Chapter 12: Management of Increased Intracranial Pressure

Table 12-1 lists different classifications of medical conditions that are associated with high ICP. In a neurologic intensive care unit (NeuroICU, NICU) setting, common conditions that are frequently associated with elevated ICP include acute aneurysmal high-grade SAH, severe traumatic brain injury (TBI), large intraparenchymal hemorrhage either spontaneous (such as hypertensive bleed) or in the setting of underlying coagulopathy (atrial fibrillation on warfarin therapy or coronary artery disease with cardiac stents on dual-antiplatelet therapy), malignant middle cerebral artery infarction with herniation, and severe meningitis and/or encephalitis. Although medical centers around the country have different patient populations, a great majority of all NICU cases with high ICP would fall into one of these conditions, with severe TBI being the most common etiology for intracranial hypertension¹ at an annual incidence estimated to be 200 cases per 100,000 people.²

Monro-Kellie Hypothesis

The Monro-Kellie hypothesis is a widely accepted concept for explaining the elevation of ICP. In 1783, Alexander Monro first articulated this in his Observations on the Structure and Function of the Nervous System, and later was supported by Kellie in 1824 by his observation in two humans: “Appearances observed in the dissection of two individuals; death from cold and congestion of the brain.” The hypothesis explains that the human cranium has a fixed amount of space with three main components: cerebrospinal fluid (CSF), brain parenchyma, and blood. If any space-occupying lesion or any of these constituents' volumes are increased beyond the compliance, an elevation of ICP is inevitable.³

Intracranial Compliance

Intracranial compliance is defined as the change in volume over the change in pressure (ΔV/ΔP). The relationship between pressure and volume in the brain is linear (Figure 12-1) in the beginning but exponential in the later phase. As volume increases, the ICP rises slowly (Figure 12-1, Point A), CSF is displaced into the spinal thecal sac, and blood is decompressed from the distensible cerebral veins. Once these compensatory and compliant redistribution mechanisms no longer continue, ICP can increase much more rapidly with just small increments of additional volume (Figure 12-1, Point B). The amplitude of the ICP pulse wave may provide a clue that compliance is reduced; as compliance falls, the ICP pulse amplitude increases (Figure 12-1, Point B, inset). During the high ICP crisis, ICP waveforms may show that the second peak (P2) may rise as high as (increased amplitude compared to normal waveforms) the first peak or even higher (Figure 12-2).

Cerebral Perfusion Pressure

The main negative consequence of elevated ICP is reduced cerebral blood flow (CBF) and secondary hypoxic-ischemic injury due to poor flow. Cerebral perfusion pressure (CPP) is calculated as mean arterial pressure (MAP) – ICP. CPP along with cerebral blood volume determines CBF; normally, autoregulation of the cerebral vasculature maintains CBF at a constant level between a CPP of approximately 50 and 100 mm Hg. Brain injury leading to impaired cerebral autoregulation may cause CBF to approximate a more straight-line relationship with CPP.⁴ Although the optimal CPP for a given patient may vary, in general, CPP optimization should be greater than 60 (to avert ischemia) and below 110 mm Hg (to avoid breakthrough hyperperfusion and cerebral edema).^5,6

Pathologic Pressure Waves

In patients with raised ICP, pathologic ICP waveforms may occur (Figure 12-3). Lundberg A waves (or plateau waves) represent prolonged periods of profoundly high ICP.⁷ These waves do not refer to the individual ICP waveform described in Figure 12-2, but rather a graphic representation of ICP values plotted over time. A plateau wave is considered a high risk of further (or ongoing) brain injury, with critically reduced perfusion as a result of a prolonged period of high ICP crisis. It often occurs abruptly when either CPP or intracranial compliance is low. Their duration may vary from minutes to hours, and pressures as high as 50 to 100 mm Hg may be seen and considered ominous in the setting of acute brain injury. Lundberg B waves are of shorter duration, lower amplitude elevations in ICP that indicate that intracranial compliance reserves are simply compromised. The important thing is not to memorize which is Lundberg A or B per se, but to understand the trend of the ICP elevations and figure out the overall picture as to why there is intracranial hypertension and how brain compliance is coping with the injury. Accurate assessment of the underlying etiology will help guide the clinicians to choose the right type and method of treatment.

