OBSERVATIONAL REPORTS COMPARING SEDATION AND GENERAL ANESTHESIA FOR ENDOVASCULAR THROMBECTOMY
Limitations of Observational Reports
At the present (literature review current through April 2018), there are at least 24 observational reports comparing sedation and general anesthesia for endovascular thrombectomy1–24 with regard to one or more of the following outcome categories: (1) neurologic/functional (n = 191–6,8–12,14,16,17,19–21,23,24); (2) mortality (n = 201–14,17,19,21–24); (3) workflow (n = 201–12,14,15,17–20,23,24); and (4) intra– and post–endovascular thrombectomy complications (n = 201–14,18,19,21–24). In none of these 24 observational reports does general anesthesia appear to be superior to sedation with respect to neurologic outcome or mortality. In 14 of 19 reports regarding neurologic outcome,1–6,8–12,14,17,19 data collection was completed before publication of the 2014 Society for Neuroscience in Anesthesia and Critical Care guidelines for the anesthetic management of endovascular thrombectomy (April 2014).25 In the remaining 5 neurologic outcome reports, approximately 80% of the patients underwent endovascular thrombectomy before the Society for Neuroscience in Anesthesia and Critical Care guidelines were published.16,20,21,23,24 Thus, at present, observational reports almost entirely reflect anesthetic practice predating the 2014 Society for Neuroscience in Anesthesia and Critical Care guidelines. There are 3 observational reports that provide data that are included in other larger reports. Sixty-four of 66 patients reported in Sugg et al2 were reported in Abou-Chebl et al.3 All patients reported in Berkhemer et al19 and Bracard et al20 were included in the subsequent report by Campbell et al.24 For a tabular summary of these data, see Supplemental Digital Content 1, Table 1, http://links.lww.com/AA/C717.
Many of these observational reports (at least 10 of 24) considered all patients who were intubated for any reason to have received general anesthesia and considered all patients who were not intubated to have received sedation.1–3,5–7,11–13,22 In other words, all intubated patients were assigned to the general anesthesia group for observational study, and all nonintubated patients were assigned to the sedation group for study. Yet, of course, a patient may receive general anesthesia but not be intubated. In addition, and very importantly, endovascular thrombectomy patients are sometimes intubated before endovascular thrombectomy for reasons other than the need for general anesthesia; see Supplemental Digital Content 2, Table 3, http://links.lww.com/AA/C718. Nevertheless, this generalization (intubation defines general anesthesia) is so pervasive in the observational literature that one is forced to accept this limitation to even begin to consider the studies’ findings. In many of these observational reports (11 of 24), the sedatives, analgesics, anesthetics that were administered were not reported at all.1,3,11–14,17,19,20,22,24 Likewise, in many of these observational reports (17 of 24), intra–endovascular thrombectomy hemodynamics were not reported at all.1–4,6–8,11–14,17–20,22,24 Consequently, it is often impossible to compare their findings to any other report or to directly apply their findings to clinical practice—ie, these reports provide no insight as to what might be done differently.
Even accepting these substantive limitations, the findings of few observational reports can be accepted at face value because of identifiable biases against general anesthesia. Of the 24 observational reports, risk factors for unfavorable outcome that were more commonly present in patients selected for general anesthesia included: (1) greater National Institutes of Health Stroke Scale scores (12 reports1–6,8,10–12,15,23); (2) greater incidence of intracranial internal carotid artery occlusion (5 reports1,3,9,12,23); and (3) greater incidence of posterior circulation occlusion (5 reports5,7,8,17,21). Additional risk factors and biases in patients selected to receive general anesthesia included: (1) all patients intubated before endovascular thrombectomy were exclusively assigned to the general anesthesia group for analysis (13 reports1–3,5,6,9–14,21,22) (also see Supplemental Digital Content 2, Table 3, http://links.lww.com/AA/C718); (2) patients selected for sedation but who required conversion to general anesthesia during endovascular thrombectomy were assigned to the general anesthesia group for analysis (9 reports1,3,5–7,11–13,22); (3) sedation and general anesthesia groups were noncontemporaneous, with general anesthesia patients undergoing endovascular thrombectomy earlier in the series (4 reports4,8,9,21); and (4) other risks or comorbidities were greater in patients selected to received general anesthesia (12 reports2,4–6,10,12,14,15,18,21,23,24). For a tabular summary of these data, see Supplemental Digital Content 1, Table 1, http://links.lww.com/AA/C717.
Failure to adjust adequately for risk factor imbalance between patients selected to receive sedation instead of general anesthesia is the pervasive limitation of this observational literature. Accordingly, it is virtually certain many observational reports overestimate the apparent adverse effect of general anesthesia with endovascular thrombectomy. Meta-analysis of the unadjusted outcome data cannot overcome such selection bias. Although some meta-analyses have attempted to adjust for differences in National Institutes of Health Stroke Scale scores between patients selected for sedation and general anesthesia,26,27 National Institutes of Health Stroke Scale is only one of several determinants of endovascular thrombectomy effectiveness; see part 1 of this review.28 Hence, adjusting only for National Institutes of Health Stroke Scale score differences is not a sufficient adjustment for bias between endovascular thrombectomy patients who were selected to receive sedation instead of general anesthesia.
Twelve of the 19 observational reports regarding neurologic/functional outcome statistically adjusted sedation versus general anesthesia comparisons using either: (1) observed differences in risk factors between groups or (2) preplanned risk factors based on prior reports. Table 1 summarizes these 12 neurologic outcome reports, all of which considered a 90-day modified Rankin Scale score ≤2 to represent a good outcome. The reported odds ratios >1 indicate more favorable outcome with sedation as compared to general anesthesia. Data from Berkhemer et al19 are not included in Table 1 because data from this report are included in Campbell et al.24 A meta-regression of data in Table 1 is shown graphically in the Figure.29 The apparent risk-adjusted benefit of sedation over general anesthesia has progressively decreased over time; P = .009. This progressive decrease of the apparent benefit of sedation for endovascular thrombectomy strongly suggests the apparent adverse effect of general anesthesia is not fixed or immutable. Over time, the conduct of general anesthesia for endovascular thrombectomy has changed in ways that have lessened (or even nearly eliminated) the previous associations between general anesthesia and less favorable outcome such that, now, the outcome differences between sedation and general anesthesia appear to be rather small.
