Research

Original Investigation

Endovascular Thrombectomy for Acute Ischemic Stroke

A Meta-analysis
Jetan H. Badhiwala, MD; Farshad Nassiri, MD; Waleed Alhazzani, MD; Magdy H. Selim, MD; Forough Farrokhyar, PhD; Julian Spears, MD;
Abhaya V. Kulkarni, MD; Sheila Singh, MD; Abdulrahman Alqahtani, MD; Bram Rochwerg, MD; Mohammad Alshahrani, MD; Naresh K. Murty, MD;
Adel Alhazzani, MD; Blake Yarascavitch, MD; Kesava Reddy, MD; Osama O. Zaidat, MD; Saleh A. Almenawer, MD

Editorial page 1803

IMPORTANCE Endovascular intervention for acute ischemic stroke improves revascularization. Supplemental content at
But trials examining endovascular therapy yielded variable functional outcomes, and the effect jama.com
of endovascular intervention among subgroups needs better definition.
CME Quiz at
jamanetworkcme.com and
OBJECTIVE To examine the association between endovascular mechanical thrombectomy CME Questions page 1868
and clinical outcomes among patients with acute ischemic stroke.

DATA SOURCES We systematically searched MEDLINE, EMBASE, CINAHL, Google Scholar, and
the Cochrane Library without language restriction through August 2015.

STUDY SELECTION Eligible studies were randomized clinical trials of endovascular therapy
with mechanical thrombectomy vs standard medical care, which includes the use of
intravenous tissue plasminogen activator (tPA).

DATA EXTRACTION AND SYNTHESIS Independent reviewers evaluated the quality of studies
and abstracted the data. We calculated odds ratios (ORs) and 95% CIs for all outcomes using
random-effects meta-analyses and performed subgroup and sensitivity analyses to examine
whether certain imaging, patient, treatment, or study characteristics were associated with
improved functional outcome. The strength of the evidence was examined for all outcomes
using the GRADE method.

MAIN OUTCOMES AND MEASURES Ordinal improvement across modified Rankin scale (mRS)
scores at 90 days, functional independence (mRS score, 0-2), angiographic revascularization
at 24 hours, symptomatic intracranial hemorrhage within 90 days, and all-cause mortality at
90 days.

RESULTS Data were included from 8 trials involving 2423 patients (mean [SD] age, 67.4 [14.4]
years; 1131 [46.7%] women), including 1313 who underwent endovascular thrombectomy and
1110 who received standard medical care with tPA. In a meta-analysis of these trials,
endovascular therapy was associated with a significant proportional treatment benefit across
mRS scores (OR, 1.56; 95% CI, 1.142.13; P = .005). Functional independence at 90 days (mRS
score, 0-2) occurred among 557 of 1293 patients (44.6%; 95% CI, 36.6%-52.8%) in the
endovascular therapy group vs 351 of 1094 patients (31.8%; 95% CI, 24.6%-40.0%) in the
standard medical care group (risk difference, 12%; 95% CI, 3.8%-20.3%; OR, 1.71; 95% CI,
1.18-2.49; P = .005). Compared with standard medical care, endovascular thrombectomy was
associated with significantly higher rates of angiographic revascularization at 24 hours
(75.8% vs 34.1%; OR, 6.49; 95% CI, 4.79-8.79; P < .001) but no significant difference in rates
of symptomatic intracranial hemorrhage within 90 days (70 events [5.7%] vs 53 events
[5.1%]; OR, 1.12; 95% CI, 0.77-1.63; P = .56) or all-cause mortality at 90 days (218 deaths
[15.8%] vs 201 deaths [17.8%]; OR, 0.87; 95% CI, 0.68-1.12; P = .27). Author Affiliations: Author
affiliations are listed at the end of this
article.
CONCLUSIONS AND RELEVANCE Among patients with acute ischemic stroke, endovascular
Corresponding Author: Saleh A.
therapy with mechanical thrombectomy vs standard medical care with tPA was associated with
Almenawer, MD, Division of
improved functional outcomes and higher rates of angiographic revascularization, but no Neurosurgery, Department of Clinical
significant difference in symptomatic intracranial hemorrhage or all-cause mortality at 90 days. Epidemiology and Biostatistics,
McMaster University, 1280 Main St W,
Hamilton, ON, Canada L8S 4L8
JAMA. 2015;314(17):1832-1843. doi:10.1001/jama.2015.13767 (Dr_menawer@hotmail.com).

