Cardiovascular system
Stroke and TIA are markers of general atherosclerosis. Many patients presenting for CEA will have concomitant coronary artery disease and up to 20% have a history of MI. In one reported series following over 9000 patients, cardiac complications (either myocardial, unstable angina, pulmonary oedema or ventricular tachycardia) occurred in 3.9% and was associated with a greater risk of post- operative stroke or death. Te cardiac risk is further increased by other associated medical conditions such as hypertension and obesity. Te high prevalence of coronary artery disease, as determined by history, ECG or cardiac catheterization present in over 55% of these patients, is responsible for the increased risk of post-operative MI (5%) when compared with those patients without coronary artery disease (0.5%). Evidence of cardiac disease should be sought by careful history and thorough examination, noting the presence of angina and its severity, previous MI, and symptoms and signs of cardiac failure. Te ECG should be examined for abnorm alities of rhythm and evidence of previous infarction and ischaemia. When indicated, a chest radiograph is examined for evidence of cardiac failure. Further cardiac work-up, including an exercise ECG, radionuclide studies or coronary angiography, may be necessary and is best coordinated with a cardiologist. In order to assess the risk to patients with coronary artery disease having non-cardiac surgery, a number of risk indexes have been proposed, the most common of which are the American Society of Anesthesiologists (ASA) index, the Goldman index, the Detsky index and the Revised Cardiac Risk (RCR) index. However, there are two scores specific to CEA – the Tu score and the Halm score. Tese scores have been compared for their ability to predict complications specifically following CEA. Te Halm score CEA-specific risk model and the RCR index were found to be most favourable. In addition the Vascular Anaesthesia Society of Great Britain and Ireland has produced an on-line risk calculator to help in the assessment of risk to your patient.
Hypertension must be well controlled in the pre-and post-operative period. Poorly controlled hypertension (blood pressure >170/95) is associated with post-operative hypertension and transient neurological deficits. Terefore, unless urgent, surgery should not be performed in poorly controlled hypertensive patients, and sudden normalization of blood pressure should be avoided to reduce the risk of hypoperfusion and stroke. In some unstable patients, combined coronary artery bypass and CEA may be necessary and are discussed later in this topic.
Neurological system
Evaluation of the cerebrovascular system should document carefully the presence of transient or permanent neurological deficit. Tis is essential for assessing postoperative progress as well as quantifying perioperative risk of stroke. Frequent daily TIAs, multiple neurological deficits secondary to cerebral infarctions or a progressive neurological deficit increase the risk of a new post-operative neurological deficit. Te results of tests assessing the cerebral vascular system such as a Duplex ultrasound scan, cerebral angiography and carbon dioxide reactivity should be available.
Respiratory system
Chronic obstructive pulmonary disease is often present in these patients and needs optimal medical treatment pre-operatively, which may include bronchodilators, corticosteroids, physiotherapy and incentive spirometry. Cigarette smoking should be stopped 6-8 weeks pre-operatively. If necessary, pre-operative pulmonary function tests such as peak expiratory flow rate, FVC/ FEV1 (forced vital capacity/forced expiratory volume in 1 s) ratio and a baseline arterial blood gas analysis with the patient breathing air should be carried out to guide perioperative care of the patient.
Endocrine system
Diabetes mellitus has been shown to exist in about 20% of patients presenting for CEA and most of these patients are insulin dependent. Adequate blood glucose control with absence of ketoacidosis pre-oper-atively must be established. In experimental studies, even modest elevations in blood glucose have been shown to augment post-ischaemic cerebral injury. Manifestations of diabetes mellitus such as renal failure, silent MI, autonomic and sensory neuropathy, and ophthalmic complications must be looked for.
Medications
It is very important that the patient’s pre-operative medications are reviewed. hese patients are often receiving cardiac and antihypertensive drugs, antiplate-let agents, antacids, steroids, insulin and anticoagulants. Antihypertensives should be given, but the anaesthetist needs to be aware of and anticipate the hypotension that may occur on induction of anaesthesia. A sliding scale may be required to ensure tight control of blood glucose levels. Antiplatelet and anticoagulant agents are common in this population; usually aspirin is continued, but warfarin is discontinued. he need for clopidogrel should be discussed with the surgeon, but it should be continued for carotid stenting .
