INTRODUCTION — Carotid atherosclerosis is a common cause of stroke. Carotid endarterectomy (CEA) and carotid artery stenting (CAS) are both established revascularization interventions [1-3]. This topic will review the anesthetic management of patients undergoing elective CEA or CAS, utilizing the options of either general anesthesia or local/regional anesthesia [4,5].

The indications for carotid revascularization and the surgical considerations in selecting treatment with CEA or CAS are reviewed elsewhere [1,2,6]. (See "Management of asymptomatic extracranial carotid atherosclerotic disease" and "Management of symptomatic carotid atherosclerotic disease" and "Carotid endarterectomy" and "Overview of carotid artery stenting".)

PREANESTHESIA CONSULTATION

Cardiovascular comorbidity — A major focus of the preanesthesia evaluation is to detect and optimize cardiac conditions. Many patients presenting for carotid interventions have coexisting coronary atherosclerosis and increased risk for perioperative morbidity due to myocardial ischemia [7]. A 12-lead electrocardiogram should be routinely performed for all patients undergoing carotid endarterectomy (CEA) or carotid artery stenting (CAS). Further consultation with a cardiologist may be warranted in patients with severe or unstable heart disease. (See "Evaluation of cardiac risk prior to noncardiac surgery" and "Noncardiac surgery in adults with aortic stenosis".)

Medications — Appropriate perioperative medication management reduces cardiovascular risk and minimizes thrombotic complications.

●Statins and beta blockers should be continued in patients already receiving these therapies. (See "Management of cardiac risk for noncardiac surgery", section on 'Statins' and "Management of cardiac risk for noncardiac surgery", section on 'Beta blockers'.)

●Recommendations regarding starting new statin therapy prior to carotid endarterectomy are discussed separately. (See "Carotid endarterectomy", section on 'Statins' and "Overview of carotid artery stenting", section on 'Statin therapy'.)

●Aspirin and/or clopidogrel therapy are recommended prior to CEA or CAS. (See "Carotid endarterectomy", section on 'Antiplatelet therapy' and "Overview of carotid artery stenting", section on 'Dual antiplatelet therapy'.)

GENERAL VERSUS LOCAL/REGIONAL ANESTHESIA — Carotid revascularization can be performed using either general anesthesia or local/regional anesthesia. For carotid artery stenting (CAS), most procedures are performed with local anesthesia at the arterial puncture site. For carotid endarterectomy (CEA), ideally, surgical and anesthetic teams should be competent in both techniques because a patient might prefer, or there might be a medical reason to choose, one anesthetic technique rather than another. In an analysis of 26,070 cases in the American College of Surgeons (ACS) National Surgical Quality Improvement Program (NSQIP) database, general anesthesia was used in 84.6 percent and regional anesthesia was used in 15.4 percent of cases [8,9]. Local/regional anesthesia may be more beneficial for some patients but can be uncomfortable for the patient and may necessitate urgent conversion to general anesthesia or urgent shunt placement.

Anesthetic goals — Regardless of anesthetic technique, the most important anesthetic goals for carotid revascularization are to:

●Ensure that the patient is promptly awake and able to fully cooperate with a neurologic examination for early stroke detection during the procedure with the local/regional anesthetic technique and at the end of the procedure for both anesthetic techniques.

●Avoid wide variations in blood pressure and/or heart rate throughout the procedure.

Advantages of local/regional anesthesia — The major advantage of using a local/regional anesthetic technique in an awake patient is the ability to continuously monitor neurologic function by simply talking to the awake patient and asking him to perform basic tasks with the contralateral hand, rather than using EEG or other continuous neuromonitoring techniques to detect brain ischemia during a carotid intervention [10,11]. If the patient develops a new neurologic deficit, the surgeon or interventionalist can intervene to mitigate neurologic injury. However, sedation must be minimized so that the patient is able to give a reliable neurologic response. (See 'Local/regional anesthesia technique' below.)

Another advantage of local/regional anesthesia for CEA is a lower incidence of hypotension during and after the procedure. Also, these techniques allow for selective surgical placement of a carotid shunt, rather than the routine shunting that many surgeons employ during general anesthesia. However, selective shunting based on continuous neuromonitoring techniques is also possible with general anesthesia. Finally, local/regional anesthesia may incur lower costs, with shorter intraoperative times and hospital stays [12-14]. (See "Carotid endarterectomy", section on 'Routine versus selective shunting' and 'Neuromonitoring' below.)

Advantages of general anesthesia — General anesthesia is more comfortable for the patient and may be favored if the preoperative assessment reveals anxiety, reluctance to be awake, inability to cooperate or communicate, neurocognitive dysfunction, or inability to lie supine comfortably (eg, congestive heart failure or severe chronic obstructive pulmonary disease).

In addition, use of general anesthesia at the outset avoids the need for urgent conversion from a local/regional to a general anesthetic technique, which can be problematic for the patient undergoing CEA.

Evidence review — The choice of general anesthesia versus local/regional anesthesia is determined in consultation with the surgeon, considering the individual patient's characteristics and preferences. Most available evidence for CEA suggests that the choice of anesthetic technique has no significant impact on clinically important outcomes, although some data suggest that local or regional anesthesia is beneficial [4,8,12,15-19].

●A 2018 pooled meta-analysis that included 21 observational and 12 randomized trials with a total of >58,000 patients, local anesthesia was employed in >14,000 patients and was associated with lower incidences of stroke (odds ratio [OR] 0.66, 95% CI 0.55-0.80), transient ischemic attack (OR 0.52, 95% CI 0.38-0.70), myocardial infarction (OR 0.55, 95% CI 0.41-0.74), and mortality (OR 0.73, 95% CI 0.56-0.94), compared with general anesthesia in >44,000 patients [15]. However, a separate meta-analysis that included only the 4453 patients enrolled in the 12 randomized trials noted no differences between local/regional versus general anesthesia in stroke, myocardial infarction, mortality, or any other outcome. Some studies have noted potential advantages of local/regional anesthesia with respect to other outcomes, such as transient postoperative cognitive decline (POCD) and duration of hospital stay [4].