1832 (Reprinted) jama.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

 Endovascular Thrombectomy for Acute Ischemic Stroke
T
he current standard therapy for acute ischemic stroke
is intravenous administration of tissue plasminogen ac-
tivator (tPA).1 Although intravenous tPA improves sur-
vival and functional outcomes when administered as early as
possible after onset of ischemic stroke,2 its use is limited by
the narrow therapeutic time window (<4.5 hours)3 and by im-
portant contraindications, including coagulopathy, recent sur-
gery, or stroke or head injury within the past 3 months.4 Ulti-
mately, as few as 10% of patients presenting with ischemic
stroke can be eligible for treatment with intravenous tPA.5
Moreover, intravenous fibrinolysis is associated with long
recanalization times and poor revascularization rates in
proximal large arterial occlusions, and the prognosis of these
patients remains poor.6,7
The limitations of intravenous tPA have led to interest in
endovascular therapy for acute ischemic stroke. Compared
with intravenous tPA, endovascular intervention can recana-
lize large arterial occlusions earlier and more frequently.8-10
Whether this translates into more favorable clinical out-
comes was assessed in randomized clinical trials that evalu-
ated outcomes of endovascular therapy vs intravenous tPA for
ischemic stroke. Results of these trials yielded a varied effect
of endovascular treatment, warranting further examination.
Prior systematic reviews examining this treatment11,12 have
had limitations, such as inclusion of trials that did not use in-
travenous tPA as the standard medical therapy in the control
group and pooling results from small pilot and observational
studies with those of large phase 2 and 3 clinical trials. In this
study, we conducted a meta-analysis including complete re-
sults from recently published multicenter randomized clini-
cal trials to assess the association between endovascular me-
chanical thrombectomy and clinical outcomes, including
functional outcomes, revascularization, intracranial hemor-
rhage, and mortality, among patients with acute ischemic
stroke. We also aimed to examine whether certain imaging-,
patient-, treatment-, or study-related factors were associated
with improved outcomes to help define the optimal setting for
using endovascular therapy.
Methods
Search Strategy
We undertook a systematic review and meta-analysis based on
a predefined protocol (see the Supplement) in accordance with
the Preferred Reporting Items for Systematic Reviews and Meta-
Analyses (PRISMA) Statement13 and the Cochrane Handbook,14
and used the Grading of Recommendation, Assessment, Devel-
opment, and Evaluation (GRADE)15 to examine the level of evi-
dence for all outcomes of interest. We searched, without lan-
guage restriction, MEDLINE (PubMed and Ovid), EMBASE,
CINAHL, Google Scholar, and the Cochrane Library through
August 2015. We used, in various relevant combinations, key-
words and MeSH terms pertinent to the intervention of inter-
est: endovascular, intra-arterial, tissue plasminogen activator, al-
teplase, fibrinolysis, intervention, embolectomy, thrombolysis, and
thrombectomy; and terms pertinent to the medical condition of
interest: ischemia, stroke, cerebrovascular accident, and infarct
jama.com
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017
Original Investigation Research
(see the Search Strategy in the Supplement). References of
studies with potential relevance and review articles were manu-
ally screened to identify any eligible resources that were not
previously identified.
Study Selection
Three reviewers (J.H.B., F.N., and S.A.A.) independently evalu-
ated studies for eligibility. Disagreements between the review-
ers concerning the decision to include or exclude a study were
resolved by consensus, and if necessary, consultation with a
fourth reviewer (W.A.). We regarded studies as eligible for in-
clusion if they were published randomized clinical trials of
adult participants (18 years) with acute ischemic stroke man-
aged with endovascular therapy compared with the standard
medical care, which includes the use of intravenous tPA. En-
dovascular therapy was defined as the intra-arterial use of a
microcatheter or other device for mechanical thrombec-
tomy, with or without the use of a chemical thrombolytic agent.
Studies were included if they reported on functional out-
comes using the modified Rankin scale (mRS). This 7-score or-
dinal scale (eTable 1 in the Supplement) ranges from 0 (no
symptoms) to 6 (death).16 We excluded duplicate reports and
post hoc analyses. Pilot studies, defined as preliminary inves-
tigations assessing the safety and feasibility of endovascular
thrombectomy in order to guide the design of a future study
or trial, were excluded given concerns related to their small
sample size, short follow-up, and study design. We excluded
abstracts from meeting proceedings, unless published as full-
text reports in a peer-reviewed journal. Moreover, we ex-
cluded studies that did not include intravenous tPA in the con-
trol standard of care group and studies that did not examine
mechanical thrombectomy in the intervention group.
Data Extraction
Three investigators (J.H.B., F.N., and S.A.A.) independently ex-
tracted data from the trials primary texts, supplementary ap-
pendixes, and protocols using data abstraction forms that con-
tained fields for: trial name, year of publication, number and
country of centers, sources of funding, recruitment period,
number of patients in each treatment group, details regard-
ing trial design (eg, randomization, blinding, allocation con-
cealment), eligibility criteria, intervention, control therapy,
baseline patient demographics and comorbidities, efficacy out-
comes (eg, mRS scores, revascularization [eTables 2 and 3 in
the Supplement], quality of life indices), safety outcomes (eg,
intracranial hemorrhage, morbidity, mortality), and out-
comes among relevant subgroups of patients. Disagreements
between the 3 reviewers were resolved by consensus, and if
necessary, consultation with a fourth reviewer (W.A.).
Quality Assessment
Three reviewers (J.H.B., W.A., and S.A.A.) independently per-
formed quality assessment. We used the Cochrane Collabora-
tions tool17 to assess the risk of selection, performance, detec-
tion, attrition, and reporting biases among the included
randomized trials. We judged trials with more than 2 high-risk
components as having a moderate risk of bias, and trials with
more than 4 high-risk components as having a high risk of bias.
(Reprinted) JAMA November 3, 2015 Volume 314, Number 17 1833
 Research Original Investigation
Figure 1. Flowchart of Literature Search and Study Selection
4193 Resources identified through systematic
review of literature for title and abstract screen
4155 Citations excluded
1067 Duplicates
3088 Irrelevant, nonrandomized
studies
38 Articles selected for full-text review
30 Studies excluded
9 No standard medical therapy
control group
5 Pilot studies
4 Protocols
4 Post hoc analyses
4 Nonrandomized studies
3 No mechanical endovascular
therapy (pharmacologic only)
1 Extracranial occlusion only
8 Randomized trials included in meta-analysis
Data Synthesis and Analysis
We calculated, and subsequently pooled in independent meta-
analyses, odds ratios (ORs) with corresponding 95% confi-
dence intervals for each outcome of interest. The primary out-
come was mRS score at 90 days. We calculated the proportional
OR of mRS score for each study by ordinal logistic regression.
The proportional OR expresses the common odds for treat-
ment benefit at each cut-off across all 7 scores of the mRS.
We also evaluated the secondary outcomes of functional in-
dependence (mRS score, 0-2) at 90 days, angiographic revas-
cularization at 24 hours, all-cause mortality at 90 days, and
symptomatic intracranial hemorrhage within 90 days. We de-
termined the risk difference for functional independence and
calculated the number needed to treat (NNT) as 1/risk differ-
ence. Outcomes were analyzed as reported by eligible trials fol-
lowing the intention-to-treat method (ie, based on including
all randomized patients and analyzing them according to the
groups into which they were randomly assigned).
Furthermore, we examined whether important imaging,
patient, or treatment characteristics were associated with im-
proved functional outcome (reduced disability at 90 days) and
accounted for between-study heterogeneity. Outcomes of en-
dovascular therapy were compared with standard medical care
in key subgroups defined by age, sex, National Institutes of
Health Stroke Scale (NIHSS) score,18 Alberta Stroke Program
Early Computed Tomography Score (ASPECTS),19 time to ran-
domization, routine use of computed tomography angiogra-
phy or magnetic resonance angiography to confirm proximal
arterial occlusion prior to intervention, arterial location of oc-
clusion on angiographic imaging, use of intravenous tPA, and
type of endovascular thrombolysis whether chemical or me-
chanical. To further evaluate heterogeneity, we conducted sen-
sitivity analyses of the method of thrombectomy by stratify-
ing trials into low (<20%) vs high (>80%) rate of use of stent
retrievers (ie, first-generation methods, with thrombolytics,
1834 JAMA November 3, 2015 Volume 314, Number 17 (Reprinted)
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017
Endovascular Thrombectomy for Acute Ischemic Stroke
coil retrievers, and aspiration devices vs second-generation
methods and techniques, with stent retrievers), and also of
older (published in 2013) vs newer studies (published in 2015)
to account for important methodological differences be-
tween older and newer studies.20
For all meta-analyses, outcomes were pooled using the
DerSimonian and Laird random-effects model,21 with weights
calculated by the inverse variance method. Heterogeneity
across trials was investigated by the Cochran Q test and mea-
sured by the I2 statistic, with I2 values exceeding 25%, 50%,
and 75% representing low, moderate, and high heteroge-
neity, respectively.22 Publication bias was evaluated visually
by funnel plots, and quantified by the Egger regression,23 the
Begg-Mazumdar test,24 and the Copas selection model.25
Interactions in subgroup and sensitivity analyses were evalu-
ated by random-effects analysis, which assumes the study-
to-study variance is the same for all subgroups. A 2-tailed
P value of <.05 was considered a criterion for statistical sig-
nificance. We did not impute any missing data because the data
provided in the trials primary texts, supplementary appen-
dixes, and protocols were sufficiently granular for robust analy-
ses. Comprehensive Meta-Analysis version 2.2 (Biostat Inc) was
used to conduct all statistical analyses.
Results
The search strategy identified 4193 studies, of which data from
8 trials26-33 were used, comprising 2423 patients (mean [SD]
age, 67.4 [14.4] years), of whom 1131 (46.7%) were women
(Figure 1). A total of 1313 patients underwent endovascular
therapy and 1110 received standard medical treatment. In-
cluded trials and patient characteristics are summarized in
Table 1 and detailed in Table 2 and eTable 4 in the Supple-
ment. All 8 eligible studies were multicenter randomized trials,
results of which were published between 2013 and 2015. There
were similar distributions of patient characteristics, includ-
ing demographics and comorbidities. Locations of ischemic
strokes were all within the anterior circulation distribution, ex-
cept for 44 patients from 2 trials.26,28 The upper limit of time
from stroke onset to endovascular treatment among these trials
varied from 5 to 12 hours (mean, [SD], 3.8 [1.2] hours). Overall
risk of bias was rated as low in all eligible studies, as assessed
using the Cochrane Collaborations tool (eFigure 1 in the Supple-
ment). We did not observe significant publication bias based
on the Egger regression, the Begg-Mazumdar test, or the Copas
selection model (eFigure 2 in the Supplement).
Table 3 presents the distribution of mRS scores separated
by trial. Figure 2A provides a graphical summary of pooled
shifting across the 7 scores of the mRS between both groups
at 90 days. Distribution of mRS scores was more favorable with
endovascular therapy relative to standard therapy, with greater
proportions of patients in each category of favorable out-
come (0, 1, or 2), and smaller proportions in unfavorable cat-
egories (4, 5, or 6). In meta-analysis of all 8 trials, endovascu-
lar intervention was associated with significant proportional
treatment benefit across the mRS scores (OR, 1.56; 95% CI, 1.14-
2.13; P = .005; Figure 2B). In addition, endovascular therapy
jama.com
 Table 1. Descriptive Summary of Included Patients and Randomized Trials Characteristicsa

Treatment Group

jama.com
Baseline
Time Frame for CTA or Intervention, No. (%) Control
Recruitment Endovascular MRA, No. of No. of Follow-up,
Sourceb Period Therapy No. (%) Patients Stroke Locations IV tPA Endovascular Intervention Patients Stroke Locations IV tPA days
SYNTHESIS,26 2008- Initiate within 0 181 Anterior circulation: 0 IA tPA + micro-guidewire: 181 Anterior circulation: 174 90
2013 2012 6 h of onset 160 (88.4) 109/165 (66.1) 170 (93.9) (96.1)
Posterior circulation: Solitaire FR: 18/165 (10.9) Posterior circulation:
18 (9.9) Penumbra: 9/165 (5.5) 11 (6.1)
Both: 1 (0.6) Trevo: 5/165 (3) Both: 0
Merci: 5/165 (3)
MR RESCUE,27 2004- Initiate within 118 64 ICA: 13 (20.3) 28 Merci: 37/61 (60.7) 54 ICA: 7 (13) 16 90
2013 2011 8 h of onset (100) M1 MCA: 39 (60.9) (43.8) Penumbra: 14/61 (23) M1 MCA: 39 (72.2) (29.6)
M2 MCA: 12 (18.8) Merci + Penumbra: 10/61 (16.4) M2 MCA: 8 (14.8)
Merci/Penumbra + IA tPA:
8/61 (13.1)
IMS III,28 2006- Initiate within 306 434 Left hemisphere: 434 IA tPA + Mechanical thrombectomy: 222 Left hemisphere: 222 90
Endovascular Thrombectomy for Acute Ischemic Stroke