Pre-medication
A good rapport should be established with the patient in the pre-operative period. his will help to reduce anxiety, which may exacerbate perioperative blood pressure abnormalities with increased risk of myocar-dial ischaemia and cardiac arrhythmias. An anxiolytic pre-medicant is especially important in those patients undergoing the procedure under regional or local blockade. Regional anaesthesia allows neurological assessment during and immediately following the procedure but necessitates judicious use of pre-operative sedation. A balance must be struck between adequate sedation and ‘oversedation’ as the latter depresses neurological function. Oversedation often leads to hypoventilation with carbon dioxide retention and blood pressure abnormalities, often with detrimental effects on the cerebral circulation. Benzodiazepine is used routinely in our institution for pre-medication.
Regional or local versus general anaesthesia
The type of anaesthetic used seems to depend on individual practice rather than hard evidence. Local anaesthesia or a cervical plexus block allows evaluation of neurological status during carotid cross-clamping to assess the need for shunting and therefore prevention of stroke from hypoperfusion. However, periopera-tive strokes are more likely to be embolic than low flow in origin. Other potential advantages include a lower incidence of post-operative hypertension and a reduced need for vasoactive drugs with a shorter stay in the ICU. Unfortunately, this technique has numerous disadvantages. It requires patient cooperation and the ability to remain supine for the duration of the procedure. Many of the patients presenting for CEA are unable to lie flat and not cough for the duration of surgery. he procedure may be uncomfortable for the patient, many ofwhom would prefer to be unaware during surgery. Anxiety, especially with the proximity of the surgical drapes, may lead to hyperventilation with a concomitant reduction in CBF and increased risk of cerebral ischaemia. Autonomic responses to surgical manipulation of the carotid bulb may be excessive, resulting in hypotension, hypertension or bradycardia. Tere is also an ever-present risk of airway obstruction, as well as the occurrence of nausea and vomiting. Uncontrolled haemorrhage or sudden neurological deterioration may require general anaesthesia with rapid tracheal intubation. Nevertheless, when used properly in carefully selected patients by experienced surgeons, regional anaesthesia has a good safety record and is not associated with any increase in the rate of perioperative MI.
Regional or local anaesthesia
The patient is attached to all the standard monitors as for general anaesthesia. An appropriate dose of sedation is given. Regional anaesthesia is achieved with a deep cervical plexus block. Tis may be performed by a single injection or a multiple injection technique (performed by the surgeon). For the single injection technique, the patient is placed supine with the head turned to the opposite side. Te area is prepped and draped. Te lateral margin of the clavicular head of the sternocleidomastoid muscle is identified at the level of C4 (level with the superior margin of thyroid cartilage). Te middle and index fingers are rolled laterally over the anterior scalene muscle until the interscalene groove, between the anterior and middle scalene muscle, is palpated. Asking the patient to lift the head off the table slowly may further enhance the groove. After raising a skin wheal with 1% lidocaine, a short-bevel needle is inserted between the palpating fingers, perpendicular to all levels and slightly caudad in direction until paresthesia is elicited. Ultrasound can be used to confirm position and identify the nerves. After careful aspiration, 5-6 ml of local anaesthetic suitable for the duration of surgery is injected (1% lidocaine or 0.5% bupivacaine with 1:200,000 epinephrine). Te local anaesthetic should spread in the fascial sheath extending from the cervical transverse processes to beyond the axilla, investing the cervical plexus in between the middle and anterior scalene muscles. Te slight cau-dad direction is important, because, should the nerve not be encountered, advancing the needle in this direction is less likely to result in epidural or subarach-noid puncture, as this complication is prevented by the transverse process of the cervical vertebra. Tere is no need to perform a superficial cervical plexus block with this technique, as the nerve roots are already anaesthetized. It may be more comfortable for the patient, who is going to have his or her head turned laterally intraop-eratively, if 5 ml of local anaesthetic is deposited below the attachment of the sternocleidomastoid muscle, thus anaesthetizing the accessory nerve. Local infiltration by the surgeon may be required if the upper end of the incision is in the trigeminal nerve area or if the midline is crossed. Judicious administration of intravenous midazolam or propofol can provide sedation without compromising the ability to evaluate the patient’s neurological function.
