determine whether hemodynamic augmentation with vasopressors results in ICP reduction or ICP elevation.

Treatment of Elevated Intracranial Pressure

The first step in ICP treatment should be to determine whether ICP-based treatment is actually justified based on the etiology of the neurologic injury. For patients with focal cerebral edema or compression of brain structures from a space-occupying lesion, ICP monitoring may not be indicated in the first place. Current guidelines do not recommend routine ICP monitoring or ICP-based treatment algorithms in patients with spontaneous ICH, brain tumors, meningitis, or ischemic stroke. ICP monitoring and treatment continues to be recommended in patients with severe TBI (initial Glasgow Coma Scale score of ≤8). Implicit in this recommendation, however, is that patients with TBI who are noncomatose do not necessarily require ICP monitoring, and ICP-based treatment may be unhelpful or detrimental in these patients. In neurocritical care, this situation commonly arises when patients with TBI who are initially comatose are subsequently recovering and develop periods of intracranial hypertension associated with pain, agitation, or other causes. In this situation, the risk of overtreatment is high, and providers should consider removing the ICP monitor. Current Brain Trauma Foundation guidelines recommend maintaining an ICP of 22 mm Hg or less and a CPP of at least 50 mm Hg to 60 mm Hg. The recent BEST:TRIP (Benchmark Evidence From South American Trials: Treatment of Intracranial Pressure) clinical trial, however, failed to show improved outcomes in patients with severe TBI treated with an ICP-based protocol compared to patients treated with a protocol based on imaging and clinical findings without ICP monitoring.

Best practice for ICP-based management is the creation and institution of a multimodality treatment algorithm that reinforces consistent, evidence-based practices for neurocritical care nurses and physicians. Although no single algorithm can account for every patient, ICP-based treatment algorithms improve the consistency of care, unburden physicians and nurses from focusing on individual treatment decisions, and improve patient outcomes. Institutional treatment protocols should be formulated by a multidisciplinary team that includes nurses, intensive care providers, neurologists, and neurosurgeons. These protocols should be reviewed and revised annually and on an ad hoc basis. Most ICP-based treatment protocols include first-line therapies (noninvasive maneuvers such as repositioning, ventilator changes, sedation, analgesia), second-line therapies (osmotic agents, hyperventilation, CSF diversion), and third-line therapies (metabolic suppression with anesthetic agents, induced hypothermia, surgical decompression). See FIGURE 1-3 for the Emergency Neurological Life Support (ENLS) ICP treatment algorithm.

ANESTHETIC AGENTS AND METABOLIC SUPPRESSION

Although propofol and high-dose benzodiazepine infusions can be used for metabolic suppression, the mainstay for induction of “pharmacologic coma” in refractory intracranial hypertension is pentobarbital. Although strong evidence indicates that pentobarbital lowers ICP, there is a paucity of high-quality evidence for improvement of neurologic outcomes. Pentobarbital is typically delivered as a 20 mg/kg IV bolus followed by continuous infusion at 0.5 mg/kg/h to 5 mg/kg/h. Although the primary titration parameter should be targeted to ICP control, EEG monitoring with at least a limited electrode montage should also be undertaken to trend the suppression ratio. Pentobarbital is typically titrated to achieve burst-suppression anesthesia with a suppression ratio of 0.8 to 0.9, where suppression ratio is the ratio of the duration of EEG suppression to the duration of EEG bursts. Suppression ratio may be monitored with a commercially available quantitative EEG product typically used to measure depth of anesthesia that also displays raw EEG on bedside monitors.