Clinical signs of intracranial hypertension may vary and depend on the underlying etiology. In general, the clinical manifestations are suggestive of global, or bilateral, hemispheric cerebral dysfunction rather than a focal finding such as arm weakness. Depressed level of consciousness, blurred vision, confusion, disorientation, nausea, vomiting, diplopia, and sixth cranial nerve palsy (false localizing sign) may be seen, especially if the rise in ICP is acute rather than chronic. It is important to once again emphasize that ICP as an absolute value by itself may not have much clinical significance. More importantly, it is the CPP and brain compliance. The Cushing triad, a well-known phenomenon of hypertension and bradycardia in the setting of critically elevated ICP, is more commonly seen with the late phase of intracranial hypertension such as near brain dead/herniation syndrome rather than in the beginning of an acute injury. ICP elevation may become a local phenomenon and compartmentalized as a result of the rigid boundaries formed by the falx and tentorium cerebelli. Compartmentalized mass effect and pressure differentials, in turn, can lead to herniation of brain tissue from the area of higher to lower pressure. Different herniation syndromes are each marked by characteristic signs (Table 12-2).

Currently, the best level of evidence is the “level 2” recommendation of the Brain Trauma Foundation (2007 guidelines), which states that “treatment should be initiated with ICP >20 mm Hg.” This is heavily driven by the large amount of observation data that showed poor outcome in TBI subjects with hypotension and ICP greater than 20 mm Hg (http://tbiguidelines.org). Systemic hypotension and high ICP leading to poor outcome further supports the theory that failure to optimize CPP is associated with an increased risk of poor flow and perfusion and, therefore, further injury.

Indications for ICP Monitoring

The diagnosis of increased ICP should not be made on clinical grounds alone as clinical signs are not reliable and vary. In acute brain injury, in order to accurately diagnose intracranial hypertension, ICP must be directly measured. Because of the invasive nature of ICP monitoring, and the need for ICU management, patients should generally meet three criteria prior to placement of an ICP monitor: (1) brain imaging reveals a space-occupying lesion and severe cerebral edema suggesting that the patient is at risk for high ICP; (2) the patient has a depressed level of consciousness (this does not mean the patient must be comatose in order to meet the criteria, but clearly those with a Glasgow Coma Scale (GCS) score of 8 or higher would meet the criteria); and (3) the prognosis is such that aggressive ICU treatment is indicated. Although the literature supporting the use of invasive ICP monitoring originates from patients with severe TBI, it is not unreasonable to consider monitoring for other illnesses such as SAH, ICH, large ischemic strokes, and meningoencephalitis cases despite the lack of data.

ICP Monitoring Devices

Several types of ICP monitors exist (Figure 12-4). The external ventricular drainage (EVD) catheter is believed to be the gold standard; it consists of a catheter that is placed through a burr hole into the ventricle and connected to a pressure transducer set at ear level. It allows for both ICP monitoring and therapeutic CSF drainage. Its major drawback is the risk of infection such as potentially life-threatening ventriculitis, which occurs in approximately 10% to 15% of patients and steadily increases until the 10th day of use.^8,9 As such, it is important to place the catheter with care using sterile technique as well as maintain the sterility thereafter. It is usually placed through a long subcutaneous tunnel in order to minimize the rate of infection. The best alternatives to EVD include fiberoptic transducers (Integra LifeSciences, Plainsboro, NJ) or pressure microsensors (Codman & Shureleff, Inc, Raynham, MA) placed through a burr hole into either the parenchyma or ventricle. These devices carry less risk of infection but do not allow therapeutic drainage of CSF.

