In ischemic stroke patients, general anesthesia (GA) for endovascular thrombectomy has been reported to result in a worse outcome compared with local anesthesia or conscious sedation (CS).1,2 This could reflect a true signal of harm or result from selection bias, treatment delay or unmeasured confounding. An early meta-analysis used post hoc analyses of randomized controlled trials (RCTs) of endovascular thrombectomy versus standard therapy, in which the decision to perform the procedure under GA or CS was not randomized.2 The analysis adjusted for baseline differences and found that patients treated with CS had better outcomes at 3 months compared with those treated with GA. Other meta-analyses combined the results of studies in which patients were randomized to GA or CS with nonrandomized cohort studies, and concluded that thrombectomy patients treated with GA had a worse outcome at 3 months.3,4 In contrast, a recent study presented a meta-analysis limited to 3 trials where patients were randomized to GA or CS, and found that GA was associated with higher rates of successful recanalization and functional independence than CS.5 Since this time a further small RCT has been published. We therefore performed an updated systematic review and meta-analysis with the aim of comparing the efficacy and safety of GA or CS in endovascular thrombectomy.
Literature Search Strategy and Selection Criteria
This study was performed in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) statement.6 Literature published before April 1, 2019 was systematically searched on Medline, Embase, and the Cochrane Central Register of Controlled Trials with no restrictions of the source language. The keywords for the population, intervention, and trial design for the detailed search strategy are outlined in the Supplementary Material (Supplementary Material 1, Supplemental Digital Content 1, http://links.lww.com/JNA/A207: Detailed Search Strategy). Two investigators (D.C., W.K.D.) independently reviewed each retrieved literature item. Studies were considered eligible if they were RCTs comparing GA and CS during endovascular thrombectomy and reported at least one of: successful recanalization, defined as Thrombolysis in Cerebral Infarction (TICI) score of 2b to 3; early neurological improvement, defined as a change in National Institutes for Health Stroke Scale (NIHSS) at 24 hours from baseline; intracerebral hemorrhage, defined as any postprocedure asymptomatic or symptomatic intracerebral hemorrhage; good functional outcome, defined as a modified Rankin Scale (mRS) score of 0 to 2 at 3 months; and 3-month mortality. Disagreements were resolved by consensus.
Two independent investigators (D.C., W.K.D.) reviewed and extracted baseline data and outcomes from published manuscripts for each eligible study. This included demographic data, baseline NIHSS and Alberta Stroke Program Early CT Score (ASPECTS), the number of patients treated with intravenous (IV) alteplase, time from symptom onset to groin puncture, door to groin puncture, groin puncture to recanalization, successful recanalization, early neurological improvement, intracerebral hemorrhage, good functional outcome at 3 months, and 3-month mortality.
Risk of Bias Assessment
Two investigators (D.C., W.K.D.) independently assessed the risk of bias in individual studies at study level using the Revised Cochrane risk-of-bias tool for randomised trials (RoB 2.0).7 Studies were assessed in 5 domains: risk of bias arising from randomization, risk of bias due to deviations from intended interventions, missing outcome data, risk of bias in outcome measurement, and risk of bias in selection of reported result.
All statistical analyses were performed using Review Manager 5 (RevMan 5.3. Copenhagen: The Nordic Cochrane Centre, The Cochrane Collaboration, 2014). Efficacy outcomes were successful recanalization and good functional outcome at 3 months. Safety outcomes were intracerebral hemorrhage and 3-month mortality. Summary measures for these categorical variables are reported as odds ratio (OR) and illustrated using forest plots. Measures of consistency were summarized by an I2 statistic to assess heterogeneity. A value for I2 >50% was considered to indicate moderate to high heterogeneity. All analyses were performed using random effect models.
Characteristics of Included Studies
A systematic search of the 3 databases identified 206 endovascular thrombectomy studies, of which 4 studies—Sedation versus Intubation for Endovascular Stroke Treatment (SIESTA),8 Anesthesia during Stroke (AnStroke),9 General or Local Anesthesia in Intra Arterial Therapy (GOLIATH),10 and Choice of Anesthesia for Endovascular Treatment of Acute Ischemic Stroke (CANVAS) pilot study11—randomized patients to GA or CS and were included in this analysis (Fig. 1). Publication bias was not assessed qualitatively or quantitatively as the pooled results are from 4 small studies and analysis would unlikely detect publication bias in this setting.12 There was no evidence of selective reporting of outcomes. All studies had a low risk of bias according to the Revised Cochrane risk-of-bias tool for randomized trials and so further analysis of bias risk was not performed (Supplementary Material 2: Supplementary Table 1, Supplemental Digital Content 2, http://links.lww.com/JNA/A208). Sensitivity analyses were not performed as all outcomes had low heterogeneity.