●A subsequent randomized trial in 210 consecutive patients who had baseline magnetic resonance imaging (MRI) before undergoing CEA noted a higher risk of new postoperative cerebral infarctions detected on MRI scans in patients who had general anesthesia (17.1 percent), compared with those who had local anesthesia (6.7 percent) [16].

●A later 2019 retrospective study that included 18,945 patients who had general anesthesia during CEA reported a higher incidence of postoperative pneumonia and greater need for blood transfusions compared with 3,809 propensity-matched patients who had regional anesthesia [17].

●An older 2008 Cochrane review, which was updated in 2013, identified 14 randomized trials comparing local/regional with general anesthesia and included 4596 operations, of which 3526 were from the General Anesthesia versus Local Anesthesia (GALA) trial [5,18]. The incidence of stroke was not significantly different between the groups (local: 3.2 percent; general: 3.5 percent). There was a trend toward lower perioperative mortality with local anesthetic (0.9 versus 1.5 percent, OR 0.62, 95% CI 0.63-1.07). The NSQIP analysis mentioned above also found no significant differences for the composite outcome of perioperative (30 day) stroke/myocardial infarction (MI)/death among the 4046 patients receiving local anesthesia and 20,670 patients receiving general anesthesia (2.4 versus 2.2 percent). Patients undergoing general anesthesia were more likely to stay in the hospital longer than one day after surgery.

The results of the Cochrane meta-analysis are weighted heavily by the results of the GALA trial [8]. Important findings of this study include:

•The primary outcome measure of the GALA trial was a composite of stroke (including retinal infarction), MI, or death between randomization and 30 days after surgery.

•There was no significant difference in the primary outcome between the local and general anesthesia groups (4.4 versus 4.8 percent). However, 9.5 percent of patients randomized to one arm received the opposite treatment. An as-treated analysis was performed removing the crossovers for the composite outcome (no differences), but individual outcomes (stroke, death, MI) were not assessed.

•There was no difference between the groups (intention-to-treat analysis) in the risk of postoperative stroke within 30 days of surgery. Local/regional anesthesia was associated with a trend toward decreased mortality (OR 0.62, 95% CI 0.36-1.07). No significant differences were found for other measures, including myocardial infarction, postoperative bleeding, pulmonary complications, or length of stay. In the NSQIP study above, adjusted analysis identified general anesthesia as a risk factor for postoperative myocardial infarction (OR 2.18, 95% CI 1.17-4.04) [9].

•A post-hoc analysis of the GALA trial found that local anesthesia may be more cost effective than general anesthesia [13]. The difference in cost was primarily due to differences in the length of intensive care unit stay and the use of consumables, with a mean cost difference of ₤178 ($285 US). Two other nonrandomized studies have similarly suggested a cost difference between the two anesthesia techniques [12,14]. The larger of these evaluated 24,716 patients from the NSQIP database, finding modestly shorter operative and anesthesia times, and a larger proportion of patients discharged on the first postoperative day in the local compared with general anesthesia groups (77 versus 64 percent) [12]. However,

Perioperative blood pressure is affected by anesthetic technique. One systematic review identified nine trials that recorded blood pressure during and after carotid endarterectomy [20]. Blood pressure dropped significantly in the general anesthesia group after induction of anesthesia, and, in one trial, more patients in the general anesthesia group had significant hypotension during or after the operation (25 versus 7 percent). The GALA trial found that more patients undergoing general anesthesia required manipulation of blood pressure compared with patients receiving local anesthetic (72 versus 54 percent) [8]. However, intraoperative and postoperative blood pressure responses are highly variable with hypertension and hypotension reported for local and general anesthesia.

Overall, general anesthesia is more commonly used in the United States for carotid endarterectomy (CEA) [9,21]. Carotid artery stenting (CAS) is more typically performed with local/regional anesthesia [9-11,21,22]. Ideally, the anesthesiology team is competent and comfortable with either technique because there might be specific medical, surgical, or patient-centered reasons that make one anesthetic technique more advantageous than the other [10]. (See "Carotid endarterectomy", section on 'Anesthesia'.)

A 2017 meta-analysis that included five randomized controlled clinical trials and 12 cohort studies demonstrated that short- and long-term costs are similar for CAS versus CEA in contemporary practice [23]. Higher procedural costs for CAS were balanced by higher postprocedural costs for CEA. Taken together, the evidence suggests that cost considerations should not be a significant factor in choosing anesthetic technique or type of carotid revascularization technique.

GENERAL ANESTHESIA TECHNIQUE — Carotid endarterectomy (CEA) and carotid artery stenting (CAS) are typically relatively short procedures, lasting less than 90 minutes, but duration depends on the technique selected, and the surgeon. Thus, short-acting medications are selected when general anesthesia is administered in order to facilitate a rapid anesthetic emergence and prompt awakening for participation in a neurologic examination [4].

Induction and maintenance of general anesthesia

●Induction – Induction of anesthesia should use a short-acting agent such as etomidate (0.15 to 0.3 mg/kg) or propofol (1 to 2.5 mg/kg) that is slowly injected and titrated for effect. The addition of a low-dose, short-acting opioid (eg, fentanyl [1 to 2 mcg/kg] or remifentanil [1 mcg/kg]) and/or lidocaine (50 to 100 mg) blunts the hypertension and tachycardia response associated with sympathetic stimulation during endotracheal intubation [24,25]. (See "General anesthesia: Intravenous induction agents".)

