Maintenance of normal intracranial pressure (ICP) is determined and closely regulated by well-defined structural components of the intracranial compartment. The introduction of intracranial pathology will ultimately exhaust the mechanisms that maintain cerebral homeostasis and lead to elevated ICP and abnormal intracranial elastance. The concept of abnormal ICP and its management goals are not static, and change as we gain a better appreciation of the underlying pathophysiologic processes involved, the impact of secondary neuronal injury, and the information obtained from cerebral monitoring available today. It is important to recognize the impact of hemodynamic manipulations, pharmacologic interventions, anesthetic agents, and ventilation on ICP and cerebral perfusion pressure (CPP). The goals of this chapter are to provide an understanding of : (i) ICP and its determinants; (ii) the regulation of ICP and its effects on CPP in normal and pathologic states; (iii) the impact of anesthetic agents on ICP; and (iv) the management of intracranial hypertension.

The intracranial compartment is defined by its contents: Brain tissue, cerebrospinal fluid (CSF), blood, and meninges. They are encased by the calvarium and communicate with the spinal axis through the foramen magnum. In the absence of pathology, the intracranial volume remains constant within the neuraxis. The calvarium is a bony, nondistensible, semiclosed container that strictly limits both intracranial volume and any expansion by acute or chronic pathologic processes.

Brain tissue alone accounts for 88% of the total intracranial volume. Eighty percent of the brain volume is water, and 20% of this is sequestered extracellularly.¹ This extracellular environment is tightly regulated by an intact blood-brain barrier, which selectively permits diffusion and active transport of limited substances.

CSF accounts for 9% of the total intracranial volume.² The total CSF volume is 150 mL in a normal adult, with daily production of 500 to 600 mL and replacement every 8 hours.³ Fifty percent of the total CSF volume (75 mL) is contained within the intracranial space.⁴ CSF is secreted primarily by the choroid plexus lining portions of the ventricular system and is principally absorbed from the subarachnoid space into the venous blood at the arachnoid granulations bordering the superior sagittal sinus. Obstructions of any portion of the ventricular system, including the arachnoid granulations, results in hydrocephalus and, occasionally, interstitial edema.

The cerebral vascular system is the smallest component of the intracranial space, contributing only 2% to 3% of its volume.² The brain receives 15% of the cardiac output, with a rapid transit time. Cerebral blood volume (CBV) is only 75 mL, with three quarters residing within the low-pressure venous system.⁵ Despite its small volume, the vascular space is an important determinant of ICP and the parameter most affected by ventilation, anesthetic agents, and vasodilating drugs.

Intracranial volume remains constant until a new volume is added or expansion of an existing intracranial volume occurs in the nondistensible system. For ICP to remain normal, a mandatory reciprocal reduction in volume of one of the intracranial components (i.e., blood, CSF) must occur to offset the effects of the new volume added. These concepts constitute the principles of the Monroe-Kellie doctrine and its modifications.⁶

Normal ICP measured in adults in the supine position is 5 to 15 mmHg; normal values are 1.5 to 6 mmHg in infants and 3 to 7 mmHg in young children.⁷ In the supine position, pressures are considered equal within the craniospinal axis. Lumbar subarachnoid pressures reflect ICP as long as the two compartments freely communicate and no obstruction exists. It is now recognized that ICP is not always uniform within the craniospinal axis when pathology is present. Local differences in ICP and cerebral blood flow (CBF) often exist in areas of pathology (i.e., brain tumors or traumatic brain injury [TBI]), although, global ICP may be recorded as normal and is misleading.⁸,⁹ ICP is considered abnormal with levels of 15 mmHg in infants, 18 mmHg in children younger than 8 years, and 20 mm Hg in older children and adults.⁷ Neurologic outcome, especially in head-injured patients, appears to correlate with the degree and duration of intracranial hypertension.¹⁰ Sustained elevations in ICP that exceed 25 to 30 mmHg are frequently associated with poor outcome or are fatal. The detrimental effects of intracranial hypertension are the result of cerebral ischemia and the direct compressive effects on cerebral structures.