The 2 most recent observational reports are instructive because they highlight the limitations of observational work. Adjusting for pre–endovascular thrombectomy differences between patients selected for sedation and general anesthesia, Slezak et al23 concluded sedation did not improve outcome relative to general anesthesia: odds ratio = 1.145 (95% CI, 0.661–1.980); P = .629. In apparent contrast, controlling for preselected variables, Campbell et al24 concluded sedation did improve outcome relative to general anesthesia: odds ratio = 1.65; 95% CI, 1.14–2.48; P = .007. Although these 2 reports appear to differ in their conclusions, because their CIs substantially overlap (between 1.14 and 1.98), they should be considered to be 2 sides of the same coin. The absence of heterogeneity among all the studies in the above meta-analysis also supports this conclusion; see the Figure legend. Both reports suggest more favorable outcome with sedation, but both reports have methodological limitations that preclude unreserved acceptance of their findings. In the report by Slezak et al,23 anesthetic methods were well described and the same blood pressure goal (systolic blood pressure [SBP] = 140–180 mm Hg) was applied to all endovascular thrombectomy patients. However, the Slezak et al23 report (401 patients) came from a single center, raising the question of generalizability, and the CIs were so wide it is possible that outcomes could be nearly 2-fold better with sedation. Campbell et al24 used pooled data from 7 endovascular thrombectomy trials (797 patients, 127 centers), which may reflect more general practice. However, in Campbell et al,24 there was no standardization or characterization of anesthesia methods or hemodynamics whatsoever; we simply do not know the essential details of how sedation and general anesthesia were managed. With 127 centers, there could be substantial heterogeneity in anesthesia methods.
Hypotheses Based on Observational Data
In the authors’ opinion, these observational reports have served their purpose by providing data for hypothesis generation regarding the key aspects of anesthetic management for endovascular thrombectomy. This is distinct from hypothesis testing, which requires randomized controlled trials. The reader is invited to consider the evidence for 7 hypotheses regarding anesthesia management for endovascular thrombectomy derived from the 24 observational reports1–24; see Supplemental Digital Content 3, Document, http://links.lww.com/AA/C719. In large measure, these hypotheses are the basis for the methods and outcome measures used in 3 recent randomized controlled trials that are reviewed in the following section.
- Hypothesis 1: General anesthesia increases the time between the decision for endovascular thrombectomy and achieving reperfusion by about 20 minutes.
- Hypothesis 2: Outcome differences between sedation and general anesthesia are due to blood pressure differences. When minimum mean arterial pressure (MAP) in the sedation and/or general anesthesia groups is <70–80 mm Hg, there is an association between minimum MAP and neurologic outcome. Sedation versus general anesthesia outcome differences appear to be related to the magnitude of the sedation versus general anesthesia minimum MAP difference until minimum MAP in the general anesthesia group exceeds 80 mm Hg. When minimum MAP in the sedation and general anesthesia groups exceed 80 mm Hg, there is not a detectable relationship between minimum MAP and outcome, nor is there outcome difference between sedation and general anesthesia.
- Hypothesis 3: Endovascular thrombectomy patients selected to receive sedation require vasopressors less often and/or at lesser doses than patients selected to receive general anesthesia.
- Hypothesis 4: The majority (>50%) of endovascular thrombectomy patients selected to receive sedation but who were converted to general anesthesia during the procedure were because of “agitation.”
- Hypothesis 5: General anesthesia may slightly increase the incidence of good reperfusion, but the difference is probably too small to detect statistically without a very large randomized controlled trial.
- Hypothesis 6: General anesthesia may slightly decrease the incidence of endovascular complications, but the difference is probably too small to detect statistically without a very large randomized controlled trial.
- Hypothesis 7: Because stroke-associated pneumonia is related to National Institutes of Health Stroke Scale score, the greater incidence of post–endovascular thrombectomy pneumonia in patients selected for general anesthesia may be due to selection bias.
RANDOMIZED CLINICAL TRIALS COMPARING SEDATION AND GENERAL ANESTHESIA FOR ENDOVASCULAR THROMBECTOMY
To date, 3 single-center randomized controlled trials of sedation versus general anesthesia for endovascular thrombectomy for patients with anterior circulation strokes have been completed: Sedation vs Intubation for Endovascular Stroke Treatment (SIESTA),30,31 Anesthesia during Stroke (AnStroke),32 and General or Local Anesthesia in Intra Arterial Therapy (GOLIATH).33,34 As summarized in Table 2, in each of the 3 trials, sedation did not result in more favorable 3-month functional outcome or lesser mortality. In fact, outcomes might appear to be slightly more favorable with general anesthesia. However, this is unknown. There is substantive heterogeneity among SIESTA, AnStroke, and GOLIATH in the incidence of good outcomes with sedation (18%, 40%, and 52%, respectively) and with general anesthesia (37%, 42%, and 67%, respectively). Adjusting for these differences in a meta-analysis of these 3 trials, the pooled relative risk for good outcome with sedation is 0.74; P = .19. Thus, collectively, these 3 trials do not establish better outcome with either sedation or general anesthesia for endovascular thrombectomy.
