Maintaining the balance between cerebral metabolic supply and demand underpins much of neurocritical care and neuroanaesthetic practice. Cerebral perfusion pressure (CPP) and metabolic rate can be manipulated by a number of means, but these are not benign interventions and do not guarantee the absence of regional oxygenation deficits. A bedside technique for the quantitative, continuous measurement of cerebral oxygen supply and demand is clearly highly desirable.
Such a system should alert the clinician to the presence of regions of failing oxygenation and ideally be sensitive, specific, minimally invasive, robust and easy to use. A number of technologies aimed at detecting oxygen supply/demand imbalance have been developed of which jugular bulb oximetry is the most mature. More recently, near-infrared spectroscopy and direct brain tissue oxygen measurement have become clinically available. hese techniques employ very different measurement principles and, as we shall see, all have their individual shortcomings. However, by combining them with other parameters in a synergistic, multimodal approach, some of these difficulties may be mitigated.
Jugular bulb oximetry
Percutaneous sampling and analysis of human cerebral venous blood from the jugular bulb was first described by Myerson and colleagues in 1927. Arteriovenous differences between oxygen, glucose and lactate content were later studied by Gibbs and colleagues, who proposed them as global measures of the balance between cerebral metabolic supply and demand. Since its introduction as a monitoring technique, the technology for jugular venous sampling and oximetry has been refined, most recently with the introduction of reliable fibre-optic catheters capable of continuous oxygen saturation measurement, making the technique a routine element of multimodal monitoring in neurocritical care.
Deoxygenated blood from the brain is collected by the major cranial venous sinuses. he majority of this blood then drains into the left or right sigmoid sinuses. hese follow a curved path in the posterior fossa, passing caudally and forwards to form an internal jugular vein (IJ V) in the posterior jugular foramen on each side (Fig. 6.1). he paired inferior petrosal sinuses are an exception to this common pathway, emptying instead directly into the IJV. he occipital sinuses are of variable anatomy but typically drain into the vertebral venous plexus rather than the IJV, although this represents only a small contribution to the total blood flow when supine. he venous outflow from left and right hemispheres is mixed with approximately 30% of the IJV blood arising from the contralateral side. In most cases, one side (usually the right) is dominant, although subcortical areas tend to drain preferentially to the left . A small percentage of the blood is extracranial in origin, from anatomically variable emissary veins and the cavernous sinus.
At its origin in the posterior part of the jugular foramen, the IJV is dilated and known as the (superior) jugular bulb. Several veins carrying extracerebral blood, of which the facial vein is the most significant, empty into the IJV a few centimetres below. From its origin, the vein runs down the neck within the carotid sheath, lateral to the vagus nerve and carotid artery, where it can be conveniently cannulated. Caudally, it dilates again, forming the inferior jugular bulb before joining the subclavian vein to become the brachio-cephalic vein behind the medial clavicle at the thoracic inlet. he vein is related anterolaterally to the superficial cervical fascia, platysma, deep cervical fascia and sternomastoid muscle. he transverse processes of the cervical vertebrae, cervical plexus, phrenic nerve and (on the left) thoracic duct lie posteriorly to the vein. A bicuspid valve is typically found just above the inferior bulb, the central venous system cephalad being valveless.
Fig. 6.1. Schematic diagram of the right jugular bulb and internal jugular vein. The majority of the cerebral venous blood drains into the superior jugular bulb either indirectly via the sigmoid sinus or directly in the case of the inferior petrosal sinus. Significant extracerebral contamination of the internal jugular venous blood occurs just below the jugular bulb, most notably from the facial vein.
The brain extracts oxygen from arterial blood at a rate to supply its global metabolic requirements leaving an oxygen-poor venous effluent. he oxygen saturation of this venous blood is related to the cerebral metabolic rate for oxygen (CMRO2) and cerebral blood flow (CBF) by the Fick equation:
he quantitiesrepresent the arterial and jugular venous oxygen contents, respectively. If we rearrange this equation, it is clear that the arterio-venous oxygen difference (AVDO2) is a measure of the ratio
In the healthy brain, flow-metabolism coupling results in the AVDO2 being relatively constant in the range of 4-9 ml dl-1. High values, however, suggest a situation where blood flow is low relative to metabolic requirements. Conversely, low values represent states ofhyper-aemia, where CBF is in relative excess.
For practical reasons, it is helpful to work with blood oxygen saturation rather than content, as this can be measured straightforwardly by optical techniques. Assuming fixed oxyhaemoglobin dissociation characteristics, the arterial and venous saturations, SaO2 and SjO2 . are related to oxygen content and haemoglobin concentration [Hb] by:
The contribution from dissolved oxygen, given by terms involving arterial and jugular venous partial pressures of oxygen PaO2 and PjO2 in this equation, are small and can safely be neglected. Combining these and the two previous equations, we find that the venous oxygen saturation is equivalently given by:
Again assuming constant arterial oxygenation, SjO2 represents the balance between oxygen supply and demand. In health, low-metabolism coupling maintains SjOt between about 55 and 75%. hese values are somewhat lower than mixed venous saturations, reflecting the brain’s relatively high metabolic requirements. Indeed, measured values down to 45% may be normal in health if extracerebral contamination is carefully avoided by angiography-guided catheterization of the jugular bulb.
In circumstances of hypo-perfusion without a corresponding decrease in CMRO2, the brain must extract a greater proportion of arterial oxygen, and SjOt will be observed to fall. However, there is a limit to how far oxygen extraction can be increased. Values of SjO2 below 50% imply that oxygen supply may be critically low for the metabolic demand, and the brain is at risk of ischaemic injury.