General Measures for ICP Control

These general measures apply to all patients at risk for or ongoing intracranial hypertension.

Head position

Elevation of the head to at least 30 degrees is advised in patients with raised ICP, with one study confirming that this degree of elevation produces consistent reduction in ICP.¹⁰ Although some investigators advocate a head-flat position to preserve CPP, moderate elevation is safe as long as CPP is continuously maintained at greater than 60 mm Hg. For patients with large abdominal girth, it is also important to pay attention so that excessive head elevation is not causing abdominal distress, as increased abdominal pressure and pain may exacerbate ICP elevation.

Body temperature

Fever may potentiate ischemic brain injury and contribute to elevated ICP. It has been demonstrated that increases in brain temperature are correlated with ICP elevations, as CBF and cerebral blood volume (CBV) increase disproportionately in relation to the cerebral metabolic rate of oxygen consumption (CMRo₂).¹¹ Acetaminophen and cooling blankets are the first line of therapy and should be instituted when temperature is sustained over 101°F (38.3°C). Whether the prophylactic maintenance of mild hypothermia (34^°C to 35°C) can reduce the number of ICP crises remains unknown at this time. Early mild-to-moderate hypothermia (32^°C to 34°C) within 8 hours of onset has not been found to be effective for improving outcome after severe TBI even though a modest benefit on ICP was seen.¹² If fever is not adequately controlled with conventional methods, it would be wise to use advanced temperature-modulation devices—either surface or intravascular cooling catheters.

Seizures

Seizures cause an increase in CMRo₂ and hence CBV. These alterations tend to be abrupt and marked and, in patients with reduced cerebral compliance, can trigger plateau waves due to suddenly elevated flow and blood volume. For patients at risk for raised ICP, it is reasonable to administer prophylaxis with an intravenous anticonvulsant. This is not an absolute recommendation, but convulsion in the setting of ICP crisis would be devastating. In TBI patients, prophylaxis was shown to reduce the frequency of seizures during the first 7 days (with no evidence to continue beyond the first week) from 14% to 4%. Fosphenytoin or phenytoin (15- to 20-mg/kg bolus followed by 300 mg daily) has been recommended in this situation, but that is not because phenytoin is superior to other antiepileptic drugs. In fact, owing to multiple side effects, difficulty maintaining a therapeutic level, and numerous drug-to-drug interactions, it would be reasonable to use a newer agent such as levetiracetam or lacosa-mide, both of which are available in oral and intravenous formulations without the need for checking levels (as a prophylactic use anyway: this may be different in refractory status epilepticus management) or significant concerns of tolerability or drug-to-drug interactions.

Fluid management

Historically, raised ICP had been managed by fluid restriction. It was later found that this effort to dehydrate the brain actually worsened hypoxic-ischemic injury because the hypovolemia led to reduced CPP.¹³ Patients with high ICP states should be kept euvolemic with isotonic saline. Hypotonic solution in any form (eg, 0.45% saline, D₅W, or enteral water) must be avoided because it will accumulate through an osmotic gradient in regions of injured brain and aggravate brain swelling. Some centers use a continuous infusion of 2% or 3% saline at 1 mL/kg as an alternative to normal saline for patients at risk for ICP, which is directed toward establishing and maintaining a hypernatremic (goal Na⁺ approximately 155 mEq/L) hyperosmolar (goal osmolality approximately 320 mOsm/L) environment for the injured brain. However, whether this strategy is effective for reducing ICP crises or improving outcome is unknown. Continuing maintenance fluid intake at 2% to 3% hypertonic saline may not produce a sustained effect in controlling ICP crisis, and, if used, continuous infusion should only be used for a finite period of time. A higher concentration given in boluses (eg, 23.4% hypertonic saline in 30-mL boluses given as rapid IV push over 1 to 2 minutes) may be more effective in reducing high ICP.

