Does the Ischemic Core Really Matter? An Updated Systematic Review and Meta‐Analysis of Large Core Trials After TESLA, TENSION, and LASTE

Abstract

Background

The available evidence supporting the use of endovascular thrombectomy (EVT) in acute ischemic stroke patients with large core has increased with the recent release of the Thrombectomy for Emergent Salvage of Large Anterior Circulation Ischemic Stroke (TESLA), Efficacy and Safety of Thrombectomy In Stroke with Extended Lesion and Extended Time Window (TENSION), and Large Stroke Therapy Evaluation (LASTE) trials, providing critical information on additional subgroups not included in initial trials. We aimed to study the efficacy and safety of EVT in patients with acute ischemic stroke with large core and stratify by several subgroups including core infarct at presentation, using a comprehensive meta‐analysis of aggregate data.

Methods

We executed a systematic search to identify randomized controlled trials that compared EVT to medical management (MM) for the treatment of patients with acute ischemic stroke with large core, defined as Alberta Stroke Program Early CT [Computed Tomography] Score ≤5 on noncontrast CT and/or estimated ischemic core ≥50 mL on CT‐perfusion/MR diffusion. The primary outcome was the shift analysis in the 90‐day modified Rankin scale (mRS) score. Secondary outcomes included functional independence (mRS score 0–2), independent ambulation (mRS score 0–3), 90‐day mortality, and symptomatic intracranial hemorrhage. Pooled odds ratios were calculated for shift mRS score through the random‐effects meta‐analyses, and risk ratios (RRs) were used for the other outcomes, comparing EVT with MM alone.

Results

Out of 3402 titles and abstracts screened, 6 randomized controlled trials with 1886 patients were included. The EVT group had a higher shift toward a lower mRS than MM alone (odds ratio [OR], 1.49 [95% CI, 1.24–1.79]). Furthermore, the use of EVT was associated with higher rates of functional independence (19.5% versus 7.5%, RR, 2.49 [95% CI, 1.92–3.24]), independent ambulation (36.5% versus 19.9%, RR, 1.91 [95% CI, 1.51–2.43]), and symptomatic intracranial hemorrhage (5.5% versus 3.2%, RR, 1.73 [95% CI, 1.01–2.95]) compared with MM. There was no difference between the 2 groups regarding mortality (31.5% versus 36.8%, RR, 0.86 [95% CI, 0.72–1.02]). Importantly, EVT was consistently associated with a shift toward a lower mRS score in both Alberta Stroke Program Early CT Score 3–5 (OR, 1.60 [95% CI, 1.10–2.32]) and Alberta Stroke Program Early CT Score 0–2 (OR, 1.45 [95% CI, 1.17–1.80]) when compared with MM alone.

Conclusion

Our results confirm the efficacy of EVT for acute ischemic stroke with large core and suggest a consistent benefit across all Alberta Stroke Program Early CT Score categories. These results represent an important shift in the current large vessel occlusion selection paradigm that currently considers core as an effect modifier for EVT selection.

Nonstandard Abbreviations and Acronyms

AIS

acute ischemic stroke

ASPECTS

Alberta Stroke Program Early CT Score

EVT

endovascular thrombectomy

ICV

ischemic core volume

LASTE

Large Stroke Therapy Evaluation

LKW

last known well

LVO

large vessel occlusion

MM

medical management

mRS

modified Rankin scale

NCCT

noncontrast CT

NIHSS

National Institutes of Health Stroke Scale

sICH

symptomatic intracranial hemorrhage

TENSION

Efficacy and Safety of Thrombectomy In Stroke with Extended Lesion and Extended Time Window

TESLA

Thrombectomy for Emergent Salvage of Large Anterior Circulation Ischemic Stroke

Clinical Perspective

What Is New?

• This is the most updated systematic review and aggregate meta‐analysis that explores the efficacy and safety of endovascular thrombectomy for patients with acute ischemic stroke‐large vessel occlusion with large ischemic core.

• Endovascular thrombectomy was more effective than medical management alone for patients with acute ischemic stroke‐large vessel occlusion with large core.

• This benefit remained consistent despite a significant increase in the rate of symptomatic intracranial hemorrhage and is independent of age, clot location, time from last known well to randomization, National Institutes of Health Stroke Scale score, Alberta Stroke Program Early CT Score, ischemic core volume, cause of stroke, and thrombolysis prior to endovascular thrombectomy.

What Are the Clinical Implications?

• These results represent an important shift in the current large vessel occlusion selection paradigm that currently considers ischemic core as an effect modifier for endovascular thrombectomy selection.

