If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
This article reviews cerebral metabolism and blood flow, and the pressure dynamics within the cranial cavity. The brain functions within the confines of the cranial cavity and it is important to understand the dynamics of the parenchyma, cerebrospinal fluid and blood in relation to intracranial pressure (ICP) and metabolic needs. It requires an uninterrupted supply of oxygen and glucose to maintain its basal energy requirements and these are increased during periods of enhanced activity. Cerebral blood flow (CBF) is therefore critical for normal cerebral function. Its control is dictated by local intrinsic metabolic needs as well as extraneous factors such as arterial blood pressure, arterial carbon dioxide and oxygen tension, temperature and neural factors; all of which can be measured to guide therapy.
After reading this article, you should understand the:
importance of an uninterrupted supply of oxygen and glucose for normal cerebral metabolism
fundamentals of the regulation and measurement of cerebral blood flow
concept of intracranial pressure and how it affects cerebral perfusion pressure
The contents of the cranium are brain parenchyma (80%), blood (9%), cerebrospinal fluid (CSF; 6%) and interstitial fluid (5%). The brain's primary function is the generation of neuronal action potentials in response to stimuli. This function is mediated by ionic movement against electrical gradients and the release of neurotransmitters at synaptic junctions. The normal physiology of the brain is energy intensive requiring large amounts of adenosine triphosphate (ATP). Glucose is the primary metabolic fuel and requires a sufficient supply of oxygen for the oxidative process.
This article explores the pressure dynamics within the cranium and the physiological mechanisms that maintain supply of oxygen and glucose to the brain.
Monro-Kellie doctrine and intracranial pressure
After closure of the cranial sutures, the cranial cavity functions as a rigid box with no room for expansion. The intracranial pressure (ICP) is maintained between 7 and 12 mmHg in normal circumstances. The Monro-Kellie doctrine describes the pressure–volume relationship within the cranial cavity. It states the sum of the volumes of brain, CSF and intracranial blood is constant. Since the intracranial volume is fixed, an increase in one of the three intracranial components will result in a rise in ICP unless compensated for by a reduction in volume of another component. Initially, a rise in ICP is compensated for by CSF migration into the spinal compartment accompanied by an increase in CSF absorption and decrease in production, and a reduction in cerebral blood flow. As these compensatory mechanisms are overwhelmed, intracranial compliance falls and ICP rises dramatically with even a small increase in intracranial volume. This could lead to brainstem compression if untreated, and manifest as hypertension, bradycardia and irregular respiration (Cushing's reflex).
ICP is a dynamic pressure with fluctuations occurring with arterial pulsation, position, respiration, coughing and straining. It may be measured using an intracranial pressure bolt or external ventricular drain (Figure 1).
Managing raised ICP
Surgically, raised ICP may be controlled by CSF diversion using an external ventricular drain or by a decompressive craniectomy. Medical measures for ICP control aim at reducing intracranial blood volume or interstitial fluid volume. Reduction of intracranial blood volume can be achieved by optimizing ventilation to reduce arterial carbon dioxide tension (PaCO2) and improve oxygenation, increasing venous drainage with a head-up position, and providing adequate sedation and muscular relaxation to reduce cerebral metabolic rate and intrathoracic pressure. Interstitial fluid volume reduction can be achieved by fluid restriction or by the administration of diuretics (e.g. mannitol and furosemide) or corticosteroids.
Cerebrospinal fluid (CSF)
CSF is an ultrafiltrate of plasma that circulates through the cerebral ventricles and the central canal of the spinal cord. It is formed (and reabsorbed) at the rate of 500 ml/day by energy dependent perfusion-related processes in the choroid plexus and the ependymal lining of the lateral ventricles. It flows via the foramina of Monro to the third ventricle, and then to the fourth ventricle through the aqueduct of Sylvius. CSF then flows into the central canal of the spinal cord and the subarachnoid spaces through the median foramen of Magendie and lateral foramina of Luschka. CSF is ultimately absorbed by the subarachnoid villi into the cerebral venous sinuses. If the rate of CSF formation exceeds that of absorption, hydrocephalus ensues.