There were 408 endovascular thrombectomy patients (231 male and 177 female) randomized to GA (203) or CS (205). Study and baseline characteristics are shown in Table 1. Mean patient age ranged from 63.5 to 72.5 years, mean baseline NIHSS from 13.5 to 18.5, and mean ASPECTS score from 8 to 10 in the 2 studies that reported ASPECTS. The proportion of patients treated with IV alteplase ranged from 50% to 76.7%. The mean stroke onset to groin puncture time ranged from 119 to 318 minutes. Successful recanalization was achieved in 68.6% to 90% of participants. The median improvement in NIHSS from baseline to 24 hours ranged from 4.5 to 8.5, and the proportion of patients experiencing good functional outcome at 3 months ranged from 27.6% to 59.7%. Intracerebral hemorrhage occurred in 2.0% to 5.5%, and mortality at 3 months in 10.2% to 24.6% participants.
TABLE 1 -
Baseline Characteristics and Main Outcomes of the Individual Studies
|Sex, male, n (%)
|Hypertension, n (%)
|Atrial fibrillation, n (%)
|Dyslipidemia, n (%)
|Diabetes, n (%)
|Smoking, n (%)
|IV tPA, n (%)
|Onset to door time (min)
|Door to groin time (min)
|Groin puncture to reperfusion (min)
|TICI 2b-3 recanalization, n (%)
|Change in NIHSS at 24 hours, n (IQR)
||5 (−2 to 10)
||4 (−2 to 10)
|mRS 0-2 at 3 mo, n (%)
|Intracerebral hemorrhage, n (%)
|Mortality at 3 mo, n (%)
|Conversion to GA, n (%)
ASPECTS indicates Alberta Stroke Program Early CT Score; CS, conscious sedation; GA, general anesthesia; IV tPA, intravenous tissue plasminogen activator; mRS, modified Rankin Score; NA, not applicable; NIHSS, National Institutes of Health Stroke Scale; NR, not reported; TICI, Thrombolysis in Cerebral Infarction.
Pooled outcomes are summarized in Table 2. Forest plots of the pooled estimates of treatment effects, confidence intervals (CI), and measures of heterogeneity for the main outcomes are presented in Figure 2. Early neurological improvement was dropped as an outcome as one trial reported change in NIHSS from baseline as mean and 95% CI and the other studies as the median and interquartile range. We considered imputing mean values from medians using the Wan method,13,14 but believe that the potential for data distortion was too high.
TABLE 2 -
||OR (95% CI)
| Successful recanalization (%)
| Good functional outcome* (%)
| Intracerebral hemorrhage (%)
| 3-mo mortality (%)
*Modified Rankin Scale score 0 to 2 at 3 months.
CI indicates confidence interval; CS, conscious sedation; GA, general anesthesia; OR, odds ratio.
Successful Recanalization (TICI 2b to 3)
All 4 RCTs reported successful recanalization (Fig. 2A). One of the 4 RCTs (GOLIATH)10 showed a significantly higher proportion of patients having successful recanalization with GA compared with CS. In the pooled analysis, patients treated with GA achieved a higher proportion of successful recanalization than CS (GA 86.2% vs. CS 74.6%; OR: 2.14, 95% CI: 1.26-3.62; P=0.005). There was no significant heterogeneity detected between studies (P=0.47, I2=0%).
Good Functional Outcome (mRS 0 to 2)
All 4 RCTs reported the functional outcome at 3 months (Fig. 2B). The SIESTA trial8 showed a significantly higher proportion of patients demonstrating good functional outcome in the GA group. In the pooled analysis, GA showed a higher proportion of patients exhibiting good functional outcome compared with CS (GA 49.3% vs. CS 36.6%; OR: 1.71, 95% CI: 1.13-2.59; P=0.01). There was no significant heterogeneity detected between studies (P=0.45, I2=0%).
All 4 RCTs reported intracerebral hemorrhage (Fig. 2C). The overall event rate in each individual RCT was low. In the pooled analysis, there was no difference in intracerebral hemorrhage rates between GA and CS (GA 2.5% vs. 4.9% CS; OR: 0.61, 95% CI: 0.20-1.85; P=0.38). There was no significant heterogeneity detected between studies (P=0.46, I2=0%).