Pentobarbital includes propylene glycol, which can accumulate to cause fatal lactic acidosis with elevated osmolar gap, acute kidney failure, and circulatory collapse, so the osmolar gap should be monitored in patients being treated with pentobarbital. Other toxicities include suppression of gastrointestinal motility, immunosuppression, bone marrow suppression, and distributive shock. The half-life of pentobarbital ranges from 15 to 50 hours, so patients may not recover consciousness for several days after medication discontinuation, and caution should be exercised in brain death declaration for patients who were treated with pentobarbital. Very high-dose pentobarbital can mimic every clinical sign of brain death, including bilateral fixed and dilated pupils and diabetes insipidus. After prolonged use of pentobarbital, rapid discontinuation may produce withdrawal symptoms including seizures and rebound ICP elevation, so phenobarbital can be administered and tapered during pentobarbital weaning.

INDUCED HYPOTHERMIA

Ample evidence indicates induced hypothermia lowers ICP, but the evidence that it improves outcomes is lacking. The recent Eurotherm3235 trial tested whether an ICP treatment protocol that included early induction of hypothermia resulted in improved outcomes and reduced the need for other third-line therapies in patients with severe TBI. Although fewer ICP-directed interventions were required in the hypothermia group, neurologic outcomes were actually worse. This clinical trial demonstrated that early induced hypothermia does not improve outcomes in TBI as a neuroprotective strategy, but hypothermia continues to be employed as a third-line therapy for refractory ICP elevation. Most ICP treatment protocols that include hypothermia target a core temperature of 32°C to 34°C (89.6°F to 93.2°F) using either surface cooling or intravascular cooling devices with a continuous automated feedback mechanism. Successful induction and maintenance of hypothermia, especially in patients not concurrently treated with anesthetic agents, requires institution of a multimodality shivering protocol. Shivering causes failure of temperature maintenance and paradoxical ICP elevation from a combination of cerebral hypermetabolism and systemic hypercarbia. Antishivering protocols typically include a combination of surface counterwarming, magnesium infusion, buspirone, meperidine, and sedative infusion. During the induction phase, hypothermia can produce peripheral vasoconstriction leading to skin ischemia, severe hypokalemia, and diuresis from shunting blood flow to the kidneys. Electrolytes should be monitored frequently during the hypothermia induction and rewarming phases. Rewarming should occur in a tightly controlled fashion to avoid severe rebound hyperkalemia and distributive shock from peripheral vasodilation. For this reason, rewarming should be undertaken at a rate of 0.1°C (32.2°F) per hour or slower in patients who have been cooled for long intervals.

DECOMPRESSIVE SURGERY

For patients with ICP elevation due to obstructive hydrocephalus, placement of an EVD for CSF diversion is considered a first-line treatment. For patients with elevated ICP due to a focal compressive brain lesion, decompressive surgery is considered a first-line treatment rather than global ICP-directed treatments (CASE 1-1). For patients with elevated ICP due to global or multifocal brain injuries such as severe TBI, however, decompressive craniectomy should be considered as a third-line treatment for refractory ICP elevation. The appropriate sequence of ICP treatment protocols has come into better focus with the recent DECRA (Decompressive Craniectomy in Diffuse Traumatic Brain Injury) and RESCUEicp (Randomised Evaluation of Surgery With Craniectomy for Uncontrollable Elevation of Intracranial Pressure) clinical trials. DECRA tested whether early bifrontal decompressive craniectomy would improve outcomes and lessen the need for other ICP-directed interventions in patients with severe TBI. Although surgical decompression lowered ICP, patients had worse neurologic outcomes in the early surgery arm. The RESCUEicp trial randomly selected patients with severe TBI with refractory ICP elevation to undergo decompressive craniectomy (unilateral or bilateral) as a third-line treatment or continued medical management with anesthetic agents. This clinical trial demonstrated modest benefits to decompressive craniectomy with higher odds of survival and lower odds of severe disability, but there was no effect on the odds of good recovery, and more patients survived in the vegetative state in the surgical arm. These study results clarify that decompressive surgery should be reserved for third-line management of refractory ICP elevation, and careful discussion of expected outcomes should be undertaken with surrogate decision makers, including the higher chance of survival in a vegetative state.