Endovascular thrombectomy (EVT) is both effective and safe in treating selected patients with acute ischemic stroke (AIS) with anterior circulation large vessel occlusions (LVOs). ¹ , ² , ³ , ⁴ , ⁵ The current selection criteria for EVT are composed of various factors, including the time since symptoms started, the severity of symptoms, the individual's functional status before the event, extent of blood flow blockage, and stroke infarct core at time of presentation. The latest guidelines from the American Heart Association and the European Stroke Organization recommend using EVT for the treatment of patients with AIS‐LVO whose Alberta Stroke Program Early CT [Computed Tomography] Score (ASPECTS) is 6 or above, provided that the treatment is administered within a 24‐hour time window. ⁶ , ⁷ This led to lack of indication based on the guidelines statement regarding the benefit of EVT for patients with a large ischemic core, who currently comprise ∼20% to 25% of the population of patients with AIS‐LVO. ⁸

Three initial large core randomized controlled trials (RCTs) and meta‐analyses of those RCTs demonstrated EVT benefit in patients with large ischemic core that were selected based on noncontrast CT (NCCT), CT perfusion, or diffusion‐perfusion magnetic resonance imaging (MRI). ⁹ , ¹⁰ , ¹¹ , ¹² , ¹³ Recently, the Thrombectomy for Emergent Salvage of Large Anterior Circulation Ischaemic Stroke (TESLA) study enrolled patients with AIS‐LVO with large infarct core (ASPECTS 2–5) that were selected by NCCT up to 24 hours after last known well (LKW) and showed no significant difference in functional outcomes between EVT and the medical management (MM) alone groups. ¹⁴ However, the results from the efficacy and safety of the Thrombectomy In Stroke with Extended Lesion and Extended Time Window (TENSION) study up to 12 hours showed significant improvement in functional outcomes and mortality in the EVT group compared with those who received MM alone. ¹⁵ Finally, the Large Stroke Therapy Evaluation (LASTE) trial, which enrolled more than half of the patients with ASPECTS 0–2 selected by MRI or NCCT, showed additional efficacy and safety of EVT for patients with AIS‐LVO with large and extra‐large infarct core. ¹⁶

In light of the results from the TESLA, TENSION, and LASTE trials, we aimed to perform an updated systematic review and aggregate meta‐analysis to better evaluate the overall efficacy and safety of EVT for patients with AIS‐LVO with large infarct core. We further sought to evaluate the benefits and risks of EVT in important subpopulations based on clinical and radiological factors with an emphasis on ASPECTS at presentation.

Methods

Data are available upon reasonable request to the corresponding author. This systematic review and meta‐analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta‐analysis guidelines. ¹⁷ The study is registered with International Prospective Register of Systematic Reviews., ID code: CRD42023399752.

Search Strategy and Studies Selection

We performed a systematic electronic literature search of PubMed, Embase, Scopus, and Google Scholar through October 2023 (Table S1). Titles and abstracts were retrieved and reviewed for eligibility by 3 independent reviewers (M.A., M.D., and M.G.) using an online application for systematic reviews (https://rayyan.qcri.org/). To identify further studies, we screened the reference list of the relevant records. The reports that were included in the initial screening were further assessed for eligibility in full text. Three independent reviewers (M.A., M.G., and M.D.) then extracted data from the finally included studies using a predesigned standardized data collection sheet. Any disagreements were resolved through discussion with the senior reviewer (S.O.G.). No language restriction was applied.

Inclusion and Exclusion Criteria

This study included all RCTs that enrolled patients with AIS‐LVO with anterior circulation large infarct core and were randomized to EVT and MM or MM alone, within 24 hours of the time the patient was LKW. Large infarct core was defined as ASPECTS of 0–5 and/or ischemic core volumes (ICVs) ≥50 mL prior to intervention. All non‐RCTs and observational studies were excluded.

Baseline Data and Variables of Interest

We extracted the following data, if available: authors, year of publication, affiliated institutions, type of study, and patient demographics and characteristics, including the number of participants, age, sex, comorbidities, smoking status, ischemic events, stroke workflow metrics, National Institutes of Health Stroke Scale (NIHSS) score, medications received during an intervention, and the type of intervention (EVT versus MM alone).

Imaging biomarkers were evaluated by the imaging core labs for individual trials. Baseline neuroimaging findings, including CT/MRI ASPECTS, occlusion location, and neuroimaging findings including follow‐up infarct volume, hemorrhagic transformation, and grade were collected for all patients.

Intervention

EVT was performed using a stent retriever, aspiration catheter, a combination of both, or direct angioplasty. Standard medical care, including thrombolysis when indicated, antiplatelets or anticoagulation as indicated, blood pressure management, and monitoring and treatment of cerebral edema, as determined by each individual trial, was also provided to all patients.