The blood--brain barrier (BBB) exists between the bloodstream and the central nervous system (CNS). It is a semi-permeable membrane consisting of three layers: the vascular endothelium with its basement membrane, the astrocyte foot processes and pericytes. The endothelial cells have very few pinocytic vesicles and are sealed by tight junctions with no anatomical gaps. This provides a high electrical resistance barrier. The BBB explains the difference in constitution of the CSF and plasma. CSF has a very low protein content when compared to plasma (0.2 versus 60 g/L). Increased levels of protein in CSF would indicate a disruption to the BBB. Concentrations of potassium, calcium, glucose, urea and lymphocytes are lower in CSF.
The passage of substances across the BBB is directly proportional to their lipid solubility but is inversely proportional to molecular weight, ionic charge and degree of plasma-protein binding. It is facilitated by active transport mechanisms that require energy. Lipophilic substances (carbon dioxide, oxygen, volatile anaesthetic agents) pass freely, unlike large-molecular-weight molecules (e.g. proteins) and highly charged moieties (e.g. sodium ions). Proteins and drugs (e.g. penicillin) cannot cross the barrier unless it is inflamed (e.g. in meningitis). The integrity of the barrier can be examined by the intravenous injection of radioactive isotopes bound to protein; scanning techniques can then be used to determine whether the radioactive label is escaping from cerebral vessels. Ruptured aneurysms and increased permeability at tumour sites may be detected using such techniques.
Water moves freely across the BBB via astrocytes expressing aquaporin-4 (AQP4). This depends on osmotic gradients. Sudden changes in plasma osmolality secondary to changes in electrolyte or glucose concentrations can therefore lead to potentially problematic fluid shifts in the brain. This emphasizes the importance of correcting sodium and glucose abnormalities slowly. Changing osmotic gradients may also be used to reduce ICP via fluid shifts (e.g. using hypertonic saline). The BBB is disrupted by a number of processes such as hypertension, stroke, trauma, status epilepticus, hypercarbia, hypoxia, and especially inflammation (chemical, infective or autoimmune). When disruption occurs, fluid movement becomes largely dependent on hydrostatic gradients.
Cerebral blood flow
Blood flow to the brain is primarily by the paired internal carotid arteries anteriorly and the paired vertebral arteries posteriorly. About 70% of cerebral blood flow (CBF) is supplied by the internal carotid arteries. The anterior and posterior circulations anastomose at the base of the brain to form the Circle of Willis. There are numerous anatomical variations in the Circle of Willis with an incomplete anastomosis in around 50% individuals.
Although the brain constitutes only 2% of the total body mass, it receives 15% of the cardiac output (750 ml/min in adults). Resting CBF is approximately 50 ml/100 g/min. The flow is not evenly distributed. Grey matter, which is metabolically more active, receives approximately 90 ml/100 g/min and in these regions the rate of oxygen consumption, termed the cerebral metabolic rate for oxygen (CMRO2), is about 3 ml/100 g/min. White matter receives about 20 ml/100 g/min and its CMRO2 is approximately 1 ml/100 g/min. The level of CBF is critical. Complete interruption of CBF produces loss of consciousness within seconds as does a reduction of CBF to approximately 20 ml/100 g/min. Neuronal conversion to anaerobic metabolism occurs below 18 ml/100 g/min and the electroencephalogram becomes flat. Brain cell death (infarction) takes place at about 3 h with flows of 10 ml/100 g/min and after 30 minutes at flows of 5 ml/100 g/min (Table 1).