2013 2012 5 h of onset (46.6) 224 (51.6) (100) 266/334 (79.6) 106 (47.7) (100)
Right hemisphere: Mechanical thrombectomy alone: Right hemisphere:
197 (45.4) 68/334 (20.4) 109 (49.1)
Brainstem/cerebellum: Micro-catheter: 142/334 (42.5) Brainstem/cerebellum:
10 (2.3) Merci: 95/334 (28.4) 4 (1.8)
Unknown or multiple: Penumbra: 54/334 (16.2) Unknown or multiple:
3 (0.7) EKOS: 22/334 (6.6) 3 (1.4)
Solitaire FR: 5/334 (1.5)
Other: 16/334 (4.8)
MR CLEAN,29 2010- Initiate within 500 233 ICA: 60 (25.8) 203 Retrievable stent: 190/195 (97.4) 267 ICA: 78/266 (29.3) 242 90
2015 2014 6 h of onset (100) M1 MCA: 154 (66.1) (87.1) Other: 5/195 (2.6) M1 MCA: 165/266 (62.0) (90.6)
M2 MCA: 18 (7.7) Mechanical thrombectomy + IA M2 MCA: 21/266 (7.9)
A1 or A2 ACA: 1 (0.4) Thrombolytic agent: 24/195 (12.3)a A1 or A2 ACA: 2/266 (0.8)
IA Thrombolytic agent alone:
1 (0.4)a
ESCAPE,30 2013- Randomization 315 165 ICA: 45/163 (27.6) 120 Retrievable stent: 130/151 (86.1) 150 ICA: 39/147 (26.5) 118 90
2015 2014 within 12 h; (100) M1 MCA or all M2s: (72.7) Solitaire FR: 100/151 (66.2) M1 MCA or all M2s: (78.7)
initiate within 111/163 (68.1) 105/147 (71.4)
1 h of CT Single M2 MCA: Single M2 MCA: 3/147 (2)

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

6/163 (3.7)
EXTEND-IA,31 2012- Initiate within 70 35 ICA: 11 (31.4) 35 Solitaire FR: 27/27 (100) 35 ICA: 11 (31.4) 35 90
2015 2014 6 h, complete (100) M1 MCA: 20 (57.1) (100) M1 MCA: 18 (51.4) (100)
within 8 h M2 MCA: 4 (11.4) M2 MCA: 6 (17.1)

Copyright 2015 American Medical Association. All rights reserved.

of onset
SWIFT-PRIME,32 2012- Initiate within 196 98 ICA: 17/93 (18.3) 98 Solitaire FR or Solitaire 2: 98 ICA: 15/94 (16) 98 90
2015 2014 6 h of onset (100) M1 MCA: 62/93 (66.6) (100) 87/98 (88.8) M1 MCA: 72/94 (76.6) (100)
and 1.5 h M2 MCA: 13/93 (14) M2 MCA: 6/94 (6.4)
of imaging
REVASCAT,33 2012- Initiate within 206 103 ICA: 26/102 (25.5) 70 Solitaire FR: 103 (100) 103 ICA: 28/102 (27.7) 80 90
2015 2014 8 h of onset (100) M1 MCA: 66/102 (64.7) (68) M1 MCA: 65/101 (64.4) (77.7)
and 1 h of M2 MCA: 10/102 (9.8) M2 MCA: 8/101 (7.9)
imaging
a
Abbreviations: ACA, anterior cerebral artery; CTA, computed tomography angiography; IA, intra-arterial; Use of tPA or urokinase was permitted for intra-arterial thrombolysis.
ICA, internal carotid artery; IV, intravenous; MCA, middle cerebral artery; MRA, magnetic resonance angiography; b
All studies were multicenter, open-label randomized clinical trials with a blinded end point.
tPA, tissue plasminogen activator.

(Reprinted) JAMA November 3, 2015 Volume 314, Number 17

Original Investigation Research

1835
 Research Original Investigation Endovascular Thrombectomy for Acute Ischemic Stroke

Table 2. Baseline Patient Characteristics and Treatment Parameters by Treatment Group Among Included Randomized Trials

SYNTHESIS,26 2013 MR RESCUE,27 2013 IMS III,28 2013 MR CLEAN,29 2015

Endovascular Standard Endovascular Standard Endovascular Standard Endovascular Standard
Characteristics (n = 181) (n = 181) (n = 64) (n = 54) (n = 434) (n = 222) (n = 233) (n = 267)
Age, mean (SD) or 66 (11) 67 (11) 64.2 (12.8) 67.1 (16.5) 69 (23-89) (23-84) 65.8 65.7
median (IQR), y (54.5-76.0) (55.5-76.4)
Risk factors, No. (%)
Women 75 (41.4) 78 (43.1) 34 (53.1) 27 (50.0) 216 (49.8) 100 (45.0) 98 (42.1) 110 (41.2)
Hypertension 102 (56.4) 105 (58.0) 54 (84.4) 41 (75.9) 319 (73.5) 171 (77.0) 98 (42.1) 129 (48.3)
Atrial fibrillation 14 (7.7) 29 (16.0) 16 (25.0) 20 (37.0) 153 (35.3) 70 (31.5) 66 (28.3) 69 (25.8)
Coronary artery NS NS 102 (23.5) 72 (32.4)
disease
Myocardial 10 (15.6) 14 (25.9) 33 (14.2) 42 (15.7)
infarction
Antiplatelet 73 (40.3) 59 (32.6) NS NS 186 (42.9) 108 (48.6) 64 (27.5) 80 (30.0)
therapy
Congestive NS NS 5 (7.8) 14 (25.9) 50 (11.5) 31 (14.0) NS NS
heart failure
Hyperlipidemia NS NS 36 (56.3) 32 (59.3) 215 (49.5) 112 (50.5) 58 (24.9) 71 (26.6)
Diabetes mellitus 20 (11.0) 19 (10.5) 12 (18.8) 14 (25.9) 94 (21.7) 54 (24.3) 34 (14.6) 34 (12.7)
Past stroke NS NS 10 (15.6) 8 (14.8) NS NS 29 (12.4) 25 (9.4)
Smoking NS NS 27 (42.2) 20 (37.0) NS NS 65/225 (28.9) 78/252 (31.0)
NIHSS score, 13 (9-17) 13 (9-18) NS NS 17 (7-40) 16 (8-30) 17 (14-21) 18 (14-22)
mean (SD) or
median (IQR)a
ASPECTS NS NS NS NS NS NS 9 (7-10) 9 (8-10)
Site of occlusion NS NS NS NS
ICA/carotid 13 (20.3) 7 (13.0) 60 (25.8) 78/266 (29.3)
terminus
M1 MCA 39 (60.9) 39 (72.2) 154 (66.1) 165/266 (62.0)
M2 MCA 12 (18.8) 8 (14.8) 18 (7.7) 21/266 (7.9)
Time from
stroke onset,
median (IQR),
min
To randomization 148 (124-190) 145 (119-179) 315 (90) 346 (69) NS NS 204 (152-251) 196 (149-266)
To IV tPA NA 165 (140-200) NS NS 122.4 (33.7) 121.2 (33.8) 85 (67-110) 87 (65-116)
To groin puncture 225 (194-260) NS 208 (46.7) 260 (210-313)
IV tPA, No. (%) 0 (0) 174 (96.1) 28 (43.8) 16 (29.6) 434 (100) 222 (100) 203 (87.1) 242 (90.6)

(continued)

was associated with significantly higher rates of functional in-
dependence at 90 days (557 of 1293 patients [44.6%]; 95% CI,
36.6%-52.8%) than standard treatment (351 of 1094 patients
[31.8%]; 95% CI, 24.6%-40.0%), for a risk difference of 12.0%
(95% CI, 3.8%-20.3%; OR, 1.71; 95% CI, 1.18-2.49; P = .005;
Figure 3A). The number needed to treat for endovascular in-
tervention relative to usual medical care for achieving the out-
come of functional independence was 8 (95% CI, 5-26).
Secondary outcomes among both examined groups for all
trials are detailed in eTable 5 in the Supplement. Rates of an-
giographic revascularization at 24 hours for endovascular
therapy was 75.8% (95% CI, 68.1%-82.2%) vs 34.1% (95% CI,
29.8%-38.7%) for standard therapy (OR, 6.49; 95% CI, 4.79-
8.79; P < .001; Figure 3C). There was no significant difference
in rates of symptomatic intracranial hemorrhage within 90 days
between groups: 5.7% (95%, CI; 4.4%-7.3%) for endovascular
therapy vs 5.1% (95%, CI; 3.9%-6.6%) for standard therapy (OR,
1.12; 95% CI, 0.77-1.63; P = .56; Figure 3D). Nor was there sig-
nificant difference in the rates of all-cause mortality at 90 days:
15.8% (95% CI, 12.7%-19.3%) for endovascular therapy vs 17.8%
(95% CI, 14.4%-21.8%) for standard therapy (OR, 0.87; 95% CI,
0.68-1.12; P = .27; Figure 3B). Overall morbidity, including rates
of in-hospital medical complications (eg, deep venous throm-
bosis, myocardial infarction, pneumonia), were not signifi-
cantly different between the endovascular intervention group
and standard treatment group (eTable 5 in the Supplement). The
overall quality of evidence, as examined using GRADE, was high
(eTable 6 in the Supplement). The outcomes of proportional
treatment benefit across the mRS scores, functional indepen-
dence, and revascularization were rated as high quality; whereas
mortality and symptomatic intracranial hemorrhage were rated
as moderate quality due to concerns related to imprecision, as
detailed in eTable 6 in the Supplement.
Substantial heterogeneity (I2 = 75.4%) was detected in the
outcome of functional improvement. Therefore, we con-
ducted multiple subgroup and sensitivity analyses to exam-
ine the relative efficacy of endovascular therapy vs standard
medical care stratified by key imaging-, patient-, treatment-,
and study-related factors. Functional outcomes were signifi-
cantly better among patients with angiographic imaging con-

