Mechanical thrombectomy versus medical care alone in large ischemic core: An up-to-date meta-analysis

Abstract

Objective

We compared outcomes and adverse events of thrombectomy versus medical management in acute ischemic stroke (AIS) patients with baseline large infarct core.

Methods

We searched Ovid MEDLINE(R) ALL, Cochrane Library Clinical Controlled Trials and EMBASE from inception to January 2021 for studies comparing thrombectomy and medical management alone in AIS patients who had ASPECTS <=7 or ischemic core volume >=50 ml. Imaging modalities to valuate ASPECTS and core volume were without restriction. The functional outcome was measured by mRS (modified Rankin Scale) score 0-2 at 90 days or discharge. The safety end point included the rates of mortality and sICH (symptomatic intracranial hemorrhage) or PH2 (parenchymal hematoma type 2).

Results

Fourteen studies with a total of 2547 patients (thrombectomy n = [1197]; medical care alone [n = 1350]) fulfilled our criteria. As for patients with low ASPECTS, pooled results indicated a higher odds of good functional outcome (OR = 3.47; 95% CI 1.99 to 6.07; P < 0.0001, I²=66%) and a lower risk of mortality (OR = 0.62; 95% CI 0.46 to 0.83; P = 0.001, I²=32%) in thrombectomy group compared with no thrombectomy group, but the risk of sICH or PH2 did not differ between two groups. As for patients with large core volume, both functional outcome and safety end point between two groups showed no statistically significant difference.

Conclusion

Thrombectomy remained safe and effective by careful selection in patients with low ASPECTS. More studies were warranted to explore contraindications for mechanical thrombectomy in AIS patients, especially in patients with large core volume.

Introduction

Using rigorous selection criteria, patients with large vessel occlusion clearly benefit from mechanical thrombectomy. ¹ In recent guidelines,^2,3 mechanical thrombectomy has been strongly recommended for patients who had small core sizes, such as ASPECTS (Alberta Stroke Program Early Computed Tomography Score) >=6 and ischemic core volume of lower than 70 cm³ (level I, class A evidence). However, the benefits of mechanical thrombectomy in patients with large baseline infarcts are still controversial. Large infarct core is a common exclusion criterion for mechanical thrombectomy in most randomized controlled trials. For instance, the EXTEND-IA (Endovascular Therapy for Ischemic Stroke with Perfusion-Imaging Selection) trial ⁴ ruled out the patients involving core infarct size of more than 70 ml. Likewise, patients with initial ASPECTS lower than 6 were ineligible for SWIFT-PRIME (Stent-Retriever Thrombectomy after Intravenous t-PA vs. t-PA Alone in Stroke) trial. ⁵ Nevertheless, several multicenter studies have indicated different results. The HERMES (Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke trials) collaboration^6,7 indicated that ischemic core volume was independently associated with functional improvement but did not modify the treatment benefit of endovascular thrombectomy over standard medical therapy for improved functional outcome. Moreover, the SELECT (Selection for Endovascular Treatment in Acute Ischemic Stroke) trial ⁸ suggested potential benefits in patients with large cores who received thrombectomy, though this conclusion was more hypothesis generating than definitive because it was just a cohort study of 105 patients.

This raised the question, regarding to mechanical thrombectomy versus medical care alone in patients with large core infarct, which one is better? Although there has been some quantitative synthesis touched on this subject, they did not make a comparison with the no thrombectomy group^9,10 or had a small sample size. ¹¹ As a consequence, the objective of this meta-analysis was to investigate that in comparison to medical treatment alone, whether thrombectomy obtained better functional outcome and safety in large ischemic core patients by both semi-quantitative (ASPECTS 0-7,0-5,5-7) and quantitative (core volume >=50 or 70 ml) imaging modalities.