All 3 trials had similar intra–endovascular thrombectomy blood pressure goals for both sedation and general anesthesia groups: SIESTA (SBP = 140–160 mm Hg); AnStroke (SBP = 140–180 mm Hg before reperfusion); and GOLIATH (SBP >140 mm Hg and MAP ≥70 mm Hg). These blood pressure goals were consistent with the 2014 Society for Neuroscience in Anesthesia and Critical Care guidelines.25 As summarized in Table 3, many patients—in both the sedation and general anesthesia groups—required vasopressors to maintain blood pressure. In all 3 trials, vasopressor doses were 1.5- to 3-fold greater in general anesthesia patients. Despite well-defined blood pressure goals and intensive vasopressor use, in both AnStroke and GOLIATH, the percentage of patients who had >20% decrease in MAP at any time was 1.5- to 2-fold greater in general anesthesia patients; P = .0003 and P < .001, respectively.32,34 In post hoc analyses of relationships between various measures of intra-endovascular thrombectomy blood pressure and neurologic outcome, there were no significant relationships in either SIESTA36 or GOLIATH.37 These findings are consistent with observational reports that suggest when MAPs are maintained in the 90–100 mm Hg range during endovascular thrombectomy, intra–endovascular thrombectomy blood pressure is not a significant outcome determinant; see Supplemental Digital Content 3, Hypothesis 2, http://links.lww.com/AA/C719.
As summarized in Table 4, in all 3 trials, general anesthesia increased the mean time between evaluation and arterial puncture by 10 minutes or less, an interval consistent with the minimum time required to induce general anesthesia and intubate the patient. However, after arterial puncture, mean time to achieve reperfusion was numerically less with general anesthesia in both SIESTA and AnStroke, and was comparable between groups in GOLIATH. In all 3 trials, adequate reperfusion (modified Thrombolysis in Cerebral Infarction 2b–3) was numerically greater with general anesthesia.
As summarized in Table 5, a substantive percentage (6%–33%) of patients randomized to sedation had problematic movement during endovascular thrombectomy. Nevertheless, there was a low incidence of vascular complications, without a significant difference between patients randomized to sedation or general anesthesia. In these 3 trials, 6%–16% of patients randomized to sedation required conversion to general anesthesia during endovascular thrombectomy. In SIESTA, sedation-to-general anesthesia conversions were because of patient agitation (7/11 [64%]) and respiratory/airway problems (3/11 [27%]). In AnStroke, sedation-to-general anesthesia conversions were because of patient movement (2/7 [29%]) and airway problems (1/7 [14%]). In GOLIATH, sedation-to-general anesthesia conversions appeared to be less common (4/63 [6%]) than in the other 2 trials, and were due to movement in all cases (2 of 4 patients also had emesis, and 1 aspirated and desaturated). Why sedation-to-general anesthesia conversion in GOLIATH was apparently less common than in the other 2 trials is not obvious. It also is not obvious why, in SIESTA, general anesthesia was associated with such a high incidence of delayed extubation when compared to AnStroke and GOLIATH; 49% vs 7% vs 3%, respectively. It is possible that, in SIESTA, delayed extubation may have contributed to a greater rate of pneumonia in general anesthesia patients. However, this hypothesis cannot be reconciled with AnStroke findings, in which the incidence of pneumonia in both the sedation and general anesthesia groups was comparable to the general anesthesia pneumonia rate in SIESTA.
Collectively, these 3 trials demonstrate when: (1) general anesthesia is integrated into routine workflow; and (2) blood pressure goals in sedation and general anesthesia patients are the same (eg, SBP >140 mm Hg), patients who receive general anesthesia for endovascular thrombectomy do not have less favorable outcome than patients who receive sedation. In these trials, general anesthesia: (1) delayed the start of endovascular thrombectomy on average by 10 minutes or less; and (2) resulted in slightly lower blood pressure despite intensive vasopressor administration. Counterbalancing these limitations, general anesthesia created slightly better working conditions for the interventionist (less patient movement), which may have facilitated both more rapid workflow after arterial puncture and restoration of good reperfusion. General anesthesia decreases delays associated with failed sedation, which occurred in a substantive fraction (~10%) of patients who were randomized to receive sedation. With this information, anesthesiologists and interventionists can now decide to use general anesthesia for endovascular thrombectomy when conditions require it, with far less concern that patients will necessarily be adversely affected.
Before applying the findings of these 3 trials to clinical practice, their limitations must be considered. Each was a single center trial in which expert protocol-driven care was provided. Whether these results would be reproduced in other medical centers with different personnel and with different workflow, endovascular, and/or anesthesia protocols is unknown. There are multiple reasons why application of trial protocols to all endovascular thrombectomy patients could result in different findings. In both SIESTA31 and AnStroke,32 approximately 40% of patients who were eligible and consented to participate were excluded before randomization. Excluded patients were considered at high risk of complications/failures with either sedation or general anesthesia. In both SIESTA and AnStroke, at least 20% of otherwise eligible patients were excluded because they were judged to be poor candidates for sedation because of pre–endovascular thrombectomy agitation, vomiting, airway compromise, and/or the inability to follow commands. Although, in GOLIATH, patients were not excluded based on pre–endovascular thrombectomy movement or agitation, 94 patients who were excluded from randomization (because of National Institutes of Health Stroke Scale score, posterior circulation stroke, or pre–endovascular thrombectomy intubation) underwent endovascular thrombectomy outside of the trial. Thus, in GOLIATH (94/[94 + 128]), 42% of all endovascular thrombectomies were performed outside of the trial. Overall, in all 3 trials, approximately 40% of all endovascular thrombectomy patients were not randomized.
In clinical practice: (1) exclusion of nonideal patients is not an option; and (2) equipoise is often not present. Therefore, with understanding of: (1) the pathophysiology of acute ischemic stroke; (2) the determinants of endovascular thrombectomy effectiveness; and (3) the potential limitations and advantages of sedation and general anesthesia for endovascular thrombectomy, the anesthesiologist can make a decision for each individual patient—specifically, sedation or general anesthesia?
