Keywords
Anesthesia, General - Conscious Sedation - Ischemic Stroke - Thrombectomy
Palavras-chave
Anestesia Geral - Sedação Consciente - AVC Isquêmico - Trombectomia
INTRODUCTION
Mechanical thrombectomy is indicated for patients with acute ischemic stroke due to
a large artery occlusion in the anterior or posterior circulation who can be treated
within 24 hours of the time last known to be well, whether or not they receive IV
tPA for the same ischemic stroke event.[1]
Mechanical thrombectomy can be performed with general anesthesia (GA) or conscious
sedation (CS). The choice between anesthesia management is usually individualized
based on patient and procedural factors and resource availability,[2] highlighting the selection's uncertainty. GA offers a still patient and a secure
airway, the ability to institute controlled apnea, and the ability to fully control
procedural pain. In contrast, CS has a shorter time for treatment initiation and allows
neurologic examination during and after the procedure. The best evidence suggests
that GA with optimal hemodynamic control may improve technical success[3] and functional outcomes.[4]
The last meta-analysis included seven randomized trials of patients who underwent
mechanical thrombectomy for anterior circulation ischemic stroke with GA versus non-GA
techniques (local anesthesia, CS).[4] GA improved the recanalization rate and increased the functional independence rate
(modified Rankin Scale 0 to 2) at three months.
However, these findings come from observational and preliminary studies[5]
[6] and several randomized trials that reported either no difference or a slight improvement
in infarct size or other clinical outcomes with GA, while some retrospective and prospective
observational studies have reported worse outcomes with GA for mechanical thrombectomy.
A recent 2023 randomized clinical trial explored the impact of anesthesia or sedation
on periprocedural complications and functional outcomes.[7] This study revealed that neither anesthesia nor procedural sedation significantly
influenced these outcomes. The best anesthetic strategy during mechanical thrombectomy
is still debatable in this context.
Therefore, this study presents a systematic review and meta-analysis comparing the
effectiveness and safety of GA versus CS in patients who underwent a mechanical thrombectomy.
METHODS
The Systematic review and meta-analysis were performed following the Cochrane Collaboration's
tool for assessing the risk of bias and the Preferred Reporting Items for Systematic
Reviews and Meta-Analysis (PRISMA) statement.[8] The study protocol was registered in the International Prospective Register of Systematic
Reviews (PROSPERO) with the registration number CRD42023439944.
Eligibility criteria and data extraction
Studies with the following eligibility criteria were included for analysis:
-
randomized controlled trials (RCTs);
-
comparing GA with CS;
-
enrolling patients with AIS; and
-
reporting at least one of the outcomes of interest.
Non-randomized studies and trials without a control group were excluded; studies with
overlapping populations with the most significant number of patients were included.
Two authors (A.C.F.F.S and L.L.S.C.) independently extracted data following prespecified
criteria for search, data extraction, and quality assessment methods. Disagreements
were resolved by consensus between the two authors and the senior author (A.C.F.F.S.,
L.L.S.C. and L.A.C.).
Search strategy
PubMed, Embase, and Cochrane Library were systematically searched on June 22, 2023,
for studies published solely in English. The search strategy included the following
terms: 'acute ischemic stroke,' 'ischemic stroke,' 'posterior circulation stroke,'
'endovascular therapy,' 'endovascular treatment,' 'thrombectomy,' 'vertebrobasilar
stroke,' 'general anesthesia,' 'general anesthesia,' 'Non-general anesthesia,' 'conscious
sedation,' 'moderate sedation, ''local anesthesia,' 'anesthesia care.' Additionally,
the reference lists of all included studies and meta-analyses were manually assessed
for additional studies. The search strategy is detailed in [Supplementary Material] (https://www.arquivosdeneuropsiquiatria.org/wp-content/uploads/2024/04/ANP-2023.0290-Supplementary-Material-atualizado.docx), [Supplementary Material Table S1].
Endpoints
Outcomes of interest were:
Quality assessment
The Cochrane Collaboration's tool for risk of bias in randomized trials (ROB 2) was
used to assess the quality assessment of individual studies.[9] Two authors (A.C.F.F.S. and V.D.B.C) conducted the quality assessment independently.
