PATHOPHYSIOLOGY OF CEREBRAL ISCHEMIA
The brain is very vulnerable to ischemic injury because of its relatively high oxygen consumption and near-total dependence on aerobic glucose metabolism (see earlier discussion). Interruption of cerebral perfusion, metabolic substrate (glucose), or severe hypoxemia rapidly results in functional impairment; reduced perfusion also impairs clearance of potentially toxic metabolites. If normal oxygen tension, blood flow, and glucose supply are not quickly reestablished, under most conditions ATP stores are depleted, and irreversible neuronal injury begins. When CBF decreases below 10 mL/100 g/min, cell function is deranged, and ion pumps fail to maintain cellular vitality. The ratio of lactate to pyruvate is increased secondary to anaerobic metabolism. During ischemia, intracellular K+ decreases and intracellular Na+ increases. More importantly, intracellular Ca2+ increases because of failure of ATP-dependent pumps to either extrude the ion extracellularly or into intracellular cisterns, increased intracellular Na+ concentration, and release of the excitatory neurotransmitter glutamate. Glutamate acts at the NMDA receptor, further enhancing Ca2+ entry into the cell, hence the potential benefit of NMDA blockers for neuroprotection.
Sustained increases in intracellular Ca2+ activate lipases and proteases, which initiate and propagate structural damage to neurons. Increases in free fatty acid concentration and cyclooxygenase and lipoxygenase activities result in the formation of prostaglandins and leukotrienes, some of which are potent mediators of cellular injury. Accumulation of toxic metabolites impairs cellular function and interferes with repair mechanisms. Lastly, reperfusion of ischemic tissues can cause additional tissue damage due to the formation of oxygen-derived free radicals. Likewise, inflammation and edema can promote further neuronal damage, leading to cellular apoptosis.
STRATEGIES FOR BRAIN PROTECTION
Ischemic brain injury is usually classified as focal (incomplete) or global (complete). Global ischemia may result from total circulatory arrest as well as global hypoxia. Cessation of perfusion may be caused by cardiac arrest or deliberate circulatory arrest, whereas global hypoxia may be caused by severe respiratory failure, drowning, and asphyxia (including anesthetic mishaps). Focal ischemia includes embolic, hemorrhagic, and atherosclerotic strokes, as well as blunt, penetrating, and surgical trauma.
In some instances, interventions aimed at restoring perfusion and oxygenation are possible; these include reestablishing effective circulation, normalizing arterial oxygenation and oxygen-carrying capacity, or reopening and stenting an occluded vessel. With focal ischemia, the brain tissue surrounding a severely damaged area may suffer marked functional impairment but still remain viable. Such areas are thought to have very marginal perfusion (<15 mL/100 g/min), but, if further injury can be limited and normal flow is rapidly restored, these areas (the “ischemic penumbra”) may recover completely. When these interventions are not applicable or available, the emphasis must be on limiting the extent of brain injury.
From a practical point of view, efforts aimed at preventing or limiting neuronal tissue damage are often similar whether the ischemia is focal or global. Clinical goals are usually to optimize CPP, decrease metabolic requirements (basal and electrical), and possibly block mediators of cellular injury. Clearly, the most effective strategy is prevention, because once injury has occurred, measures aimed at cerebral protection become less effective.
Hypothermia is a suggested method for protecting the brain during focal and global ischemia. Indeed, profound hypothermia is often used for up to 1 h of total circulatory arrest. Unlike anesthetic agents, hypothermia decreases both basal and electrical metabolic requirements throughout the brain; metabolic requirements continue to decrease even after complete electrical silence. Additionally, hypothermia reduces free radicals and other mediators of ischemic injury. Induced hypothermia has shown benefit following cardiac arrest and is a routine part of most post-arrest protocols for comatose patients.
Barbiturates, etomidate, propofol, isoflurane, desflurane, and sevoflurane can produce burst suppression, and all but desflurane and sevoflurane can produce complete electrical silence of the brain and eliminate the metabolic cost of electrical activity. Unfortunately, these agents have no effect on basal energy requirements. Furthermore, with the exception of barbiturates, their effects are nonuniform, affecting different parts of the brain to variable extents.
Ketamine may also have a protective effect because of its ability to block the actions of glutamate at the NMDA receptor.
No anesthetic agent has consistently been shown to be protective against global ischemia. Studies highlighting the potential neurotoxicity of anesthetics (especially in infants) also question the role of volatile anesthetics in neuroprotection.
Nimodipine plays a role in the in the treatment of vasospasm associated with subarachnoid hemorrhage.
General patient management techniques are the neuroanesthesia interventions most likely to improve patient outcomes.
Maintenance of a satisfactory CPP is critical. Hypotension, increases in venous pressure, and increases in ICP should be avoided. Oxygen-carrying capacity should be maintained and normal arterial oxygen tension preserved. Hyperglycemia aggravates neurological injuries following either focal or global ischemia, and blood glucose should be maintained at less than 180 mg/dL. Normocarbia should be maintained, as both hypercarbia and hypocarbia have no beneficial effect on cerebral ischemia; hypocarbia-induced cerebral vasoconstriction may aggravate the ischemia, whereas hypercarbia may induce a steal phenomenon (with focal ischemia) or worsen intracellular acidosis.
