Chapter 25. Management of Acute Intracranial Hypertension

The cranial vault is a rigid structure containing brain, blood, and cerebrospinal fluid (CSF). According to the Monro–Kellie doctrine, the volume of this chamber is unchangeable, and any addition of contents must be matched by a displacement of volume elsewhere. The goal of this chapter is to briefly outline the pathophysiologic processes that result in volume shifts in the cranium and measures that can be taken to identify and treat these conditions.

Intracranial pressure (ICP) is defined as the pressure exerted on the dura mater by the intracranial contents.1 It comprises the sum of three partial pressures:

Any increase in the partial pressure of one compartment will cause a decrease in the pressure of another to maintain a constant ICP. The change in volume divided by a change in ICP is defined as intracranial compliance. Initially, the addition of volume is easily accommodated in the vault without a corresponding increase in pressure. Once this “compensatory reserve”2 is exhausted, the pressure rises rapidly in response to an increase in volume (Figure 25-1).

Normal ICP ranges between 5 and 15 mm Hg or 7.5 and 20 cm H2O.3

An increase in ICP can jeopardize the cerebral perfusion pressure (CPP) that is defined as follows:

where MAP is the mean arterial pressure, and hence the cerebral blood flow (CBF) = CPP/cerebral vascular resistance (CVR).

Although transient ICP elevations of up to 100 cm H2O have been tolerated by the human brain under experimental conditions,4 sustained ICP values above 20 mm Hg are associated with worse outcomes in brain trauma patients. CPP is less predictive of neurologic outcome as long as it is maintained above 60 mm Hg.5

In normal physiologic states, CBF remains stable or “autoregulated” over a wide range of CPP through cerebral arteriolar vasodilation and vasoconstriction.6 CPP and ICP are clinical surrogates for CBF and therefore are utilized as clinical diagnostic and therapeutic indices. CVR is increased or decreased based on alterations in CPP when autoregulation is intact. As CBF decreases beyond the limits of autoregulation, the brain increases its oxygen extraction fraction (OEF) to compensate for reduced blood flow. In the neurologically injured brain, the concept of autoregulation may be disrupted and therefore the normal compensatory measures may not exist. Any therapeutic measures directed toward improving the injured brain or other organ systems should consider this pathophysiologic process/concept prior to instituting therapy.

The clinical presentation of elevated ICP depends on the etiology and varies in reliability. Signs include somnolence, papilledema, a symptom complex of headache, nausea/vomiting, and blurry/double vision, or a complex of bradycardia, irregular respirations, and widened pulse pressure (Cushing's triad).7 Cushing proposed that these findings in the presence of profound intracranial hypertension were a sign of medullary ischemia. However, this constellation can be seen anytime there is distortion of the brainstem even in the setting of normal ICP8 and should be considered an ominous trend without a specific clinical correlation to intracranial hypertension.

Acute increases in ICP such as those due to epidural hematomas, subarachnoid hemorrhage (SAH), or severe brain trauma usually present with a more global impairment in cerebral function such as low Glasgow Coma Scale (GCS) score, headache, nausea, and vomiting. Venous bleeds, subdural hematomas, brain tumors, and malignant strokes are more likely to present as focal neurologic deficits progressing to one of the herniation syndromes and elevated ICP. In these patients, it is important to monitor for worsening paresis, cranial nerve palsies (especially third and sixth nerves), and pupillary changes.

Any patient with suspected intracranial hypertension should undergo emergent neuroimaging. Concerning findings include:

1. Presence of acute intraventricular, subarachnoid, epidural, or subdural blood

2. Obliteration of the third ventricle or basal cisterns9

3. Dilation of the contralateral temporal horn10

4. Obstructive hydrocephalus with enlarged lateral ventricles and transependymal flow3

5. Midline shift

6. Diffuse or focal cerebral edema—loss of gray–white junction, large-vessel ischemia, or large areas of vasogenic edema resulting in sulcal effacement

The shifting of intracranial contents from one intracranial compartment to the next because of mass effect is called herniation.11 There are different types of herniation syndromes: (1) transtentorial herniation; (2) central herniation; (3) tonsillar herniation; and (4) subfalcine herniation (Figure 25-2).