●Maintenance – Anesthetic maintenance can be achieved utilizing either a volatile or intravenous anesthetic technique [26]. The use of volatile versus intravenous techniques (or their combination) is usually based on the preference and experience of the anesthesiologist. If neuromonitoring is employed, the neuromonitoring team may request a ceiling concentration for a volatile inhalation anesthetic agent or a total intravenous anesthetic (TIVA) technique. (See 'Neuromonitoring' below and "Neuromonitoring in surgery and anesthesia", section on 'Anesthetic strategy'.)

•Volatile anesthetic agents – When a volatile anesthetic is selected for carotid revascularization, we suggest desflurane or sevoflurane because the low solubility of these agents facilitates a more rapid emergence from anesthesia than isoflurane [27,28] or propofol infusion [29]. (See "Inhalation anesthetic agents: Clinical effects and uses".)

If neuromonitoring is used, the volatile anesthetic is typically administered at a dose that is ≤0.5 to 1 minimum alveolar concentration (MAC) to avoid signal suppression that may interfere with detection of brain ischemia [30] (see "Neuromonitoring in surgery and anesthesia", section on 'Volatile inhalation agents' and "Neuromonitoring in surgery and anesthesia", section on 'Anesthetic strategy'). Such low concentrations are maintained and significant adjustments in volatile anesthetic dose are avoided in the treatment of hemodynamic aberrations. (See 'Hemodynamic management' below.)

•Intravenous anesthetic agents – Maintenance of general anesthesia using a TIVA technique (eg, propofol and remifentanil in combination) is also a common practice for carotid interventions. Clinical studies demonstrate equivalence of TIVA with volatile anesthetics for maintenance of general anesthesia, but no major advantages [26,31-33]. (See "Neuromonitoring in surgery and anesthesia", section on 'Intravenous agents'.)

•Nitrous oxide – While nitrous oxide (N₂O) can be used as an adjunct for general anesthesia in patients undergoing carotid revascularization, the following clinical issues have been raised [34-36] (see "Inhalation anesthetic agents: Clinical effects and uses", section on 'Nitrous oxide'):

-N₂O significantly increases the risk of postoperative nausea and vomiting [37-39]. It is particularly important to avoid nausea and vomiting after CEA because retching may increase the risk of postoperative neck hematoma. If N₂O is deemed essential to the anesthetic plan, then the risk of nausea and vomiting is reduced by limiting its total dose and use of nausea and vomiting prophylaxis [40,41].

-N₂O administration results in mild myocardial depression and mild sympathetic nervous system stimulation. Heart rate and blood pressure are usually unchanged, but pulmonary vascular resistance may be mildly increased. However, these changes are not associated with increased risk of cardiac complications in noncardiac surgery [39,42-46].

-The effect of N₂O on the EEG is usually to increase frequency, but the effects vary depending on the other agents also being administered. The effect of N₂O on other neuromonitoring modalities is similar to that of the volatile inhalation agents, and it is synergistic when administered with a volatile agent. (See "Neuromonitoring in surgery and anesthesia", section on 'Nitrous oxide'.)

Airway management — We generally prefer endotracheal intubation in patients undergoing CEA with general anesthesia because of limited access to the airway during this procedure. We use a laryngeal mask airway (LMA) for patients undergoing CAS with general anesthesia.

●An endotracheal tube secures the airway and enables reliable controlled mechanical ventilation throughout the procedure. However, placement of an endotracheal tube, as well as manipulation of the head during surgery in an intubated patient, may induce sympathetic stimulation and consequent tachycardia, hypertension, and myocardial ischemia in patients at high risk [7]. Furthermore, during anesthetic emergence and tracheal extubation, coughing and hypertension may cause suture line disruption, carotid hematoma, and life-threatening airway compression after CEA, or may cause femoral hematoma after CAS [47]. Hemodynamic lability during anesthetic emergence can be minimized by extubation under deep anesthesia, followed by gentle mask ventilation, which is continued until the return of full consciousness and adequate spontaneous ventilation.

●An LMA is an alternative airway management device for CEA or CAS. In comparison with endotracheal intubation, this technique has been associated with less hemodynamic lability during anesthetic induction and emergence, because the supraglottic position of the LMA minimizes tracheal irritation [48]. However, during CEA, there is a greater chance of dislodging an LMA than an endotracheal tube.

Ventilation management — Normocapnia should be maintained during general anesthesia in patients undergoing CEA.

●Permissive hypercapnia is avoided. While permissive hypercapnia, achieved by hypoventilating patients under general anesthesia, had been proposed as a method to increase cerebral blood flow and mitigate ischemia caused by cross-clamping the carotid [49], adverse reactions have been reported in patients who experienced intracerebral vascular "steal" during hypercapnia, due to increased blood flow to normally perfused brain tissue [50,51]. In addition, the cerebral vasodilation caused by hypercapnia may increase the risk of cerebral embolization.

●Hypocapnia is avoided because it increases cerebrovascular tone, thereby adversely decreasing cerebral blood flow.

LOCAL/REGIONAL ANESTHESIA TECHNIQUE — Local/regional anesthesia is performed with or without a nerve block, and is usually supplemented with intravenous sedation to maintain patient comfort during the procedure. However, sedation is minimized in order to allow neurologic exams to be performed at frequent intervals during the procedure, as well as at the end of the procedure.