Jugular venous oxygenation can be measured intermittently by serial blood sampling or continuously by fibre-optic oximetry. Intermittent sampling has the advantage of low cost and additionally allows AVDO2 and the arteriovenous glucose and lactate difference to be measured by oximetry, co-oximetry and blood biochemical analysis. As venous flow velocities are low, it is essential to draw the sample slowly (<2 ml min-1) to avoid contamination with extracranial blood (Fig. 6.2).
Spectrophotometric catheters measure SjO2 continuously, avoiding the difficulties of blood sampling and also allowing more transient supply/demand mismatch events to be detected. Two systems for continuous oximetry are available: Oximetrix (Abbott Laboratories, North Chicago, IL, USA) and Edslab II (Baxter Healthcare Corporation, Irvine, C A, USA). Both systems consist of narrow-gauge double-lumen catheters. One lumen contains transmit and receive optical fibres by means of which the SjO2 is measured optically. he other lumen is available for blood sampling and may be continuously flushed with saline at 2-4 ml h-1 to maintain patency. he spectrophotometric determination of SjO2 relies on the different absorption properties of oxygenated and deoxygenated haemoglobin in the red and near-infrared range. he Edslab II system alternately sends two different wavelengths down one fibre at millisecond intervals and measures the reflected light. As the absorption also depends on haemoglobin concentration, this must be entered manually into the machine. he device then calculates SjO2 by averaging the reflection measurements over several seconds. he Oximetrix system is similar except that three different optical wavelengths are used, which obviates the need for prior determination of the haemoglobin concentration and may make it more stable under situations where this is rapidly changing.
The current generation of catheters are relatively stiffer and are thus less prone to kinking, which has been a problem in the past. Modern catheters also have an antimicrobial and antithrombotic coating to minimize the reduction in signal that occurs with fibrin deposition over the fibre end, and should require recalibration only every 24-48 h. However, inaccurate or unreliable behaviour can still be observed if the catheter abuts the vessel wall, which can be a particular problem if the patient’s head is not in the neutral position. Despite the advances in continuous SjO2 technology, careful blood sampling and co-oximetry remain the ‘gold standard’ and should be performed before instituting a change of therapy based on the continuous spectrophotometric results alone.
Fig. 6.2. The speed ofwithdrawal from jugular bulb catheters affects accuracy of readings. SjO2 values are higher with faster rates of blood withdrawal due to contamination with extracranial blood. Optimal rate appears to be 2 ml min-1.
Placement of a jugular venous bulb catheter is a quick and safe procedure in experienced hands. Continuous oximetry catheters are introduced through an intravenous sheath, introduced by retrograde cannulation of the IJV using a Seldinger technique. he site is selected, prepared and infiltrated with local anaesthetic using a strict aseptic technique, as for conventional central venous cannulation. Ultrasound visualization, which has been demonstrated to reduce complications in central venous access, may be particularly helpful in avoiding the need for significant head-down tilt, which may otherwise have deleterious effects on intracranial pressure. Conventional anterograde central venous access may also be obtained at the same time by a second venepuncture a short distance caudal to the jugular venous catheter insertion site. In this case, both guidewires can be placed first to avoid accidentally damaging the first catheter while positioning the second line (Fig. 6.3).
Fig. 6.3. The technique used in our unit for inserting jugular bulb catheters. The junction between the sternal and clavicular heads of the sternocleidomastoid muscle is localized and the skin is punctured under ultrasound guidance with a 16 gauge 5.25 inch Angiocath catheter (Becton and Dickinson) mounted with a 5 ml syringe. During gentle aspiration, the needle is passed in a cranial direction for 1-2 cm, at an angle of 15-20° in the sagittal plane. Once the vein is entered, the catheter is advanced over the needle until a slight elastic resistance is felt, or when the tip of the catheter is estimated to be just behind the mastoid process. SCM, sternocleidomastoid muscle; IJV, internal jugular vein; carotid A, carotid artery.
Complications are similar in frequency and nature to those seen in standard central venous access and specifically include those due to the venepuncture itself (such as carotid artery puncture and haematoma formation) and late complications (such as infection and thrombosis). A rise in intracranial pressure (ICP) is rarely if ever seen as a result of jugular venous cannulation. However, subclinical thrombosis is an ultrasound finding in up to 40% of patients with catheters after 6 days.
A number of considerations are noteworthy when choosing which side to monitor from. In health, similar saturations are usually found between the left and right sides. However, the venous drainage is variably lateralized between left and right hemispheres and even between subcortical and cortical regions. It is therefore not surprising that measurements of SjO2 may become asymmetrical with highly focal or unilateral pathology and that this asymmetry is unpredictable from patient to patient.
The dominant side for venous drainage is generally chosen and so must first be determined. Transient manual compression of a dominant IJV leads to a greater rise in ICP compared with the non-dominant side. Alternatively, ultrasonography of internal jugular flow or CT determination of the larger jugular foramen have also been suggested. he dominant system is usually the right, and this offers an alternative pragmatic solution where better information is not available.
Catheter position is important if the measurements are to be reliable and robust. In particular, if the catheter tip is too caudal, significant error may occur as a result of admixture with extracranial blood from the facial vein: a 2 cm displacement may result in up to 10% contamination. hus, once the catheter has been placed and secured, its position should be confirmed with an anteroposterior (AP) radiograph or appropriately penetrated lateral film. he catheter tip is usually positioned cephalad of the level of the inferior border of C1 on the lateral view. CT studies suggest that the tip is optimally placed if it is cephalad of both a line connecting the tips of the mastoid processes and of the level of the atlanto-occipital joint but caudal to the transverse plane through the level of the inferior margin of the orbit on the AP view (Fig. 6.4).