CASE 1-1

A 47-year-old man presented to the neurocritical care unit because of a gunshot wound to the left occipital lobe that was initially managed with surgical debridement and placement of an intracranial pressure (ICP) monitor. The patient initially had purposeful movements of the left side, but he had aphasia and right hemiplegia.

On hospital day 3, he developed refractory ICP elevation that was treated with sedation and analgesia followed by hypertonic saline boluses according to the institutional ICP management protocol. Despite these interventions, he developed a sustained ICP of more than 30 mm Hg and a worsening neurologic examination with extensor posturing. Repeat head CT demonstrated an evolving infarct of the left occipital lobe with hemorrhagic transformation and 8-mm midline shift. He was treated with mannitol 1 g/kg IV as a temporizing measure and then was taken for decompressive craniectomy. After surgery, the ICP normalized, and he began to have purposeful movements of the left side again. The postoperative head CT is shown (FIGURE 1-4).

COMMENT

This case illustrates the importance of tailoring interventions for ICP elevation to the underlying pathophysiology. In this case, rather than escalating to third-line medical management of ICP elevation with pentobarbital coma or induced hypothermia, the team recognized that ICP elevation was caused by an enlarging focal brain lesion that required surgical decompression.

CEREBRAL HERNIATION SYNDROMES

Cerebral herniation occurs when brain structures are displaced into an adjacent compartment with compression of surrounding neurologic structures. Cerebral herniation typically occurs when enlarging space-occupying lesions result in displacement or expansion of brain tissue. Herniation is often but not universally associated with elevated ICP. In some cases, global ICP values are normal, but compartmentalized ICP elevation contributes to pressure gradients that displace brain tissue. For this reason, global ICP monitoring is not always a sensitive indicator of impending cerebral herniation. This is especially true of posterior fossa lesions, which may result in upward herniation with little change in ICP. Clinical monitoring for impending herniation relies on serial neurologic examinations.

Transtentorial (Uncal) Herniation

Transtentorial herniation occurs when an expanding supratentorial mass lesion displaces the medial temporal lobe (uncus) through the tentorial incisura with compression of the ipsilateral oculomotor nerve followed by the midbrain. The clinical hallmark of transtentorial herniation is ipsilateral dilation of the pupil and, less commonly, deviation of the eye laterally and inferiorly. As herniation progresses, patients develop extensor posturing and depressed mental status due to dysfunction of the corticospinal tract and reticular activating system in the midbrain. Bilateral transtentorial herniation, also known as central herniation, typically occurs with global cerebral edema and presents with coma, extensor posturing, and bilateral fixed and dilated pupils. With expanding extraaxial lesions (typically subdural hematoma), a characteristic pattern of ipsilateral pupil dilation and ipsilateral hemiplegia may develop. This so-called Kernohan notch phenomenon occurs with lateral displacement of the midbrain such that the contralateral cerebral peduncle is compressed against the tentorial edge at Kernohan notch with relative sparing of the ipsilateral cerebral peduncle. For patients with supratentorial space-occupying lesions, the side of the pupil dilation is ipsilateral to the source of mass effect, while the side of hemiplegia may be either ipsilateral or contralateral (ie, “the pupil doesn’t lie”).

Pupil dilation is the sine qua non manifestation of transtentorial herniation, so clinical monitoring for pupillary abnormalities is the hallmark of bedside neuromonitoring in patients with supratentorial mass lesions at risk for herniation. ICP elevation is often a late finding. Prior to dilation of the pupil, however, the velocity of pupillary constriction diminishes with early compression of the oculomotor nerve. This can be monitored with serial bedside pupillometry with quantitative measurement of the pupillary constriction velocity using commercially available bedside monitors. Observational studies have demonstrated that the diminution of the constriction velocity may precede frank clinical signs of transtentorial herniation by hours. In some cases, early recognition of impending herniation may provide time for temporizing medical interventions and definitive surgical decompression.