The most commonly used nerve block techniques for carotid endarterectomy (CEA) are superficial or deep cervical plexus blocks [4,52]. We recommend a superficial cervical plexus block rather than a deep plexus block. Evidence from randomized and non-randomized trials suggests that superficial plexus blocks provide adequate anesthesia for CEA, while avoiding potentially serious complications associated with a deep plexus block (eg, subarachnoid injection, Horner syndrome, unwanted blockade of the phrenic, recurrent laryngeal and vagus nerves) [4,53-55]. A 2007 systematic review of 69 studies noted a higher risk of need to convert to general anesthesia (see 'Conversion from local/regional to general anesthesia' below) or development of serious complications (eg, intravascular injection, respiratory distress [due to presumed or confirmed diaphragmatic or vocal cord paralysis]) in 514/7558 patients who had a deep plexus block, compared with 126/2533 patients who had superficial plexus block [54]. Details regarding performance of these blocks, including use of ultrasound guidance, are found in a separate topic [56,57]. (See "Scalp block and cervical plexus block techniques", section on 'Superficial cervical plexus block technique' and "Scalp block and cervical plexus block techniques", section on 'Deep cervical plexus block technique'.)

We use dexmedetomidine (typically 1 mcg/kg initial loading dose, followed by an infusion at 0.3 mcg/kg/hour) to provide sedation in patients undergoing CEA or CAS with local/regional anesthesia. Although dexmedetomidine may theoretically decrease cerebral blood flow by inducing a degree of vasoconstriction within the brain [58], evidence suggests that it is safe and effective in this setting [59-62]. Dexmedetomidine may also result in hypotension, which can be treated with a vasopressor (eg, phenylephrine or ephedrine) (see 'Hemodynamic management' below). Prolonged sedation due to a long duration of action may limit its suitability in some patients. (See "Monitored anesthesia care in adults", section on 'Dexmedetomidine'.)

Although limited data suggest that the quality and duration of a cervical plexus block may be enhanced with addition of fentanyl or clonidine to the local anesthetic, we do not typically use these additives [63,64].

CONVERSION FROM LOCAL/REGIONAL TO GENERAL ANESTHESIA — Indications for conversion from local/regional anesthesia to general anesthesia include patient request, severe agitation, or seizure. The need to convert is a relatively uncommon event, occurring in 4 percent of the patients in the general anesthesia versus local anesthesia (GALA) trial [8]. In a systematic review, the risk of conversion to general anesthesia was significantly higher with deep compared with superficial cervical plexus block (odds ratio [OR] 5.35), possibly due to a higher risk of complications and/or inadequate analgesia with the deep plexus block [54]. (See 'Local/regional anesthesia technique' above.)

Airway access may be compromised by an open incision for carotid endarterectomy (CEA) that is partially covered by surgical drapes on the head and neck. Thus, preoxygenation may not be possible. In such cases, a rapid intravenous induction allows the anesthesiologist to promptly secure the airway. An assistant or the surgeon may aid the anesthesiologist by allowing the head to return a neutral position and partially lifting the drape while maintaining surgical site sterility.

For carotid artery stenting (CAS), the anesthesiologist has free access to the airway if conversion to general anesthesia becomes necessary.

NEUROMONITORING — Neurologic injury from stroke associated with carotid endarterectomy (CEA) and carotid artery stenting (CAS) may be minimized if ischemia is detected in a timely fashion and treated promptly [65,66]. A variety of techniques to monitor cerebral perfusion and detect emboli have been suggested for carotid revascularization surgery.

The gold standard for assessment of brain perfusion is the neurologic examination in an awake patient (ie, a patient receiving little or no sedation). When general anesthesia is used, a variety of neuromonitoring modalities are available; these are classified in the categories of brain activity monitoring, cerebral perfusion monitoring, and brain oxygen saturation monitoring.

●In CEA, neuromonitoring may detect cerebral ischemia (eg, a change in mental status or new weakness in an awake patient or severe EEG changes in a patient under general anesthesia) due to hypoperfusion after carotid clamping or embolism shortly after carotid clamping or unclamping. In such cases, surgical placement of a carotid shunt is the usual treatment. (See "Carotid endarterectomy", section on 'Carotid shunting'.)

●In CAS procedures, neuromonitoring may detect cerebral ischemia due to carotid vasospasm, emboli, or dissection. In such cases, the interventionalist would then perform cerebral angiography to identify a potentially treatable condition, which may be amenable to neurovascular intervention such as thrombus aspiration. (See "Overview of carotid artery stenting" and "Percutaneous carotid artery stenting", section on 'Techniques'.)

Neurologic examination in awake patients — A baseline assessment is performed at the beginning of the procedure and repeated every 10 to 15 minutes during exposure of the carotid arteries, immediately prior to carotid clamping, and continuously during carotid clamping. This assessment consists of noting the answers to simple questions directed to the patient, and asking him to squeeze a team member's hand or a noise-making toy to ensure that contralateral handgrip strength is normal [10,11,67]. Agitation, slurred speech, disorientation, or extremity weakness indicate possible cerebral ischemia, and indicate the need for shunt placement.

Electroencephalography (EEG) with general anesthesia — The aim of brain activity monitoring is to detect decreased perfusion due to carotid artery clamping, embolism, or dissection. Brain activity is typically monitored by continuous electroencephalography (EEG). (See "Neuromonitoring in surgery and anesthesia", section on 'Electroencephalography'.)

EEG monitoring is most commonly used for neurologic assessment in patients undergoing CEA with general anesthesia [68-70]. A meta-analysis of 4664 measurements from 29 studies examined the ability of various monitors to detect cerebral ischemia during CEA performed in awake patients, using clinical neurological assessment as the ideal comparison [71]. The cerebral monitors that were analyzed included EEG, SSEPs, stump pressure, transcranial Doppler (TCD), cerebral oximetry, jugular venous bulb saturation, and jugular venous lactate levels. The best predictor of cerebral ischemia was a combination of a low stump pressure and EEG (or TCD) indicators.