ICP should be measured whenever elevation of ICP is suspected in patients who will benefit from the procedure. There is insufficient level 1 evidence to support a standard for ICP monitoring or to confirm that it improves outcomes. According to the most recent traumatic brain injury (TBI) guidelines,12 there is level II evidence that all salvageable patients with a GCS score of 3–8 and an abnormal computed tomography (CT) scan should have some form of ICP monitoring. Level III evidence supports ICP monitoring in all patients with severe TBI and a normal CT who have two of the following: age >40 years, unilateral or bilateral motor posturing, or a systolic blood pressure <90 mm Hg. Beyond these guidelines, indications are less well defined, although ICP monitoring is used in poor-grade SAH, intraventricular hemorrhage, intraparenchymal hemorrhage, meningitis, acute liver failure, hydrocephalus, etc.

The gold standard for ICP monitoring is direct measurement in the lateral ventricle (Figure 25-3). This allows for continuous monitoring of the ICP and CSF drainage for ICP control. External ventricular drains (EVDs) are inserted into the ventricle through a burr hole. They are connected to a transducer and a drainage bag, which is positioned at a set level above the tragus to maintain the desired ICP. It is important to remember that the height of the collecting bag relative to the tragus is often measured in cm H2O, while ICP is measured in millimeters of mercury. The biggest complications of EVDs are malfunction and infection. Infection rates from 5% to 20% have been documented in the literature13 and are related to surgical technique, duration of EVD placement, frequency of manipulation, and catheter flushing. In general, more than three placement attempts and flushing more than twice for malfunction should be avoided.1 Neither routine catheter exchange nor the use of prophylactic antibiotics is recommended to reduce infections.14

Intraparenchymal monitors are less invasive and independent of head position. They cannot be rezeroed once inserted, although the newer models have less drift, making this less of an issue.15 These devices measure pressure in the anatomic compartment that they are placed in, which may not be an accurate assessment of global (ventricular) ICP.

Subarachnoid bolts are hollow saline-filled bolts that are screwed into the burr hole. The fluid in the lumen is continuous with the CSF in the subarachnoid space and the transmitted pressure is considered the ICP. The main advantages of this device are the ease of insertion and the low risk of infection and bleeding. However, they do not allow for CSF drainage, are less accurate than EVDs, and tend to get occluded by swollen brain.16

Epidural devices are fiberoptic catheters that are placed in the space between the skull and dura. Although they have a low risk of infection and bleeding, they are often inaccurate.

Once the diagnosis of intracranial hypertension has been established, the treatment can be directed to the cause: CSF drainage for hydrocephalus, steroids, and resection for intracranial tumors and craniectomy for strokes. General principles of ICP management continue until definitive treatment can be implemented, or if the patient is not a candidate for any of the above. As with all emergencies, the airway, breathing, and circulation must be stabilized before any further steps are taken.

Position

A change in head position from 0° to 60° is associated with a significant decrease in ICP17 as it improves venous return and decreases CSF hydrostatic pressure. Unfortunately, this is coupled with a drop in MAP and CPP18 that may adversely affect patients with impaired cerebral autoregulation. A midline head position ensures that both jugular veins are open and draining. Special attention should be given to collars and endotracheal tube (ETT) holders, which may be constrictive and impair venous return.

Hyperventilation

A decrease in Pco2 effectively decreases the ICP19 by causing cerebral arteriolar constriction and reducing cerebral blood volume. The effect generally lasts for less than 24 hours, and prolonged hyperventilation should be avoided. Level II evidence discourages maintaining Pco2 levels below 25 mm Hg in TBI patients given the risk for global ischemia.20 In general, eucapnia should be maintained and hyperventilation should be avoided or only used as a temporizing measure.

Hemodynamics

As the ICP rises, the MAP reflexively increases to maintain CPP. More protocols are now incorporating CPP-targeted therapies with a lower limit ≥60 mm Hg. This allows for less use of vasopressors and fewer pulmonary complications than ICP-driven management of brain injury.21

Hyperosmolar Therapy

Mannitol is the most commonly used osmotic agent in the treatment of intracranial hypertension. It is usually dosed as a bolus infusion at 0.25–1.0 g/kg body weight. It does not cross the blood–brain barrier (BBB) in uninjured brain tissue but can in areas where the BBB is compromised, creating a reverse osmotic effect. Mannitol acutely expands the intravascular volume, increasing CBF. This in turn enhances oxygen delivery to the brain and causes vasoconstriction in areas of the brain where autoregulation is intact, resulting in a decrease in ICP. Mannitol also creates an osmotic gradient between the cells and plasma, resulting in reduction of intracerebral volume and a drop in ICP. There is subsequent urinary osmotic diuresis, which should be replaced by intravenous fluids to avoid dehydration, hypotension, and renal failure. These side effects are more common when the drug is dosed frequently, continuously, or in large volumes, especially at serum osmolarities greater than 320 mOsm.22 Renal toxicity is one of the primary concerns with usage of mannitol especially when given in scheduled dosing regimens or in continuous intravenous formats. It is related to mannitol accumulation, and therefore osmolar gap should be calculated in multiple-dose formats:

There have been no randomized controlled trials to prove mannitol's superiority over other agents or an improvement in outcomes with its use.