Table 1Normal cerebral physiological values
750 ml/min or 15% of cardiac output
50 ml/100 g/min
90 ml/100 g/min
20 ml/100 g/min
CMRO2 (grey matter)
3 ml/100 g/min
CMRO2 (white matter)
1 ml/100 g/min
30 mg/100 g/min or 25% of total body consumption
CBF, cerebral blood flow; CMRO2, cerebral metabolic rate for oxygen; CMRGl, cerebral metabolic rate for glucose.
The perfusion pressure (i.e. the arteriovenous pressure gradient) in the brain is more complex than that of other organs because it is confined within an incompressible vault. It is dependent on the pressure difference between the mean arterial pressure (MAP) or the driving pressure (measured at brain level) and ICP or the pressure that needs to be overcome to supply adequate blood to the brain. This pressure difference is known as the CPP. A normal CPP is 70–80 mmHg; the threshold for critical ischaemia is 30–40 mmHg. As can be seen from the equation below, even at normal levels of MAP, an elevated ICP of more than 20 mmHg will compromise CPP and therefore reduce cerebral blood flow. This emphasizes the importance of maintaining an adequate MAP in circumstances such as head injury to ensure adequate perfusion.
Control of cerebral blood flow
Various mechanisms exist to maintain an adequate basal supply of blood and therefore substrate to meet the energy demands of the brain. Physiological variables, such as arterial carbon dioxide and oxygen levels, influence CBF and local regulatory mechanisms direct blood flow to regions of the brain that are particularly active (i.e. blood flow is coupled to local metabolic needs) (Box 1).
Carbon dioxide is a potent vasodilator of cerebral blood vessels. As PaCO2 rises, there is a linear increase in CBF between 3.5 kPa (26 mmHg) and 8 kPa (60 mmHg) (Figure 2).
Above 8 kPa cerebral vessels are maximally dilated and no further increase in vessel diameter is possible; conversely, at a PaCO2 of 3 kPa cerebral vessels are maximally constricted. The effect of hypocarbia on cerebral vasculature is achieved by an increase in brain hydrogen ion (H+) concentration and takes place in minutes, reflecting the time taken for the conversion of CO2 to HCO3− and H+ in the perivascular space. The vasoconstricting effect of a low PaCO2 is progressively attenuated by a fall in the brain's bicarbonate level, which normalizes pH. The opposite is true for prolonged hypercarbia.
Arterial oxygen tension (PaO2) and oxygen content
CBF is directly responsive to changes in oxygen delivery and remains unaltered until a threshold PaO2 of 6.8 kPa (50 mmHg) is reached. Below this threshold CBF dramatically rises (Figure 3).
This corresponds to the steep part of the oxyhaemoglobin dissociation curve (i.e. CBF is responsive not to PaO2 but to oxygen content). This effect explains the modest increase in CBF (an increase of 10%) when breathing 100% oxygen.
Autoregulation is the maintenance of a constant CBF despite variations in CPP. Under normal conditions when both ICP and cerebral venous pressure are low, systemic arterial perfusion pressure (i.e. MAP) is the primary determinant of CPP. Between a MAP of 50 mmHg and 150 mmHg the mean CBF remains constant at 50 ml/100 g/min. However, autoregulation has its physiological limits, above and below which CBF is directly related to perfusion pressure (see Figure 4).
Autoregulation is achieved by alterations in cerebrovascular resistance (CVR) (occurring over 10–60 s) caused by myogenic reflexes to transmural tension in the resistance vessels; as CPP increases from 50 to 150 mmHg, cerebral arterioles constrict and therefore restrict increases in CBF. Autoregulation may be modified by sympathetic nervous system activity. Chronic hypertension or sympathetic stimulation shifts the autoregulatory curve to the right, whereas sympathetic blockade or cervical sympathectomy shifts the curve to the left. Symptoms of ischaemia occur only when the MAP falls below 60% of the lower autoregulatory limit. Above the upper autoregulatory limit, mechanisms such as forced cerebral arteriolar dilatation, reversal of hydrostatic gradients and cerebral oedema result in increases of cerebral blood volume and ICP. Autoregulation is disrupted in the presence of intracranial pathology, hypoxaemia, hypercarbia, fixed vascular obstructions (e.g. carotid atheroma), and volatile anaesthetic agents.