Corticosteroids

Corticosteroids such as dexamethasone are not effective as a general measure to treat elevated ICP, with associated risk of developing or promoting nosocomial infection, hyperglycemia, impaired wound healing, muscle catabolism, and psychosis/delirium. Steroids are effective only for reducing the volume of mass lesions related to abscess or neoplasm, and the mechanism is based on its effect on vasogenic edema. A cytotoxic edema process such as acute ischemic stroke, therefore, would be a wrong indication to use this therapy.

Cranial Decompression

Whenever a sustained ICP elevation of 20 mm Hg or more occurs, the clinician should consider (or reconsider) surgical decompression. A CT scan should be considered, and if increased mass effect or CSF volume is detected, surgical intervention such as CSF drainage, mass evacuation, or hemicraniectomy may be performed.

By opening the cranial vault, hemicraniectomy can reverse brain tissue displacement and herniation, and effectively improve and usually normalize ICP. Hemicraniectomy is increasingly being used as a final salvage therapy for patients with malignant middle cerebral artery area infarction and other space-occupying mass lesions. It is considered as a definitive therapeutic option as an alternative to instituting barbiturate therapy or mild-to-moderate hypothermia. Several studies have found that hemicraniectomy definitely improves survival after malignant middle cerebral artery (MCA) infarction.¹⁴ A meta-analysis of hemicraniectomy for MCA infarction found that survival with a good functional outcome is most likely among younger patients.¹⁵

Sedation

Agitation must be avoided because it can aggravate ICP elevation through straining (increasing thoracic, jugular venous, and systemic blood pressure) and increased CMRo₂. During an ICP spike, sedation must be maximized and may be all that is necessary to control the ICP. For this reason, adequate sedation must be the first pharmacologic intervention in managing ICP crisis. The preferred regimen is the combination of a short-acting opioid such as fentanyl (1 to 3 μg/kg per hour) or remifentanil (0.03 to 0.25 μg/kg per minute) to provide analgesia and propofol (0.3 to 3 mg/kg per hour) because of its extremely short half-life (despite its well-known side effect of systemic hypotension), which makes it ideal for periodic interruption for neurologic assessments. Bolus injections of opioids, however, should be used with caution in patients with elevated ICP because they can transiently lower MAP and increase ICP due to autoregulatory vasodilation of cerebral vessels.¹⁶ Compared to an opioid-based sedation regimen, in one trial propofol was associated with lower ICP and fewer ICP interventions in patients with severe TBI.¹⁷

CPP Optimization

After adequate sedation, if ICP remains elevated, attention should be directed to optimizing CPP. CPP low enough to induce ischemia can trigger reflex vasodilation and aggravate ICP elevation. Conversely, a high CPP (greater than 110 mm Hg) can sometimes cause breakthrough cerebral edema and also potentially elevate ICP. For these reasons, in general, CPP should be maintained between 60 and 110 mm Hg. Appropriate vasopressors to raise blood pressure and CPP include phenylephrine (2 to 10 μg/kg per minute), dopamine (5 to 30 μg/kg per minute), or norepinephrine (0.01 to 0.6 μg/kg per minute). Useful agents to lower blood pressure and CPP include labetalol (5 to 150 mg per hour) and nicardipine (5 to 15 mg per hour); nitroprusside should be avoided because of its dilating effects on all cerebral vasculature, which may exacerbate the ICP. Sodium nitroprusside is also difficult to titrate compared to other agents.