ANESTHESIA FOR ENDOVASCULAR THROMBECTOMY: DECISION-MAKING
Who among nonintubated endovascular thrombectomy patients should undergo endovascular thrombectomy with sedation and who should receive general anesthesia? At present, there is no evidence-based algorithm that can be used to make the decision. Because of the appropriate emphasis on rapid workflow, endovascular thrombectomy patients often undergo their acute stroke evaluation and treatment so quickly that their medical record lags behind. Thus, anesthesiologists are often asked to care for endovascular thrombectomy patients with only a few minutes of notice and with little objective information. Consequently, the sedation versus general anesthesia decision will often need to be made very rapidly with information obtained within the first few minutes after the patient arrives to the interventional suite. Because as many as 50% of endovascular thrombectomy patients have aphasia, many will not able to effectively communicate their medical history, medications, and allergies, or even their fasting interval; see part 1 of this review.28 Pre–endovascular thrombectomy anesthesia assessments often need to occur while other activities are taking place—transferring the patient to the angiography table, establishing anesthesia monitors, and prepping the groin for arterial access. With little information and little time, endovascular thrombectomy anesthesia decision making is often based more on general principles than on specific knowledge of patient comorbidities.
Endovascular thrombectomy is a time-limited emergency procedure. Getting the procedure started (femoral arterial puncture for angiographic access) as soon as possible is a cardinal management goal; see part 1 of this review.28 What is unknowable is when rapid workflow becomes unsafe.
Collateral perfusion maintains penumbral viability until reperfusion is accomplished. Collateral perfusion is probably blood pressure dependent in many endovascular thrombectomy patients. Thus, maintaining patient blood pressure at presentation values until reperfusion occurs is another cardinal management goal; see part 1 of this review.28 If the patient has received tissue plasminogen activator, the upper limits are SBP ≤180 mm Hg, diastolic blood pressure ≤105 mm Hg.39 If the patient has not received tissue plasminogen activator, greater blood pressure values may be reasonable before reperfusion.39
Most endovascular thrombectomy patients can be managed with sedation. Medical centers that have a strong preference and extensive experience with sedation for endovascular thrombectomy use it in >90% of patients. For example, in the Endovascular Treatment for Small Core and Anterior Circulation Proximal Occlusion with Emphasis on Minimizing CT to Recanalization Times (ESCAPE)40 and Randomized Trial of Revascularization with Solitaire FR Device versus Best Medical Therapy in the Treatment of Acute Stroke Due to Anterior Circulation Large Vessel Occlusion Presenting within Eight Hours of Symptom Onset (REVASCAT)41 trials, sedation was used in 91% and 93% of the patients, respectively. However, both observational reports and randomized controlled trials suggest at least 20% of patients with anterior circulation strokes are not favorable candidates for sedation.5,12,31,32 Because patients with posterior circulation (basilar artery) ischemia often have impaired consciousness, dysphagia, and cranial nerve dysfunction,42 they may more often be candidates for general anesthesia than patients with anterior circulation strokes. A recent observational report of 176 nonintubated patients undergoing endovascular thrombectomy for vertebrobasilar ischemia reported 63 patients (36%) were selected to receive dexmedetomidine-based sedation.43 Conversion to general anesthesia during endovascular thrombectomy was necessary in 8 (13%) of these patients. The most common reasons for sedation-to-general anesthesia conversion were “agitation” and excessive patient movement. When sedation is selected, one should expect that at least 5%–10% of the patients will require conversion to general anesthesia during the procedure.
When patients are obviously restless, agitated, and/or inconsolable, one must consider whether any sedative or hypnotic can reliably alter that without inducing patient unresponsiveness. Alternatively, or in addition, if the patient has a receptive aphasia, he/she will not be able to understand requests to be motionless. Other endovascular thrombectomy patients may show signs of severe dysphagia (swallowing dysfunction), or abnormal breathing patterns, with or without hypoxemia and/or hypoventilation. Some patients may be obtunded and/or have early brain swelling and increased intracranial pressure.
Some rapid bedside assessments may include: (1) is the patient responsive to verbal or tactile stimuli?; (2) can the patient lie supine without respiratory difficulty, from airway obstruction, secretions (dysphagia), or a pathologic breathing pattern?; (3) does the patient have acceptable oxygen saturation (≥94%) with or without supplemental oxygen (nasal cannula or facemask)?; (4) can the patient understand and follow verbal commands (close eyes, open mouth, squeeze hand, and be still)?; (5) if sedation-to-general anesthesia conversion were to be needed, can the patient’s airway be managed safety under nonideal conditions? In the authors’ opinion, if the answer to one or more of these 5 questions is “no,” then greater consideration should be given to general anesthesia.
The Planned Procedure and the Interventionist
In the authors’ opinion, in the absence of clear indication for either sedation or general anesthesia from the anesthesiologist’s perspective, the recommendation of the interventionist should be favorably considered and implemented. The interventionist may know the clot likely will be easy to retrieve and the procedure will be brief (30 minutes)—this may favor sedation. Conversely, the interventionist may know the occlusion is more difficult and/or there is ipsilateral extracranial carotid disease (“tandem” lesion) such that endovascular thrombectomy will take more time and be more difficult—this may favor general anesthesia. In addition, each interventionist has their individual tolerance for patient movement during endovascular thrombectomy, which may vary depending on the specifics of the occlusion. In the authors’ opinion, in the absence of an indication (anesthesiologist) or recommendation (interventionist) regarding method of anesthesia for endovascular thrombectomy, sedation should be considered as the default anesthetic method, especially when there is little information regarding patient comorbidities. However, even with rational decision-making and vigilant execution, neither sedation nor general anesthesia can be guaranteed to have a satisfactory result.
ANESTHESIA FOR ENDOVASCULAR THROMBECTOMY: METHODS
When caring for endovascular thrombectomy patients, the anesthesiologist must be aware of the potential of acute coagulopathy (from tissue plasminogen activator) and/or chronic coagulopathy (from warfarin, oral thrombin inhibitors, antiplatelet agents, etc). Thus, if the endovascular thrombectomy patient presents with nonfunctioning IV access, it should not be removed because of the risk of bleeding and/or large hematoma formation. Potential advantages of establishing additional IV access must be weighed against this same risk in the event that attempts are not successful. Because of potential acute/chronic coagulopathy, the clinical author of this review (B.J.H.) avoids placing nasal airways and/or nasal temperature monitoring probes.