The brain requires more energy than any other organ in the body. Glucose is the main energy source. Cerebral metabolic rate for glucose (CMRGl) is about 30 mg/100 g/min, approximately 25% of the body's total glucose consumption. The majority of this energy is used to maintain ion gradients across neuronal membranes via the Na+/K+-ATPase ion pumps. Glucose crosses the blood–brain barrier via the GLUT 1 transporter and then enters cells via the appropriate glucose transporter (i.e. GLUT 1 to astrocytes, GLUT 3 to neurons and GLUT 5 to microglial cells).
Under normal aerobic conditions, glucose contributes to tricarboxylic acid (TCA) cycle and oxidative phosphorylation to provide ATP for energy. The remainder is converted to amino acids, proteins and lipids. Under hypoxic conditions the glucose receptors are upregulated. Glucose is anaerobically metabolized by glycolysis to form lactate, which is converted to pyruvate in neurons and can then be used in the TCA cycle. Lactate can be actively removed across the blood–brain barrier via a monocarboxylate transporter. During fasting the brain utilizes ketone bodies (exported from the liver) which are broken down to acetyl coenzyme A (acetyl-CoA). This is oxidized via the TCA cycle to yield energy. Gluconeogenesis can also occur in such conditions. If CBF ceases, glycogen reserves can be exhausted within 2 min. Hypoglycaemia results in cerebral cellular dysfunction, manifesting as anxiety and confusion, convulsions and eventually coma.
Flow metabolism coupling and local chemical regulators
Local neuronal activity causes an increase in CMRO2 and CMRGl, and is accompanied by an increased regional CBF to match glucose and oxygen use with delivery. The parallel change in CBF with CMRO2 and CMRGl is known as flow-metabolism coupling. There is evidence to suggest that CBF may be modulated by changes in glucose consumption rather than oxygen consumption under hypoxic conditions. Vasoconstriction occurs by the action of free calcium ions, thromboxane (a product of arachidonic acid/endoperoxidase metabolism) and endothelin (secreted by endothelial cells by endothelin-converting enzyme acting on endothelin A receptors in vascular smooth muscle). Some calcium-channel blockers blunt hypoxic vasodilation and prevent adenosine release. Potent vasodilators include perivascular potassium (released in high concentrations during seizures, hypoxia, and electrical stimulation), adenosine (an ATP metabolite produced in response to arterial hypotension and hypoxia), as well as prostaglandins (e.g. PGE2 and prostacyclin (PGI2)), lactate, acetylcholine, serotonin, substance P and nitric oxide.
Other factors affecting cerebral blood flow and intracranial pressure
Raised cerebral venous pressure
An elevated cerebral venous pressure reduces cerebral venous drainage, expands cerebral blood volume, disrupts capillary Starling's forces leading to cerebral oedema, raises intracranial pressure and therefore reduces CBF. It may be caused by obstruction of venous drainage by neck compression from neck collars or tracheal tube ties, from head-down positioning (e.g. during central line insertion) or from increased intrathoracic pressure (e.g. from coughing, straining, incomplete muscle relaxation and the application of positive end-expiratory pressure during positive pressure ventilation).
Haematocrit is the main determinant of blood viscosity and of oxygen content (and therefore oxygen delivery). CBF changes inversely with whole blood viscosity. Within the normal range of haematocrit, this has a minimal effect on CBF. However, in certain situations when CBF is pathologically decreased (e.g. cerebral vasospasm following subarachnoid haemorrhage), a decrease in haematocrit by haemodilution may improve CBF.