Transtentorial herniation is caused by focal displacement and compression of brain structures rather than global elevation of ICP or cerebral edema. For this reason, therapies directed purely at reduction of cerebral edema (eg, corticosteroids) or global reduction in ICP (eg, anesthetic agents, hypothermia) are ineffective in reversing clinical signs of transtentorial herniation. Without rapid treatment to reverse clinical signs of transtentorial herniation, the natural history is progression to central herniation, which is nearly universally fatal. Although observational studies have demonstrated restoration of the pupillary light reflex—and in some cases survival—after treatment with hypertonic saline and mannitol, radiographic evidence of reversal of herniation is lacking. For this reason, osmotic therapy and hyperventilation should be considered temporizing measures to reverse clinical evidence of transtentorial herniation prior to definitive surgical decompression. Decompressive surgery requires either excision of the space-occupying lesion (extraaxial hematoma or temporal lobectomy) or decompressive craniectomy.

Transtentorial herniation used to be considered a universally fatal event. Depending on the etiology of herniation and availability of definitive surgical treatment, patients with unilateral transtentorial herniation have been reported to survive and recover independent function in some series. Even if clinical signs of herniation are reversed with medical or surgical interventions, three common sequelae of the herniation event may occur. Compression of venous drainage from the central midbrain and pons may lead to venous congestion and hemorrhagic infarction within the medial brainstem structures, which is termed Duret hemorrhage. Compression of the contralateral posterior cerebral artery against the tentorial edge may cause posterior cerebral artery territory stroke. Compression of the ipsilateral anterior cerebral artery against the inferior edge of the falx cerebri due to associated subfalcine herniation may cause anterior cerebral artery territory stroke.

Posterior Fossa Herniation Syndromes

Expanding posterior fossa lesions such as cerebellar stroke can result in compression of the fourth ventricle with acute obstructive hydrocephalus, compression of the brainstem and cranial nerves, and two distinct herniation syndromes.

Cerebellar tonsillar herniation occurs with downward displacement of the cerebellar tonsils through the foramen magnum with compression of the cervicomedullary junction. This herniation syndrome may not be accompanied by elevated ICP or pupillary changes until obstructive hydrocephalus develops as a late event. For this reason, tonsillar herniation is commonly missed with neuromonitoring that is focused on ICP and pupillary changes. Tonsillar herniation typically causes quadriparesis due to compression of the pyramidal decussation, respiratory insufficiency due to bulbar dysfunction and compression of the upper cervical spinal cord, and depressed level of consciousness. As with transtentorial herniation, osmotic therapy and hyperventilation are temporizing measures, but definitive treatment requires surgical decompression, either decompressive suboccipital craniectomy, partial cerebellectomy, or hematoma evacuation.

Upward herniation occurs with upward displacement of the cerebellum through the tentorial incisura with compression of the dorsal midbrain. This uncommon herniation syndrome typically occurs in patients with expanding posterior fossa mass lesions in combination with an EVD. Excessive CSF diversion from the lateral ventricle creates a pressure gradient resulting in upward displacement of the cerebellum. Patients typically present with new-onset vertical gaze palsy and depressed level of consciousness. Treatment requires clamping the EVD in addition to treatment with osmotic agents and hyperventilation. As with tonsillar herniation, upward herniation is not commonly associated with pupil dilation. Because of the mechanism of herniation, the ICP is typically normal or low rather than elevated. For these reasons, diagnosis of upward herniation is often delayed or missed (CASE 1-2). Because of the risk of upward herniation, current AHA guidelines recommend against placement of an EVD as the primary intervention for obstructive hydrocephalus caused by cerebellar strokes. The current recommendation is to consider decompressive surgery prior to EVD insertion in this setting.

CASE 1-2

A 55-year-old man presented to the emergency department after he fell down a flight of stairs, during which he struck his head on the pavement and experienced brief loss of consciousness. The initial head CT showed convexity subarachnoid hemorrhage, small bifrontal cerebral contusions, and a 20-mL contusion of the cerebellar vermis. He was admitted to the neurocritical care unit.