1836 JAMA November 3, 2015 Volume 314, Number 17 (Reprinted) jama.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

 Endovascular Thrombectomy for Acute Ischemic Stroke Original Investigation Research

Table 2. Baseline Patient Characteristics and Treatment Parameters by Treatment Group Among Included Randomized Trials (continued)

ESCAPE,30 2015 EXTEND-IA,31 2015 SWIFT-PRIME,32 2015 REVASCAT,33 2015

Endovascular Standard Endovascular Standard Endovascular Standard Endovascular Standard
Characteristics (n = 165) (n = 150) (n = 35) (n = 35) (n = 98) (n = 98) (n = 103) (n = 103)
Age, mean (SD) or 71 (60-81) 70 (60-81) 68.6 (12.3) 70.2 (11.8) 65.0 (12.5) 66.3 (11.3) 65.7 (11.3) 67.2 (9.5)
median (IQR), y
Risk factors, No. (%)
Women 86 (52.1) 79 (52.7) 18 (51.4) 18 (51.4) 44 (44.9) 51/96 (53.1) 48 (46.6) 49 (47.7)
Hypertension 105 (63.6) 108 (72.0) 21 (60.0) 23 (65.7) 66 (67.4) 56 (57.7) 62 (60.2) 72 (69.9)
Atrial fibrillation 61 (37.0) 60 (40.0) 12 (34.3) 11 (31.4) 35 (35.7) 38 (39.2) 35 (34.0) 37 (35.9)
Coronary artery 40 (24.2) 31 (20.7) NS NS NS NS
disease
Myocardial 8 (8.2) 11 (11.3)
infarction
Antiplatelet NS NS 16 (45.7) 12 (34.3) NS NS 26 (25.2) 31 (30.1)
therapy
Congestive 24 (14.6) 24 (16.0) NS NS NS NS NS NS
heart failure
Hyperlipidemia 58 (35.2) 66 (44.0) NS NS 24 (24.5) 22 (22.7) 54 (52.4) 62 (60.2)
Diabetes mellitus 33 (20.0) 39 (26.0) 2 (5.7) 8 (22.9) 12 (12.2) 15 (15.5) 22 (21.4) 19 (18.4)
Past stroke 17 (10.3) 17 (11.3) NS NS 3 (3.1) 1/97 (1.0) NS NS
Smoking 80 (48.8) 73 (49.3) 12 (34.3) 15 (42.9) 41/96 (42.7) 39/93 (41.9) 26 (25.2) 23 (22.5)
NIHSS score, 16 (13-20) 17 (12-20) 17 (13-20) 13 (9-19) 17 (13-20) 17 (13-19) 17 (14-20) 17 (12-19)
mean (SD) or
median (IQR)a
ASPECTSb 9 (8-10) 9 (8-10) NS NS 9 (7-10) 9 (8-10) 7 (6-9) 8 (6-9)
Site of occlusion
ICA/carotid 45/163 (27.6) 39/147 (26.5) 11 (31.4) 11 (31.4) 17/93 (18.3) 15/94 (16.0) 26/102 (25.5) 28/101 (27.7)
terminus
M1 MCA 111/163 (68.1) 105/147 (71.4) 20 (57.1) 18 (51.4) 62/93 (66.6) 72/94 (76.6) 66/102 (64.7) 65/101 (64.4)
M2 MCA 6/163 (3.7) 3/147 (2.0) 4 (11.4) 6 (17.1) 13/93 (14.0) 6/94 (6.4) 10/102 (9.8) 8/101 (7.9)
Time from
stroke onset,
median (IQR),
min
To randomization 169 172 NS NS 190.5 188 223 226
(117-285) (119-284) (141-249) (130-268) (170-312) (168-308)
To IV tPA 110 (80-142) 125 (89-183) 127 (93-162) 145 (105-180) 110.5 (85-156) 117 (80-155) 117.5 (90-150) 105 (86-137.5)
To groin puncture NS 210 (166-251) 224 (165-275) 269 (201-340)
IV tPA, No. (%) 120 (72.7) 118 (78.7) 35 (100) 35 (100) 98 (100) 98 (100) 70 (68) 80 (77.7)

Abbreviations: ASPECTS, Alberta Stroke Program Early Computed Tomography
Score; CT, computed tomography; ICA, internal carotid artery; IV, intravenous;
MCA, middle cerebral artery; NA, not applicable; NS, not specified; NIHSS, National
Institutes of Health Stroke Scale; tPA, tissue plasminogen activator
a
The National Institutes of Health Stroke Scale (NIHSS)18 evaluates the clinical
severity of stroke and ranges from 0 to 42, with higher values indicating more
severe neurological deficit.
b
ASPECTS19 is a 10-point topographic score evaluating the presence and
severity of early ischemic change on standard CT scan in patients with early
acute ischemic stroke of the anterior circulation, with a normal CT scan
receiving 10 points and a score of 0 indicating diffuse involvement throughout
the MCA territory.

firming proximal arterial occlusion (OR, 2.24; 95% CI, 1.72-
2.90, P for interaction <.001), among patients who received
the combined therapy of intravenous tPA and endovascular
intervention (OR, 2.07; 95% CI, 1.46-2.92; P for interac-
tion = .018), and when stent retriever devices were used for
mechanical thrombectomy (OR, 2.39; 95% CI, 1.88-3.04; P for
interaction <.001). In addition, comparison of newer vs older
trials to evaluate if study-related factors (eg, patient popula-
tion, thrombectomy technique, advances in imaging) were re-
lated to functional outcome revealed statistically significant
differences in treatment outcomes stratified by the trial year
of publication (P for interaction <.001). Examination of endo-
vascular (chemical vs mechanical) thrombolysis was not fea-
sible, because only 1% of all patients received pure chemical
endovascular therapy and less than 17% had combined me-
chanical and chemical endovascular intervention. In sub-
group analyses, there were no differences in treatment out-
comes based on age, sex, NIHSS score, time to randomization,
ASPECTS or location of arterial occlusion (Figure 4; eFigures
3 and 4 in the Supplement).
Discussion
This study reports detailed analyses of 8 recently published mul-
ticenter randomized clinical trials that compared endovascular
therapy to the current standard therapy for patients with acute
ischemic stroke. The results of this meta-analysis show that com-
pared with standard medical management, endovascular in-
tervention with mechanical thrombectomy was associated with

jama.com (Reprinted) JAMA November 3, 2015 Volume 314, Number 17 1837

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

 Research Original Investigation Endovascular Thrombectomy for Acute Ischemic Stroke

Table 3. Distribution of 90-Day Modified Rankin Scale Scores by Treatment Group

No. (%)

Modified SYNTHESIS,26 2013 MR RESCUE,27 2013 IMS III,28 2013 MR CLEAN,29 2015
Rankin Endovascular Standard Endovascular Standard Endovascular Standard Endovascular Standard
Scale Therapy Therapy Therapy Therapy Therapy Therapy Therapy Therapy
Scorea (n = 181) (n = 181) (n = 64) (n = 54) (n = 415) (n = 214) (n = 233) (n = 267)
0 22 (12.2) 28 (15.5) 2 (3.1) 3 (5.6) 53 (12.8) 19 (8.9) 7 (3.0) 1 (0.4)
1 33 (18.2) 35 (19.3) 7 (10.9) 4 (7.4) 69 (16.6) 39 (18.2) 20 (8.6) 15 (5.6)
2 21 (11.6) 21 (11.6) 3 (4.7) 4 (7.4) 55 (13.3) 28 (13.1) 49 (21.0) 35 (13.1)
3 37 (20.4) 28 (15.5) 10 (15.6) 12 (22.2) 71 (17.1) 35 (16.4) 43 (18.5) 44 (16.5)
4 32 (17.7) 38 (21.0) 15 (23.4) 12 (22.2) 64 (15.4) 30 (14.0) 52 (22.3) 81 (30.3)
5 10 (5.5) 13 (7.2) 15 (23.4) 6 (11.1) 20 (4.8) 15 (7.0) 13 (5.6) 32 (12.0)
6 26 (14.4) 18 (9.9) 12 (18.8) 13 (24.1) 83 (20.0) 48 (22.4) 49 (21.0) 59 (22.1)