Hypothermia (body temperature of less than 35°C) reduces CMRO2 and CMRGl, and as a result of flow-metabolism coupling, CBF decreases. The converse is true for hyperthermia up to a body temperature of 42°C, above which neuronal damage occurs with a corresponding reduction in oxygen uptake. For every 1°C change in body temperature CBF changes by 5%. At 18°C, the cerebral metabolic rate is so low that safe circulatory arrest (e.g. during cardiac surgery) is possible without causing ischaemic cerebral damage. This effect forms the basis of clinical trials of hypothermia in head injury and cerebrovascular surgery.
Autonomic nervous system regulation
Cerebrovascular vessels have a rich innervation. Large intracranial and pial vessels have a nerve supply originating from autonomic and sensory ganglia. These contain many vasoactive transmitters, which seem to have a role in the regulation of CBF. Although the exact function of the various neurons is difficult to define, it seems that parasympathetic stimulation produces cerebral vasodilation and sympathetic stimulation produces vasoconstriction. Similarly, smaller intracerebral arterioles have rich innervation with many neurotransmitters.
The cerebral physiology described forms the basis of management of patients undergoing intracranial neurosurgery and on neurocritical care. In both situations, maintenance of normal intracranial physiology and the prevention of brain injury are key. For example, to prevent secondary injury after traumatic brain injury careful maintenance of MAP, oxygenation and control of PaCO2 are required. ICP, cerebral blood flow and cerebral oxygenation and metabolic status may be measured for diagnosis and to optimize therapy and develop individualized treatment plans for patients. For example, in cerebral vasospasm following aneurysmal subarachnoid haemorrhage, brain tissue oxygenation may be measured and used to guide therapy (oxygenation and blood pressure management) and as a prognostic indicator. Similarly, ICP monitoring and brain tissue oxygenation may be used in the management of traumatic brain injury.
Monitoring of cerebral blood flow and function
(see article on Principles of intraoperative neurophysiological monitoring and anaesthetic considerations on pp 39–44 of this issue)
A variety of monitoring techniques are used singularly or together to assess and monitor cerebral physiology in patients who cannot be clinically assessed due to intracranial pathology such as traumatic brain injury or intracranial haemorrhage.
Measurement of cerebral blood flow
A number of techniques for the measurement of CBF have emerged since the pioneering method of Kety and Schmidt in 1945. Many techniques now allow measurement of regional blood flow, giving useful information about changes in blood flow in diseased parts of the brain.
Kety-Schmidt technique: The Fick principle states that the blood flow through an organ can be measured by determining the amount of an inert substance (Q) removed from the bloodstream by the organ per unit time, and dividing that value by the difference between the concentration of the substance in arterial blood [A] and the concentration in the venous blood [V] from the organ.
Kety and Schmidt used nitrous oxide (N2O) as the inert substance. Patients breathed 15% N2O for 10 minutes while serial samples were taken from a peripheral artery and the jugular venous bulb and analysed for N2O content until equilibrium was reached. The total value of N2O taken up was determined by calculating the N2O content of jugular venous blood at equilibrium. This technique has a number of disadvantages: the N2O assay is time consuming and tedious and, when low CBF exists, the technique underestimates CBF. Most importantly, it gives a value for global CBF but not regional CBF. A variation of the Kety-Schmidt technique substitutes heat for the inert substance allowing a thermodilution principle to assess CBF.
Xenon-133 wash-out: Regional cortical blood flow can be measured by monitoring the decay of inhaled radioactive isotope xenon-133 by positioning scintillation counters over the head. The slope of the wash-out curve of the radioactive tracer is proportional to the CBF under the detector. The technique provides a two-dimensional analysis of regional CBF but primarily evaluates cortical blood flow. Three-dimensional resolution can be achieved using CT reconstruction in a technique called single-photon emission CT (SPECT).
Other imaging techniques: Since metabolism is so tightly coupled to blood flow, the uptake of 2-deoxyglucose can be used to estimate regional blood flow. If 2-deoxyglucose is labelled with a positron emitter (e.g. oxygen-15, fluorine-18, carbon-11), its uptake can be followed using positron emission tomography (PET) scanning and CBF can be estimated.