On hospital day 2, he became more lethargic with new onset of quadriparesis and acute respiratory insufficiency requiring intubation. Repeat head CT (FIGURE 1-5) demonstrated increasing brainstem compression, cerebellar tonsillar herniation, and acute obstructive hydrocephalus. The neurosurgeon placed an external ventricular drain (EVD), but the patient had no significant clinical improvement after CSF diversion. The patient was subsequently noted to be obtunded with vertical ophthalmoplegia characterized by bilateral inferior-medial eye deviation (“sunset eyes”). Suspecting upward herniation, the neurocritical care team clamped the EVD, and the patient was taken for decompressive suboccipital craniectomy. He had gradual neurologic improvement, and the EVD was removed during the subsequent hospital course.

COMMENT

This case demonstrates the clinical presentation and treatment of posterior fossa herniation syndrome, which is distinct from the more commonly recognized transtentorial herniation syndrome. Because upward herniation is uncommon and the progression is insidious, signs are often overlooked. Current American Heart Association guidelines recommend surgical decompression prior to or concomitant with placement of an EVD for posterior fossa lesions with mass effect in order to lower the chances of upward herniation. This case illustrates the potential to worsen upward herniation when CSF diversion is undertaken without surgical decompression.

CONCLUSION

Cerebral edema, elevated ICP, and cerebral herniation syndromes are a major cause of morbidity and mortality in neurocritical care. Although these conditions frequently overlap, they are distinct pathophysiologic entities that require tailored neurologic monitoring and treatment. For cerebral edema, both the mechanism and context are important to determine best practices for treatment. While vasogenic edema surrounding brain tumors may respond to corticosteroids, this same treatment strategy is contraindicated in treatment of cerebral edema from TBI. Our understanding of ICP as a global measure and treatment target is changing. Management strategies are typically directed to global reduction of ICP, including osmotic therapies, metabolic suppression, and hypothermia, whereas in many focal neurologic conditions, ICP can be compartmentalized and contribute to displacement of brain structures. In these situations, surgical decompression should be the mainstay of treatment rather than global ICP reduction strategies.

Although many observational studies in TBI have shown that sustained ICP elevation is associated with poor outcomes, a recent clinical trial showed that ICP-targeted management resulted in outcomes that were no better than management based on imaging and clinical examination, a result that could almost be considered heretical to the last few decades of TBI management. In addition, improved understanding of the physiology of ICP-volume relationships and cerebral autoregulation have demonstrated that acceptable ICP thresholds at the level of an individual patient are not as clear-cut as we imagined. ICP thresholds may be best understood in the context of accompanying brain perfusion and tissue oxygenation.

Monitoring for and treating cerebral herniation syndromes have also evolved over time. What was once considered a uniformly fatal event may now be reversible in the right clinical context. Routine bedside pupillometry in the neurocritical care setting has demonstrated the ability to detect impending transtentorial herniation and may buy time for definitive surgical management in some cases.