Modified ESCAPE,30 2015 EXTEND-IA,31 2015 SWIFT-PRIME,32 2015 REVASCAT,33 2015

Rankin Endovascular Standard Endovascular Standard Endovascular Standard Endovascular Standard
Scale Therapy Therapy Therapy Therapy Therapy Therapy Therapy Therapy
Scorea (n = 164) (n = 147) (n = 35) (n = 35) (n = 98) (n = 93) (n = 103) (n = 103)
0 24 (14.6) 11 (7.5) 9 (25.7) 6 (17.1) 17 (17.3) 8 (8.6) 7 (6.8) 6 (5.8)
1 34 (20.7) 15 (10.2) 9 (25.7) 4 (11.4) 25 (25.5) 10 (10.8) 18 (17.5) 7 (6.8)
2 29 (17.7) 17 (11.6) 7 (20.0) 4 (11.4) 17 (17.3) 15 (16.1) 20 (19.4) 16 (15.5)
3 27 (16.5) 22 (15.0) 6 (17.1) 4 (11.4) 12 (12.2) 16 (17.2) 19 (18.4) 20 (19.4)
4 22 (13.4) 36 (24.5) 1 (2.9) 6 (17.1) 15 (15.3) 20 (21.5) 8 (7.8) 17 (16.5)
5 11 (6.7) 18 (12.2) 0 (0) 4 (11.4) 3 (3.1) 12 (12.9) 12 (11.7) 21 (20.4)
6 17 (10.4) 28 (19.0) 3 (8.6) 7 (20.0) 9 (9.2) 12 (12.9) 19 (18.4) 16 (15.5)
a 16
Data on mRS score at 90 d was not available for 19 patients in the follow-up in these trials. The modified Rankin scale measures functional
endovascular therapy group and 8 patients in the standard medical treatment outcome on a 7-point ordinal scale: 0, no symptoms at all; 1, no significant
group in IMS III,28 1 patient in the endovascular therapy group and 3 patients in disability despite symptoms; 2, slight disability; 3, moderate disability;
the standard medical treatment group in ESCAPE,30 and 5 patients in the 4, moderately severe disability; 5, severe disability; 6, death.
standard medical treatment group in SWIFT-PRIME32 due to losses to

Figure 2. Functional Outcomes of Endovascular Therapy vs Standard Therapy

A Degree of disability at 90 d (modified Rankin Scale [mRS])

The modified Rankin scale16
mRS score 0 1 2 3 4 5 6 measures functional outcome on a
Endovascular therapy (n = 1293) 10.9% 16.6% 15.5% 17.4% 16.2% 6.5% 16.9% 7-point ordinal scale: 0, no symptoms
at all; 1, no significant disability
despite symptoms; 2, slight disability;
3, moderate disability; 4, moderately
mRS score 0 1 2 3 4 5 6
severe disability; 5, severe disability;
Standard therapy (n = 1094) 7.5% 11.8% 12.8% 16.5% 21.9% 11.1% 18.4%
6, death. Data on modified Rankin
scale score at 90 days was not
0 10 20 30 40 50 60 70 80 90 100 available for 19 patients in the
Percentage endovascular therapy group and 8
patients in the standard medical
treatment group in IMS III,28 1 patient
B Reduced disability at 90 d
in the endovascular therapy group
Favors Favors and 3 patients in the standard
Odds Ratio Standard Endovascular
medical treatment group in
Source (95% CI) Therapy Therapy P Value Weight, %
ESCAPE,30 and 5 patients in the
SYNTHESIS,26 2013 0.86 (0.601.23) .40 14.2
standard medical treatment group in
MR RESCUE,27 2013 0.86 (0.451.63) .65 10.1
SWIFT-PRIME32 due to losses to
IMS III,28 2013 1.17 (0.881.57) .28 15.3 follow-up in these trials.
MR CLEAN,29 2015 1.66 (1.222.28) .001 14.9
A, Pooled distribution of modified
ESCAPE,30 2015 2.53 (1.703.79) <.001 13.6
Rankin scale scores at 90 days
EXTEND-IA,31 2015 3.22 (1.367.61) .008 7.5 stratified by treatment group.
SWIFT-PRIME,32 2015 2.55 (1.534.26) <.001 11.9
B, Meta-analysis of endovascular
REVASCAT,33 2015 1.57 (0.972.55) .07 12.4
therapy vs standard therapy for the
Overall 1.56 (1.142.13) .005 100.0
outcome of proportional treatment
I2 = 75.9%, P<.01 benefit across modified Rankin scale
0.1 1.0 10 scores at 90 days. Size of data marker
Odds Ratio (95% CI) for each study is proportional to its
weight.

1838 JAMA November 3, 2015 Volume 314, Number 17 (Reprinted) jama.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

 Endovascular Thrombectomy for Acute Ischemic Stroke Original Investigation Research

Figure 3. Secondary Efficacy and Safety Outcomes of Endovascular Therapy vs Standard Therapy

A Functional independence (modified Rankin Scale score 02) at 90 d

Endovascular Standard Favors Favors
Therapy Therapy Standard Endovascular
Source Events/No. Events/No. Odds Ratio (95% CI) Therapy Therapy P Value Weight, %
SYNTHESIS,26 2013 76/181 84/181 0.84 (0.551.27) .40 14.6
MR RESCUE,27 2013 12/64 11/54 0.90 (0.362.25) .82 8.6
IMS III,28 2013 177/415 86/214 1.11 (0.791.55) .55 15.6
MR CLEAN,29 2015 76/233 51/267 2.05 (1.363.09) .001 14.7
ESCAPE,30 2015 87/164 43/147 2.73 (1.714.37) <.001 13.9
EXTEND-IA,31 2015 25/35 14/35 3.75 (1.3810.17) .009 7.8
SWIFT-PRIME,32 2015 59/98 33/93 2.75 (1.534.94) .001 12.4
REVASCAT,33 2015 45/103 29/103 1.98 (1.113.53) .02 12.5
Overall 557/1293 351/1094 1.71 (1.182.49) .005 100.0
I2 = 75.4%, P<.01
0.1 1.0 10
Odds Ratio (95% CI)
B Mortality at 90 d
Endovascular Standard Favors Favors
Therapy Therapy Endovascular Standard
Source Events/No. Events/No. Odds Ratio (95% CI) Therapy Therapy P Value Weight, %
SYNTHESIS,26 2013 26/181 18/181 1.52 (0.802.88) .20 12.5
MR RESCUE,27 2013 12/64 13/54 0.73 (0.301.76) .48 7.2
IMS III,28 2013 83/434 48/222 0.86 (0.581.28) .45 25.3
MR CLEAN,29 2015 49/233 59/267 0.94 (0.611.44) .77 23.1
ESCAPE,30 2015 17/164 28/147 0.49 (0.260.94) .03 12.3
EXTEND-IA,31 2015 3/35 7/35 0.38 (0.091.59) .18 2.9
SWIFT-PRIME,32 2015 9/98 12/97 0.72 (0.291.79) .47 6.8
REVASCAT,33 2015 19/103 16/103 1.23 (0.592.55) .58 10.1
Overall 218/1312 201/1106 0.87 (0.681.12) .27 100.0
I2 = 17.7%, P= .29
0.1 1.0 10
Odds Ratio (95% CI)
C Revascularization at 24 h
Endovascular Standard Favors Favors
Therapy Therapy Standard Endovascular
Source Events/No. Events/No. Odds Ratio (95% CI) Therapy Therapy P Value Weight, %
MR CLEAN,29 2015 141/187 68/207 6.27 (4.039.74) <.001 47.3
ESCAPE,30 2015 113/156 43/138 5.81 (3.519.60) <.001 36.4
EXTEND-IA,31 2015 33/35 15/35 22.00 (4.55106.43) <.001 3.7
SWIFT-PRIME,32 2015 53/64 21/52 7.11 (3.0316.70) <.001 12.6
Overall 340/442 147/432 6.49 (4.798.79) <.001 100.0
I2 = 0.0%, P= .46
0.1 1.0 10
Odds Ratio (95% CI)
D Symptomatic intracranial hemorrhage within 90 d
Endovascular Standard Favors Favors
Therapy Therapy Endovascular Standard
Source Events/No. Events/No. Odds Ratio (95% CI) Therapy Therapy P Value Weight, %
SYNTHESIS,26 2013 10/181 10/181 1.00 (0.412.46) >.99 17.4
MR RESCUE,27 2013 3/64 2/54 1.28 (0.217.95) .79 4.2
IMS III,28 2013 27/434 13/222 1.07 (0.542.11) .85 30.4
MR CLEAN,29 2015 18/233 17/267 1.23 (0.622.45) .55 30.0
ESCAPE,30 2015 6/165 4/150 1.38 (0.384.98) .63 8.6
EXTEND-IA,31 2015 0/35 2/35 0.19 (0.014.08) .29 1.5
SWIFT-PRIME,32 2015 1/98 3/97 0.32 (0.033.16) .33 2.7
REVASCAT,33 2015 5/103 2/103 2.58 (0.4913.59) .27 5.1
Overall 70/1313 53/1109 1.12 (0.771.63) .56 100.0
I2 = 0.0%, P= .82
0.1 1.0 10
Odds Ratio (95% CI)