Possible complications of interscalene cervical plexus block include epidural, subarachnoid and intervertebral artery injection, which can be minimized by the caudad direction of the needle and by repeated aspiration before injecting the local anaesthetic. Hoarseness may occur if the recurrent laryngeal nerve is blocked, and Horner’s syndrome if the cervical sympathetic chain is blocked. Te lower roots of the brachial plexus may also be blocked by spread of local anaesthetic. Local infiltration with or without superficial cervical plexus block has been used. A large volume of local anaesthetic is required and the results are not as satisfactory as deep cervical plexus block.
General anaesthesia
These patients in general have a tendency for extreme blood pressure lability under general anaesthesia. However, general anaesthesia reduces cerebral metabolic demand and may offer some degree of cerebral protection. It also allows for the precise control and manipulation of systemic blood pressure and arterial carbon dioxide tension to optimize CBF. Several techniques are available and the precise one used depends on the experience and preference of the anaesthetist. A balanced general anaesthesia that maintains the blood pressure at the pre-operative level is preferred to ‘deep’ general anaesthesia that may necessitate the use of vasopressors to maintain blood pressure, as the risk of myocardial ischaemia may be increased in the latter.
Induction
The aim is to maintain cerebral and myocardial per-fusion as close to baseline values as possible. A pre-induction intra-arterial line is useful to monitor blood pressure during and after induction. Anaesthesia can be induced in several ways. After pre-oxygenation, fentanyl or remifentanil, and thiopentone or propo-fol are given in incremental doses, titrated against the patient’s haemodynamic responses. Muscle relaxation is achieved using a cardiostable non-depolarizing agent such as vecuronium, and a peripheral nerve stimulator is used to monitor the neuromuscular junction. To obtund the intubation response, lidocaine 1-1.5 mg kg-1 may be given 2-3 min before laryngoscopy and intubation. When muscle relaxation is complete, laryn-goscopy and intubation are performed. After confirmation of tracheal tube placement by breath sounds and end-tidal capnometry, the tube is secured away from the operative side. Some surgeons may prefer nasotra-cheal placement of the tube to allow maximum extension of the neck and therefore better exposure. Te lungs are ventilated to maintain adequate arterial oxygen saturation and normocarbia.
Maintenance
Once again, the aim is to provide stable cerebral perfu-sion while minimizing stress to the myocardium. Tis may be provided with either total intravenous anaesthesia or inhalational anaesthesia, and muscle relaxation. Te use of isoflurane is associated with a lower critical CBF needed to maintain a normal EEG, as well as a lower incidence of ischaemic EEG changes compared with halothane and enflurane. It may also serve to offer myocardial protection from ischaemic damage. Sevoflurane is another inhalational agent that can be used successfully in this patient population. In addition to its low blood gas solubility coefficient allowing early awakening, sevoflurane maintains cerebral autoregulation, and has minimal direct cerebral vascular effects , Total intravenous anaesthesia with propo-fol and a fentanyl or remifentanil infusion can also be used. Remifentanil is an ultra-short-acting synthetic opioid that can cause hypotension and bradycardia in overdose. It has been used successfully in combination with propofol to provide anaesthesia and allows a rapid change of anaesthetic depth as needed. Even though it is a powerful analgesic while administered, there is a concern about hyperalgesia and subsequent hypertension following its use. Adequate longer-acting analgesics must be given early in the anaesthetic to avoid this problem. Alternatively, fentanyl can be used. Whichever technique is chosen, the regimen should allow early awakening so that neurological function can be assessed.
Before carotid cross-clamping, heparin (75-100 units kg-1) is administered intravenously. Application of the carotid cross-clamp is often associated with an increase in blood pressure. Mild increases in blood pressure up to about 20% above pre-operative levels are acceptable, but excessive increases should be controlled. Heparin is generally not reversed after closure of the artery, but if the surgeon is not satisfied with haemostasis at the time of wound closure, a small dose of protamine (0.5 mg kg-1) maybe given.