Methods

Databases and sources

This meta-analysis was firstly registered at the International Prospective Register of Systematic Reviews (CRD42020180819) and conducted in adherence with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) Statement. ¹² We searched Ovid MEDLINE(R) ALL, Cochrane Library Clinical Controlled Trials and EMBASE from inception to January 2021 for studies comparing thrombectomy and medical management alone in patients with ASPECTS <=7 or ischemic core volume >=50 ml without language restriction. ASPECTS could be measured by NCCT (Non-contrast Computed Tomography), CTA (Computed Tomography Angiography), DWI (Diffusion Weighted Imaging) or any other modalities. Similarly, both MRP (Magnetic Resonance Perfusion) and CTP (Computed Tomography Perfusion) could be used to evaluate ischemic core volume. The predefined search strategy terms were “thrombectomy” OR “endovascular” OR “interventional” OR “intra-arterial” AND “ischemic core” OR “large infarct*” OR “low ASPECTS” OR “large baseline infarct*” OR “large core*” OR “core volume” OR “Diffusion-Weighted Imaging Stroke Lesions” OR “Imaging Lesions” OR “Alberta Stroke Program Early Computed Tomography Score” OR “massive cerebral infarction” OR “large hemispheric infarction” OR “malignant middle cerebral artery infarction”. In addition, we also manually searched the bibliographies of the relevant studies to identify additional studies that were not included in the initial literature searches.

Study selection

Two investigators (QMJ and HSW) independently reviewed each entry. Discrepancies were resolved by consensus or with the help of the senior investigator (GDX). Matched randomized controlled trials, cohort studies and case‐control studies were included. Case reports and reviews were excluded. The identified studies were then subjected to the following inclusion criteria: (1) Studies selecting patients with acute ischemic stroke who had ASPECTS <=7 or ischemic core volume >=50 ml. Imaging modalities to valuate ASPECTS or core volume, such as DWI and CTP, were without restriction; (2) The studies compared the outcomes of thrombectomy group with no thrombectomy group; (3) Recorded baseline characteristics information, including study design, first author, study locations, mean age and proportion of male; (4) Sufficient outcomes of mRS (modified Rankin Scale) scores at discharge or 90 days, the rates of mortality, as well as sICH (symptomatic intracranial hemorrhage) or PH2 (parenchymal hematoma type 2).

Data extraction

Two investigators (QMJ and JG) independently extracted all baseline data and main outcomes from each eligible study: 1. General data: name of author, year of publication, patient demographics, baseline NIHSS score, proportion of patients who treatment with intravenous recombinant tissue plasminogen activator, proportion of patients who have successful recanalization. 2. Main outcomes: Numbers of patients with good functional outcome defined as mRS 0 to 2 at discharge or 90-day. Numbers of patients with mortality, as well as sICH or PH2.

Quality assessment

To evaluate the risk of bias, two investigators (JH and ML) independently assessed each study according to Newcastle-Ottawa Quality Assessment Scale (NOS) for non-randomized studies. ¹³ According to NOS scale, case control studies were assessed basing on three aspects: selection of cases, comparability of cases, and ascertainment of exposure. Cohort studies were assessed basing on patient selection, study comparability and outcome. Disagreements were resolved by consensus or with the help of the senior investigator (ZCH).

Data synthesis and analysis

Meta-analysis was carried out by RevMan 5.3 (Review Manager 5.3 Software Package) provided by Cochrane collaboration network. The heterogeneity included in the study was tested by i² test. If P < 0.1 and I² > 50%, random-effect model was used. If not, the fixed-effect model was employed for combined analysis. Subgroup analysis were used to detect the sources of heterogeneity based on patients only with ASPECTS 0-5, only with ASPECTS 5-7, only with core volume >=70ml and only prospective studies. We applied Odds Ratio (OR) to represent the enumeration data. Measurement data will be represented by mean difference (MD) and 95% confidence interval (95% CI). Sensitivity analysis were conducted by sequential exclusion of one study at a time to assess for a significant change in the summary OR. When more than ten studies are included, publication bias would be evaluated by the symmetry of the funnel plot.

Results

Characteristics of included studies

Among 1166 studies obtained, finally 14 studies with 2547 patients fulfilled our eligibility criteria (Figure 1). Of the total 14 studies consisting of 19 trials, 11 studies were cohort studies (2 obtained 7 scores in NOS,^14,15 5 obtained 8 scores in NOS^16–20 and 4 obtained 9 scores in NOS^8,21–23) 1 study ²⁴ was matched case-control study (9 scores in NOS) and the remaining 2 were secondary analysis of 7 randomized controlled trials.^6,7 The study characteristics were shown in Table 1. The selected studies were conducted between 2016 and 2020 with the sample size ranging from 56 to 743. Overall, 9 studies^{7,15–19,21–23} considered large ischemic core as low aspects and 3 studies^6,14,24 regarded it as large core volume, along with 2 studies^8,20 defined it as low aspects and/or large core volume (Table 1). In addition, 3 studies^8,18,23 used NCCT to evaluate ASPECTS, 3 studies^19–21 used DWI, 2 studies^7,22 used NCCT or DWI,1 study ¹⁶ used CBV, 1 study ¹⁵ used CTASI, and one study had no clear definition to it. As for core volume, 2 studies^8,24 measured it by CTP, 1 study ²⁰ measured it by MRP, along with 2 studies^6,14 measured it by CTP or MRP (Table 1).