EEG monitoring is initiated prior to induction of anesthesia. Because anesthetic agents may impact the EEG recording, the American Society of Neurophysiological Monitoring recommends a baseline EEG prior to induction of anesthesia. After induction, a second baseline EEG is obtained under general anesthesia before any carotid manipulation has occurred [30]. Subsequently, EEG waveforms obtained from scalp electrodes in multi-channel sets are continuously evaluated throughout the procedure by trained neuromonitoring personnel [30,69]. These personnel will modify aspects of the EEG monitoring technique according to the anesthetic regimen, in an attempt to optimize detection of signal changes that may indicate cerebral ischemia [72]. Continuous communication between the anesthesiologist and the neuromonitoring team is necessary to minimize anesthetic interference with the neuromonitoring goals for an individual patient. (See "Neuromonitoring in surgery and anesthesia", section on 'Anesthetic strategy'.)

Changes in EEG related to ischemia may be classified as mild, moderate, or severe [69]. Severe changes (ie, a greater than 50 percent decrease in waveform amplitude in a generalized or lateralized distribution) indicate a need for safe augmentation of the systemic blood pressure by the anesthesiologist or insertion of a shunt by the surgeon.

Limitations of EEG monitoring include the inability to monitor subcortical structures, difficulties with interpretation, and limited sensitivity [32].

Unprocessed EEG monitoring is preferred over processed EEG (eg, bispectral index [BIS]) for monitoring during carotid revascularization. Processed EEG has limited ability to detect cerebral ischemia because only the frontal cortex is monitored [73]. Use of processed EEG for monitoring depth of anesthesia is discussed separately. (See "Accidental awareness after general anesthesia", section on 'Brain monitoring'.)

Other neuromonitoring modalities — The EEG is the most commonly used neurologic monitor, but other techniques are used according to available expertise and clinician preference.

●Cerebral perfusion monitoring – Measurement of internal carotid artery stump pressure can be used to assess cerebral perfusion during CEA. Transcranial Doppler may be used during either CEA or CAS.

•Carotid stump pressure – Carotid stump pressure obtained during CEA is less commonly used than in the past. One reason is that only a single pressure is obtained, rather than continuous pressure monitoring. However, in one meta-analysis, the combination of EEG monitoring (or TCD monitoring) with carotid stump pressure appeared to be the best combined predictive technique for cerebral ischemia monitoring [71].

•Transcranial Doppler (TCD) – Transcranial Doppler (TCD) uses pulsed wave Doppler to measure blood velocities in the middle cerebral artery. The sensitivity and specificity of this technique is unreliable for detection of cerebral ischemia, since changes in blood flow velocity may reflect changes in arterial diameter rather than blood flow changes.

However, specialized TCD techniques can provide a reliable method to detect and quantify emboli [74]. A higher embolic burden detected by TCD correlates with increased stroke risk [75-77]. In addition, the instantaneous audio feedback that occurs with embolization may guide surgical manipulation of the carotid artery [78,79]. However, the use of this technique has not been associated with improved outcomes.

A disadvantage of TCD is the position of the probes, which may interfere with the surgeon's access to the patient's neck as well as the anesthesiologist's access to the patients' airway.

●Brain oxygen saturation monitoring – Modalities for monitoring brain oxygen saturation during CEA or CAS include jugular venous bulb monitoring and cerebral oximetry.

•Jugular venous bulb monitoring – Jugular venous oxygen saturation (SjVO₂) in blood returning from the cerebral circulation is sampled via a catheter inserted into the internal jugular vein ipsilateral to the carotid surgical site. In a modification of this technique, jugular venous lactate levels are followed as a surrogate marker for developing brain ischemia.

A major limitation of monitoring SjVO₂ is that it is a global rather than a regional monitor of ischemia, due to intracranial mixing of cerebral venous blood. Furthermore, there is a wide range of normal SjVO₂ values (55 to 75 percent) due to inter-individual variability, as well as possible extracerebral blood sample contamination. Evidence suggesting benefit of jugular venous bulb monitoring in the setting of CEA or CAS is lacking [80].

•Cerebral oximetry – Cerebral oximetry utilizes near-infrared spectroscopy (NIRS) to detect regional cerebral oxygen saturation (rSO₂) via an adhesive pad applied to the forehead [81-83]. In one study of 466 patients undergoing CEA, a decrease in rSO₂ ≥20 percent below baseline during temporary carotid clamping predicted perioperative stroke in seven patients, with a sensitivity of 86 percent and a specificity of 57 percent [83]. Also, absolute baseline rSO₂ values that were ≤50 percent before induction of anesthesia predicted stroke with a sensitivity of 91 percent and a specificity of 67 percent [83]. Although NIRS can guide selective shunting during general anesthesia for CEA, further trials are required to determine whether a 15 to 20 percent decrease from baseline values after clamping represents the best trigger for placement of a carotid shunt [84].

Limitations of cerebral oximetry monitoring in carotid revascularization surgery include the wide variation in baseline readings in a single individual and between patients, lack of agreement about the threshold indicating a need for shunt placement or other interventions, and the multiple factors that cause decreased rSO₂ (eg, systemic and regional hemodynamics, blood oxygen transport, tissue metabolism) [82,83,85]. Furthermore, the sensitivity of cerebral oximetry for detection of ischemia is likely limited by its small sampling window in a region of the frontal lobe cortex.

●Somatosensory evoked potentials (SSEPs) – Somatosensory evoked potential (SSEP) monitoring is rarely used for neuromonitoring during CEA. Results from studies of the sensitivity of SSEP monitoring for detection of cerebral ischemia are not consistent [86-88]. A 1998 meta-analysis of 15 studies including a total of 3136 patients concluded that isolated SSEP monitoring poorly predicts perioperative neurologic deficits [89]. A subsequent retrospective study noted that multimodal neuromonitoring with both SSEP and EEG reduced use of selective shunting without increased neurologic risk [90]. (See "Neuromonitoring in surgery and anesthesia", section on 'SSEP Monitoring' and "Neuromonitoring in surgery and anesthesia", section on 'Anesthetic strategy'.)