At present, when sedation is selected instead of general anesthesia, there is not sufficient clinical evidence to support the use of one sedative, hypnotic, or analgesic over any other. Observational reports in which sedation was provided with a variety of agents show all can result in substantive decreases in blood pressure and, in approximately half of the patients, vasopressors will be needed to maintain blood pressure; see Supplemental Digital Content 3, Hypothesis 3, http://links.lww.com/AA/C719. Thus, just like with general anesthesia, a proactive and intensive approach toward maintaining blood pressure at values present immediately before endovascular thrombectomy (SBP <180 mm Hg if tissue plasminogen activator has been given39) is a cardinal goal of sedation. Maintaining blood pressure goals may be somewhat easier with sedation than with general anesthesia.5,10,21,31,32,34 At present, there is not sufficient clinical evidence to suggest 1 agent/method is superior to another in reducing sedation-to-general anesthesia conversions; see Supplemental Digital Content 3, Hypothesis 4, http://links.lww.com/AA/C719. To preserve the best possible neurologic examination during and after endovascular thrombectomy, the clinical author of this review attempts to avoid agents with a prolonged duration of action and/or agents that cannot be reversed. In some endovascular thrombectomy patients, it may not be necessary to administer any sedative or analgesic at all.
At present, when general anesthesia is selected instead of sedation, there is not sufficient clinical evidence to support the use of any induction or maintenance agent over any other.31,32,34 Regardless of agents used to induce and maintain general anesthesia, a proactive and intensive approach toward maintaining blood pressure at values present immediately before endovascular thrombectomy (SBP <180 mm Hg if tissue plasminogen activator has been given39) is a cardinal goal of general anesthesia.
The 2014 Society for Neuroscience in Anesthesia and Critical Care Consensus guidelines state an intraarterial cannula is useful if it can be placed without delaying the procedure.25 However, the clinical author of this review does not routinely attempt to start an arterial line before starting endovascular thrombectomy, even when general anesthesia is selected. Instead, to facilitate workflow, the clinical author very frequently cycles the noninvasive blood pressure cuff until femoral arterial access is obtained by the interventionist (usually <5 minutes after induction) and femoral arterial pressure is continuously monitored thereafter. If a radial arterial line is desired after endovascular thrombectomy, it is placed before leaving the interventional suite. With general anesthesia, it is virtually certain the patient will require large dose vasopressors to maintain blood pressure.31,32,34 At the present, there is not sufficient clinical evidence to support the use of a specific vasopressor during endovascular thrombectomy. In principle, a vasopressor that most favorably affects collateral flow to the ischemic penumbra would be preferred. However, at present, there is no clinical evidence that such a difference exists among vasopressors. Therefore, treatment of hypotension is most logically based on what is known of the patient’s pre–endovascular thrombectomy medical condition (eg, vasopressin for patients taking agents that inhibit angiotensin; inotropes for patients with poor contractile function, etc).
Two observational reports of endovascular thrombectomy patients managed only with general anesthesia indicated group mean end-tidal (not arterial) carbon dioxide values were less in endovascular thrombectomy patients who had unfavorable outcomes.44,45 Although these observations must be interpreted with caution, in general, hyperventilation (hypocarbia) is not likely to benefit most endovascular thrombectomy patients. The effects of hypercarbia have not been reported in endovascular thrombectomy patients. However, approximately 10% of patients with acute stroke show evidence of decreased cerebral blood flow in the affected vascular territory with 15–30 seconds of breath-holding.46 This is thought to be due to hypercarbia-induced blood flow redistribution away from penumbral tissue, referred to as a “steal.” Carbon dioxide “steal” physiology has been observed in patients with chronic (nonacute) intracranial cerebral occlusive disease.47,48 Thus, at present, maintenance of arterial normocarbia during endovascular thrombectomy appears to be reasonable.
When reperfusion is restored, or the interventionist determines additional procedures are not worthwhile, endovascular thrombectomy is very nearly completed. Only a few minutes are needed to remove the arterial access catheter and manage the puncture site. Although there is no direct evidence to support it, some authors recommend endovascular thrombectomy patients who have good reperfusion (modified Thrombolysis in Cerebral Infarction ≥2b) should have moderate blood pressure reductions (SBP ≤ 140–150 mm Hg) in an attempt to decrease reperfusion-related adverse events (eg, hemorrhagic transformation and/or cerebral edema).49–51 In contrast, for endovascular thrombectomy patients who are not well reperfused (modified Thrombolysis in Cerebral Infarction ≤2a), some authors recommend maintenance of hypertension (but SBP ≤ 180 mm Hg) for at least another 24 h to support collateral flow.50,51
Hyperoxia may exacerbate reperfusion-related brain injury. Observational reports indicate intubated stroke patients who are hyperoxic have less favorable outcome than patients who are not hyperoxic.52 However, with risk adjustment, the association between hyperoxia and outcome is no longer significant.53 Therefore, in endovascular thrombectomy patients who have good reperfusion, the authors of this review suggest considering decreasing the inspired oxygen concentration to achieve arterial hemoglobin saturation in the 94%–97% range, but only if doing so appears to be safe otherwise.