Blood pressure and arterial carbon dioxide management
Arterial blood gases are checked after tracheal intubation and when necessary during the surgical procedure to ensure adequate oxygenation and normocapnia. Hypercapnia should be avoided, as it will only vasodi-late blood vessels supplying the normal brain without affecting those supplying the ischaemic regions, which are presumably already maximally dilated. Tis diverts blood from the ischaemic areas to the normal areas (intracranial steal), thus further aggravating cerebral ischaemia. Hypocapnia, on the other hand, may increase flow to the ischaemic area by constricting the blood vessels in the normal areas (Robin Hood effect) and could be beneficial. However, hypocapnia causes a global reduction in CBF, and it is generally accepted that the changes in CBF associated with changes in carbon dioxide in these individuals are unpredictable. Normocapnia is therefore preferred.
The pre-operative blood pressure of individual patients should give guidance to a ‘target’ mean arterial blood pressure in the perioperative period. Cerebral autoregulation in these patients may be impaired, and the autoregulation curve is shifted to the right in uncontrolled or poorly controlled hypertensive patients. As autoregulation may be lost completely in ischaemic areas, maintaining an adequate blood pressure is a critical factor in the maintenance of CBF. If the blood pressure decreases below the individual patient’s normal level, ‘lightening’ anaesthesia by reducing isofl urane concentration or decreasing the propofol infusion rate within acceptable limits should be done before using a vasopressor. Use of a vasopressor to elevate blood pressure during the cross-clamp may be necessary, but it has been shown to induce ventricular dysfunction and may increase the incidence of MI. If necessary, a phenylephrine (0.1-0.5 (g kg-1 min-1) infusion can be administered judiciously. On the other hand, patients who remain hypertensive during anaesthesia may require intravenous hypotensive agents (nitroglycerine infusion 1-5 (g kg-1 min-1) for control. Surgical manipulation of the carotid sinus may cause a marked alteration in heart rate and blood pressure. Tese reflexes can be minimized by prior local infiltration with lidocaine. Should EEG changes occur shortly after infiltration, one must be aware that the injection may have been into the carotid artery, resulting in transient CNS lidocaine toxicity.
Monitoring
Cardiovascular and respiratory monitoring in addition to routine monitoring, an intra-arterial cannula is placed under local anaesthesia, before induction, to continuously monitor blood pressure throughout the perioperative period and facilitate the sampling of arterial blood gases. In low-risk patients, a rapidly cycling non-invasive blood pressure device can be used during induction, and invasive monitoring sited once the patient is anaesthetized but before the start of surgery. Central venous pressure, pulmonary capillary wedge pressure, cardiac output and urine output may be indicated in high-risk cardiac patients. End-tidal carbon dioxide, checked against an arterial sample obtained after induction, helps to maintain normocapnia in ventilated patients.
Central nervous system monitoring
No special monitoring is required in awake patients operated on under regional anaesthesia. Under general anaesthesia, brain function has been monitored in a number of ways, summarized in Table 13.5 below.
Electrophysiological monitoring
The unprocessed EEG displays voltage as a function of time. Proper use of this technique is tedious and time-consuming, and interpretation of the raw EEG is not easy – certainly in the setting of an operating theatre. Furthermore, at the usual rate of 25 mm s-1. a 270 m strip of paper is produced for a 3 h case. Nevertheless, intraoperative neurological complications have been shown to correlate well with EEG changes indicative of ischaemia. Ipsilateral or bilateral attenuation of high-frequency amplitude or development of low-frequency activity seen during carotid cross-clamping is indicative of cerebral hypoperfusion. he computer-processed EEG and somatosensory evoked potential have also been found to be useful.
This allows less experienced observers to concentrate on how the parameters are changing with respect to time instead of trying to mentally analyse them. Although computer-processed EEGs are easier to interpret, they have been shown to be less accurate than the 16-channel EEG. Nonetheless, the use of EEG monitoring has still not decreased the incidence of perioperative stroke.
