The intraoperative recording of electroencephalographic (EEG) activity began soon after the development and introduction of the EEG as a neurodiagnostic technique. One of the initial clinical applications of the intraoperative recording of cerebral electrical activity was electrocorticography (ECoG) at the time of focal corticectomy for intractable epilepsy. This technique was used before the development of extraoperative EEG monitoring to localize the epileptogenic zone in patients with focal seizures. ECoG is still used at selected epilepsy centers before and after excision of the epileptic brain tissue. The use of ECoG during epilepsy surgery is addressed in Chapter 7 . Other potential indications for intraoperative EEG recordings include the monitoring of cerebral function during carotid endarterectomy and cardiopulmonary bypass surgical procedures. The rationale for intraoperative EEG monitoring is the detection of electrophysiologic alterations that are associated intimately with cerebral ischemia or cerebral hypoperfusion before the development of cerebral infarction. Extracranial or scalp-recorded EEG monitoring during carotid endarterectomy has been shown to be a reliable indicator of cerebral ischemia. The findings obtained with extracranial EEG monitoring at the time of an endarterectomy may lead to an alteration in operative strategy (i.e., may indicate the need for placement of a carotid artery shunt). Intraoperative EEG recordings have also been performed during cardiac bypass surgery. The effect of hypothermia, however, significantly limits the potential utility of EEG monitoring during the latter procedure. Profound hypothermia produces a suppression of EEG activity that reduces the diagnostic yield of intraoperative recordings in identifying alterations resulting from cerebral ischemia. Reversible causes of ischemia during cardiac bypass surgery occur much less commonly than during a carotid endarterectomy for carotid artery stenosis.
Intraoperative EEG recordings were the first monitoring technique used during carotid and cardiac surgical procedures to minimize the likelihood of a postoperative neurologic deficit that might affect significantly the individual’s quality of life. The potential adverse effects of cardiovascular and cerebrovascular surgery include stroke, cognitive impairment, and a postoperative encephalopathy. Importantly, additional neurodiagnostic studies may be used in the operating room to monitor cerebral perfusion and the effect of anesthesia. Transcranial Doppler, carotid ultrasound, and xenon blood flow studies are performed in some centers during carotid endarterectomy and cardiac surgery. The use of transcranial Doppler in selected patients may be predictive of cerebral ischemia in the absence of appropriate EEG changes. Intraoperative carotid ultrasound is being used increasingly to validate the patency of the internal carotid artery after endarterectomy. Monitoring the level of anesthesia by the bispectral index may reduce the hemodynamic changes that occur in some patients undergoing cardiopulmonary bypass.
This chapter provides an overview of some of the aspects of intraoperative EEG monitoring for the identification of changes related to cerebral ischemia during carotid endarterectomy and cardiac surgery. In particular, the clinical applications and potential limitations of such cerebral function monitoring are considered.
Carotid endarterectomy is the cerebrovascular surgical procedure most commonly performed to reduce the risk of stroke in patients with symptomatic carotid artery stenosis related to atherosclerosis. Even before the publication of rigorous scientific studies confirming the efficacy of carotid endarterectomy, the use of this operative technique to protect against ipsilateral stroke had become popular. The increase in the number of operative procedures performed annually in the United States over several years reflects the interest in identifying a protective surgical approach to stroke. Fewer than 15,000 carotid endarterectomies were performed in 1970, but by 1985 an estimated 100,000 carotid endarterectomies had been performed in the United States. The rationale for this operative procedure is the removal of atherosclerotic thrombotic material that may come to restrict flow and lead to either an occlusion of the carotid artery or an artery-to-artery embolus. Carotid artery surgery is the direct result of observations in the 1950s that established a relationship between extracranial internal carotid artery disease and stroke.