Figure 1.

Flow diagram of literature screening and selection process.

Table 1.

Characteristics of included studies.

Publication year, first author	Origin	Study design	Definition of large ischemic core	Cases (I/C)	Time to thrombectomy (hours)	Previous use of IV-r tPA (I/C)	Successful reperfusion rate (I/C)	Proportion of anterior circulation	Age (mean) (I/C)	Male (%) (I/C)	Baseline NIHSS score, mean (I/C)
2020, Gabriel B	Germany	Multicenter cohort study	NCCT-ASPECTS 0-5	99/71	NA	49.50%/45.07%	NA/69.70%	100%	NA/76	57.6/51.0	NA/18.4
2020, Kerleroux B	France	Multicenter retrospective cohort study	DWI-ASPECTS 0-6 and DWI volume >70 mL	130/42	NA	48.5%/ 100%	79.3%/NA	100%	66.2 ± 15/77.7 ± 13.5	59.2/78.6	18.7 ± 4.2/18 ± 5.1
2020, Kim-Tenser	America	Multicenter randomized trial	NCCT-ASPECTS 5-7	28/29	6-16 h	NA/NA	NA/NA	100%	NA/NA	NA/NA	NA/NA
2019, Schnieder M	Germany	Retrospectively cohort study	CBV-ASPECTS 0-7	20/40	0-6 h	70%/ 72.5%	NA/NA	100%	76 (66–86)/ 80.5 (72.2–88.6)	55/37.5	18 (14–22)/ 18 (13.5–22.5)
2019, Sarraj A	America	Prospective multicenter cohort study	NCCT-ASPECTS < 6 or a volume with a relative CTP-CBF<30% of more than 50 cm3	62/43	0-24h	69%/ 60%	81%/NA	100%	66 (59-74)/ 66 (60-81)	60/53	20 (16-23) /21 (17-23)
2019, Kakita H	Japan	Prospective multicenter cohort study	NCCT or DWI-ASPECTS 0-5	172(15 patients with ASPECTS 0-2, 157 patients with ASPECTS 3-5)/332(120 patients with ASPECTS 0-2, 212 patients with ASPECTS 3-5)	0-24 h	39.0%/12.4%	83.7% /NA	100%	74.6 (10.8)/ 80.4 (11.3)	57.0/36.8	21 (16–24) /22 (18–27)
2019, Campbell BCV	Multicenter	Meta	greater than 70 ml on CTP or DWI	48/61	0-24h	88.9%/92.4%	NA/NA	95%	64.7/65.0	52.3/52.6	17/17
2018, Román LS	Multicenter	Meta	NCCT or DWI-ASPECTS 0-4 and 5-7	378(57 patients with ASPECTS 0-4, 321 patients with ASPECTS 5-7)/365(69 patients with ASPECTS 0-4, 296 patients with ASPECTS 5-7)	0-24 h	88%/91%	NA/NA	100%	67·4(57·0–76·0)/ 67·8(58·0–76·0)	53/53	17 (14–20)/ 17 (13–21)
2018, Ohta T	Japan	Retrospectively cohort study	DWI-ASPECTS 0-6	15/23	0-4.5 h	39% /100%	NA/NA	100%	85(78–90)/84(67–86)	40/48	18 (14–22) /19 (16–24)
2018, Mourand I	France	Retrospectively cohort study	DWI-ASPECTS 0-5	60/48	0-8 h	57%/NA	75%/NA	100%	66 (22–86)/67 (41–87)	66.7/62.5	20 (9–28) /22 (5–40)
2018, Shaowei J	China	Retrospectively cohort study	ASPECTS 0-5	36/53	0-6 h	NA/40%	NA/NA	100%	60.83 ± 14.22/NA	53/53	19.67 ± 3.41/NA
2018, Zhicai C	China	Retrospectively cohort study	greater than 70 ml on MRP or CTP	28/76	0-6h	NA/100%	84.0%/58.5%	100%	68.4 ± 14.0/73.8 ± 11.8	78.6/65.8	16 (12–18)/16 (13–19)
2017, Rebello LC	USA	A matched case-control study	greater than 50 or 70 mL on CTP	28(28 patients with core volume greater than 50 ml, 12 patients with core volume greater than 70 ml)/28(28 patients with core volume greater than 50 ml, 12 patients with core volume greater than 70 ml)	0-12 h	32%/ 54%	79%/NA	100%	62.25 (13.92)/ 58.32 (14.79)	46/50	19.5 (15.2-24.0)/ 18.0 (15.0-23.0)
2016, Wasser K	Germany	Prospectively cohort study	CTASI-ASPECTS 0-5 and 6-7	93(41 patients with ASPECTS 0-5, 52 patients with ASPECTS 6-7)/139(80 patients with ASPECTS 0-5, 52 patients with ASPECTS 6-7)	0-3 h	NA/100%	NA/NA	100%	72/74	54/47	16/16