ICP can be maintained at normal levels and regulated, even in the presence of a space-occupying lesion, as long as compensatory mechanisms are operational and the pathologic process evolves slowly. If rapid volume expansion occurs (i.e., epidural hematoma), ICP rises acutely because of inadequate compensatory mechanisms. A chronic pathologic process such as a slow-growing tumor will allow adequate reductions in cerebral blood and CSF volume to maintain normal intracranial dynamics until the mechanisms for spatial compensation are exhausted. The CSF system has the greatest buffering capacity of the intracranial contents, largely because of CSF absorption. Intracranial hypertension from a chronic pathologic process occurs when the buffering capacity of the CSF pathways is exhausted.

The intracranial pressure-volume curve was defined by Langfitt et al. in 1965¹¹ (see Fig. 22.1). During the period of
compensation (horizontal axis), ICP increases minimally, even with relatively large changes in intracranial volume. When the buffering capacity of the intracranial space is exhausted, small increases in volume are followed by abrupt rises in ICP. If a mass expands rapidly, the additional volume cannot be accommodated, and ICP rises acutely.

The slope dP/dV of the pressure-volume curve defines the elastance of the system and reflects its stiffness or resistance to the deformation exerted by the intracranial contents with the addition of volume.¹² Compliance is also used to define the pressure-volume curve; however, elastance more accurately defines this relation. Compliance is defined as the slope dV/dP and reflects the distensibility of the intracranial space.

The relation between ICP and CPP is defined by the following formula: CPP = MAP-ICP (MAP = mean arterial pressure). The lower limit of CPP considered acceptable in normotensive adults is 50 mmHg. Under normal conditions, ICP is usually 5 mmHg or less, and CPP changes reflect alterations in MAP. Central venous pressure has been used when ICP is unknown or not measured. In the presence of intracranial pathology, effective perfusion pressure declines as ICP rises. CPP should be maintained at >60 mmHg in head-injured patients, and perhaps higher when there is documented global or focal ischemia.¹³ A decrease in MAP or an elevation in ICP will deleteriously alter the effective perfusion pressure.

Cerebral autoregulation involves the maintenance of a constant CBF when MAP (or CPP if ICP is normal) is between 50 and 150 mm Hg and is accomplished by adjusting the cerebrovascular resistance (CVR) (see Fig. 22.2). In recent years, the lower limit of autoregulation has been questioned and is likely higher than previously depicted by Lassen.¹⁴ There is evidence that the lower limit of autoregulation in nonanesthetized, normotensive adults is as high as 70 mmHg and that the lower limit depicted in the autoregulation curve may be more applicable to anesthetized patients or those receiving vasodilating agents that lower the limit of autoregulation (i.e., sodium nitroprusside).¹⁵ When the limits of autoregulation are impaired or exceeded, CBF passively follows blood pressure. In chronic hypertension, the autoregulatory curve is shifted to the right; consequently, in hypertensive individuals, cerebral ischemia can occur at a CPP that generally would ensure adequate CBF in normotensive subjects. Intracranial hypertension alone does not appear to abolish autoregulation. Autoregulation of CBF occurs within seconds, and, at least in nonanesthetized humans, depends on the underlying PACO₂.¹⁶ The exact mechanism of autoregulation is not known, but it is speculated that the resistance vessels may be responsive to the local metabolic environment or to transmural pressure.¹⁷

Normal global CBF is approximately 50 to 55 mL/100 gm brain tissue/minute.¹⁸ Grey matter receives 75% of CBF, reflecting its increased metabolic activity compared to the white matter.¹⁹ Local increases in CBF in the nonpathologic state indicate the use of a specific region of the cortex during a task or activity.²⁰ CBF is normally tightly coupled to cerebral oxygen consumption (CMRO₂) and is expressed as CMRO₂ = CBF × AVDO₂ (arteriovenous difference of oxygen).²¹ Normally CMRO₂ is constant at 3.2 to 3.5 mL/100gm brain tissue/minute.²¹ CBF and CMRO₂ do not always remain coupled, as often observed after head injury and during the use of potent inhalational agents.²²

Cerebral ischemia occurs when CBF falls below the critical level required to meet the metabolic demands of the cerebral tissue. The threshold for cerebral ischemia is increased under most general anesthetic techniques, especially isoflurane, and predominantly results from reduction in CMRO₂.