Numerous studies have attempted to define optimal CPP management in acute TBI, with various conclusions. In recent years, two distinct approaches have developed with differing views on whether CPP should be maintained at a higher or lower level. The high-CPP approach, popularized by Rosner et al, focuses on pharmacologic means to elevate MAP and CPP in order to maintain adequate CBF.^18,19 Support for this method comes from case series demonstrating good clinical outcomes and higher brain tissue oxygen levels with this management approach, and clinical examples demonstrating that induced hypertension can lead to the termination of plateau waves, presumably by causing reflex vasoconstriction.²⁰ The main argument against the high-CPP approach comes from a randomized trial by Robertson et al that found no clinical benefit with CPP-targeted therapy (CPP greater than 70 mm Hg) compared with traditional ICP-targeted therapy (CPP greater than 50 mm Hg). In this study, the high-CPP approach led to fewer jugular venous desaturations, but a fivefold increase in the risk of acute respiratory distress syndrome (ARDS).²¹

The low-CPP approach, popularized in Lund, Sweden, concentrates on ICP reduction by minimizing CPP and reducing CBV and intravascular hydrostatic pressure.²² The fundamental principles of the Lund approach include the maintenance of normal colloidal pressure to prevent extravascular fluid shifts, reduction of intracapillary hydrostatic pressure through systemic blood pressure reduction, and minimization of CBV by suppressing CMRo₂ with thiopental and promoting precapillary vasoconstriction with dihydroergotamine. Evidence in support of the Lund approach includes case series managed according to this protocol with good outcomes,²² and cerebral microdialysis studies demonstrating that significant oxidative stress in the form of increased lactate/pyruvate ratios does not consistently occur until CPP falls below 50 mm Hg.²³

It seems most likely that both the high- and low-CPP strategies described above are valid depending on the individual circumstances of the patient. In this view, no single approach should be generalized to all patients; instead, CPP should be optimized based on individualized physiologic monitoring. Advanced multimodality monitoring techniques such as brain tissue oxygen monitoring, jugular bulb oximetry, signal-processed electroencephalography (EEG), and micro-dialysis may eventually allow clinicians to fine tune CPP and MAP goals based on the specific physiologic circumstances in a particular patient at any given point in time during acute injury. This approach makes better physiologic sense and may eventually be the ultimate goal in the future.

Hyperventilation

Hyperventilation has long played a role in ICP therapy. A decrease in Pco₂ causes vasoconstriction, which lowers CBV and thus ICP. The effect is almost immediate, but often transient, with a steady loss of effect occurring within the first hour. In conditions characterized by excessive vasodilation and cerebral hyperemia, the effect of hyperventilation may be sustained for days. Intracranial pressure is directly related to CBV. Hyperventilation works by directly reducing CBF through vasoconstriction. A reduction of CBF could potentially limit blood flow to ischemic areas of the brain, with only a modest reduction in ICP through its indirect effects on CBV.

Studies have demonstrated a risk of exacerbation of cerebral ischemia with ongoing hyperventilation. As such, the use of hyperventilation should occur in the setting of ICP crisis and emergency as a transient method before other therapies are available. The routine application of extreme hyperventilation (less than 30 mm Hg) within the first few hours of TBI is generally considered harmful because of the risk of exacerbation of ischemia.²⁴ If necessary, jugular venous oxygen saturation or brain tissue oxygen monitoring can be used as a guide during hyperventilation to ensure adequate brain oxygen delivery.^25,26