Continuous variables are expressed as mean ± standard deviation or as median (interquartile range). Categorical variables are expressed as frequency (percent).

Abbreviations: NCCT=Non-contrast computed tomography; NA=Not available; NIHSS=National Institutes of Health Stroke Scale; IV-rtPA=intravenous recombinant tissue plasminogen activator; I/C=Intervention group (Thrombectomy group)/Control group (No thrombectomy group); ASPECTS= Alberta Stroke Program Early Computed Tomography Score; CTA= Computed Tomography Angiography; CTP= Computed Tomography Perfusion; DWI= Diffusion Weighted Imaging; CTASI= Source Images of Computed Tomography Angiography; MRP= Magnetic Resonance Perfusion; MCA= Middle Cerebral Artery.

Functional outcome

Eleven studies reported the proportion of good functional outcome in patients with ASPECTS 0-7 and indicated higher odds of good functional outcome in thrombectomy group compared with no thrombectomy group (OR = 3.47; 95% CI 1.99 to 6.07; P < 0.0001, I²=66%, Figure 2(a)). The empty left side of the funnel plot indicated potential publication bias (online Supplementary Figure 1). The subgroup analyses included studies with only ASPECTS 0-5 group, only ASPECTS 5-7 group and only prospective studies group, which confirmed the results of the primary analyses (Figure 2(b), (c), and (f)). However, in patients with large core volume, no significant difference existed between those with and without thrombectomy in patients with large core size (Figure 2(d) and (e)).

Figure 2.

Forest plot for good functional outcome of thrombectomy in patients with large ischemic cores compared with those without thrombectomy.

Safety outcome

As Figure 3 illustrated, all of the included studies reported mortality outcome with no publication bias (online Supplementary Figure 2). In patients with ASPECTS 0-7 and 0-5, there was a lower risk of 90-day mortality in patients underwent thrombectomy (Figure 3(a) and (b)). However, we did not find a significant difference in patients with ASPECTS 5-7 and the large volume subgroup between those with and without thrombectomy (Figure 3(c) to (e)). Moreover, when we only involved prospective studies, the thrombectomy group was associated with a significantly lower risk for mortality (OR = 0.53; 95% CI 0.39 to 0.72; P < 0.0001, I²=0%, Figure 3(f)).

Figure 3.

Forest plot for mortality of thrombectomy in patients with large ischemic cores compared with those without thrombectomy.

This meta-analysis did not find a significant difference in the odds of sICH or PH2 whether in low ASPECTS group or in large core volume group (Figure 4). The symmetrical funnel plot did not show publication bias (online Supplementary Figure 3).

Figure 4.

Forest plot for sICH or PH2 of thrombectomy in patients with large ischemic cores compared with those without thrombectomy.

Discussion

This meta-analysis summarized AIS patients with ASPECTS <=7 or ischemic core volume >=50 ml for a total of 14 studies, suggesting that patients with low ASPECTS might not be the absolute contraindication for thrombectomy. More studies were warranted to explore contraindications for mechanical thrombectomy, especially in patients with large core volume.