CBF is influenced by pathology disrupting autoregulation, potent inhalational agents, intravenous anesthetics, and alterations in PACO₂ and PaO₂. In general, changes in CBF are accompanied by parallel changes in CBV at a ratio of 7:1.²³ This relation does not always exist, and each factor affecting CBF and CBV must be considered independently. For example, inhalational agents and hypercarbia produce elevations in CBF and CBV by cerebral vasodilation, but hypotension with a MAP <50 mmHg (or higher in nonanesthetized individuals) results in decreased CBF but increased CBV, reflecting changes in the tone of the cerebral resistance vessels.

The cerebral resistance vessels, predominantly end-arterioles, are exquisitely sensitive to alterations in PACO₂ and rapidly alter ICP. Hypocapnea is characterized by vasoconstriction of arterioles, resulting in decreased CBV, CBF, and ICP, whereas hypercapnea influences the vascular tone in opposite manner. CBF is increased by approximately 2 mL/100 gm brain tissue/minute or a 4% change for each mm Hg increase in PACO₂.²⁴ The pH of the extracellular fluid of the brain influences the tone of the cerebral vasculature, but within 24 to 36 hours, the CSF pH begins to normalize, and benefits of hyperventilation are lost.²⁵ The responsiveness of the resistance vessels to carbondioxide (CO₂) diminishes as the upper and lower limits of cerebral autoregulation are approached.²⁴ A linear relation between CBF and PACO₂ exists when PACO₂ ranges from 20 to 80 mmHg.²⁴ The potent inhalational anesthetic agents do not interfere with the reactivity of the cerebral vasculature to CO₂ at a 1.0 minimum alveolar concentration (MAC) of anesthesia. CBF is not affected until the PaO₂ falls below 50 mmHg, and reflects a response to tissue acidosis, but rapidly reverses when PaO₂ is corrected.

The symptoms and signs of elevated ICP are relatively nonspecific and depend largely on the chronicity and location of the underlying pathologic process. The clinical presentation of a patient with a chronic, slow-developing, intracranial process is vastly different from an acute process with early spatial exhaustion and impending herniation of brain tissue. The classic clinical manifestations of elevated ICP are headache, nausea, vomiting, visual disturbances, altered mentation, papilledema, and ocular palsies. Focal neurologic deficits reflect mass effect or ischemia of surrounding brain tissue by an expanding lesion. Headache frequently occurs and is located bifrontal or bioccipital; it is worse in the morning upon awakening, exacerbated by coughing or Valsalva maneuver, paroxysmal in nature, and relieved by emesis. Dural stretching and traction on vessels at the base of the brain caused by the mass lesion contribute to headache. Headache suggests elevated ICP but is not diagnostic. Nausea and projectile vomiting occur usually in the morning or at night, and emesis often relieves headache. Papilledema is present bilaterally 24 to 48 hours after ICP becomes elevated, and results from axoplasmic stasis as ICP is transmitted through the subarachnoid space to the optic disc. Focal neurologic deficits may be evident if an intracranial mass lesion compresses or distorts underlying sensory, motor, or cranial nerve pathways.

Deterioration in the neurologic examination progresses in a rostral-to-caudal fashion with supratentorial mass lesions. Brainstem dysfunction is a late finding with a slowly expanding mass, but occurs early with a rapidly enlarging mass and signals herniation.²⁶ Conversely, the initial clinical manifestation of an infratentorial mass lesion may be brainstem dysfunction (i.e., cranial nerve palsy, cerebellar findings, respiratory abnormalities). Loss of consciousness associated with a supratentorial mass is ominous in the absence of a metabolic etiology and implies bilateral cerebral hemispheric or diencephalic dysfunction.