Meta-analyses of endovascular therapy vs standard therapy for outcomes of
functional independence (modified Rankin scale score 0-2), mortality at 90 days,
revascularization at 24 hours, and symptomatic intracranial hemorrhage within
90 days. Size of data marker for each study is proportional to its weight.
Revascularization was defined as angiographic restoration of blood flow at the
site of arterial occlusion within 24 hours of stroke. Revascularization was assessed
at 27 hours in the SWIFT-PRIME32 trial, and this was considered equivalent to
24 hours for the purposes of the present analysis. The revascularization outcome
in this trial was based on successful reperfusion (reperfusion ratio 90%) on
computed tomographic or magnetic resonance perfusion imaging.

jama.com (Reprinted) JAMA November 3, 2015 Volume 314, Number 17 1839

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

 Research Original Investigation Endovascular Thrombectomy for Acute Ischemic Stroke

Figure 4. Subgroup and Sensitivity Analyses for Favorable Functional Outcome Reduced Disability at 90 Days

Favors Favors
Odds Ratio Standard Endovascular P for
Subgroup No. of Trials (95% CI) Therapy Therapy P Value Interaction
Age, y 626,28-30,32,33
<70 1.62 (0.972.72) .06
.66
70 1.94 (1.063.54) .03
Sex 328,30,32
Women 1.44 (0.842.47) .18
.69
Men 1.69 (0.992.87) .05
NIHSS score a 626,28-30,32,33
<20 1.33 (0.981.82) .07
.28
20 1.84 (1.123.02) .02
ASPECTS score b 528-30,32,33
0-7 1.52 (1.002.31) .05
.80
8-10 1.63 (1.182.24) .003
Time to randomization, h 526,29,30,32,33
3 1.30 (0.732.33) .37
.70
>3 1.52 (0.882.63) .13
Year of publication 826-33
2013 0.98 (0.771.26) .90
<.001
2015 2.39 (1.883.04) <.001
Confirmed arterial occlusion c 826-33
No 0.98 (0.711.35) .91
<.001
Yes 2.24 (1.722.90) <.001
Location of occlusion 429,30,32,33
Cervical ICA 1.94 (1.01-3.73) .05
ICA terminus 2.07 (1.18-3.61) .01 .96
Proximal MCA d 1.86 (1.23-2.81) .003
IV tPA e 826-33
No 0.86 (0.451.62) .63
.02
Yes 2.07 (1.462.92) <.001
Method of thrombectomy f 826-33
First generation devices 0.98 (0.771.26) .90
<.001
Stent retrievers 2.39 (1.883.04) <.001

0.1 1.0 10
Odds Ratio (95% CI)

Favorable functional outcome was defined as reduced disability at 90 days.
Odds ratios and corresponding confidence intervals among patient subgroups
from individual trials were pooled and interactions were evaluated by
random-effects meta-analyses.
a
The National Institutes of Health Stroke Scale (NIHSS)18 evaluates the clinical
severity of stroke and ranges from 0 to 42, with higher values indicating more
severe neurological deficit. An NIHSS score of 20 points was used as a cut-off
because scores higher than 20 are considered severe impairment and
correspond to a significantly greater risk of intracranial hemorrhage and
unfavorable outcome.34-36
b
The Alberta Stroke Program Early Computed Tomography Score (ASPECTS)19
is a 10-point topographic score evaluating the presence and severity of early
ischemic change on standard computed tomographic (CT) scan in patients
with early acute ischemic stroke of the anterior circulation, with a normal CT
scan receiving 10 points and a score of 0 indicating diffuse involvement
throughout the middle cerebral artery (MCA) territory. An ASPECTS of 7 points
was used as a cutoff because scores higher than 7 are associated with poorer
functional outcomes and greater risk of intracranial hemorrhage.19
c
Confirmed arterial occlusion refers to use of CT angiography or magnetic
resonance angiograph (MRA) to confirm arterial occlusion prior to treatment.
d
Proximal MCA occlusion refers to occlusion of the M1 MCA or 2 or more M2
MCA segments; data on outcomes of M1 vs M2 MCA occlusion were not
available.
e
Intravenous tissue plasminogen activator (tPA) refers to use of combination
therapy (endovascular intervention plus intravenous tPA).
f
Method of thrombectomy was separated based on high (>80%) vs low
(<20%) rate of use of retrievable stent devices.

improved functional outcomes and higher rates of functional
independence at 90 days. In addition, endovascular interven-
tion, compared with standard medical therapy with tPA, was
associated with higher rates of angiographically demon-
strated revascularization at 24 hours but was associated with
no significant differences in symptomatic intracranial hemor-
rhage or all-cause mortality at 90 days. Furthermore, sensitiv-
ity analyses suggested that the presence of proximal arterial oc-
clusion on angiographic imaging, intravenous tPA combined
with endovascular intervention, and use of stent retriever de-
vices for mechanical thrombectomy were important factors
associated with improved functional outcomes related to
endovascular thrombectomy.
The results of trials included in this study were variable,
with a high degree of detected heterogeneity for the primary
outcome of mRS score (I2 = 75.9%), justifying use of random-
effects models. We qualitatively and quantitatively investi-
gated and addressed this heterogeneity in our analysis. Simi-
lar to initial reports of percutaneous coronary intervention for
myocardial infarction, the initial clinical trials of endovascu-

1840 JAMA November 3, 2015 Volume 314, Number 17 (Reprinted) jama.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

 Endovascular Thrombectomy for Acute Ischemic Stroke
lar therapy for acute ischemic stroke, in particular Intra-
arterial Versus Systemic Thrombolysis for Acute Ischemic
Stroke (SYNTHESIS),26 Mechanical Retrieval and Recanaliza-
tion of Stroke Clots Using Embolectomy (MR RESCUE),27 and
Interventional Management of Stroke III (IMS III),28 failed to
show a significant benefit of endovascular strategies.
However, these trials had several well-recognized limita-
tions, including inconsistent use of vascular imaging to
confirm vessel occlusion prior to randomization, variable use
of intravenous tPA in the endovascular therapy group, and re-
liance on less effective and older-generation mechanical
devices.20,37,38 These factors were important contributors to
heterogeneity in our meta-analysis. These limitations of early
trials were addressed in the more recent trials, beginning with
Multicenter Randomized Clinical Trial of Endovascular Treat-
ment for Acute Ischemic Stroke in the Netherlands (MR
CLEAN).29 The results of MR CLEAN,29 favoring endovascu-
lar intervention, prompted interim analyses of the Endovas-
cular Treatment for Small Core and Anterior Circulation
Proximal Occlusion with Emphasis on Minimizing CT to
Recanalization Times (ESCAPE),30 Extending the Time for
Thrombolysis in Emergency Neurological DeficitIntra-
Arterial (EXTEND-IA),31 and Solitaire with the Intention for
Thrombectomy as Primar y Endovascular Treatment
(SWIFT-PRIME)32 trials. The Randomized Trial of Revascular-
ization with Solitaire FR Device versus Best Medical Therapy
in the Treatment of Acute Stroke Due to Anterior Circulation
Large Vessel Occlusion Presenting within Eight Hours of
Symptom Onset (REVASCAT) 33 trial was stopped for a
preplanned interim analysis. These 4 trials were subse-
quently halted due to observed benefits in the endovascular
therapy group.
In our analysis, the relative benefit associated with endo-
vascular therapy was increased by concomitant use of intra-
venous tPA. Evidence in support of this combination therapy
is present in the current literature with several possible
explanations.39,40 A combination approach takes advantage of
the speed of intravenous tPA administration and the greater
recanalization potential of endovascular therapy. In addi-
tion, early initiation of treatment with intravenous tPA may
reduce clot burden, restore a critical amount of blood flow, and
facilitate subsequent arterial recanalization by endovascular
mechanical thrombectomy.41,42 However, patients who re-
ceive tPA and those who do not receive tPA can be different
populations. Patients who did not receive intravenous tPA may
have had contraindications or late presentation. These may
have contributed to the relatively smaller treatment effect re-
ported with mechanical thrombectomy among patients not re-
ceiving intravenous tPA that we observed in our study.
In our study, the improvement in functional outcomes as-
sociated with endovascular therapy was significantly greater
when computed tomography angiography or magnetic reso-
nance angiography were used to confirm proximal arterial oc-
clusion prior to trial enrolment. This is intuitive because the
absence of preprocedural vascular imaging may lead to the
catheterization of patients without a proximal occlusive clot,
and therefore patients who are unlikely to benefit from neuro-
interventional treatment.
jama.com
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017
Original Investigation Research
In the context of acute ischemic stroke, endovascular
therapy is often considered, and evaluated as, a single treat-
ment modality. However, a high degree of variability exists in
the inherent nature of this therapy, and in reality, endovascu-
lar intervention may include a number of different but re-
lated strategies, some of which may be more effective than oth-
ers. Endovascular strategies include chemical clot dissolution
with local delivery of tPA, or recanalization of arterial occlu-
sion by clot disruption, aspiration, or retrieval using a micro-
catheter or one of many mechanical devices. The Merci re-
triever was the first thrombectomy device to receive US Food
and Drug Administration approval in 200443 and was widely
used in early trials evaluating endovascular treatments for
acute ischemic stroke, including SYNTHESIS,26 MR RESCUE,27
and IMS III.28 Nevertheless, high-quality evidence exists from
2 trials in support of the improved efficacy of newer retriev-
able stent devices, including the Solitaire Flow Restoration de-
vice and Trevo retriever, compared with older devices, such
as the Merci retriever. 44,45 Thrombectomy therapy was
achieved by stent retriever devices in the majority of patients
in the MR CLEAN29 and ESCAPE30 trials and in all patients in
the EXTEND-IA,31 SWIFT-PRIME,32 and REVASCAT33 trials. In
our meta-analysis, the use of stent retrievers for mechanical
thrombectomy was a significant source of heterogeneity re-
lated to treatment outcomes and significantly affected the rela-
tive benefit associated with endovascular therapy compared
to optimal medical treatment.
The limitations of our meta-analysis include variability in
the design and reporting of included trials that we investi-
gated using prespecified subgroup and sensitivity analyses.
However, we could not evaluate some factors due to lack of
reported data (eg, general vs local anesthesia, time to treat-
ment, use of intra-arterial thrombolytic agents). Of these, time
to treatment may have an important effect on the efficacy of
endovascular therapy. Delays of even less than 30 minutes can
significantly reduce the probability of functional indepen-
dence after endovascular therapy.41,42 However, endovascu-
lar intervention for stroke is not universally available, and
therefore, some of these patients may require transfer to a re-
gional stroke center with neurointerventional capabilities. For
this reason and others, it would be prudent to define a pre-
cise maximal time window after which treatment is consid-
ered relatively futile, similar to what exists for intravenous tPA
(<4.5 hours). Previous clinical studies evaluating the impor-
tance of time to endovascular therapy have provided mixed
results,46-48 with current included trials using a variable thera-
peutic time window, ranging from 5 to 12 hours. Most eligible
studies used a time window up to 6 hours from stroke onset.
Although REVASCAT33 included a 6-to-8-hour group and
ESCAPE30 enrolled 49 patients at 6 to 12 hours, a definitive posi-
tive treatment effect has not been demonstrated in these
subgroups.20 However, trials evaluating treatment windows
extending beyond 6 hours and up to 24 hours are under way,
including the Trevo and Medical Management Versus Medi-
cal Management Alone in Wake Up and Late Presenting Strokes
(DAWN, ClinicalTrials.gov NCT02142283) and Perfusion
Imaging Selection of Ischemic Stroke Patients for Endovascu-
lar Therapy (POSITIVE, ClinicalTrials.gov NCT01852201) trials.
(Reprinted) JAMA November 3, 2015 Volume 314, Number 17 1841
 Research Original Investigation Endovascular Thrombectomy for Acute Ischemic Stroke