Somatosensory evoked potentials
Somatosensory evoked potentials (SSEPs) have been shown to be useful during CEA, yet the need for computer averaging means that this technique does not provide a continuous real-time monitor. Stable anaesthesia must be maintained to minimize the influence of anaesthetic agents on the amplitude. In general, >50% reduction or complete loss of amplitude of the cortical component is considered to be a significant indicator of inadequate cerebral perfusion. In contrast to conventional EEG, SSEPs monitor the cortex as well as the subcortical pathways in the internal capsule, an area not reflected in the cortical EEG. Not all studies agree that the use of SSEPs is either sensitive or specific for cerebral ischaemia and they are not therefore universally used.
Measurement of stump pressure (internal carotid artery back pressure)
As one important determinant of CBF is perfusion pressure, it seems reasonable to assume that the distal arterial pressure in the ipsilateral hemisphere during carotid occlusion would provide some indication of collateral CBF. Stump pressure represents the mean arterial pressure measured in the carotid stump (the ICA cephalad to the common carotid cross-clamp) after cross-clamping of the common and external carotid arteries, and the retrograde pressure transmitted along the ipsilateral carotid artery from the vertebral and contralateral carotid arteries. Stump pressure has been found useful to indicate intraoperative cerebral ischaemia and direct the use of shunt placement in a large case series and also when compared with near-infrared spectroscopy (NIRS) and transcranial Doppler ultrasonography (TCD).
Intraoperative measurement of cerebral blood flow
Intraoperative CBF measurement has also been used to determine the need for placement of shunts, but the associated cost makes it prohibitive for general use. his involves the intra-arterial injection of 20 mCi of the inert radioactive gas 1 33Xe and measuring the wash-out of p emissions by an extracranial collimated sodium iodide scintillation counter focused on the parietal cortex. Te initial slope or fast component of the wash-out curve relates directly to regional blood flow. Newer measurement techniques involve single-photon emission CT (SPECT) of inhaled xenon. Both techniques are useful as research tools, but very few centres have the equipment and expertise required to produce accurate results .
Table 13.5 Summary of available central nervous system monitoring
|
Monitor |
Advantages |
Disadvantages |
|
Awake patient |
Continuous neurological assessment Avoids the risks of general anaesthesia Lower incidence of post-operative hypertension |
Requires patient cooperation, ability to lie flat, anxiety, hyperventilation with potential risk of cerebral ischaemia Risk of autonomic disturbances, nausea, |
|
Shorter ICU stay |
vomiting and airway obstruction |
|
|
EEG (16-channel) |
Gold standard |
Cumbersome, difficult to interpret Not suitable for theatre environment |
|
EEG (computer processed), cerebral function monitor, digital subtraction angiography, etc. |
Easier to use than 16-channel |
More than one channel needed for reasonable detection of ischaemia |
|
Less cumbersome set-up |
||
|
Embolic events not easily detectable |
||
|
Somatosensory evoked potentials |
Can detect subcortical ischaemia |
Cumbersome |
|
Intermittent monitor with ‘time lag’ |
||
|
Affected by anaesthetic agents |
||
|
Stump pressure |
Measures retrograde perfusion pressure |
Unreliable |
|
Does not reflect regional blood flow |
||
|
Easy to perform |
||
|
Cheap |
||
|
Regional cerebral blood flow |
Measures cerebral blood flow |
Expensive |
|
Invasive |
||
|
Requires steady state |
||
|
Intermittent |
||
|
Transcranial Doppler ultrasound |
Continuous |
Not as sensitive as EEG |
|
Non-invasive |
Measures flow velocity and not cerebral blood flow |
|
|
Relatively easy to use |
||
|
Can be used pre-, intra- and post-operatively |
Failure rate of 5-10% due to lack of ultrasonic window |
|
|
Detects emboli |
||
|
Detects shunt malfunction |
||
|
Near-infrared spectroscopy |
Continuous |
Extracranial contamination a problem |
|
Non-invasive |
No defined ischaemic thresholds yet |
|
|
Easy to use |