General anesthesia agents and methods that increase the likelihood of complete recovery from neuromuscular blockade, rapid and reliable emergence from general anesthesia, and optimal conditions for extubation at the end of endovascular thrombectomy would appear to be reasonable. In observational reports, the greater incidence of post–endovascular thrombectomy pneumonia in patients selected to receive general anesthesia appears to be due, at least in part, to selection bias; see Supplemental Content 3, Hypothesis 7, http://links.lww.com/AA/C719. However, it is also possible that general anesthesia and intubation, either alone or in combination, predispose patients to post–endovascular thrombectomy pneumonia. In particular, continued intubation after endovascular thrombectomy might contribute to less favorable outcomes on the basis of ventilator-associated pneumonia54 or lung injury55 and/or when sedatives are administered to facilitate continued intubation.56,57 Two observational reports of endovascular thrombectomy conducted under general anesthesia reported 60%45 to 70%58 of the patients were not extubated at completion of endovascular thrombectomy. In both reports, compared with extubation at the end of endovascular thrombectomy, continued (post–endovascular thrombectomy) intubation was associated with less favorable outcome. However, in both reports, it could not be determined if continued intubation after endovascular thrombectomy was the cause of less favorable outcomes or was the effect. The 3 randomized controlled trials report widely differing post–endovascular thrombectomy intubation incidences; see Table 5. In SIESTA, 49% of the patients randomized to general anesthesia were still intubated 2 h after completion of endovascular thrombectomy.31 The incidence of delayed extubation in general anesthesia patients was much less in AnStroke and GOLIATH, 7%32 and 3%,34 respectively; see Table 5. These differences among these 3 trials cannot be readily explained. Notably, in none (0/24) of the observational reports comparing outcomes between sedation and general anesthesia1–24 and none (0/3) of the randomized trials30–34 were predefined extubation criteria reported or apparently utilized.
The authors of this review speculate use of conventional extubation criteria after general anesthesia for endovascular thrombectomy could result in a relatively large percentage of general anesthesia patients remaining intubated at the end of endovascular thrombectomy. In particular, endovascular thrombectomy patients who present with relatively severe strokes may not emerge from general anesthesia with a level of consciousness or ability to follow commands that would otherwise be required for extubation. A similar problem—extubation of obtunded patients—is common in neurocritical care. In an observational report of 41 intubated patients who had middle cerebral artery strokes, and who were moderately obtunded at the time of extubation (Glasgow Coma Score ~9–10), patients who exhibited spontaneous eye opening were significantly more likely to be successfully extubated than patients who opened their eyes only in response to speech or pain; P < .01.59 In a different study of intubated head trauma patients who had an impaired level of consciousness (mean Glasgow Coma Score of ~9), classical respiratory and general parameters like respiratory rate, rapid shallow breathing index, and heart rate were not significantly associated with extubation success.60 However, the presence of a gag reflex, ability to swallow, and ability to cough were each independently associated with extubation success. Intubated obtunded patients who had 2 or 3 of these airway functions had (34/51) [67%] and (53/59) [90%] extubation success, respectively. In this latter study, a key predictor of extubation success was patient ability to sustain visual pursuit before extubation.60 Unfortunately, it does not seem likely that one could reliably test visual pursuit during emergence from general anesthesia. However, if (1) visual pursuit were present before endovascular thrombectomy; (2) the procedure was free of apparent complication; and (3) other indications for continued intubation are absent (eg, early brain swelling), the return of 2 or 3 airway reflexes on emergence from general anesthesia might be sufficient criteria for extubation, regardless of the patient’s level of consciousness. These signs and criteria for successful extubation in patients with neurologic injury/obtundation in critical care settings have not been tested in endovascular thrombectomy patients. Thus, at present, there are no validated criteria on which to base extubation decisions in endovascular thrombectomy patients who receive general anesthesia.
The authors thank Dr Pia Löwhagen Hendén and Dr Mads Rasmussen for reviewing this review and their helpful suggestions.
Name: Bradley J. Hindman, MD.
Contribution: This author helped perform the literature reviews, review and analyze data, and write the manuscript.
Name: Franklin Dexter, MD, PhD, FASA.
Contribution: This author helped analyze and interpret the data and critically revised the drafted manuscript.
This manuscript was handled by: Gregory J. Crosby, MD.
1. Nichols C, Carrozzella J, Yeatts S, Tomsick T, Broderick J, Khatri P. Is periprocedural sedation during acute stroke therapy associated with poorer functional outcomes? J Neurointerv Surg. 2010;2:67–70.
2. Sugg RM, Jackson AS, Holloway W, Martin CO, Akhtar N, Rymer M. Is mechanical embolectomy performed in nonanesthetized patients effective? AJNR Am J Neuroradiol. 2010;31:1533–1535.
3. Abou-Chebl A, Lin R, Hussain MS, et al. Conscious sedation versus general anesthesia during endovascular therapy for acute anterior circulation stroke: preliminary results from a retrospective, multicenter study. Stroke. 2010;41:1175–1179.
4. Jumaa MA, Zhang F, Ruiz-Ares G, et al. Comparison of safety and clinical and radiographic outcomes in endovascular acute stroke therapy for proximal middle cerebral artery occlusion with intubation and general anesthesia versus the nonintubated state. Stroke. 2010;41:1180–1184.
5. Davis MJ, Menon BK, Baghirzada LB, et al; Calgary Stroke Program. Anesthetic management and outcome in patients during endovascular therapy for acute stroke. Anesthesiology. 2012;116:396–405.
6. Hassan AE, Chaudhry SA, Zacharatos H, et al. Increased rate of aspiration pneumonia and poor discharge outcome among acute ischemic stroke patients following intubation for endovascular treatment. Neurocrit Care. 2012;16:246–250.
7. Langner S, Khaw AV, Fretwurst T, Angermaier A, Hosten N, Kirsch M. Endovascular treatment of acute ischemic stroke under conscious sedation compared to general anesthesia - safety, feasibility and clinical and radiological outcome. Rofo. 2013;185:320–327.
8. Li F, Deshaies EM, Singla A, et al. Impact of anesthesia on mortality during endovascular clot removal for acute ischemic stroke. J Neurosurg Anesthesiol. 2014;26:286–290.
9. John S, Thebo U, Gomes J, et al. Intra-arterial therapy for acute ischemic stroke under general anesthesia versus monitored anesthesia care. Cerebrovasc Dis. 2014;38:262–267.