Each trial was rated as having a high, low, or unclear risk of bias in the five domains:
randomization process, deviations from the intended interventions, missing outcomes,
measurement of the outcome, and selection of reported results. The quality of evidence
was analyzed according to the Grading of Recommendations, Assessment, Development,
and Evaluation (GRADE).[10] The outcomes were labeled with very low, low, moderate, or high-quality evidence
based on five domains: risk of bias, inconsistency of results, imprecision, publication
bias, and magnitude of treatment effects. Funnel plots of individual study weights
versus point estimates were used to detect evidence of publication bias.
Sensitivity analysis and meta-regression
We performed a leave-one-out sensitivity analysis for recanalization success, good
functional recovery, three-month mortality, and cerebral hemorrhage. We removed each
study from the outcome assessment to determine whether the results depended on a single
study. Furthermore, we conducted a meta-regression to investigate any correlation
between the National Institute of Health Stroke Scale (NIHSS) score and recanalization
success.
Statistical analyses
Odds ratio (OR) with 95% confidence intervals (CI) was computed to compare the incidence
of treatment effect for binary endpoints. Cochran Q test and I2 statistics were used
to analyze heterogeneity; p-values were considered low heterogeneity if p >10 and I2 <25%. DerSimonian and Laird random-effects models were used.
Review Manager 5.4 (Nordic Cochrane Centre, The Cochrane Collaboration, Copenhagen,
Denmark) and R statistical software, version 4.3.0 (R Foundation for Statistical Computing),
were used to perform the statistical analysis.
RESULTS
Study selection and characteristics
As detailed in [Figure 1], 868 studies were found. After removing duplicates and abstract screening, 21 studies
were fully assessed for inclusion. A total of 8 studies and 1,300 patients were included,
of whom 650 (50%) underwent GA.[3]
[7]
[11]
[12]
[13]
[14]
[15]
[16] Additional population characteristics are reported in [Table 1].
Figure 1 Preferred Reporting Items for Systematic Reviews and Meta-Analysis (PRISMA) flow
diagram of study screening and selection.
Table 1
Baseline characteristics of included studies
|
Study
|
n
|
Follow-up
|
Population
|
Age, years (SD)
|
Female, n (%)
|
ASPECTS, median
|
NIHSS score, median (IQR)
|
mRS 0, n (%)
|
M1, n (%)
|
Hypertension n, (%)
|
|
Chabanne 2023[7]
|
273
|
90 days
|
Large vessel, AC
|
GA: 72.0 (13.2)
CS: 71.3 (14.4)
|
GA: 70 (51.9)
CS: 72 (52.2)
|
GA: 8 (7–9)
CS: 8 (7–9)
|
GA: 16 (11–20)
CS: 15 (11–20)
|
NA
|
GA: 86 (63.7)
CS: 84 (60.9)
|
NA
|
|
Hendén 2017[14]
|
90
|
90 days
|
Large vessel, AC
|
GA: 73 (65–80)
CS: 72 (66–82)
|
GA: 19 (42)
CS: 22 (49)
|
GA: 10 (8–10)
CS: 10 (9–10)
|
GA: 20 (15.5–23)
CS: 17 (14–20.5)
|
NA
|
GA: 26 (58)
CS: 26 (58)
|
GA: 27 (60)
CS: 22 (49)
|
|
Hu 2021[16]
|
139
|
90 days
|
Vertebro
basilar stroke
|
GA: 72.1 (6.8)
CS: 71.9 (7.5)
|
GA: 34 (47.2)
CS: 33 (49.25)
|
NA
|
NA
|
GA: 55 (76.39)
CS: 50 (74.63)
|
NA
|
GA: 34 (47.22)
CS: 31 (46.27)
|
|
Liang 2023[11]
|
87
|
90 days
|
Vertebro
basilar stroke
|
GA: 64 (11)
CS: 60 (13)
|
GA: 10 (23.3)
CS: 6 (13.6)
|
NA
|
GA: 16 (12–21)
CS: 15 (12–18)
|
NA
|
NA
|
GA: 32 (74.4)
CS: 31 (70.5)
|
|
Maurice 2022[3]
|
345
|
6 months
|
Large vessel, AC
|
GA: 70.8 (+13.0)
CS: 72.6 (+12.3)
|
GA: 80 (47)
CS: 77 (44)
|
NA
|
GA: 16 ± 6
CS: 16 ± 5
|
NA
|
GA: 99 (59)
CS: 109 (62)
|
GA: 97 (57)