Primary and Secondary Outcomes

The primary outcome of this study was the shift analysis in the 90‐day modified Rankin scale (mRS) score as reported by a decrease in ≥1‐point in mRS grade. The mRS score is an ordinal scale that reflects the degree of disability after an AIS (0, no symptoms; 6, death). Secondary outcomes included functional independence (mRS 0–2) and independent ambulation (mRS 0–3), 90‐day all‐cause mortality, and symptomatic intracranial hemorrhage (sICH). sICH was defined according to the criteria by each enrolled RCT (Table S2). Moreover, the data related to decompressive‐hemicraniectomy in both groups and the modified treatment in cerebral infarction reperfusion grade of 2b or higher (indicating antegrade reperfusion of more than half of the previously occluded target‐artery ischemic territory) for the EVT group were also extracted.

Risk of Bias Assessment

Two independent reviewers (M.A., M.D.) evaluated the quality of the studies using the revised Cochrane risk‐of‐bias tool for randomized trials. ¹⁸ The tool assesses potential risk of bias in different components of clinical trial design and execution. Any disagreements between the authors were solved by discussion and consensus. The risk of bias was not assessed for the LASTE trial, as the full report has not yet been published. ¹⁶

Statistical Analysis

For the primary outcome, we obtained the effect estimate and standard error from the generalized odds ratio (OR) reported in each included study and calculated the pooled OR and 95% CIs. For secondary outcomes, we extracted the raw data for both groups and then calculated the risk ratio (RR) and the 95% CI. For modified thrombolysis in cerebral infarction 2b–3, we pooled rates and obtained a weighted overall proportion with the 95% CI using generalized linear mixed models. We used a random‐effects model and calculated prediction intervals for all the analyses. We performed subgroup analyses based on age, clot location, time from LKW to randomization, NIHSS score, ASPECTS, ICV, cause of stroke, and receipt of thrombolysis prior to EVT by extracting the effect estimates and standard errors from the generalized OR from each study. We used the I² statistic to measure percentage of variability due to the between‐study heterogeneity. Values of 25%, 50%, and 75% were considered low, moderate, and high risk of heterogeneity, respectively. ¹⁹ Publication bias was not studied due to having <10 studies in each outcome. All statistical analyses and graphs included in this study were performed using the R Statistical Software (version 4.1.3) and R Studio.

Certainty of the Evidence

Following Cochrane recommendations, 2 reviewers (M.D., M.G.C.) assessed the quality of the body of evidence using the grading of recommendation, assessment, development, and evaluation approach and online tool. ²⁰ For each outcome, we downgraded the evidence for serious study limitations in risk of bias, indirectness of evidence, inconsistency, imprecision of effect estimates, publication bias, and analyses in which only pooled proportions were calculated.

Results

Study Selection

Our search retrieved 3402 titles and abstracts for initial screening (Figure 1). In the full‐text evaluation, we excluded 32 documents and included 5 RCTs. The LASTE trial data were presented at SLICE Worldwide 2023 and were included in this study. ¹⁶

Figure 1.

Flow diagram. RCT indicates randomized controlled trial.

Study Characteristics

Six RCTs with 1886 participants were included. The patients’ baseline characteristics of the included studies are found in Table 1. The median age ranged in the EVT and MM alone groups from 66 to 73 and 67 to 74 years, respectively. The EVT group comprised 514 males (54.4%), and the MM group contained 474 (50.3%) males. The median ASPECTS ranged between 3 and 4. The LASTE trial included the largest group of patients who had ASPECTS 0–2 with 86 (54%) in the EVT group, compared with 95 (58%) in the MM. Intravenous thrombolysis was given in 265 (28%) patients in the EVT group, compared with 254 (27.0%) in the MM group. Imaging modalities and key outcomes are summarized in Table 2.

Table 1.