KEY POINTS

  • Corticosteroids are ineffective for the treatment of cytotoxic edema and are contraindicated in the treatment of patients with severe traumatic brain injury.
  • Frequent administration of hypertonic saline may cause hyperchloremic metabolic acidosis, which is associated with higher mortality in neurocritical care. Buffering hypertonic solutions with acetate may lower the chance of developing hyperchloremic metabolic acidosis.
  • Serum osmolality should be monitored in patients treated with frequent doses of mannitol. Mannitol should be held when serum osmolality exceeds 320 mOsm/kg to 340 mOsm/kg or when the osmolar gap exceeds 20 mOsm/kg.
  • Osmotic agents can be used as a temporizing measure to treat mass effect from cytotoxic edema related to stroke and intracerebral hemorrhage, but evidence for efficacy is weak. These lesions may require surgical decompression.
  • In states of poor intracranial compliance due to global cerebral edema, small changes in intracranial volume related to flat head position, hyperemia, hypercarbia, fever, or pain may result in exaggerated increases in intracranial pressure.
  • Spontaneous oscillations in intracranial pressure called Lundberg A and B waves may cause self-limited increases in intracranial pressure that last several minutes. These oscillations are an indicator of poor intracranial compliance, but they typically resolve spontaneously without treatment.
  • Augmentation of cerebral perfusion pressure with systemic vasopressors may lower intracranial pressure by causing reflex cerebral vasoconstriction, thereby lowering the intracranial volume of blood. Excessive cerebral perfusion, however, may contribute to vasogenic edema in regions with a disrupted blood-brain barrier.
  • Current Brain Trauma Foundation guidelines recommend maintaining an intracranial pressure of 22 mm Hg or less and a cerebral perfusion pressure of at least 50 mm Hg to 60 mm Hg in patients with severe traumatic brain injury.
  • A recent clinical trial of an intracranial pressure–based treatment protocol for severe traumatic brain injury compared to a treatment protocol based on clinical examination and imaging without intracranial pressure monitoring showed no difference in outcomes between the two groups.
  • Pentobarbital infusion for intracranial pressure reduction can result in severe medication side effects such as propylene glycol toxicity, immunosuppression, impaired gastrointestinal motility, and distributive shock.
  • A recent clinical trial of early induced hypothermia as a neuroprotective strategy in severe traumatic brain injury showed reduction of intracranial pressure–directed interventions, but neurologic outcomes were worse in patients treated with hypothermia.
  • Induced hypothermia continues to be a useful third-line intervention for refractory intracranial pressure elevation, but an effective hypothermia protocol that includes multimodality treatment of shivering is required.
  • A recent clinical trial demonstrated that decompressive craniectomy for refractory intracranial pressure elevation in severe traumatic brain injury improves survival and reduces the chance of severe disability, but more patients survived in a vegetative state compared to medical management alone.
  • With transtentorial herniation, the pupil dilation is ipsilateral to the mass lesion, but hemiplegia may be contralateral because of direct corticospinal tract involvement or ipsilateral because of compression of the contralateral cerebral peduncle against the tentorial edge (Kernohan notch phenomenon).
  • Serial bedside quantitative pupillometry may detect reduction in the pupillary constriction velocity hours prior to frank clinical signs of transtentorial herniation, which may afford time for preemptive treatment.
  • Although patients can survive and recover from transtentorial herniation in some cases, the sequelae of herniation can include Duret brainstem hemorrhage, ipsilateral anterior cerebral artery stroke, and contralateral posterior cerebral artery stroke.
  • Posterior fossa mass lesions can cause cerebellar tonsillar herniation or upward herniation of the cerebellum through the tentorial incisura, which may not be accompanied by intracranial pressure elevation or pupillary changes.