All 4 RCTs reported 3-month mortality or 3-month mortality could be calculated from the tabulated mRS scores (Fig. 2D). None of the 4 RCTs reported a significant difference in this outcome. In the pooled analysis, there was no difference in 3-month mortality between GA and CS (GA 14.8% vs. CS 21.5%; OR: 0.62, 95% CI: 0.33-1.17; P=0.14). There was no significant heterogeneity detected between studies (P=0.28, I2=21%).
This systematic review and meta-analysis of pooled data from four RCTs showed that endovascular thrombectomy patients treated with GA had higher rates of successful recanalization and good functional outcome at three months compared with patients treated with CS. GA treated patients also had non-significant trends to lower 3-month mortality. The proportion of patients with the good functional outcome at 3 months was 49.3% in GA patients and 36.6% in CS patients, an absolute difference of 12.7%. This means that for every 7.9 thrombectomy patients treated with GA, one more patient achieved a good functional outcome at 3 months compared with CS. The overall conversion rate from CS to GA was 12.7%, suggesting the true treatment effect may be higher than seen in the intention to treat analyses.
These findings are in contrast to observational studies and meta-analyses of observational studies, where endovascular thrombectomy patients treated with GA were reported to have a worse outcome than those treated with CS.1,2 Previous meta-analyses including 3 of the 4 RCTs presented here found similar results to ours, with patients randomized to GA having better outcomes than those treated with CS.3–5 Following the publication of a fourth RCT, we have performed a further meta-analysis that provides the most up to date data and gives greater precision in the estimates of benefit of GA compared with CS.
In each of the 4 RCTs, GA and CS regimes were protocolized and summary descriptions of drugs, doses and associated cardiovascular and respiratory physiology were reported, with physiological parameters largely equivalent between groups.8–11 In contrast, the observational studies did not report the anesthetic drugs or doses, and physiological parameters are largely unreported.1–3 We speculate that high quality, specialized neuroanesthesia care with protocol driven management of GA may have contributed to the superior outcomes in the RCTs included in this analysis.
The finding that GA results in improved outcome are clinically plausible. Greater patient immobilization and controlled apnea at critical times during thrombectomy likely confer superior procedural conditions. During GA, oxygenation, ventilation, heart rate, systolic blood pressure, and temperature are physiological parameters that can be modified and may improve outcome in brain injured patient.15 For example, protocolized management of procedural BP provided by a specialized team avoids sudden drops in BP and allows the other potential benefits of GA to be manifest.16,17 We suggest that the unreported and unadjusted BP differences may have been a major contributor to the signal of harm from GA compared with CS in the observational studies.
There are plausible pharmacological and physiological reasons why drugs used in GA may be protective in humans. Common agents used for maintenance of GA such as propofol and sevoflurane can reduce the cerebral metabolic rate by 60%,18–20 and show evidence of neuroprotection in animal models of neurological injury.21 Inhalational anesthetic agents, such as isoflurane, sevoflurane and desflurane, impair cerebral autoregulation in a dose-dependent manner.22 Hence, the maintenance anesthetic agent could increase the potential for harm due to relative hypotension, or the potential for benefit through increased penumbral perfusion from augmented BP during endovascular thrombectomy.15,23 This study suggests that the effects of different GA techniques in patients undergoing endovascular thrombectomy, such as intravenous or inhaled agents, warrants further investigation.
This study has a number of limitations. The 4 RCTs were small single-center studies with a total of only 408 participants in the pooled analysis. None of the individual trials were designed primarily to investigate functional outcomes at 3 months. Despite this, the pooled results displayed little heterogeneity. The low number of relatively small RCTs means that the pooled estimate of treatment effects may be subject to reporting bias. Patients were managed by highly specialized anesthesiology teams that are not always available in the real world, where pools of nonspecialized anesthesiologists often provide stroke care. This limits the generalizability of these results. Large, multicenter RCTs are still lacking so further data will be needed to provide better evidence. This will be available when the full CANVAS study (NCT02677415),24 General Anesthesia versus Sedation during intra-arterial treatment for Stroke (GASS; NCT02822144),25 and SEdation Versus General Anesthesia for Endovascular Therapy in Acute Ischemic Stroke (SEGA; NCT03263117)26 trials are reported.
This meta-analysis suggests that endovascular thrombectomy patients treated with GA do not have inferior outcomes compared with those treated with CS. In centers with high quality, protocolized, specialized neuro-anesthesia care, GA-treated thrombectomy patients had superior recanalization rates and better functional outcome at 3 months.
1. van den Berg LA, Koelman DLH, Berkhemer OA, et al. Type of anesthesia and differences in clinical outcome after intra-arterial treatment for ischemic stroke. Stroke. 2015;46:1257–1262.