Characteristics of the Included Studies

Trial	Arms	No.	Age (years)	Sex (male)	ASPECTS	Admission NIHSS	Intravenous thrombolysis	Occlusion location
								ICA	MCA M1	MCA M2
SELECT2 10	EVT	178	66 (58–75)	107 (60.1)	4 (3–5)	19 (15–23)	37 (20.8)	80 (44.9)	91 (51.1)	7 (3.9)
	MM	174	67 (58–75)	100 (57.5)	4 (4–5)	19 (15–22)	30 (17.3)	66 (37.9)	100 (57.5)	8 (4.6)
ANGEL‐ASPECT 9	EVT	230	68 (61–73)	135 (58.7)	3 (3–4)	16 (13–20)	66 (28.7)	83 (36.1)	145 (63.0)	2 (0.9)
	MM	225	67 (59–73)	144 (64.0)	3 (3–4)	15 (12–19)	63 (28.0)	81 (36.0)	142 (63.1)	2 (0.9)
RESCUE‐Japan LIMIT 11	EVT	100	76.6 ±10.0	55 (54.5)	3 (3–4)	22 (18–26)	27 (26.7)	47 (46.5)	74 (73.3)	0 (0)
	MM	102	75.7 ±10.2	58 (56.9)	4 (3–4)	22 (17–26)	29 (28.4)	49 (48.0)	70 (68.6)	3 (2.9)
TESLA 14	EVT	152	66 (54–74)	76 (50.0)	4 (3–5)	19 (15–23)	31 (20.4)	46 (31.5)	97 (66.4)	3 (2.1)
	MM	148	67.5 (57.5–73.5)	84 (56.8)	4 (3–5)	18 (14.5–21)	30 (20.3)	41 (28.3)	94 (64.8)	10 (6.9)
TENSION 15	EVT	125	73 (65–81)	59 (55)	0–2: 15 (12) 3–5: 103 (82) 6–10: 7 (6)	19 (16–22)	49 (39)	41 (33)	83 (66)	0 (0)
	MM	128	74 (64–80)	51 (48)	0–2: 23 (18) 3–5: 99 (78) 6–10: 5 (4)	18 (15–22)	44 (34)	37 (29)	88 (69)	1 (1)
LASTE 16	EVT	159	73 (66–79)	82 (51.6)	0–2: 86 (54) 3–5: 73 (46)	21 (18–24)	55 (34.6)	69 (43.4)	84 (52.8)	4 (2.5)
	MM	165	74 (65–80)	88 (53.3)	0–2: 95 (58) 3–5: 70 (42)	21 (18–24)	58 (35.2)	74 (44.8)	89 (53.9)	2 (1.2)

Categorical variables are described as n (%), and continuous variables are described as median (interquartile range) or mean±SD. ANGEL‐ASPECT indicates Study of Endovascular Therapy in Acute Anterior Circulation Large Vessel Occlusive Patients With a Large Infarct Core; ASPECTS,Alberta Stroke Program Early Computed Tomographic Score; EVT, endovascular thrombectomy; ICA, internal carotid artery; LASTE, Large Stroke Therapy Evaluation; MCA, middle cerebral artery; MM, medical management; NIHSS, National Institutes of Health Stroke Scale; RESCUE‐Japan LIMIT, Recovery by Endovascular Salvage for Cerebral Ultra‐Acute Embolism–Japan Large Ischemic Core; SELECT2, Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke; TENSION, Efficacy and Safety of Thrombectomy In Stroke with Extended Lesion and Extended Time Window; and TESLA, Thrombectomy for Emergent Salvage of Large Anterior Circulation Ischemic Stroke.

Table 2.

Imaging Modality and Key Outcomes from the Included Trials

	SELECT2 10	ANGEL‐ASPECT 9	RESCUE‐Japan LIMIT 11	TESLA 14	TENSION 15	LASTE 16
Imaging inclusion criteria	Presented ≤24 h ASPECTS 3–5 on NCCT or ICV ≥50 mL on CT perfusion or DWI	Presented ≤24 h ASPECTS 3–5 on NCCT with no limitation to DWI ICV ASPECTS 0–2 during 24 hours and ICV 70–100 or ASPECTS ≥5 in 6–24 h and ICV 70–100	Presented ≤24 h ASPECTS 3–5 on NCCT/DWI <6 h, if 6–24 has negative FLAIR	Presented ≤24 h ASPECTS 2–5 on NCCT	Presented ≤12 h ASPECTS 3–5 on NCCT	Presented ≤7 h ASPECTS 0–5 on NCCT or DWI and negative FLAIR
90‐d shift mRS, OR (95% CI)	1.51 (1.20–1.81)	1.37 (1.11–1.69)	2.42 (1.46–4.01)	1.40 (0.91–2.16)	2.58 (1.60–4.15)	1.63 (1.29–2.06)
90‐d functional independence, RR (95% CI)	2.97 (1.60–5.57)	2.62 (1.69–4.06)	1.79 (0.87–4.07)	1.64 (0.86–3.12)	7.16 (2.12–24.21)	2.39 (1.18–6.22)
90‐d independent ambulation, RR (95% CI)	2.97 (1.60–5.51)	1.50 (1.17–1.91)	2.43 (1.35–4.37)	1.50 (1.00–2.26)	2.84 (1.48–5.47)	2.62 (1.72–4.36)
sICH, RR (95% CI)	0.49 (0.04–5.36) *	2.07 (0.79–5.41) †	1.84 (0.64–5.29) †	2.96 (0.61–14.43) *	5% versus 5% *	1.73 (0.78–4.68)
90‐d mortality, RR (95% CI)	0.91 (0.71–1.18)	1.00 (0.65–1.54) ‡	0.77 (0.44–1.32)	1.06 (0.77–1.45)	0.67 (0.46–0.98)	0.65 (0.50–0.84)