Furthermore, the type of mechanical device used for en-
dovascular thrombectomy may have a significant influence on
revascularization and functional outcomes. Although newer
retrievable stent devices were available for use in more re-
cently published trials, the exact outcomes of all types of de-
vices were not recorded in all studies, particularly older trials,
preventing a complete comparison of outcomes stratified by
the device type. In our sensitivity analyses, stent retrievers
were associated with more favorable outcomes than other de-
vices, although results according to the exact device type (eg,
Solitaire Flow Restoration device vs Trevo retriever) were not
reported. Moreover, some of the subgroup analyses in the
present study were limited by relatively smaller sample size
and by virtue of patient selection among included trials. For
example, age, NIHSS score, and ASPECTS may be associated
with the relative benefit of endovascular over standard therapy.
However, many of the trials excluded patients with low
ASPECTS (ie, <6) and older patients (ie, >80 years) or patients
without premorbid independent function, limiting the abil-
ity of our meta-analysis to detect differences in treatment ben-
efit related to these variables. These are important consider-
ations because many patients with ischemic stroke are elderly
and many present with evidence of advanced ischemia and in-
farction. Future trials will need to delineate upper age limits
and clinical and radiological indicators of utility or futility of
endovascular treatment.
This meta-analysis synthesizes evidence from multi-
center randomized clinical trials, and may help inform the de-
sign and execution of future studies examining the efficacy of
endovascular therapy for acute ischemic stroke. Additional
trials are needed to systematically study the relationship of
patient-, disease-, and treatment-related variables with out-
comes following mechanical thrombectomy, and to identify
the ideal patient to undergo endovascular therapy. Limits on
age, ASPECTS, NIHSS score, and, perhaps most importantly,
time to treatment, need to be explored. In addition to opti-
mizing patient selection, trials should explore and define the
optimal endovascular therapy with respect to technique, de-
vice, regional vs general anesthesia, and dosage of intra-
arterial thrombolytic, if any. The relationship of these vari-
ables to safety outcomes, such as mortality and morbidity,
should also be studied. The results of such studies could in-
form the development of clinical practice guidelines. More-
over, studies are needed to evaluate the cost-effectiveness of
endovascular therapy for the treatment of ischemic stroke. In
addition, it may be beneficial for medical personnel involved
in the early care of patients with acute ischemic stroke, such
as paramedics and emergency physicians, to be trained to iden-
tify candidate patients who may benefit from endovascular
therapy, and for communication to appropriate neurointer-
ventional staff to be streamlined to mobilize neurointerven-
tional resources and reduce the time from stroke onset to
recanalization and reperfusion.
Conclusions
Among patients with acute ischemic stroke, endovascular
therapy with mechanical thrombectomy compared with
standard medical care with tPA was associated with im-
proved functional outcomes and higher rates of angiographic
revascularization but no significant difference in occurrence
of symptomatic intracranial hemorrhage or all-cause mortality
at 90 days.

ARTICLE INFORMATION
Author Affiliations: Division of Neurosurgery,
University of Toronto, Toronto, Ontario, Canada
(Badhiwala, Nassiri, Spears, Kulkarni); Division of
Neurosurgery, Department of Clinical Epidemiology
and Biostatistics, McMaster University, Hamilton,
Ontario, Canada (W. Alhazzani, Farrokhyar, Singh,
Rochwerg, Murty, Reddy, Almenawer); Department
of Neurology, Stroke Division, Beth Israel
Deaconess Medical Center and Harvard Medical
School, Boston, Massachusetts (Selim);
Department of Medicine, University of Ottawa,
Ottawa, Ontario, Canada (Alqahtani); Department
of Medicine, Dammam University, Dammam, Saudi
Arabia (Alshahrani); Department of Neurology, King
Khalid University, Abha, Saudi Arabia (A. Alhazzani);
Department of Neurological Surgery, University of
Texas Southwestern Medical Center, Dallas
(Yarascavitch); Departments of Neurology,
Radiology, and Neurosurgery, Medical College of
Wisconsin, Milwaukee (Zaidat).
Author Contributions: Dr Almenawer had full
access to all of the data in the study and takes
responsibility for the integrity of the data and the
accuracy of the data analysis.
Study concept and design: Badhiwala, Nassiri,
Alhazzani, Spears, Kulkarni, Alshahrani, Reddy,
Zaidat, Almenawer,
Acquisition, analysis, or interpretation of data:
Badhiwala, Nassiri, Alhazzani, Selim, Farrokhyar,
Singh, Alqahtani, Rochwerg, Murty, Alhazzani,
Yarascavitch, Zaidat, Almenawer,
Drafting of the manuscript: Badhiwala, Nassiri,
Kulkarni, Alqahtani, Murty, Almenawer,
Critical revision of the manuscript for important
intellectual content: Badhiwala, Nassiri, Alhazzani,
Selim, Farrokhyar, Spears, Singh, Alqahtani,
Rochwerg, Alshahrani, Murty, Alhazzani,
Yarascavitch, Reddy, Zaidat, Almenawer,
Statistical analysis: Badhiwala, Nassiri, Alhazzani,
Almenawer,
Administrative, technical, or material support:
Badhiwala, Rochwerg, Murty, Alhazzani,
Yarascavitch, Zaidat, Almenawer,
Study supervision: Badhiwala, Farrokhyar, Spears,
Kulkarni, Alshahrani, Murty, Zaidat, Almenawer,
Conflict of Interest Disclosures: All authors have
completed and submitted the ICMJE Form for
Disclosure of Potential Conflicts of Interest and
none were reported.
REFERENCES
1. Wardlaw JM, Murray V, Berge E, et al.
Recombinant tissue plasminogen activator for
acute ischaemic stroke: an updated systematic
review and meta-analysis. Lancet. 2012;379(9834):
2364-2372.
2. Emberson J, Lees KR, Lyden P, et al; Stroke
Thrombolysis Trialists Collaborative Group. Effect
of treatment delay, age, and stroke severity on the
effects of intravenous thrombolysis with alteplase
for acute ischaemic stroke: a meta-analysis of
individual patient data from randomised trials. Lancet.
2014;384(9958):1929-1935.
3. Lees KR, Bluhmki E, von Kummer R, et al; ECASS,
ATLANTIS, NINDS and EPITHET rt-PA Study Group.
Time to treatment with intravenous alteplase and
outcome in stroke: an updated pooled analysis of
ECASS, ATLANTIS, NINDS, and EPITHET trials. Lancet.
2010;375(9727):1695-1703.
4. Jauch EC, Saver JL, Adams HP Jr, et al; American
Heart Association Stroke Council; Council on
Cardiovascular Nursing; Council on Peripheral
Vascular Disease; Council on Clinical Cardiology.
Guidelines for the early management of patients
with acute ischemic stroke: a guideline for
healthcare professionals from the American Heart
Association/American Stroke Association. Stroke.
2013;44(3):870-947.
5. de Los Ros la Rosa F, Khoury J, Kissela BM, et al.
Eligibility for Intravenous Recombinant Tissue-Type
Plasminogen Activator Within a Population: The
Effect of the European Cooperative Acute Stroke
Study (ECASS) III Trial. Stroke. 2012;43(6):1591-1595.
6. Ribo M, Alvarez-Sabn J, Montaner J, et al.
Temporal profile of recanalization after intravenous
tissue plasminogen activator: selecting patients for
rescue reperfusion techniques. Stroke. 2006;37(4):
1000-1004.