10. Whalin MK, Lopian S, Wyatt K, et al. Dexmedetomidine: a safe alternative to general anesthesia for endovascular stroke treatment. J Neurointerv Surg. 2014;6:270–275.
11. Abou-Chebl A, Zaidat OO, Castonguay AC, et al. North American SOLITAIRE Stent-Retriever Acute Stroke Registry: choice of anesthesia and outcomes. Stroke. 2014;45:1396–1401.
12. Abou-Chebl A, Yeatts SD, Yan B, et al. Impact of general anesthesia on safety and outcomes in the endovascular arm of Interventional Management of Stroke (IMS) III trial. Stroke. 2015;46:2142–2148.
13. McDonald JS, Brinjikji W, Rabinstein AA, Cloft HJ, Lanzino G, Kallmes DF. Conscious sedation versus general anaesthesia during mechanical thrombectomy for stroke: a propensity score analysis. J Neurointerv Surg. 2015;7:789–794.
14. van den Berg LA, Koelman DL, Berkhemer OA, et al; MR CLEAN pretrial study group; Participating centers. Type of anesthesia and differences in clinical outcome after intra-arterial treatment for ischemic stroke. Stroke. 2015;46:1257–1262.
15. Mundiyanapurath S, Schönenberger S, Rosales ML, et al. Circulatory and respiratory parameters during acute endovascular stroke therapy in conscious sedation or general anesthesia. J Stroke Cerebrovasc Dis. 2015;24:1244–1249.
16. Sivasankar C, Stiefel M, Miano TA, et al. Anesthetic variation and potential impact of anesthetics used during endovascular management of acute ischemic stroke. J Neurointerv Surg. 2016;8:1101–1106.
17. Just C, Rizek P, Tryphonopoulos P, Pelz D, Arango M. Outcomes of general anesthesia and conscious sedation in endovascular treatment for stroke. Can J Neurol Sci. 2016;43:655–658.
18. Janssen H, Buchholz G, Killer M, Ertl L, Brückmann H, Lutz J. General anesthesia versus conscious sedation in acute stroke treatment: the importance of head immobilization. Cardiovasc Intervent Radiol. 2016;39:1239–1244.
19. Berkhemer OA, van den Berg LA, Fransen PS, et al; MR CLEAN Investigators. The effect of anesthetic management during intra-arterial therapy for acute stroke in MR CLEAN. Neurology. 2016;87:656–664.
20. Bracard S, Ducrocq X, Mas JL, et al; THRACE Investigators. Mechanical thrombectomy after intravenous alteplase versus alteplase alone after stroke (THRACE): a randomised controlled trial. Lancet Neurol. 2016;15:1138–1147.
21. Jagani M, Brinjikji W, Rabinstein AA, Pasternak JJ, Kallmes DF. Hemodynamics during anesthesia for intra-arterial therapy of acute ischemic stroke. J Neurointerv Surg. 2016;8:883–888.
22. Bekelis K, Missios S, MacKenzie TA, Tjoumakaris S, Jabbour P. Anesthesia technique and outcomes of mechanical thrombectomy in patients with acute ischemic stroke. Stroke. 2017;48:361–366.
23. Slezak A, Kurmann R, Oppliger L, et al. Impact of anesthesia on the outcome of acute ischemic stroke after endovascular treatment with the Solitaire stent retriever. AJNR Am J Neuroradiol. 2017;38:1362–1367.
24. Campbell BCV, van Zwam WH, Goyal M, et al; HERMES Collaborators. Effect of general anaesthesia on functional outcome in patients with anterior circulation ischaemic stroke having endovascular thrombectomy versus standard care: a meta-analysis of individual patient data. Lancet Neurol. 2018;17:47–53.
25. Talke PO, Sharma D, Heyer EJ, Bergese SD, Blackham KA, Stevens RD. Society for Neuroscience in Anesthesiology and Critical Care expert consensus statement. Anesthetic management of endovascular treatment for acute ischemic stroke. J Neurosurg Anesthesiol. 2014;26:95–108. (Also published in Stroke
26. Brinjikji W, Murad MH, Rabinstein AA, Cloft HJ, Lanzino G, Kallmes DF. Conscious sedation versus general anesthesia during endovascular acute ischemic stroke treatment: a systematic review and meta-analysis. AJNR Am J Neuroradiol. 2015;36:525–529.
27. Brinjikji W, Pasternak J, Murad MH, et al. Anesthesia-related outcomes for endovascular stroke revascularization: a systematic review and meta-analysis. Stroke. 2017;48:2784–2791.
28. Hindman BJ. Anesthetic management of emergency endovascular thrombectomy for acute ischemic stroke, part 1: patient characteristics, determinants of effectiveness, and effect of blood pressure on outcome. Anesth Analg. 2019;128:695–705.
29. Harbord RM, Higgins JPT. Meta-regression in Stata. Stata J. 2008;8:493–519.
30. Schönenberger S, Möhlenbruch M, Pfaff J, et al. Sedation vs Intubation for Endovascular Stroke TreAtment (SIESTA) - a randomized monocentric trial. Int J Stroke. 2015;10:969–978.
31. Schönenberger S, Uhlmann L, Hacke W, et al. Effect of conscious sedation vs general anesthesia on early neurological improvement among patients with ischemic stroke undergoing endovascular thrombectomy: a randomized clinical trial. JAMA. 2016;316:1986–1996.
32. Löwhagen Hendén P, Rentzos A, Karlsson JE, et al. General anesthesia versus conscious sedation for endovascular treatment of acute ischemic stroke: the AnStroke Trial (Anesthesia During Stroke). Stroke. 2017;48:1601–1607.
33. Simonsen CZ, Sørensen LH, Juul N, et al. Anesthetic strategy during endovascular therapy: general anesthesia or conscious sedation? (GOLIATH - General or Local Anesthesia in Intra Arterial Therapy) A single-center randomized trial. Int J Stroke. 2016;11:1045–1052.