CS: 124 (70)
|
|
Ren 2020[12]
|
90
|
90 days
|
Large vessel, AC
|
GA: 69.21 (5.78)
CS: 69.19 (6.46)
|
GA: 22(45.8)
CS: 18(42.8)
|
GA: 9 (8–10)
CS: 9 (8–10.25)
|
GA: 14 (11–16)
CS: 14 (11–16)
|
GA: 21 (35.42)
CS: 23 (47.62)
|
GA: 15(31.25)
CS: 13(30.95)
|
GA: 17 (35.42)
CS: 20 (47.62)
|
|
Schönenberger 2016[15]
|
150
|
90 days
|
Large vessel, AC
|
GA: 71.8 (12.9)
CS: 71.2 (14.7)
|
GA: 25 (34.2)
CS: 35 (45.5)
|
GA: 8 (7–9)
CS: 8 (6.25–9)
|
GA: 17 (13–20)
CS: 17 (14–20)
|
GA: 40 (54.8)
CS: 39 (50.6)
|
GA: 39 (53.4)
CS: 43 (55.8)
|
GA: 53 (72.6)
CS: 54 (70.1)
|
|
Simonsen 2018[13]
|
128
|
90 days
|
Large vessel, AC
|
GA: 71.0 (10.0)
CS: 71.8 (12.8)
|
GA: 29 (44.6)
CS: 33 (52.4)
|
NA
|
GA: 18 (13–21)
CS: 17 (15–21)
|
GA: 50 (76.9)
CS: 51 (81.0)
|
GA: 21 (32.3)
CS: 32 (50.8)
|
GA: 39 (60.0)
CS: 32 (50.8)
|
Abbreviations: AC, Anterior Circulation; ASPECTS, Alberta Stroke Program Early Computed
Tomography Score; CS, conscious sedation; GA, general anesthesia; M1, Middle cerebral
artery M1 segment; mRS, modified Rankin scale; N, number of patients; NA, not available;
NIHSS, National Institutes of Health Stroke Scale.
Notes: Scores on the NIHSS range from 0 to 42, with higher scores indicating a more
severe deficit.
† Data presented as mean (standard deviation) or median (interquartile range).
Pooled analysis of all studies
Recanalization success was significantly higher in patients treated with GA as compared
with CS (OR 1.68; 95% CI 1.26–2.24; p = 0.04; [Figure 2]). However, there was no significant difference between patients treated with GA
and CS for good functional recovery (OR 1.13; 95% CI 0.76- 1.67; p = 0.56; [Figure 3]), three-month mortality (OR 0.99; 95% CI 0.73- 1.34; p = 0.95; [Figure 4]), and cerebral hemorrhage (OR 0.97; 95% CI 0.68- 1.38; p = 0.88); [Figure 5]). Similarly, the incidence of pneumonia was not statistically significant between
groups (OR 1.23; 95% CI 0.56- 2.69; p = 0.61) ([Supplementary Material Figure S1]).
Figure 2 Recanalization success was significantly more common with general anesthesia compared
with conscious sedation.
Figure 3 There was no significant difference between groups in good functional recovery.
Figure 4 There was no significant difference between groups in Mortality at 3 months.
Figure 5 There was no significant difference between groups in cerebral hemorrhage.
Sensitivity analysis and meta-regression
We performed a leave-one-out sensitivity analysis by systematically removing each
study from the pooled estimate. After the removal of each study, the results for cerebral
hemorrhage and three-month mortality were consistent. Recanalization success increased
in the I2 heterogeneity test after omitting Chabanne et al.,[7] I2 changing from 0% to 8%. Similarly, good functional recovery was sensitive to the
removal of Hendén et al.[14] and Maurice et al.,[3] with a change in I2 heterogeneity from 48% to 61% and 61%, respectively. By omitting Schönenberger,[15] the heterogeneity in the endpoint was eliminated (OR 0.96; 95% CI 0.72–1.28; I2 = 0%).
A comprehensive result of all sensitivity analyses conducted in the primary endpoints
is shown in [Supplementary Material Tables S2–S5].
Prespecified Meta-regression showed no significant interaction between recanalization
success and the mean NIHSS score. We could only analyze seven studies due to the availability
of NIHSS scores data; hence, this adaptation is restricted and impotent. The results
of this analysis are available in the [Supplementary Material Figure S2].