Osmotherapy

If CPP is optimized, the patient is sedated, and ICP remains elevated, osmotherapy should be initiated. Mannitol, given in a 20% solution at a dose of 0.25 to 1.5 g/kg, mediates an ICP-lowering effect through two mechanisms. First, it is an osmotic diuretic that creates a concentration gradient across the blood-brain barrier and extracts free water from the brain. This decreases brain volume and lowers ICP.²⁰ Also, mannitol increases CPP through plasma expansion, and promotes vasoconstriction and CBV reduction by decreasing blood viscosity and improving CBF.²⁷ Mannitol can lower ICP within minutes. It should be given in a single rapid bolus (0.25 to 1.5 g/kg), and may be repeated as frequently as once an hour when ICP is elevated. Complications of mannitol therapy include dehydration and renal failure. Hypertonic saline (2 to 5 mL/kg of 7.5% saline or 0.5 to 2.0 mL/kg of 23.4% saline over 30 minutes) is an alternative to mannitol for treating acutely elevated ICP. It has been shown to be at least as effective^28,29 for acutely lowering ICP, and has the advantage of boosting MAP, CPP, and intravascular volume when patients are dehydrated. The main complication specific to hypertonic saline therapy is congestive heart failure due to fluid overload. For this reason, the use of a continuous IV infusion of 3% hypertonic saline (at a certain rate: typically starting at 50 mL per hour, titrating up to the serum sodium target of about 150 mEq/L) may not be the best method if volume overload is a concern. In such a case, a bolus dosing may be preferred. Once patients are on a steady dose of continuous IV 3% saline with a serum sodium level approximately 150 mEq/L, it is advised that the hypertonic saline should be weaned very slowly (over several days), if deemed appropriate, as it is not uncommon to observe rebound intracranial hypertension upon abrupt cessation of the infusion (patients with space-occupying lesions may even herniate).

Currently, there are insufficient data to suggest one concentration or method (continuous or bolus) over another. Hypertonic saline bolus therapy does appear to be as effective as mannitol in reducing refractory elevated ICP and improving CPP transiently. However, many issues remain to be clarified, including the exact mechanism of action of hypertonic saline, the best mode of administration and concentration, and its risks and complications.

Pentobarbital

Failure of hyperventilation and osmotherapy to control ICP should prompt consideration of the initiation of pentobarbital infusion.³⁰ Consideration of pentobarbital in this setting should also trigger reconsideration of performing hemicraniectomy or the application of hypothermia. The mechanism of action of pentobarbital is a profound reduction of the cerebral metabolic rate. Pentobarbital can be given in repeated 5-mg/kg boluses every 15 to 30 minutes until the ICP is controlled (usually 10 to 20 mg/kg is required), and then continuously infused at 1 to 4 mg/kg per hour. An EEG should be continuously recorded, and the pentobarbital titrated to produce a burst-suppression pattern, with approximately 6- to 8-second interbursts, to avoid overmedication (although the goal is ICP control, not burst suppression on EEG, per se). The most common complication of pentobarbital therapy is hypotension owing to its cardiac suppression, and vasopressors and inotropes are often needed for hemodynamic support. Ileus may occur as well, and feeding may have to be given parenterally during treatment. Delayed, inadequate hemodynamic support may lead to acute kidney failure (and hence multiorgan failure) and severe acid-base imbalance, leading to a much more difficult situation.

Hypothermia

If all of the above therapies fail to control ICP, induced hypothermia to 32°C to 34°C can effectively lower otherwise refractory ICP. Hypothermia reduces ICP by lowering CMRo₂ requirements and thus CBV. The recommended temperature goal is 32^°C to 34°C, which is considered mild-to-moderate hypothermia because there are fewer complications than at lower temperatures. Hypothermia can be achieved using various surface and endovascular cooling methods coupled to a rectal, esophageal, pulmonary artery, or bladder thermometer. Rapid infusion of large-volume cold fluids (30 mL/kg of 0.9% saline cooled to 4^°C to 5°C) may be suitable for core cooling, especially as an effective, cost-effective method of administering a hypothermia-induction agent. Cold saline infusion should be used from the beginning while trying to set up a more controllable temperature-modulatory device. Therapeutic hypothermia has been clinically demonstrated to control ICP in a small series of patients refractory to pentobarbital,³¹ but large controlled studies are lacking in this regard. Common potential complications of hypothermia include nosocomial infection, hypotension, cardiac arrhythmias, coagulopathy, shivering, hypokalemia, hyperglycemia, and ileus. Particular caution should be exercised when rewarming patients because rebound ICP readily occurs. Rewarming must be done slowly (0.10^°C to 0.33°C per hour) and in a controlled fashion.