REFERENCES

1. Wijdicks EF, Sheth KN, Carter BS, et al. Recommendations for the management of cerebral and cerebellar infarction with swelling: a statement for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2014;45(4):1222–1238. doi:10.1161/01.str.0000441965.15164.d6.
2. Ryken TC, McDermott M, Robinson PD, et al. The role of steroids in the management of brain metastases: a systematic review and evidence-based clinical practice guideline. J Neurooncol 2010;96(1):103–114. doi:10.1007/s11060-009-0057-4.
3. Binnahil M, Au K, Lu JQ, et al. The influence of corticosteroids on diagnostic accuracy of biopsy for primary central nervous system lymphoma. Can J Neurol Sci 2016;43(5):721–725. doi:10.1017/cjn.2016.255.
4. Hemphill JC 3rd, Greenberg SM, Anderson CS, et al. Guidelines for the management of spontaneous intracerebral hemorrhage: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke 2015;46(7):2032–2060. doi:10.1161/STR.0000000000000069.
5. Roberts I, Yates D, Sandercock P, et al. Effect of intravenous corticosteroids on death within 14 days in 10008 adults with clinically significant head injury (MRC CRASH trial): randomised placebo-controlled trial. Lancet 2004;364(9442):1321–1328. doi:10.1016/S0140-6736(04)17188-2.
6. Carney N, Totten AM, O’Reilly C, et al. Guidelines for the management of severe traumatic brain injury, 4th edition. Neurosurgery 2017;1(80):6–15. doi:10.1227/NEU.0000000000001432.
7. Ryu JH, Kahle KT, Sheth SA, et al. Induced and sustained hypernatremia for the prevention and treatment of cerebral edema following brain injury. Neurocrit Care 2013;19(2):222–231. doi:10.1007/s12028-013-9824-3.
8. Riha HM, Erdman MJ, Vandigo JE, et al. Impact of moderate hyperchloremia on clinical outcomes in intracerebral hemorrhage patients treated with continuous infusion hypertonic saline: a pilot study. Crit Care Med 2017;45(9):e947–e953. doi:10.1097/CCM.0000000000002522.
9. Palma L, Bruni G, Fiaschi AI, Mariottini A. Passage of mannitol into the brain around gliomas: a potential cause of rebound phenomenon. A study of 21 patients. J Neurosurg Sci 2006;50(3):63–66.
10. Gondim Fde A, Aiyagari V, Shackleford A, Diringer MN. Osmolality not predictive of mannitol-induced acute renal insufficiency. J Neurosurg 2005;103(3):444–447. doi:10.3171/jns.2005.103.3.0444.
11. Misra UK, Kalita J, Ranjan P, Mandal SK. Mannitol in intracerebral hemorrhage: a randomized controlled study. J Neurol Sci 2005;234(1):41–45. doi:10.1016/j.jns.2005.03.038.
12. Oshio K, Onodera H, Uchida M, et al. Assessment of brain compliance using ICP waveform analysis in water intoxication rat model. Acta Neurochir Suppl 2013;118:219–221. doi:10.1007/978-3-7091-1434-6_41.
13. Kasprowicz M, Lalou DA, Czosnyka M, et al. Intracranial pressure, its components and cerebrospinal fluid pressure-volume compensation. Acta Neurol Scand 2016;134(3):168–180. doi:10.1111/ane.12541.
14. Castellani G, Zweifel C, Kim DJ, et al. Plateau waves in head injured patients requiring neurocritical care. Neurocrit Care 2009;11(2):143–150. doi:10.1007/s12028-009-9235-7.
15. Bohman LE, Heuer GG, Macyszyn L, et al. Medical management of compromised brain oxygen in patients with severe traumatic brain injury. Neurocrit Care 2011;14(3):361–369. doi:10.1007/s12028-011-9526-7.
16. Chesnut RM, Temkin N, Carney N, et al. A trial of intracranial-pressure monitoring in traumatic brain injury. N Engl J Med 2012;367(26):2471–2481. doi:10.1056/NEJMoa1207363.
17. Clayton TJ, Nelson RJ, Manara AR. Reduction in mortality from severe head injury following introduction of a protocol for intensive care management. Br J Anaesth 2004;93(6):761–767. doi:10.1093/bja/aeh249.
18. Stevens RD, Shoykhet M, Cadena R. Emergency neurological life support: intracranial hypertension and herniation. Neurocrit Care 2015;23(suppl 2):S76–S82. doi:10.1007/s12028-015-0168-z.