HEMODYNAMIC MONITORING AND MANAGEMENT

Expected hemodynamic changes — Both carotid endarterectomy (CEA) and carotid artery stenting (CAS) may be associated with extreme lability of both blood pressure and heart rate in the perioperative period, in part due to altered baseline carotid baroreceptor function and intraoperative manipulation of these baroreceptors [91]. (See "Carotid endarterectomy" and "Overview of carotid artery stenting".)

The intraoperative periods of highest risk for hemodynamic instability and resultant myocardial or cerebral ischemia are:

●Induction of anesthesia

•Anesthetic agents used for induction may cause hypotension by removing sympathetic tone, directly decreasing systemic vascular resistance, directly depressing the myocardium, reducing venous return, or inducing bradycardia.

•Endotracheal intubation may induce sympathetic stimulation resulting in tachycardia and hypertension.

●Surgical manipulation of the carotid sinus and carotid artery

•Carotid endarterectomy (CEA) – Surgical manipulation of the carotid sinus during carotid dissection may result in sympathetic stimulation and resultant hypertension and tachycardia or may increase parasympathetic outflow with resultant bradycardia and hypotension. While injection of local anesthetic, either into the carotid body [92,93] or into the periadventitial area around the carotid sinus [94], has been proposed as a method to minimize the bradycardic reflex and reduce hemodynamic lability due to carotid manipulation, there is insufficient data to endorse this practice.

Carotid cross-clamping may precipitate ipsilateral cerebral ischemia due to mechanisms such as decreased carotid blood flow and/or plaque disruption with cerebral embolization. Thus, during cross-clamping, systolic blood pressure should be maintained in a range from the patient's baseline blood pressure to 20 percent above that baseline, to optimize collateral cerebral perfusion [10,11]. We prefer to use systolic arterial blood pressure to achieve this goal, although some clinicians prefer to use the mean arterial pressure.

Carotid unclamping and the ensuing period of reperfusion may be complicated by hypotension [11].

•Carotid artery stenting (CAS) – Balloon expansion during CAS and the resultant endovascular pressure on the baroreceptors reduces sympathetic activity and increases parasympathetic outflow, which may cause bradycardia and hypotension [95]. This usually resolves with administration of atropine (0.2 to 0.4 mg increments). Prophylactic administration of glycopyrrolate may prevent the hemodynamic response seen during carotid bulb manipulation. In some centers, 0.2 mg of glycopyrrolate is administered prior to balloon dilation, and a repeat dose of 0.2 mg is administered, if necessary. (See "Percutaneous carotid artery stenting", section on 'Stent positioning and dilation'.)

●Emergence from anesthesia

•Tracheal irritation immediately before and during extubation may cause coughing and severe hypertension.

Hemodynamic monitoring

●Electrocardiography (ECG) – Continuous ECG monitoring is necessary to detect arrhythmias and/or myocardial ischemia. Depression of the ST-segment >1 mm may indicate myocardial ischemia, prompting interventions to improve myocardial oxygen supply and/or reduce myocardial oxygen demand (eg, raising diastolic blood pressure or decreasing heart rate) [96,97]. Initial interventions may include metoprolol 1 to 5 mg to reduce heart rate or a phenylephrine bolus of 40 to 200 mcg to increase diastolic blood pressure. If evidence of ischemia persists, further management includes establishing hemodynamic stability with vasoactive medications and consultation with cardiology.

Computerized ST-segment trending is superior to visual clinical interpretation for identification of ST-segment changes, and multiple-lead monitoring is more sensitive than single-lead monitoring [7,98,99]. Despite limitations in the sensitivity of ECG monitoring, the presence of intraoperative and postoperative ST-segment changes is associated with cardiac morbidity and mortality in patients at high risk for myocardial ischemia during noncardiac surgery [7].

●Intra-arterial catheter – Invasive arterial blood pressure measurement should be used to monitor moment-to-moment changes in order to rapidly detect and treat hypotension or hypertension. The Australian Incident Monitoring Study established the superiority of direct arterial blood pressure monitoring over indirect monitoring techniques for the early detection of intraoperative hypotension [100]. (See 'Hemodynamic management' below.)

Also, an intra-arterial catheter is useful to guide management of vasoactive drugs (ie, vasopressors and vasodilators) and facilitates arterial access for blood gas measurements.

The intra-arterial catheter is inserted prior to induction of anesthesia. Patients with carotid disease usually have systemic atherosclerotic disease, which can result in blood pressure differences between the arms [101]. To determine which arm is ideal for radial or brachial arterial cannulation, the blood pressure in each arm is checked with a cuff. The arm with the highest blood pressure should be selected for arterial cannulation. If blood pressure obtained by cuff is suspiciously low in both arms, the blood pressure in the legs is checked, and femoral arterial cannulation may be selected. Also, the surgeon should be consulted regarding the possibility of upper extremity sheath placement for CAS before final selection of the artery to be cannulated for direct arterial blood pressure monitoring [102].

●Central venous catheterization – Central venous or pulmonary artery pressure monitoring is rarely indicated during carotid interventions [103].

Hemodynamic management — Blood pressure is controlled throughout the procedure and postoperative period. During carotid cross-clamping or balloon inflation, systolic blood pressure should be maintained in a range from the patient's baseline blood pressure to 20 percent above that baseline, in order to optimize collateral cerebral perfusion [10,11]. This is necessary, even if a shunt is used.