Obviously, thrombectomy could significantly improve the safety and effectiveness of low ASPECTS group, but this significance disappeared in the large core volume group. It could be explained by internal inconsistency in definitions. On the one hand, for low ASPECTS, categories for specific interval showed great discrepancy. Most studies^7,17,21,23 enrolled patients with ASPECTS of 0-5, Ohta et al. ¹⁹ for 0-6 and Kim et al. ¹⁸ for 5-7. Actually, the interval of ASPECTS 0-7 was used to define a large core in intravenous thrombolysis at the beginning. ²⁵ Instead, Kevin Phan et al. ¹⁰ suggested a cut-off of 4 may offer a viable exclusion criterion in thrombectomy eligibility to prevent futile or even harmful recanalization in a recent meta-analysis. Therefore, we synthesized the criteria of above researches and defined the large core in slightly wide range as ASPECTS <=7 in this meta-analysis, and we also made subgroup analysis of patients with ASPECTS 0-5 and 5-7. In addition, imaging modalities has given the disparity rate of up to 20% between CTA-ASPECTS and DWI-ASPECTS. ²⁶ Ozdemir et al. ²⁷ found CTASI-ASPECTS in patient selection for thrombectomy seemed to be more useful than NCCT-ASPECTS. On the other hand, for large ischemic core volume, different thresholds of ischemic core were also supposed to be taken into account. EXTEND-IA ⁴ and DEFUSE-3 ²⁸ included patients with ischemic core volume lower than 70 ml, while DAWN ²⁹ registered patients with ischemic core volume lower than 51 ml. Therefore, we set the core volume threshold as 50 ml in our meta-analysis and we also synthesized subgroup of core volume >=70ml. Actually, there was no generally accepted criteria to determine the thresholds of large ischemic core size in imaging profiles. Jelle Demeestere et al. ³⁰ found the cutoff CTP core volume of at least 50 mL most accurately matched a DWI lesion of ≥70 mL (sensitivity 0.86, specificity 0.97, PPV 0.95, NPV 0.92, Youden index 0.84). Takeshi Yoshimoto et al. ³¹ indicated the cutoff values for the redefined malignant profile were DWI-ASPECTS 4 and Vol_DWI 71 mL in the real-world clinical practice.

As for the discrepancy between semiquantitative (ASPECTS) and quantitative (core volume) imaging modalities in evaluating ischemic core, there were different views on this issue. Jelle Demeestere et al. ³⁰ suggested no significant difference between the accuracy of NCCT-ASPECTS and CTP to identify patients with a large DWI lesion (AUC 0.87 vs 0.95, respectively, p = 0.17). Whereas, James E Siegler et al. ³² considered NCCT-ASPECTS and core volume just moderately correlated (r_s=−0.37, p < 0.01). Besides, collateral status could serve as selection criterion for thrombectomy in large infarcts patients, just as the guidelines suggested (level IIB). ³³ In addition, several studies have found in combination with age, baseline infarcts could independently predict poor outcome and younger stroke patients could still benefit from endovascular treatment despite large infarcts.^34–36 While it was true that collateral status and age could affect outcomes independent of treatment, the heterogeneity of the aforementioned inclusion criteria might also play a role in the variation in outcomes seen. This was not adequately controlled for by doing subgroup analyses on low aspects studies and large core volume studies.

Our meta-analysis supplemented previous results from the HERMES collaboration and provided great insight for detecting whether thrombectomy obtained better safety and functional outcome in large core infarct in comparison to medical treatment alone. Moreover, it owned larger sample size (14 studies rather than 7), taken the patients without thrombectomy (instead of the patients without large ischemic core) as the control group and performed specific imaging subgroups analysis previously not assessed. Considering that the measurement of ischemic core differs between core volume and ASPECTS, we also divided them into the only low ASPCETS subgroup and the only large core volume subgroup.

However, it also had several limitations. Firstly, the heterogeneity of the patient group such as different time to thrombectomy, varies proportion of successful reperfusion and different modalities of evaluating core volume might confuse the truth. Secondly, some previous studies^7,10,37 on this topic have shown a difference in effect size in patients with ASPECTS of 4-5 vs ASPECTS 0-3, but it is regrettable that we could not get these patient-level data from studies we included. We hope there will be more randomized trials to further investigate the selection for endovascular therapy in large ischemic core patients. Results of the ongoing LASTE trial (NCT03811769), the TESLA trial (NCT03805308), the SELECT 2 trial (NCT03876457), the ANGEL ASPECTS trial(NCT04551664) and the RESCUE-Japan LIMIT trial (NCT03702413) will be likely to provide further data on this subset.

Conclusion

Our results indicated that treatment with thrombectomy was associated with better functional outcome and lower risk of mortality in AIS patients with low ASPECTS. More prospective studies are warranted to validate this conclusion.