ANGEL‐ASPECT indicates Study of Endovascular Therapy in Acute Anterior Circulation Large Vessel Occlusive Patients With a Large Infarct Core; ASPECTS, Alberta Stroke Program Early CT Score; DWI, diffusion‐weighted imaging; FLAIR, fluid attenuated inversion recovery; ICV, ischemic core volume; LASTE, Large Stroke Therapy Evaluation; mRS, modified Rankin scale; NCCT, non‐contrast computed tomography; OR, odds ratio; RESCUE‐Japan LIMIT, Recovery by Endovascular Salvage for Cerebral Ultra‐Acute Embolism–Japan Large Ischemic Core; RR, risk ratio; SELECT2, Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke; sICH, symptomatic intracerebral hemorrhage; TENSION, Efficacy and Safety of Thrombectomy In Stroke with Extended Lesion and Extended Time Window; and TESLA, Thrombectomy for Emergent Salvage of Large Anterior Circulation Ischemic Stroke.

^*Within 24 hours.

^†Within 48 hours.

^‡Hazard ratio.

Risk of Bias and Quality Assessment

All studies raised some concerns with regard to the overall risk of bias using the Cochrane risk‐of‐bias tool 2 (Figure S1). TESLA ¹⁴ and TENSION ¹⁵ experienced minor concerns in the randomization process due to imbalance of rates of diabetes and atrial fibrillation, respectively. Moreover, blinding to the treatment arms was not done in any of the included studies. Additionally, there were a few missing data points reported in some of the outcomes (ie, due to patients being lost to follow‐up) in SELECT2 (Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke), ¹⁰ TESLA, ¹⁴ and TENSION. ¹⁵ The LASTE trial is not yet published and was only presented in an international conference ¹⁶ ; therefore, the risk of bias was not assessed.

Certainty of evidence for the shift in mRS score, functional independence and independent ambulation, all‐cause mortality at 90 days, decompressive hemicraniectomy, and successful recanalization (modified thrombolysis in cerebral infarction 2b–3) was high, whereas certainty of evidence for sICH was moderate due to the differences in results between the trials (Table S3).

Primary and Secondary Outcome For EVT Versus MM Alone

Six studies were pooled for the primary outcome (I² = 0%; P = 0.89). The use of EVT was significantly associated with higher odds of shift toward a lower mRS scorecompared with MM alone (OR, 1.49, [95% CI, 1.24–1.79], Figure 2A). Six studies reported rates related to functional independence (I² = 8%; P = 0.37) and independent ambulation (I² = 54%; P = 0.06). The use of EVT was significantly associated with increased rates of functional independence compared with MM alone (19.5% versus 7.5%, RR, 2.49 [95% CI, 1.92–3.24], Figure 2B). Moreover, the rates of independent ambulation were higher in patients who underwent EVT versus MM alone (36.5% versus 19.9%, RR, 1.91 [95% CI, 1.51–2.43], Figure 2C).

Figure 2.

Forest plot for the efficacy of endovascular thrombectomy versus medical management alone. A, 90‐day mRS shift analysis, (B) mRS 0–2, and (C) mRS 0−3. ANGEL‐ASPECT indicates Study of Endovascular Therapy in Acute Anterior Circulation Large Vessel Occlusive Patients With a Large Infarct Core; EVT, endovascular thrombectomy; LASTE, Large Stroke Therapy Evaluation; MH, Mantel–Haenszel; MM, medical management; OR,odds ratio; RESCUE‐Japan LIMIT, Recovery by Endovascular Salvage for Cerebral Ultra‐Acute Embolism–Japan Large Ischemic Core; RR, risk ratio; SELECT2, Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke; TENSION, Efficacy and Safety of Thrombectomy In Stroke with Extended Lesion and Extended Time Window; and TESLA, Thrombectomy for Emergent Salvage of Large Anterior Circulation Ischemic Stroke.