EFFECT OF ANESTHESIA ON ELECTROPHYSIOLOGICAL MONITORING
Electrophysiological monitors are used to assess the functional integrity of the CNS. The most commonly used monitor during neurosurgical procedures is evoked potentials. EEG is less commonly used. Proper application of these monitoring modalities is critically dependent on recognizing anesthetic-induced changes. Both monitoring modalities are described in Chapter 6.
The effects of anesthetic agents on the EEG are summarized in Table 26–2.
TABLE 26–2Electroencephalographic changes during anesthesia. ||Download (.pdf) TABLE 26–2 Electroencephalographic changes during anesthesia.
|Activation ||Depression |
|Inhalational agents (subanesthetic) ||Inhalation agents (1–2 MAC) |
|Barbiturates (small doses) ||Barbiturates |
|Benzodiazepines (small doses) ||Opioids |
|Etomidate (small doses) ||Propofol |
|Nitrous oxide ||Etomidate |
|Ketamine ||Hypocapnia |
|Mild hypercapnia ||Marked hypercapnia |
|Sensory stimulation ||Hypothermia |
|Hypoxia (early) ||Hypoxia (late) Ischemia |
EEG monitoring is useful for assessing the adequacy of cerebral perfusion during carotid endarterectomy (CEA), as well as anesthetic depth (most often with processed EEG). EEG changes can be simplistically described as either activation or depression. EEG activation (a shift to predominantly high-frequency and low-voltage activity) is seen with light anesthesia and surgical stimulation, whereas EEG depression (a shift to predominantly low-frequency and high-voltage activity) occurs with deep anesthesia or cerebral compromise. Most anesthetics produce activation (at subanesthetic doses) followed by dose-dependent depression of the EEG.
Isoflurane, desflurane, and sevoflurane produce a burst suppression pattern at high doses (>1.2–1.5 MAC). Nitrous oxide is unusual in that it increases both frequency and amplitude (high-amplitude activation).
Benzodiazepines can produce both activation and depression of the EEG. Barbiturates, etomidate, and propofol produce a similar pattern and are the only commonly used intravenous agents capable of producing burst suppression and electrical silence at high doses. Opioids characteristically produce only dose-dependent depression of the EEG. Lastly, ketamine produces an unusual activation consisting of rhythmic high-amplitude theta activity followed by very high-amplitude gamma and low-amplitude beta activities.
Somatosensory evoked potentials test the integrity of the spinal dorsal columns and the sensory cortex and may be useful during resection of spinal tumors, instrumentation of the spine, and carotid artery and aortic surgery. The adequacy of perfusion of the spinal cord during aortic surgery is better assessed with motor evoked potentials (which assess the anterior part of the spinal cord). Brainstem auditory evoked potentials test the integrity of the eighth cranial nerve and the auditory pathways above the pons and are used for surgery in the posterior fossa. Visual evoked potentials may be used to monitor the optic nerve and occipital cortex during resections of large pituitary tumors.
Interpretation of evoked potentials is more complicated than that of the EEG. Evoked potentials have poststimulus latencies that are described as short, intermediate, and long. Short-latency evoked potentials arise from the nerve stimulated or the brainstem. Intermediate- and long-latency evoked potentials are primarily of cortical origin. In general, short-latency potentials are least affected by anesthetic agents, whereas long-latency potentials are affected by even subanesthetic levels of most agents. Visual evoked potentials are most affected by anesthetics, whereas brainstem auditory evoked potentials are least affected.
Intravenous agents in clinical doses generally have less marked effects on evoked potentials than do volatile agents, but, in high doses, can also decrease amplitude and increase latencies (see Chapter 6). Ketamine generally increases the amplitude of short-latency signals.
CASE DISCUSSION Postoperative Hemiplegia
A 62-year-old man has undergone a right carotid endarterectomy (CEA). Immediately following surgery, in the recovery room, he is noted to be weak on the contralateral side. How is a patient undergoing CEA evaluated preoperatively?
Patients with carotid stenosis are at very increased risk of coronary artery and peripheral arterial disease. It would be unusual for a patient to have carotid stenosis who did not have evidence of atherosclerosis elsewhere. Patients undergoing CEA, therefore, require a preoperative cardiac evaluation, according to American College of Cardiology/American Heart Association guidelines.
With respect to patient risk factors, the guidelines provide algorithms for how patients should be evaluated and managed intraoperatively. As part of this patient’s preoperative evaluation, a thorough neurological examination should have been performed with special attention paid to motor function. This patient may well have been weak on the left side prior to surgery, in which case the hemiparesis might be due to a preexisting condition. If this is a new finding, it requires aggressive management. Is general or regional anesthesia the optimal anesthetic technique for managing patients undergoing CEA?