Functional MRI (fMRI) can produce brain function maps, where changes in brain neuronal activity are reflected in changes in regional CBF. Tomographic images can also be produced without the use of external contrast agents by a technique called blood oxygen-level-dependent (BOLD) fMRI. Regional neuronal activation, especially in the cerebral cortex, triggers an influx of oxygenated haemoglobin that decreases the regional deoxy haemoglobin levels, causing an increased MRI signal intensity.
Transcranial Doppler ultrasonography (TCD) involves the application of a low-frequency (2 MHz) pulse range-gated ultrasound beam to the thin-boned trans-temporal window, allowing assessment of the middle and anterior cerebral arteries. Flow velocities within these vessels can be determined using the Doppler effect. If the angle of the ultrasound beam and the diameter of the vessel remain constant, relative changes in flow velocity correlate closely with changes in CBF.
Monitoring of cerebral oxygenation
Jugular bulb oximetry: The jugular bulb is a dilatation of the internal jugular vein just below the base of the skull and may be catheterized using the Seldinger technique. Blood may be sampled for oxygen tension and saturation, giving an indication of CBF (lower values reflecting greater uptake by the brain and therefore less blood flow, assuming O2 consumption remains constant). The major disadvantage of this technique is that only global CBF can be estimated and not regional changes. In addition, if CBF and oxygen consumption both decrease (e.g. in severe brain injury), jugular venous saturation may be unchanged.
Intracerebral microdialysis: This technique involves the insertion of a fine catheter, containing a dialysis membrane perfused with Ringer's solution, into brain parenchyma. It enables molecules involved in cerebral metabolic pathways to be directly monitored and this can provide information on the adequacy of cerebral oxygenation and blood flow. Metabolites such as glucose, pyruvate, lactate, glutamate and glycerol or drugs (e.g. phenytoin) diffuse into the solution of the probe from the interstitial fluid (extracellular space) across the membrane and are analysed. The lactate:pyruvate ratio reflects regional cerebral oxygen availability, and has been used clinically in the treatment of head injury and subarachnoid haemorrhage. A rise in the ratio suggests that anaerobic metabolism due to insufficient regional cerebral blood flow is occurring (secondary to hypoxia); treatment may then be taken to correct this impaired physiology.
Cerebral oxygen partial pressure: Sensors can be inserted into the brain parenchyma to measure the partial pressure of oxygen in the extracellular fluid of the brain (pBRO2); this reflects the availability of oxygen for oxidative metabolism. Values obtained generally reflect the balance between oxygen delivery and consumption. Correct siting of the probe is essential to measure focal pBRO2. The per haematoma site in TBI is ideal but is technically challenging and increases the risk of further injury. In routine practise, the non-dominant frontal lobe is the chosen site for probe insertion. This technique is also used in research to optimize the treatment of subarachnoid haemorrhage and traumatic brain injury.
Near infra red spectroscopy: This monitoring technique is based on the principle that light with wavelengths in the near infra red region (650–900 nm) transmits through biological tissues. It is increasingly being used to image biological events in the cerebral cortex. Photons produced by a laser photodiode are directed into the skull, and while many are reflected and dispersed, some are transmitted. Certain coloured compounds within the tissues (chromophores), especially oxyhaemoglobin, deoxyhaemoglobin and oxidized cytochrome oxidase, have characteristic absorption spectra. The emergent light intensity is detected and a computer converts the changes in light intensity into changes in chromophore concentration. Clinical applications of this technique include monitoring of cerebral oxygenation, CBF and volume.
Measuring intracranial pressure
The commonly used methods of monitoring ICP are ventricular catheters (EVD) and intracranial pressure bolts. An EVD measures global ICP and also therapeutically diverts CSF. These require high levels of training for insertion, and have the risk of infection and brain injury. Microtransducer systems or parenchymal catheters measure focal ICP. It is important to consider the location of the catheter when assessing the measured ICP.