34. Simonsen CZ, Yoo AJ, Sørensen LH, et al. Effect of general anesthesia and conscious sedation during endovascular therapy on infarct growth and clinical outcomes in acute ischemic stroke: a randomized clinical trial. JAMA Neurol. 2018;75:470–477.
35. Rasmussen LK, Simonsen CZ, Löwhagen Hendén P, Bösel J, Rasmussen M. Anaesthesia for endovascular treatment of acute ischemic stroke: still controversial? Curr Anesthesiol Rep. 2018;8:270278.
36. Schönenberger S, Uhlmann L, Ungerer M, et al. Association of blood pressure with short- and long-term functional outcome after stroke thrombectomy: post hoc analysis of the SIESTA trial. Stroke. 2018;49:1451–1456.
37. Rasmussen M, Espelund US, Juul N, et al. The influence of blood pressure management on neurological outcome in endovascular therapy for acute ischaemic stroke. Br J Anaesth. 2018;120:1287–1294.
38. Pfaff JAR, Schönenberger S, Nagel S, et al. Effect of general anesthesia versus conscious sedation for stroke thrombectomy on angiographic workflow in a randomized trial: a post hoc analysis of the SIESTA trial. Radiology. 2018;286:1016–1021.
39. Powers WJ, Rabinstein AA, Ackerson T, et al; American Heart Association Stroke Council. 2018 guidelines for the early management of patients with acute ischemic stroke: a guideline for healthcare professionals from the American Heart Association/American Stroke Association. Stroke. 2018;49:e46–e110.
40. Goyal M, Demchuk AM, Menon BK, et al; ESCAPE Trial Investigators. Randomized assessment of rapid endovascular treatment of ischemic stroke. N Engl J Med. 2015;372:1019–1030.
41. Jovin TG, Chamorro A, Cobo E, et al; REVASCAT Trial Investigators. Thrombectomy within 8 hours after symptom onset in ischemic stroke. N Engl J Med. 2015;372:2296–2306.
42. Mattle HP, Arnold M, Lindsberg PJ, Schonewille WJ, Schroth G. Basilar artery occlusion. Lancet Neurol. 2011;10:1002–1014.
43. Jadhav AP, Bouslama M, Aghaebrahim A, et al. Monitored anesthesia care vs intubation for vertebrobasilar stroke endovascular therapy. JAMA Neurol. 2017;74:704–709.
44. Takahashi CE, Brambrink AM, Aziz MF, et al. Association of intraprocedural blood pressure and end tidal carbon dioxide with outcome after acute stroke intervention. Neurocrit Care. 2014;20:202–208.
45. Athiraman U, Sultan-Qurraie A, Nair B, et al. Endovascular treatment of acute ischemic stroke under general anesthesia: predictors of good outcome. J Neurosurg Anesthesiol. 2018;30:223–230.
46. Alexandrov AV, Nguyen HT, Rubiera M, et al. Prevalence and risk factors associated with reversed Robin Hood syndrome in acute ischemic stroke. Stroke. 2009;40:2738–2742.
47. Poublanc J, Han JS, Mandell DM, et al. Vascular steal explains early paradoxical blood oxygen level-dependent cerebrovascular response in brain regions with delayed arterial transit times. Cerebrovasc Dis Extra. 2013;3:55–64.
48. Banik S, Fisher JA, McKetton L, Venkatraghavan L. Mapping intracerebral steal during a hypercapnic challenge. Can J Anaesth. 2017;64:1265–1266.
49. Patel VN, Gupta R, Horn CM, Thomas TT, Nogueira RG. The neuro-critical care management of the endovascular stroke patient. Curr Treat Options Neurol. 2013;15:113–124.
50. Al-Mufti F, Dancour E, Amuluru K, et al. Neurocritical care of emergent large-vessel occlusion: the era of a new standard of care. J Intensive Care Med. 2017;32:373–386.
51. Male S, Nickele C, Elijovich L. Critical care of brain reperfusion. Curr Neurol Neurosci Rep. 2016;16:23.
52. Rincon F, Kang J, Maltenfort M, et al. Association between hyperoxia and mortality after stroke: a multicenter cohort study. Crit Care Med. 2014;42:387–396.
53. Helmerhorst HJ, Roos-Blom MJ, van Westerloo DJ, de Jonge E. Association between arterial hyperoxia and outcome in subsets of critical illness: a systematic review, meta-analysis, and meta-regression of cohort studies. Crit Care Med. 2015;43:1508–1519.
54. Nair GB, Niederman MS. Ventilator-associated pneumonia: present understanding and ongoing debates. Intensive Care Med. 2015;41:34–48.
55. Nieman GF, Satalin J, Kollisch-Singule M, et al. Physiology in medicine: understanding dynamic alveolar physiology to minimize ventilator-induced lung injury. J Appl Physiol (1985). 2017;122:1516–1522.
56. Caroff DA, Szumita PM, Klompas M. The relationship between sedatives, sedative strategy, and healthcare-associated infection: a systematic review. Infect Control Hosp Epidemiol. 2016;37:1234–1242.
57. Klompas M, Li L, Szumita P, Kleinman K, Murphy MV; CDC Prevention Epicenters Program. Associations between different sedatives and ventilator-associated events, length of stay, and mortality in patients who were mechanically ventilated. Chest. 2016;149:1373–1379.
58. Nikoubashman O, Schürmann K, Probst T, et al. Clinical impact of ventilation duration in patients with stroke undergoing interventional treatment under general anesthesia: the shorter the better? AJNR Am J Neuroradiol. 2016;37:1074–1079.
59. Wendell LC, Raser J, Kasner S, Park S. Predictors of extubation success in patients with middle cerebral artery acute ischemic stroke. Stroke Res Treat. 2011;2011:248789.
60. Godet T, Chabanne R, Marin J, et al. Extubation failure in brain-injured patients: risk factors and development of a prediction score in a preliminary prospective cohort study. Anesthesiology. 2017;126:104–114.