A meta‐analysis of 6 studies with the random effect model (I² = 0%; P = 0.78) showed an increase in risk of sICH in the EVT group when compared with patients who received MM alone (5.5% versus 3.2%, RR, 1.71 [95% CI, 1.09–2.66], Figure 3A). Additionally, there was no significant difference in mortality in patients who underwent EVT compared to those who received MM alone (31.5% versus 36.8%, RR, 0.86 [95% CI, 0.72–1.02], Figure 3B). There was no significant difference in decompressive hemicraniectomy in patients who underwent EVT compared to those who received MM alone (12.1% versus 10.4%, OR, 1.15 [95% CI, 0.87–1.51], Figure 3C). In the EVT group, the pooled rate of successful recanalization was 82% (95% CI, 78%–85%, Figure S2).

Figure 3.

Forest plot for the safety of endovascular thrombectomy versus medical management alone. A, Symptomatic intracranial hemorrhage (B), 90‐day mortality, and (C) decompressive hemicraniectomy. ANGEL‐ASPECT indicates Study of Endovascular Therapy in Acute Anterior Circulation Large Vessel Occlusive Patients With a Large Infarct Core; EVT, endovascular thrombectomy; LASTE, Large Stroke Therapy Evaluation; MH, Mantel–Haenszel; MM, medical management; OR, odds ratio; RESCUE‐Japan LIMIT, Recovery by Endovascular Salvage for Cerebral Ultra‐Acute Embolism–Japan Large Ischemic Core; RR, risk ratio; SELECT2, Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke; TENSION, Efficacy and Safety of Thrombectomy In Stroke with Extended Lesion and Extended Time Window; and TESLA, Thrombectomy for Emergent Salvage of Large Anterior Circulation Ischemic Stroke.

Subgroup Analyses and ASPECTS Stratification

EVT treatment effect was largely persistent across various subgroups based on age, clot location, time from LKW to randomization, NIHSS score, ASPECTS, ICV, cause of stroke, and thrombolysis prior to EVT when considering ordinal distribution of mRS scores (Figure 4 and Figures S3–S9). Most important, the effect of EVT was consistently associated with a shift toward a lower mRS score in both ASPECTS 3–5 (OR, 1.60 [95% CI, 1.10–2.32]) and ASPECTS 0–2 (OR, 1.45 [95% CI, 1.17–1.80]) when compared with MM alone (Figure 5).

Figure 4.

Subgroup analyses based on the primary efficacy outcome. ACA indicates anterior cerebral artery; aOR, adjusted odds ratio; ASPECTS, Alberta Stroke Program Early CT [Computed Tomography] Score; EVT, endovascular thrombectomy; ICA, internal carotid artery; LAA, left atrial appendage; MCA, middle cerebral artery; MM, medical management; mRS, modified Rankin Scale; and NIHSS, National Institutes of Health Stroke Scale.

Figure 5.

Forest plot for EVT versus medical management for shift mRS analysis based on ASPECTS. ANGEL‐ASPECT indicates Study of Endovascular Therapy in Acute Anterior Circulation Large Vessel Occlusive Patients With a Large Infarct Core; ASPECTS, Alberta Stroke Program Early CT [Computed Tomography] Score; EVT, endovascular thrombectomy; IV, instrumental variable; LASTE, Large Stroke Therapy Evaluation; MM, medical management; mRS, modified Rankin Scale; OR, odds ratio; SELECT2, Randomized Controlled Trial to Optimize Patient's Selection for Endovascular Treatment in Acute Ischemic Stroke; and TENSION, Efficacy and Safety of Thrombectomy In Stroke with Extended Lesion and Extended Time Window.

Discussion

In this updated systematic review and meta‐analysis of 6 RCTs, we found that patients with AIS‐LVO with large infarct core treated with EVT achieved better functional outcomes than those who had MM alone. This effect persisted across different subgroups independent of age, clot location, time from LKW to randomization, NIHSS score, ASPECTS, ICV, cause of stroke, and thrombolysis prior to EVT. Most important, in patients with an ultra‐large ischemic core (ASPECTS 0–2), the effect of EVT was consistently associated with a shift toward a lower mRS score compared with MM alone. The overall rates of sICH were higher in patients treated with EVT, and we did not find differences in mortality at 90 days.