For the past several decades, the majority of patients undergoing CEAs in the United States have had general anesthesia. General anesthesia was chosen because many surgeons operating in the neck area felt more comfortable if the airway was controlled, and the patient was completely anesthetized should evidence of cerebral ischemia develop.
More recently, regional anesthesia has been advocated as providing an adequate surgical field, a comfortable and relaxed patient (if done with monitored anesthesia care), stable hemodynamics, and ideal monitoring of cerebral function during crossclamping because an awake patient provides the best evidence of adequate cerebral perfusion. The patient can indicate or be observed for evidence of aphasia, facial droop, or hemiparesis. Regional anesthesia is usually performed with superficial cervical plexus blocks. How should cerebral function be monitored intraoperatively in this patient?
When the carotid is cross-clamped, the ability to identify inadequate cerebral circulation in the ipsilateral hemisphere is critical, as there is a window of opportunity for immediate intervention and correction of any deficit.
Global and focal neurological status can continuously be assessed in awake patients if they are mildly sedated when undergoing regional anesthesia. In such a situation, practical assessment consists of frequent (every 2–5 min) examination of strength using the contralateral handgrip and maintenance of constant verbal contact (“cocktail conversation”) with the patient to assess level of consciousness.
In patients undergoing general anesthesia, indirect cerebral monitoring techniques have been used to assess the adequacy of the cerebral circulation. These techniques include stump bleeding, stump pressure, jugular venous oxygen saturation, EEG, and transcranial Doppler (TCD). Back bleeding of the distal carotid artery following crossclamp and incision of the artery suggests reasonable collateral circulation above the clamp. It is very subjective and nonquantitative.
To better qualify and quantify the adequacy of collateral perfusion (Figure 26–9), stump pressure measurements can be used. Some surgeons believe that a shunt should be used in all patients with a previous cerebrovascular accident, independent of stump pressure, and for any patient whose stump pressure is less than 25 mm Hg. However, this is controversial, as many neurosurgeons and vascular surgeons use 50 mm Hg as a cutoff. The reliability of stump pressure to predict the need for selective shunting has also been questioned. Some surgeons routinely shunt all patients, some shunt no patients, and others use selective shunting. Outcome data have not identified the best surgical approach.
The EEG is sometimes used for monitoring patients undergoing CEA under general anesthesia. In such a circumstance, inhalation or intravenous anesthesia can influence the EEG, but gross changes associated with carotid clamping can be detected. However, analyzing subtleties of the EEG is labor and technology intensive.
Evoked potentials have also been employed during CEA. Should neurophysiological studies identify cerebral ischemia, the surgeon can place a vascular shunt during the surgical repair to provide for ipsilateral cerebral perfusion. How should hemodynamics be controlled intraoperatively?
During carotid clamping and immediately afterward in the recovery room, patients are often hemodynamically labile. Bradycardia can develop during surgical manipulation of the carotid sinus because of the direct stimulation of the vagus nerve. Tachycardia may develop as a result of stress or pain or as a direct result of manipulation of the carotid sinus with release of catecholamines into the circulation.
Hypotension is also observed because of the direct vasodilating and negative ionotropic effects of anesthetic agents. Hypotension following carotid unclamping is common, particularly in patients with more severe carotid stenosis. This could be due to a cerebral protective process. Cerebral autoregulation protects the brain from reperfusion by reducing production of renin, vasopressin, and norepinephrine, which results in hypotension. Hypertension is also a frequent finding in patients undergoing CEA. Many patients have hypertension as a comorbid condition, which is often further exacerbated by the surgical stress and manipulation of the carotid body, which causes release of catecholamines and sympathetic stimulation.
Invasive arterial pressure monitoring and suitable venous access to infuse vasoactive medications are necessary during carotid surgery. What is the most likely etiology of this patient’s findings?
This patient most likely has had a cerebrovascular accident due to an arterio-to-arterial embolus; more than 95% of such patients will fit into this category. Weakness can also develop as a result of a hyperperfusion syndrome, which occurs in patients with severe carotid stenosis who have now reestablished flow to the affected cerebral hemisphere. Such patients usually have a greater than 95% carotid stenosis with a less than 1-mm channel in the affected carotid artery. Typically, the syndrome does not develop in the postoperative anesthesia care unit (PACU), but several hours afterward when the patient begins complaining of a headache, and, in severe cases, develops hemiparesis.
Because a cerebrovascular accident is most likely, when the anesthesiologist is called to see such a patient in the PACU, a thorough neurological examination quantifying any cranial nerve involvement and the degree of weakness on the contralateral side should be performed. Any hemodynamic changes need to be treated immediately, with assurance of adequate hemoglobin and oxygenation levels. The surgeon needs to be notified at once, and ultrasonic evaluation of the carotid artery is frequently required to determine whether there might be problems with the intimal suture line. It may be necessary to return to the operating room to explore the carotid artery.
The cerebral circulation.