●Vasoactive drugs – The anesthesiologist should have sympathomimetic, vasodilator, and short-acting beta blockers medications readily available throughout the perioperative period in order to rapidly treat hypertension, hypotension, and tachy- or bradyarrhythmias. We suggest that bolus doses of the following drugs be immediately available: phenylephrine (eg, 40 to 100 mcg/mL injected as 40- to 200-mcg boluses), ephedrine (eg, 5 to 10 mg/mL injected as 5 to 20 mg boluses), vasopressin (eg, 1 unit/mL injected as 1 unit boluses), nicardipine (eg, 100 mcg/mL injected as 100 to 500 mcg boluses), labetalol (eg, 5 to 10 mg/mL injected as 5 to 20 mg boluses), esmolol (10 mg/mL injected as 10 to 50 mg boluses), and atropine (0.4 mg/mL injected as 0.2 to 0.4 mg boluses).

In addition, infusion solutions of phenylephrine and nitroglycerin should be prepared and immediately available to administer (table 1 and table 2).

●Fluids – While a bolus of fluids may help achieve normovolemia and normotension, carotid surgery is not usually associated with excessive fluid loss or fluid extravasation. Thus, volume resuscitation with large amounts of fluid is rarely required.

ANTICOAGULATION MANAGEMENT — Anticoagulation during the course of carotid endarterectomy (CEA) or carotid artery stenting (CAS) is recommended and discussed in detail in separate topic reviews. (See "Carotid endarterectomy" and "Overview of carotid artery stenting".)

●Carotid endarterectomy – During CEA, a bolus of intravenous heparin is administered before clamping the carotid artery. A heparin dosing strategy that is guided by activated clotting time (ACT) values is typically employed, with a target ACT of 200 to 250 seconds [104,105]. However, there are few data supporting specific ACT targets for CEA or other major vascular procedures, and the anesthesiologist is generally not required to monitor ACT values due to the short duration of carotid clamping [106]. At the completion of a CEA procedure, reversal of heparin with protamine is suggested to decrease the incidence of serious bleeding. The anesthesiologist should always administer protamine slowly, while monitoring for an adverse reaction (eg, hypotension due to histamine-induced vasodilation or a more severe anaphylactic reaction) [107,108]. (See "Carotid endarterectomy", section on 'Endarterectomy procedure'.)

●Carotid artery stenting – During CAS, the patient is anticoagulated, typically with unfractionated heparin, prior to advancing wires into the carotid artery in order to prevent thrombosis. The ACT should be confirmed to be between 250 and 300 seconds before manipulation of the carotid artery, and it is maintained at this level until all wires and cerebral protection devices have been removed. Heparin is not reversed after CAS. (See "Percutaneous carotid artery stenting", section on 'Anticoagulation'.)

POSTOPERATIVE PROBLEMS

Blood pressure control — In the postoperative period, the disrupted baroreceptor function caused during carotid endarterectomy (CEA) or carotid artery stenting (CAS) may result in ongoing labile blood pressure and heart rate [95,109]. Also, uncontrolled postoperative pain may cause hypertension.

Continued invasive monitoring of arterial blood pressure, as well as treatment with appropriate administration of pain medications and vasoactive drugs, is necessary in patients exhibiting hemodynamic instability.

●Hypertension – Postoperative hypertension may result in abnormally increased cerebral blood flow because cerebral autoregulation may be disrupted following CEA or CAS. Thus, it is important to strictly control postoperative hypertension, maintaining systolic blood pressure at 100 to 150 mmHg.

Labetalol, nicardipine, or esmolol may be administered in bolus doses (table 2). If blood pressure is not controlled with one or more of these bolused medications, intravenous infusion of labetalol, nitroprusside, nitroglycerin, nicardipine, or esmolol is administered to control hypertension (table 2).

Occasionally, postoperative hypertension is a predecessor of a condition known as cerebral hyperperfusion syndrome [110]. This syndrome is characterized by cerebral edema, petechial or frank intracerebral hemorrhage, and seizures. (See "Complications of carotid endarterectomy", section on 'Hyperperfusion syndrome'.)

●Hypotension – Persistent hypotension may complicate the early postoperative course due to residual carotid hypersensitivity after plaque excision [111].

Systolic blood pressure should be maintained at 100 to 150 mmHg to provide adequate cerebral perfusion pressure and cerebral blood flow, and thus avoid cerebral ischemia. If necessary, an infusion of phenylephrine (10 to 200 mcg/minute; 0.1 to 2 mcg/kg/minute) is administered (table 1).

Slow emergence from anesthesia and stroke — Slow emergence from general anesthesia may be related to residual anesthetic effects, hypothermia, or hypercarbia. These causes are ruled out, and stroke is considered by evaluating the patient for new neurologic deficits. (See "Complications of carotid endarterectomy", section on 'Perioperative stroke'.)

Hematoma — Postoperative bleeding resulting in neck hematoma occasionally occurs after CEA and is more likely in patients with poorly-controlled postoperative hypertension or ongoing anticoagulation [112,113]. A significant wound hematoma may compromise the patient's airway, resulting in a need for emergency airway management and reoperation. This event is associated with higher in-hospital mortality, stroke, and myocardial infarction [113,114]. (See "Complications of carotid endarterectomy", section on 'Cervical hematoma'.)

Emergency airway management may be challenging because the neck hematoma typically compresses and displaces upper airway structures. Visualization of upper airway anatomy may be further complicated by multiple intubation attempts resulting in oropharyngeal and laryngeal edema or bleeding. The American Society of Anesthesiologists (ASA) has developed guidance for airway management in this setting, as noted in the table (table 3). (See "Anesthesia for adult trauma patients", section on 'Airway management'.)

Following CAS, inadequate closure of the femoral artery puncture site used to obtain percutaneous access may lead to bleeding and hematoma, in part because of concomitant administration of antiplatelet therapy and ongoing anticoagulation. (See "Percutaneous carotid artery stenting", section on 'Complications'.)