Studies have indicated that carotid endarterectomy is effective in reducing the risk of ipsilateral stroke in patients with high-grade stenosis (70 to 99 percent) of the internal carotid artery. Selected patients with completed strokes may also be candidates for a carotid endarterectomy, depending on their neurologic deficits and the presence of coexistent medical problems. The North American Symptomatic Carotid Endarterectomy Trial (NASCET) demonstrated a significant reduction in the risk of stroke in patients with carotid artery stenosis greater than 70 percent. Ipsilateral stroke occurred at 24 months’ follow-up in 26 percent of 328 nonsurgical patients and in 9 percent of 331 patients undergoing a carotid endarectomy. A direct correlation existed between surgical benefit and the degree of carotid artery stenosis. The benefit of carotid artery surgery in the NASCET for patients with 30 to 69 percent stenosis was less clear. The Veterans Affairs Symptomatic Stenosis Trial (VASST) also showed a significant reduction in the risk of stroke in patients with carotid artery stenosis of greater than 50 percent who underwent a carotid endarterectomy. The Asymptomatic Carotid Atherosclerosis Study (ACAS) reported in 1995 that there is an aggregate risk reduction of 53 percent with surgical treatment for asymptomatic carotid artery stenosis. A total of 1,659 patients were entered into this study. All patients had greater than 60 percent carotid artery stenosis and were randomized to surgical or medical therapy. A significant difference was evident between the two treatment groups at 3 years.
The risks of carotid endarterectomy must be considered in any discussion of the putative beneficial effects of this operative procedure. The morbidity of this treatment depends on a number of factors including surgical expertise; degree of carotid artery stenosis; the presence of cerebral infarction; and coexistent medical problems, especially ischemic heart disease. The published morbidity of carotid endarterectomy has ranged from 0 to 20 percent. In one multicenter study, the perioperative morbidity was 2.2 percent and the mortality was 3.3 percent (combined operative complication was 5.5 percent) in “relatively high-risk patients.” Importantly, angiography has a reported risk of approximately 1 percent.
The operative procedure is performed with the patient receiving general anesthesia for both comfort and safety. A combination of nitrous oxide and isoflurane is used at the Mayo Clinic. Induction is performed routinely with thiopental. The common carotid artery, internal carotid artery, and carotid bifurcation are exposed at the time of surgery. Atherosclerotic changes are most prominent in the proximal internal carotid artery and the carotid bifurcation. Clamping of the internal carotid artery, common carotid artery, and external carotid artery is necessary for an arteriotomy and endarterectomy to be performed. A shunt between the common carotid artery and the internal carotid artery may be placed if the surgical team is concerned that the period of clamping may be associated with cerebral ischemia and perioperative stroke. The attitude of neurosurgical teams concerning the use of shunt placement is variable. At the Mayo Clinic, a shunt is placed only if EEG monitoring of cerebral function or cerebral blood flow studies, or both, suggest a hemodynamic insult that may produce a significant reduction in cerebral blood flow. The routine use of shunts may be associated with an increased risk of cerebral infarction related to artery-to-artery embolus, and it prolongs the duration of the operation. Importantly, a minority of patients exhibit a significant change in cerebral function monitoring indicative of a hemodynamic insult with a diminished cerebral blood flow. The risk of shunt placement in one study was low, with embolism occurring in 2 of 511 patients (0.4 percent).
The introduction of carotid angioplasty and carotid artery stenting for carotid artery stenosis may be an alternative to carotid endarterectomy. The methodology for these therapeutic interventions is similar to the techniques introduced for the treatment of cardiovascular disease. Carotid revascularization has also been used in selected patients with restenosis following a carotid endarterectomy. EEG monitoring is not performed routinely during these procedures because the carotid artery is not occluded and there is no consideration of shunt placement. The role for these alternative techniques in the management of asymptomatic or symptomatic carotid artery disease remains under study.