The degree of midline shift of the pineal body measured by computed tomography (CT) of the brain correlates with level of consciousness: Alert with a 0 to 3mm shift; drowsy with a 3 to 4 mm shift; stuporous with a 6 to 8.5 mm shift; and comatose with a 8 to 13 mm shift.²⁷ Mental status impairment with infratentorial lesions often results from the development of noncommunicating hydrocephalus, but may be ominous and signal disruption or compression of the reticular activating system.

Assessment of brainstem function is the basis of the neurologic examination in a comatose patient. Evaluation of brainstem reflexes allows documentation of progressive neurologic deterioration and confirms brain death. The neurologic examination of brainstem function consists of assessment of pupil reactivity and size, respiratory pattern, oculomotor response, and motor response to painful stimuli.

Any patient with suspected intracranial pathology and increased ICP requires evaluation with CT or MRI of the brain. Peritumor edema, diffuse cerebral edema, subarachnoid blood, hydrocephalus, compressed ambient cisterns, midline shift, and the presence of mass lesions suggest elevated ICP but do not provide quantitative
information (see Fig. 22.3). In patients with severe head injury, the presence of midline shift and compressed ambient cisterns usually predict increased ICP or herniation.²⁸ It is important to correlate neuroradiologic findings with the neurologic examination because the clinical presentation often reflects the chronicity of the underlying disease process.

The intraventricular catheter (or ventriculostomy) was introduced in 1960 by Lundberg and has remained the gold standard for ICP monitoring.²⁹ Newer monitoring modalities have been developed; ICP can be recorded from an epidural, subdural, subarachnoid, ventricular, or intraparenchymal location, depending upon the device used. The monitoring devices presently in use include catheters, bolts, screws, and fiberoptic cables. ICP monitoring is most often achieved from a supratentorial location and is less commonly employed in the posterior fossa due to risk of cranial nerve or brainstem injury.³⁰ It is presumed that once the monitor is in position, the ICP recorded is reflective of pressure in the supratentorial and infratentorial compartments in the absence of obstruction; however, differences in ICP have been recorded not only between supratentorial and infratentorial compartments, but also within the same compartment (i.e., supratentorial).

Transcranial doppler (TCD) imaging provides an inexpensive, noninvasive bedside method to document severe
intracranial hypertension, brain death, and presence of cerebral vasospasm. TCD imaging is an ultrasound technique that determines blood flow velocity, direction of flow through a major cerebral artery, and reflects the status of the vascular resistance of the vessel. As ICP approaches diastolic arterial pressure levels, the diastolic flow velocity diminishes, approaches 0, and then reverses flow.³⁴ Brain death has been associated with one of three TCD patterns: (i) diastolic flow reversal; (ii) small systolic spikes; or (iii) absence of signal.³⁴ An absent signal should be interpreted cautiously and verified in other vessels. Information derived from TCD monitoring must be used in conjunction with the clinical examination (brainstem reflexes) when documenting brain death.

Multimodality-evoked potentials, including somatosensory and brainstem auditory-evoked potentials, can be used to diagnose brain death and predict neurologic outcome in comatose patients. Both types of evoked potential recording modalities are not accepted as the sole diagnostic criteria and should be used with knowledge of the neurologic examination. The absence of all waveforms, except wave I in brainstem auditory-evoked potential recording, accompanied by absence of cortical somatosensory-evoked potentials reliably predicts brain death.³⁵ Wave I may be absent because of destruction of the organ of Corti or due to ototoxic drug administration. In the presence of normal or abnormal brainstem auditory-evoked potentials and absence of cortical somatosensory-evoked potentials, a persistent vegetative state is predicted.³⁵

The selection of an anesthetic technique and supplemental drugs for a patient with intracranial hypertension should be tailored to perform the following functions: (i) Maintain adequate CBF and CPP, (ii) minimize ICP elevations, (iii) provide cerebral protective measures, and (iv) reduce ICP. It is important for the anesthesiologist to recognize the impact that anesthetic agents or pharmacologic interventions may have in improving or worsening the primary underlying pathology. The use of agents to prevent or attenuate secondary neuronal injury following cerebral ischemia may become as important as the treatment directed toward the primary injury.