Six RCTs have provided critical evidence regarding EVT in patients with a large core AIS‐LVO. ⁹ , ¹⁰ , ¹¹ , ¹⁴ , ¹⁵ , ¹⁶ All of these RCTs except for the TESLA trial showed that the use of EVT was significantly associated with higher odds of shift toward a lower mRS score ¹⁴ and increased rates of functional independence and independent ambulation compared with MM alone. The TESLA trial, on the other hand, showed a nonsignificant trend in functional outcomes between EVT and MM alone. ¹⁴ We hypothesize that the longer time from the onset to treatment randomization in this trial, when compared with the others, using exclusively an initial NCCT for selection of patients, might have introduced some degree of group heterogeneity at the time of randomization. In fact, the core laboratory agreed with the site investigators adjudicated ASPECTS in 85% of the cases. ²¹ Nevertheless, after pooling the 6 studies together, we found significantly higher rates of shift toward a lower mRS score, functional independence, and independent ambulation in the EVT group compared with MM alone. The results from this systematic review and meta‐analysis are in congruence with the prior systematic reviews that showed improved shift in mRS score, ²² functional independence, and independent ambulation ¹³ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ in the EVT group compared with MM alone. This evidence supports the use of EVT for patients with AIS‐LVO with large ischemic core. It also prompts the inquiry into investigating this population in future stroke trials including neuroprotection, where the objective would expand to explore the effect of drugs or interventions with mechanisms beyond mere penumbra preservation, aiming to enhance outcomes and mitigate reperfusion injuries in these patients.

Most important, this meta‐analysis has the largest pooled evidence from 6 trials, including a large cohort of 264 patients with ASPECTS 0–2. Our results suggest a consistent benefit of EVT across all low ASPECTS categories including the ultra‐large core group of ASPECTS 0–2. These findings do not align with a post hoc analysis from the RESCUE‐Japan LIMIT (Recovery by Endovascular Salvage for Cerebral Ultra‐Acute Embolism–Japan Large Ischemic Core) trial, which showed no benefit of EVT in patients with baseline ASPECTS ≤3 using MRI. ²⁷ However, it is important to recognize that the number of ASPECTS 0–2 is underrepresented with barely 4% of that total cohort. In fact, the greater statistical power due to the inclusion of more participants with ASPECTS 0–2 in the LASTE ¹⁶ trial did show significant benefit in functional outcome in favor of the EVT group. These results challenge the traditional selection paradigm focusing on “adequate” ischemic core, ASPECTS ≥6 during the early window (<6 hours after LKW) on NCCT and volume <70 cc using advanced imaging (CT perfusion or MRI) after 6 hours of LKW, for pursuing an EVT decision. It rather suggests that a simpler and more cost‐effective approach of using NCCT to rule out hemorrhage for all patients within the first 24 hours is sufficient when LVO is suspected. ²⁸ Notably, the NIHSS demonstrates high predictability for anterior circulation LVO, particularly with NIHSS scores ≥9. ²⁹ Consequently, the combined use of NIHSS and NCCT might be sufficient for guiding EVT decisions, especially in low‐ and middle‐income countries where more advanced imaging or CT angiography might not be readily available. Additionally, our subgroup analysis also revealed robust results favoring EVT over MM alone independent of age, clot location, NIHSS score, ICV, cause of stroke, and thrombolysis prior to EVT. The time elapsed since stroke onset also did not modify treatment effect, and absolute rates of favorable functional outcomes were not reduced in patients treated later. These findings supplement late window trials that included only patients with limited ICV and evidence of salvageable penumbra and it will expand EVT eligibility to a much larger group of patients presenting beyond 6 hours of stroke onset. As a result, this suggests that the patient selection for EVT should continue to be more inclusive and simpler. Furthermore, there is a need for a shift in our current selection concept toward now treating the large majority of patients with AIS‐LVO with large core while also studying the subgroups that might not benefit.

Regarding safety, we observed differences in the sICH rates among the included studies (Table 2). The overall rates of sICH were found to be higher in the EVT group compared with the MM alone group. A possible explanation is that following reperfusion, there is an increased risk of edema and hemorrhagic transformation. Alternatively, it could also be explained that, as occurs in small infarcts, patients with large infarct and non‐successful reperfusion could exhibit increased rates of sICH. Notably, sICH was defined differently in each study, as ANGEL‐ASPECT (Study of Endovascular Therapy in Acute Anterior Circulation Large Vessel Occlusive Patients With a Large Infarct Core), ⁹ TENSION, ¹⁵ and LASTE ¹⁶ used the Heidelberg bleeding classification, ³⁰ whereas the SELECT2, ¹⁰ RESCUE‐Japan LIMIT, ¹¹ and TESLA ¹⁴ used the Safe Implementation of Thrombolysis in Stroke‐Monitoring Study definitions. ³¹ Furthermore, various time cutoffs were used for sICH characterization between studies. SELECT2, ¹⁰ TESLA, ¹⁴ and TENSION ¹⁵ included sICH that occurred within 24 hours, whereas RESCUE‐Japan LIMIT ⁹ and ANGEL‐ASPECT ¹¹ used 48 hours as a cutoff. Despite variability among the trials, no heterogeneity (I² = 0%) in sICH was observed. More important, the overall estimated rate of sICH in the EVT group was 5.5%, which is consistent with the rates observed in the high ASPECTS EVT trials. Moreover, mortality was not different among the 2 groups, which supports the safety of using EVT for patients with AIS‐LVO with large infarct core.