Quality and evidence assessment
Most included studies were judged to be at low risk of bias except for three studies,[3]
[13]
[16] which were judged as having “some concerns” in 3 separate domains: randomization
process, deviations from the intended interventions, and selection of reported results
([Supplementary Material Figure S3]).
On funnel plot analyses, the studies demonstrated a symmetrical distribution by their
weight and conversion toward the pooled effect as the weight increased; thus, there
was no definitive evidence of publication bias in the funnel plots. ([Supplementary Material Figure S4]). The overall GRADE assessment certainty was high ([Supplementary Material Figure S5]).
DISCUSSION
In this systematic review and meta-analysis of 8 studies and 1300 patients, we compared
GA with CS in the mechanical thrombectomy for AIS due to LVO. We found that recanalization
rates were higher in patients with GA (OR 1.68; 95% CI 1.26–2.24; p < 0.04), but there was no difference between groups regarding functional recovery,
three-month mortality, cerebral hemorrhage, and pneumonia.
A recent meta-analysis[6] also reported significantly higher recanalization rates in patients under GA; however,
this meta-analysis included only preliminary data from the trial by Liang et al.[11] and did not include a recently published RCT. Our meta-analysis aligns with previous
prospective studies that showed worse results in patients treated with CS. It is hypothesized
that GA's significant hazards rely on inadvertent hypotension during induction and
maintenance of GA; impaired cerebral autoregulation blood flow from cerebral ischemia
potentiated by these anesthetic drug effects may also play a role,[17] which could explain our results.
Although the GA group had a higher recanalization rate, there was no statistically
significant difference in functional recovery. Considering there was no difference
in pneumonia rates, this may be related to other factors, like NIHSS, age, and time
to reperfusion. The meta-regression analysis showed no interaction between the NIHSS
score and recanalization rate. There wasn't enough data on patients' ages, so it was
not possible to evaluate it in this meta-analysis.
To our knowledge, this is the first meta-analysis to evaluate pneumonia occurrence
associated with mechanical thrombectomy. Prospective studies[18] have shown a higher rate of pneumonia in patients with LVO during acute care; moreover,
the risk could further increase in cases where mechanical thrombectomy is performed,
and patients undergo intensive care unit (ICU) treatment. Pneumonia could result in
higher mortality, as well as a higher length of stay and hospitalization costs. However,
a prospective study[19] did not find statistically significant higher pneumonia rates in patients intubated
for mechanical thrombectomy. The 4 RCTs that disclosed pneumonia rates between GA
and CS had indistinguishable statistical differences between either anesthetic choice.
This could also be related to the absence of difference in functional recovery found
in our study.
Only one trial included in this meta-analysis assessed both anterior and posterior
circulation strokes, so it was not possible to evaluate for differences between GA
and CS in these groups.
Our study has some limitations. First, there were slight differences between groups
concerning the study population; however, there was no significant heterogeneity in
the analyses. We included a sensitivity analysis which showed similar results for
cerebral hemorrhage, three-month mortality, and recanalization success; there was
moderate heterogeneity in good functional recovery, which could be explained because
one study[3] evaluated the functional status in a more extended period (from 2 to 6 months of
the stroke). Funnel plots did not encounter signs of publication bias in this meta-analysis.
Second, our study does not apply to all mechanical thrombectomy patients since most
studies included only anterior circulation strokes, and only one study of this meta-analysis
included posterior circulation strokes. Third, protocols for GA and CS differed between
studies.
GA provides higher recanalization rates than CS. However, no difference is significant
regarding good functional outcome, three-month mortality, cerebral hemorrhage, and
pneumonia, suggesting that either approach can safely be chosen based on other patient
characteristics.
Bibliographical Record
Ana Clara Felix De Farias Santos, Luciano Lobão Salim Coelho, Guilherme de Carvalho
Caldas, Luziany Carvalho Araújo, Vivian Dias Baptista Gagliardi, Leonardo Augusto
Carbonera. General anesthesia versus conscious sedation in mechanical thrombectomy for patients with acute ischemic stroke:
systematic review and meta-analysis. Arq Neuropsiquiatr 2024; 82: s00441785693.
DOI: 10.1055/s-0044-1785693