Hypertonic saline (HS) reduces ICP by creating a hyperosmolar gradient across the BBB. ICP reduction has been noted to last for ≤2 hours but may be maintained for longer with a continuous infusion.23 Side effects include electrolyte abnormalities, heart failure, and phlebitis. In a recent comparison of equiosmolar mannitol and 7.5% HS, both equally reduced ICP but mannitol had the added benefit of improving CPP. Since then, a series of patients refractory to mannitol were treated with 7.5% HS with marked reduction in ICP and improvement in brain tissue oxygen tension (PbtO2) and cerebral and systemic hemodynamics.24 Thirty- and 60-mL boluses over 15 minutes of 23.4% saline as a single osmotic agent have also been found to be safe and effective in reducing ICP and improving CPP and PbtO2.25

Temperature

Fever has been associated with adverse outcomes in all forms of brain injury, most likely secondary to an increase in brain metabolic demands. Induced moderate hypothermia (32–34°C) has been used to reduce cerebral edema but no definite benefit has been seen with the exception of postcardiac arrest anoxic injury. Outcomes are influenced by the depth and duration of hypothermia, as well as the rate of rewarming.26 Passive rewarming of patients who were hypothermic on arrival to the hospital was associated with worse outcomes than patients who were maintained at hypothermic temperatures.27 Shivering, a common side effect, elevates ICP and may require higher does of sedation or neuromuscular blockade. Other side effects include coagulopathy, arrhythmias, and suppressed immune responses. At this time, the advantages of hypothermia as a neuroprotective agent have not been proven to be greater than the risks, and treatment should be guided toward maintenance of normothermia. Questions that need to be answered include what patient populations may benefit, if any, what degree of hypothermia, and for how long hypothermia should be maintained.

Barbiturates, Analgesia, & Paralytics

Barbiturates reduce ICP by reducing brain metabolism and therefore CBF and volume. Pentobarbital is more commonly used given its intermediate half-life (approximately 20 hours) and is usually administered as a bolus of 10–30 mg/kg followed by an infusion of 0.5–3 mg/kg/h with the goal of achieving burst suppression. Barbiturates alone are seldom sufficient for the control of ICP when compared with mannitol.28 Their use is fraught with multiple side effects including profound cardiac suppression, vasodilation, and immunosuppression. The hypotension and associated drop in CPP often negates any advantages of ICP control and patients often need to be hemodynamically supported on vasopressors. The use of barbiturates should therefore be limited to patients who have elevated ICP that are refractory to standard medical and surgical management.29 Propofol has been used as an alternative to barbiturates because of its very short half-life, reduction in cerebral metabolism, and antiseizure properties.30 Its use is limited by hypotension as well as by the fact that it is lipid solvent, which can cause severe hypertriglyceridemia and increased CO2 production. The risk of propofol infusion syndrome, although rare, discourages many practitioners from long-term usage.

Pain, agitation, and shivering can increase cerebral metabolic demand and ICP. Patients should be on adequate doses of opioid analgesics to avoid this. When shivering or motor posturing is intractable, neuromuscular blockade with nondepolarizing aminosteroidal agents can be used. The pharmocokinetics of these agents may be altered in the setting of hypothermia and they should be dosed accordingly.

Decompressive Craniectomy

Removal of part of the skull for ICP control aims to negate the Monro–Kellie doctrine of fixed volume by allowing the brain to swell out of the cranial defect.31 Craniectomy has been used to treat intractable intracranial hypertension due to strokes, SAH, TBI, and intracranial hemorrhage. Although there is level 1 evidence supporting the use of decompression in malignant strokes,32 the data for TBI are restricted to case series. Two randomized control trials are underway to determine the benefit of decompression in TBI patients.33 If surgical decompression is considered, it should be done expeditiously and generously, ideally after the first-tier treatments have failed.