2. Campbell BCV, van Zwam WH, Goyal M, et al. 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.
3. Gravel G, Boulouis G, Benhassen W, et al. Anaesthetic management during intracranial mechanical thrombectomy: systematic review and meta-analysis of current data. J Neurol Neurosurg Psychiatry. 2019;90:68–74.
4. Goyal N, Malhotra K, Ishfaq MF, et al. Current evidence for anesthesia management during endovascular stroke therapy: updated systematic review and meta-analysis. J Neurointerv Surg. 2019;11:107–113.
5. Zhang Y, Jia L, Fang F, et al. General anesthesia versus conscious sedation for intracranial mechanical thrombectomy: a systematic review and meta‐analysis of randomized clinical trials. J Am Heart Assoc. 2019;8:e011754.
6. Moher D, Liberati A, Tetzlaff J, et al. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. Int J Surg. 2010;8:336–341.
7. Higgins JPT, Sterne JAC, Savović J, et al. A revised tool for assessing risk of bias in randomized trials. Cochrane Database Syst Rev. 2016;10:29–31.
8. 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.
9. Löwhagen Hendén P, Rentzos A, Karlsson J-E, 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.
10. 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.
11. Sun J, Liang F, Wu Y, et al. Choice of ANesthesia for EndoVAScular Treatment of Acute Ischemic Stroke (CANVAS): results of the CANVAS Pilot Randomized Controlled Trial. J Neurosurg Anesthesiol. 2018. [Epub ahead of print].
12. Debray TPA, Moons KGM, Riley RD. Detecting small-study effects and funnel plot asymmetry in meta-analysis of survival data: a comparison of new and existing tests. Res Synth Methods. 2018;9:41–50.
13. Wan X, Wang W, Liu J, et al. Estimating the sample mean and standard deviation from the sample size, median, range and/or interquartile range. BMC Med Res Methodol. 2014;14:135.
14. Weir CJ, Butcher I, Assi V, et al. Dealing with missing standard deviation and mean values in meta-analysis of continuous outcomes: a systematic review. BMC Med Res Methodol. 2018;18:25.
15. 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.
16. Ishida K, Berger M, Nadler J, et al. Anesthetic neuroprotection: antecedents and an appraisal of preclinical and clinical data quality. Curr Pharm Des. 2014;20:5751–5765.
17. Zwerus R, Absalom A. Update on anesthetic neuroprotection. Curr Opin Anaesthesiol. 2015;28:424–430.
18. Conti A, Iacopino DG, Fodale V, et al. Cerebral haemodynamic changes during propofol-remifentanil or sevoflurane anaesthesia: transcranial doppler study under bispectral index monitoring. Br J Anaesth. 2006;97:333–339.
19. Rudolph U, Antkowiak B. Molecular and neuronal substrates for general anaesthetics. Nat Rev Neurosci. 2004;5:709–720.
20. Oshima T, Karasawa F, Okazaki Y, et al. Effects of sevoflurane on cerebral blood flow and cerebral metabolic rate of oxygen in human beings: a comparison with isoflurane. Eur J Anaesthesiol. 2003;20:543–547.
21. Archer DP, Walker AM, McCann SK, et al. Anesthetic neuroprotection in experimental stroke in rodents: a systematic review and meta-analysis. Anesthesiology. 2017;126:653–665.
22. Strebel S, Lam AM, Matta B, et al. Dynamic and static cerebral autoregulation during isoflurane, desflurane, and propofol anesthesia. Anesthesiology. 1995;83:66–76.
23. Diprose WK, Wang MTM, Campbell D, et al. Intravenous propofol versus volatile anesthetics for stroke endovascular thrombectomy
. J Neurosurg Anesthesiol. 2019. [Epub ahead of print].
24. Peng Y, Li Y, Jian M, et al. Choice of anesthesia for endovascular treatment of acute ischemic stroke: protocol for a randomized controlled (CANVAS) trial. Int J Stroke. 2017;12:991–997.
25. NIH US National Library of Medicine. Identifier NCT02822144, General Anesthesia Versus Sedation During Intra-arterial Treatment for Stroke (GASS). ClinicalTrials.gov. 2016. Available at: https://clinicaltrials.gov/ct2/show/NCT02822144
. Accessed May 1, 2019.
26. NIH US National Library of Medicine. Identifier NCT03263117, SEdation Versus General Anesthesia for Endovascular Therapy in Acute Ischemic Stroke (SEGA). ClinicalTrials.gov. 2017. Available at: https://clinicaltrials.gov/ct2/show/NCT03263117
. Accessed May 1, 2019.