Limitations

This is the largest meta‐analysis that included data about using EVT for patients with AIS‐LVO with large infarct core. However, readers must be cautioned about the differences in the enrolled RCTs regarding the imaging selection criteria. All studies raised some concerns with regard to overall risk of bias due to minor issues with randomization, blinding, and missing data in some outcomes. ¹⁶ We were unable to ascertain the exact cause of the increased sICH rates in the EVT group due to the scarcity of data that compare in this patient population the sICH rates between patients who had successful reperfusion and those with nonsuccessful reperfusion. Future patient‐level meta‐analyses are warranted to explore this further. Another limitation is the use of unpublished data from the TESLA ¹⁴ and LASTE ¹⁶ trials, as it is possible there may be slight differences compared with the forthcoming peer‐reviewed publications and it might result in overestimation ³² of the effect. Thus, more caution is warranted in interpreting these results. Finally, although the benefit of EVT on functional outcome seems to be consistent independent of ASPECTS on aggregate data, it is important to mention that most of the patients with ASPECTS 0–2 in our analysis were part of the LASTE trial, which was the only trial that included patients with ASPECTS 0–1 in its inclusion criteria. Additionally, we cannot further evaluate the effect of EVT on the different ultra‐large subgroups. Further granularity evaluation among the different ultra‐large subgroup categories (0, 1, and 2) should be confirmed by a patient‐level pooled meta‐analysis. Given that all included RCTs enrolled patients who demonstrated functional independence at baseline, we acknowledge our limitation in ascertaining the impact of EVT in patients with large core anterior circulation LVOs and poor baseline functional status (mRS>2). This limitation extends to very old patients, particularly those beyond the age of 85 to 90 years, who were notably underrepresented in existing studies, as well as to individuals with posterior circulation large core LVOs. These specific subgroups represent critical areas where future research is necessary to explore the potential benefits of EVT more comprehensively. In the interim, it is imperative that decisions regarding EVT in these contexts be made judiciously on a case‐to‐case basis, incorporating the opinions of the patient's family and any previously expressed wishes of the patients themselves when available.

Conclusion

Our results confirm the significant benefit of EVT on functional outcome compared with MM alone for patients with AIS‐LVO with large ischemic core. This benefit remains consistent despite a significant increase in the rate of sICH and is independent of age, clot location, time from LKW to randomization, NIHSS score, ASPECTS, ICV, cause of stroke, and thrombolysis prior to EVT. These results represent an important shift in the current LVO selection paradigm that currently considers core as an effect modifier for EVT selection.

Sources of Funding

This research received no specific grant from any funding agency in public, commercial or not‐for‐profit sectors.

Disclosure

Albert Yoo: Grants: Medtronic, Cerenovus, Penumbra, Stryker, and Genentech. Consulting fees: Penumbra, Vesalio, Cerenovus, Philips Neurovascular, HCA, and the National Institutes of Health. Stock options from Nicolab; equity interest in Insera Therapeutics and Gravity Medical. Edgar Samaniego: Consulting fees: Rapid Medical, Microvention and Medtronics. Tudor Jovin: Grants: Medtronic USA, Inc, Stryker. Data and Safety monitoring: Johnson&Johnson Cerenovus, Stock: Anaconda, Freeox Biotech, Galaxy, Methinks, Route92, and Viz.ai. Consultant: Contego Medical Inc. Amrou Sarraj: Grants: Stryker Neurovascular. Consulting fees: AstraZeneca. Santiago Ortega‐Gutierrez: Grants: NIH‐NINDS (R01NS127114‐01, RO3NS126804‐01), Stryker, Medtronic, Microvention, Methinks, Viz.ai. Consulting fees: Medtronic, Stryker Neurovascular.

Tudor Jovin and Santiago Ortega‐Gutierrez serves on the Editorial Board of S:VIN. Editorial Board Members are not involved in the handling or final disposition of submissions.

Albert Yoo is an Associate Editor for S:VIN and was not involved in the handling or final disposition of this article. Disclosures provided by in compliance with American Heart Association's annual Journal Editor Disclosure Questionnaire are available at https://www.ahajournals.org/editor-coi-disclosures

Supporting information

Supplementary Figures 1–9

Supplementary Tables 1–3

References 9–11, 14–16, 30–31