19. Roberts I, Sydenham E. Barbiturates for acute traumatic brain injury. Cochrane Database Syst Rev 2012;12:CD000033. doi:10.1002/14651858.CD000033.pub2.
20. Bledsoe KA, Kramer AH. Propylene glycol toxicity complicating use of barbiturate coma. Neurocrit Care 2008;9(1):122–124. doi:10.1007/s12028-008-9065-z.
21. Andrews PJ, Sinclair HL, Rodriguez A, et al. Hypothermia for intracranial hypertension after traumatic brain injury. N Engl J Med 2015;373(25):2403–2412. doi:10.1056/NEJMoa1507581.
22. Urbano LA, Oddo M. Therapeutic hypothermia for traumatic brain injury. Curr Neurol Neurosci Rep 2012;12(5):580–591. doi:10.1007/s11910-012-0304-5.
23. Choi HA, Ko SB, Presciutti M, et al. Prevention of shivering during therapeutic temperature modulation: the Columbia anti-shivering protocol. Neurocrit Care 2011;14(3):389–394. doi:10.1007/s12028-010-9474-7.
24. Soeholm H, Kirkegaard H. Serum potassium changes during therapeutic hypothermia after out-of-hospital cardiac arrest—should it be treated? Ther Hypothermia Temp Manag 2012;2(1):30–36. doi:10.1089/ther.2012.0004.
25. Cooper DJ, Rosenfeld JV, Murray L, et al. Decompressive craniectomy in diffuse traumatic brain injury. N Engl J Med 2011;364(16):1493–1502. doi:10.1056/NEJMoa1102077.
26. Hutchinson PJ, Kolias AG, Timofeev IS, et al. Trial of decompressive craniectomy for traumatic intracranial hypertension. N Engl J Med 2016;375(12):1119–1130. doi:10.1056/NEJMoa1605215.
27. Yoo WK, Kim DS, Jang SH, et al. Kernohan’s notch phenomenon demonstrated by diffusion tensor imaging and transcranial magnetic stimulation. J Neurol Neurosurg Psychiatry 2008;79(11):1295–1297. doi:10.1136/jnnp.2007.138131.
28. Zafar SF, Suarez JI. Automated pupillometer for monitoring the critically ill patient: a critical appraisal. J Crit Care 2014;29(4):599–603. doi:10.1016/j.jcrc.2014.01.012.
29. Manley GT, Larson MD. Infrared pupillometry during uncal herniation. J Neurosurg Anesthesiol 2002;14(3):223–228.
30. Qureshi AI, Suarez JI, Bhardwaj A, et al. Use of hypertonic (3%) saline/acetate infusion in the treatment of cerebral edema: effect on intracranial pressure and lateral displacement of the brain. Crit Care Med 1998;26(3):440–446.
31. Koenig MA, Bryan M, Lewin JL 3rd, et al. Reversal of transtentorial herniation with hypertonic saline. Neurology 2008;70(13):1023–1029. doi:10.1212/01.wnl.0000304042.05557.60.
32. Ishizaka S, Shimizu T, Ryu N. Dramatic recovery after severe descending transtentorial herniation-induced Duret haemorrhage: a case report and review of literature. Brain Inj 2014;28(3):374–377. doi:10.3109/02699052.2013.864422.
33. Namura S, Kang Y, Matsuda I, Kamijyo Y. Magnetic resonance imaging of sequelae of temporal lobe herniation secondary to traumatic acute subdural hematoma: Kernohan’s notch and posterior cerebral artery territory infarctions contralateral to the supratentorial lesion—case report. Neurol Med Chir (Tokyo) 1997;37(1):32–35.
34. St Louis EK, Wijdicks EF, Li H. Predicting neurologic deterioration in patients with cerebellar hematomas. Neurology 1998;51(5):1364–1369. doi:10.1212/WNL.51.5.1364.
35. Kirollos RW, Tyagi AK, Ross SA, et al. Management of spontaneous cerebellar hematomas: a prospective treatment protocol. Neurosurgery 2001;49(6):1378–1386; discussion 1386–1387. doi:10.1097/00006123-200112000-00015.
36. Kase CS, Wolf PA. Cerebellar infarction: upward transtentorial herniation after ventriculostomy. Stroke 1993;24(7):1096–1098.
37. Wang CY, Chee CP, Delilkan AE. Upward transtentorial herniation of posterior fossa structures. Eur J Anaesthesiol 1991;8(6):469–470.
38. Gurol ME, St Louis EK. Treatment of cerebellar masses. Curr Treat Options Neurol 2008;10(2):138–150.
© 2018 American Academy of Neurology