Intraoperative EEG monitoring: methodology
Intraoperative EEG monitoring is performed commonly during carotid endarterectomy. Computer-assisted digital EEG recordings are now standard during such procedures. There is a diversity of opinion regarding the clinical application of this neurodiagnostic technique in cerebrovascular and cardiac surgery. The “hostile” environment of the operating room often makes intraoperative EEG monitoring technically difficult. Potential problems include electrical interference, which produces significant artifacts; difficulty in ensuring the stable application of the electrodes; and the use of anesthesia and pharmacotherapy that may alter the EEG recording. The technologist and electroencephalographer may also have difficulty in examining the patient and in negotiating their way through the operating room because of the surgical team and the necessary equipment. The technical factors that must be considered during intraoperative extracranial EEG monitoring include proper placement of the scalp electrodes (collodion is used at the Mayo Clinic) before anesthesia induction. Usually, 21 to 23 scalp electrodes are used for intraoperative EEG monitoring. For appropriate intraoperative recordings, at least 8 channels of EEG should be available. In most instances, 16 or 21 channels for EEG monitoring is strongly preferred. Proper grounding is essential for patient safety. The use of a 60-Hz filter is required. The linear frequency is set between 1 and 15 or 30 Hz. Sensitivities of 3 to 5 μV/mm are often necessary for recording the extracranial EEG with the patient under general anesthesia. A longitudinal bipolar anteroposterior montage is used routinely for intraoperative EEG monitoring. The Laplacian montage may also be useful in identifying a focal alteration. Subtle changes in amplitude and frequency related to cerebral ischemia can be visualized easily at the slower recording speeds. The diagnostic yield of the reduced recording speed for identifying reversible alterations associated with cerebral ischemia has been demonstrated. The recording speed can be restored to 30 mm/sec if a continuous or paroxysmal EEG pattern occurs that cannot be identified.
The use of digital EEG recordings in the operating room has removed concerns regarding paper storage during intraoperative EEG monitoring. The introduction of digital EEG has improved offline analysis and allowed individuals remote from the operating room to review the EEG as it is being acquired (i.e., “real-time” review). The current practice at the Mayo Clinic is to continue to use a recording speed of 5 mm/sec for digital EEG intraoperative recordings because of the amount of data generated. The technologists at our institution are also more familiar with the effects of cerebral ischemia displayed at this slower speed. Computer processing with a compressed spectral array can be used for data interpretation but may have no advantage over visual inspection alone.
EEG and anesthesia
It is necessary to review the relationship between anesthesia and the EEG before considering the effects of cerebral ischemia on the intraoperatively recorded EEG. Anesthetic agents may alter the normal background EEG significantly ( Fig. 9-1 ). To some extent, individual anesthetic drugs may have different effects, depending on their concentrations. The EEG patterns produced by different anesthetic agents, when used at concentrations below their MAC level (i.e., the minimal alveolar concentration necessary for preventing movement to a painful stimulus in about 50 percent of subjects), are quite similar. Selected anesthetic agents (e.g., thiopental, halothane, enflurane, isoflurane, and nitrous oxide) produce a similar subanesthetic or anesthetic effect associated with a symmetric EEG pattern.
Symmetric EEG Patterns at Subanesthetic or Minimal Anesthetic Concentrations
Thiopental produces the characteristic drug-induced beta effect at subanesthetic concentrations, which tends to be maximal in the anterior midline distribution. Halothane, enflurane, isoflurane, and 50 percent nitrous oxide, administered at subanesthetic concentrations, produce a similar pattern ( Fig. 9-2 ). The drug-induced beta activity is less prominent with these other agents than with thiopental.
Symmetric EEG Changes with Induction
Inductions for surgery usually are performed with the rapid administration of thiopental. A characteristic pattern of EEG changes occurs in relation to induction. The drug-induced beta activity noted with thiopental at subanesthetic concentrations becomes distributed more widely, gradually increases in amplitude, and slows in frequency. The background frequency ultimately slows from the beta to the alpha range. Paroxysmal bursts of high-amplitude, intermittent slowing occur; these are frontally predominant and resemble frontal intermittent rhythmic delta activity (FIRDA) (see Fig. 9-1 ; Fig. 9-3 ). Induction with other anesthetic agents may produce a similar sequence of EEG changes, but with the FIRDA-like pattern being less prominent.
Symmetric EEG Patterns at Sub-MAC Anesthetic Concentrations
The EEG pattern for all anesthetic drugs at a sub-MAC anesthetic concentration consists of generalized background slowing with anteriorly predominant rhythmic fast activity in the beta or alpha frequency range ( Fig. 9-4 ). Increasing the concentration of anesthetic drug is associated with slowing of the frontal fast activity. The anterior fast activity is nearly continuous in nature and is presumed to represent drug-induced beta activity, which is concentration dependent.