Pain — While opioids effectively treat pain, they may exacerbate respiratory depression, particularly after carotid interventions, due to disruption of the carotid chemoreceptor-mediated hyperventilation response to hypoxia [115,116]. Thus, judicious administration of opioids is necessary to minimize the side effects of respiratory depression and somnolence, which can interfere with neurologic assessment. Acetaminophen may be added to decrease opioid requirements. The safety of NSAIDs for pain control has not been established for patients undergoing carotid revascularization and we suggest avoiding their use. (See "NSAIDs: Adverse cardiovascular effects".)

Vocal cord paralysis — Injury to the recurrent laryngeal nerve during CEA (traction, compression) may result in paralysis of the ipsilateral vocal cord. This typically presents as new-onset hoarseness. In rare cases, contralateral recurrent laryngeal nerve injury related to prior neck surgery may not have been recognized preoperatively, and this combined with new ipsilateral recurrent laryngeal nerve injury during CEA may require emergency endotracheal intubation if the adduction of both vocal cords causes closure of the glottic aperture and laryngeal obstruction. (See "Carotid endarterectomy", section on 'Otolaryngologic examination' and "Complications of carotid endarterectomy", section on 'Nerve injury'.)

Other complications — Other complications occurring in the immediate postoperative period, or several days and weeks later, are discussed separately. (See "Complications of carotid endarterectomy" and "Overview of carotid artery stenting", section on 'Complications' and "Percutaneous carotid artery stenting", section on 'Complications'.)

SUMMARY AND RECOMMENDATIONS

●Preanesthesia consultation for carotid artery procedures (ie, carotid endarterectomy [CEA] and carotid artery stenting [CAS]) identifies patients with cardiovascular comorbidity who may require further investigation. (See 'Preanesthesia consultation' above.)

●Perioperative medication management includes continuing beta blockers, and continuing and/or initiating statins and antiplatelet agents. Specific recommendations are discussed separately. (See "Carotid endarterectomy", section on 'Medication management'.)

●The choice of general anesthesia versus local/regional anesthesia in an awake patient for CEA or CAS is determined by the preferences of the surgeon, anesthesiologist, and patient. Although major outcomes (death, stroke and myocardial infarction) are not impacted by anesthetic choice, length of hospital stay and secondary outcomes (eg, hemodynamic instability, delirium) may be reduced in patients receiving local/regional rather than general anesthesia. (See 'Evidence review' above.)

•A major advantage of local/regional anesthetic technique is the ability to continuously monitor neurologic function in an awake patient. However, this requires that the patient can cooperate and communicate well. (See 'Advantages of local/regional anesthesia' above.)

•Advantages of general anesthesia include greater patient comfort and elimination of the potential need for urgent conversion from a local/regional anesthetic technique. (See 'Advantages of general anesthesia' above.)

●General anesthesia for CEA or CAS is induced with a short-acting agent such as etomidate (0.15 to 0.3 mg/kg) or propofol (1 to 2.5 mg/kg). The addition of a low-dose, short-acting opioid (eg, fentanyl [1 to 2 mcg/kg] or remifentanil [1 mcg/kg]), and/or lidocaine (50 to 100 mg) blunts the hypertension and tachycardia response associated with sympathetic stimulation during endotracheal intubation. When volatile anesthetics are used to maintain anesthesia, selection of sevoflurane or desflurane facilitates a rapid emergence. (See 'Induction and maintenance of general anesthesia' above.)

●The authors prefer endotracheal intubation in patients undergoing CEA with general anesthesia because airway access is limited during this procedure. In patients undergoing CAS with general anesthesia, a laryngeal mask airway (LMA) is preferred. (See 'Airway management' above.)

●Normocapnia is maintained during general anesthesia in patients undergoing CEA or CAS. (See 'Ventilation management' above.)

●When local/regional anesthetic technique is selected for CEA, we recommend a superficial cervical plexus block; this technique provides adequate anesthesia while avoiding the complications of a deep cervical plexus block (Grade 1B).

For CAS, local/regional anesthetic is injected into the catheter insertion site.

Sedation is minimized to allow performance of neurologic exams. (See 'Local/regional anesthesia technique' above.)

●In awake patients, neuromonitoring is accomplished by frequent neurologic examination. (See 'Neurologic examination in awake patients' above.)

●When general anesthesia is selected, neuromonitoring techniques to detect cerebral ischemia include continuous electroencephalography (EEG) for CEA or CAS and/or carotid stump pressure monitoring for CEA. EEG signal suppression is avoided by administering the volatile anesthetic at a dose <1 minimum alveolar concentration (MAC). (See 'Neuromonitoring' above and 'Electroencephalography (EEG) with general anesthesia' above.)

●Blood pressure is controlled throughout the procedure, using vasoactive drugs as needed. Periods of highest risk for hemodynamic instability and resultant myocardial or cerebral ischemia include induction of general anesthesia, surgical manipulation of the carotid sinus and carotid artery, carotid cross-clamping, incision of the carotid artery, carotid unclamping with reperfusion, and emergence from anesthesia. During carotid cross-clamping, we suggest maintenance of systolic blood pressure in a range from the patient's baseline blood pressure to 20 percent above that baseline to optimize collateral cerebral perfusion (Grade 2C). (See 'Hemodynamic management' above.)

●During CEA, the patient is systemically anticoagulated prior to carotid artery clamping. At the completion of the procedure, we suggest reversal of heparin with protamine rather than no reversal (Grade 2B). (See "Carotid endarterectomy", section on 'Endarterectomy procedure'.)

●During CAS, the patient is anticoagulated with heparin to maintain the activated clotting time (ACT) at 250 to 300 seconds, prior to any manipulation of guidewires and catheters within the carotid artery. (See "Percutaneous carotid artery stenting", section on 'Anticoagulation'.)

●Problems in the immediate postoperative period requiring urgent treatment include hypertension, hypotension, slow emergence from anesthesia, stroke, hematoma at the surgical site, pain, and vocal cord paralysis. (See 'Postoperative problems' above.)