1842 JAMA November 3, 2015 Volume 314, Number 17 (Reprinted) jama.com

Copyright 2015 American Medical Association. All rights reserved.

Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017

 Endovascular Thrombectomy for Acute Ischemic Stroke
7. Saqqur M, Uchino K, Demchuk AM, et al;
CLOTBUST Investigators. Site of arterial occlusion
identified by transcranial Doppler predicts the
response to intravenous thrombolysis for stroke.
Stroke. 2007;38(3):948-954.
8. Penumbra Pivotal Stroke Trial Investigators. The
penumbra pivotal stroke trial: safety and
effectiveness of a new generation of mechanical
devices for clot removal in intracranial large vessel
occlusive disease. Stroke. 2009;40(8):2761-2768.
9. Smith WS, Sung G, Saver J, et al; Multi MERCI
Investigators. Mechanical thrombectomy for acute
ischemic stroke: final results of the Multi MERCI
trial. Stroke. 2008;39(4):1205-1212.
10. Meyers PM, Schumacher HC, Connolly ES Jr,
Heyer EJ, Gray WA, Higashida RT. Current status of
endovascular stroke treatment. Circulation. 2011;
123(22):2591-2601.
11. Osanai T, Pasupuleti V, Deshpande A, et al.
Acute endovascular reperfusion therapy in ischemic
stroke: a systematic review and meta-analysis of
randomized controlled trials. PLoS One. 2015;10(4):
e0122806.
12. Singh B, Parsaik AK, Prokop LJ, Mittal MK.
Endovascular therapy for acute ischemic stroke:
a systematic review and meta-analysis. Mayo Clin
Proc. 2013;88(10):1056-1065.
13. Moher D, Liberati A, Tetzlaff J, Altman DG;
PRISMA Group. Preferred reporting items for
systematic reviews and meta-analyses: the PRISMA
statement. BMJ. 2009;339:b2535.
14. Higgins JPT, Green S, eds. Cochrane Handbook
for Systematic Reviews of Interventions. Version
5.1.0 [March 2011]. The Cochrane Collaboration;
2011.
15. Guyatt GH, Oxman AD, Vist GE, et al; GRADE
Working Group. GRADE: an emerging consensus on
rating quality of evidence and strength of
recommendations. BMJ. 2008;336(7650):924-926.
16. van Swieten JC, Koudstaal PJ, Visser MC,
Schouten HJ, van Gijn J. Interobserver agreement
for the assessment of handicap in stroke patients.
Stroke. 1988;19(5):604-607.
17. Higgins JP, Altman DG, Gtzsche PC, et al;
Cochrane Bias Methods Group; Cochrane Statistical
Methods Group. The Cochrane Collaborations tool
for assessing risk of bias in randomised trials. BMJ.
2011;343:d5928.
18. Brott T, Adams HP Jr, Olinger CP, et al.
Measurements of acute cerebral infarction:
a clinical examination scale. Stroke. 1989;20(7):
864-870.
19. Barber PA, Demchuk AM, Zhang J, Buchan AM.
Validity and reliability of a quantitative computed
tomography score in predicting outcome of
hyperacute stroke before thrombolytic therapy.
ASPECTS Study Group. Alberta Stroke Programme
Early CT Score. Lancet. 2000;355(9216):1670-1674.
20. Campbell BC, Donnan GA, Lees KR, et al.
Endovascular stent thrombectomy: the new
jama.com
Copyright 2015 American Medical Association. All rights reserved.
Downloaded From: http://jamanetwork.com/pdfaccess.ashx?url=/data/journals/jama/934629/ on 06/01/2017
standard of care for large vessel ischaemic stroke.
Lancet Neurol. 2015;14(8):846-854.
21. DerSimonian R, Laird N. Meta-analysis in clinical
trials. Control Clin Trials. 1986;7(3):177-188.
22. Higgins JP, Thompson SG, Deeks JJ, Altman DG.
Measuring inconsistency in meta-analyses. BMJ.
2003;327(7414):557-560.
23. Egger M, Davey Smith G, Schneider M, Minder
C. Bias in meta-analysis detected by a simple,
graphical test. BMJ. 1997;315(7109):629-634.
24. Begg CB, Mazumdar M. Operating
characteristics of a rank correlation test for
publication bias. Biometrics. 1994;50(4):1088-1101.
25. Copas JB, Shi JQ. A sensitivity analysis for
publication bias in systematic reviews. Stat
Methods Med Res. 2001;10(4):251-265.
26. Ciccone A, Valvassori L, Nichelatti M, et al;
SYNTHESIS Expansion Investigators. Endovascular
treatment for acute ischemic stroke. N Engl J Med.
2013;368(10):904-913.
27. Kidwell CS, Jahan R, Gornbein J, et al; MR
RESCUE Investigators. A trial of imaging selection
and endovascular treatment for ischemic stroke.
N Engl J Med. 2013;368(10):914-923.
28. Broderick JP, Palesch YY, Demchuk AM, et al;
Interventional Management of Stroke (IMS) III
Investigators. Endovascular therapy after
intravenous t-PA versus t-PA alone for stroke.
N Engl J Med. 2013;368(10):893-903.
29. Berkhemer OA, Fransen PS, Beumer D, et al;
MR CLEAN Investigators. A randomized trial of
intraarterial treatment for acute ischemic stroke.
N Engl J Med. 2015;372(1):11-20.
30. Goyal M, Demchuk AM, Menon BK, et al;
ESCAPE Trial Investigators. Randomized
assessment of rapid endovascular treatment of
ischemic stroke. N Engl J Med. 2015;372(11):1019-
1030.
31. Campbell BC, Mitchell PJ, Kleinig TJ, et al;
EXTEND-IA Investigators. Endovascular therapy for
ischemic stroke with perfusion-imaging selection.
N Engl J Med. 2015;372(11):1009-1018.
32. Saver JL, Goyal M, Bonafe A, et al; SWIFT
PRIME Investigators. Stent-retriever thrombectomy
after intravenous t-PA vs. t-PA alone in stroke.
N Engl J Med. 2015;372(24):2285-2295.
33. Jovin TG, Chamorro A, Cobo E, et al; REVASCAT
Trial Investigators. Thrombectomy within 8 hours
after symptom onset in ischemic stroke. N Engl J Med.
2015;372(24):2296-2306.
34. Intracerebral hemorrhage after intravenous
t-PA therapy for ischemic stroke. The NINDS t-PA
Stroke Study Group. Stroke. 1997;28(11):2109-2118.
35. Kwiatkowski TG, Libman RB, Frankel M, et al;
National Institute of Neurological Disorders and
Stroke Recombinant Tissue Plasminogen Activator
Stroke Study Group. Effects of tissue plasminogen
activator for acute ischemic stroke at one year.
N Engl J Med. 1999;340(23):1781-1787.
(Reprinted) JAMA November 3, 2015 Volume 314, Number 17
Original Investigation Research
36. Adams HP Jr, Adams RJ, Brott T, et al; Stroke
Council of the American Stroke Association.
Guidelines for the early management of patients
with ischemic stroke: A scientific statement from
the Stroke Council of the American Stroke
Association. Stroke. 2003;34(4):1056-1083.
37. Arnold M, Schroth G, Gralla J. Endovascular
treatment for acute ischemic stroke. N Engl J Med.
2013;368(25):2431.
38. Cohen JE, Leker RR. Endovascular treatment
for acute ischemic stroke. N Engl J Med. 2013;368
(25):2432.
39. Bhatia R, Shobha N, Menon BK, et al.
Combined full-dose IV and endovascular
thrombolysis in acute ischaemic stroke. Int J Stroke.
2014;9(8):974-979.
40. Mazighi M, Meseguer E, Labreuche J,
Amarenco P. Bridging therapy in acute ischemic
stroke: a systematic review and meta-analysis. Stroke.
2012;43(5):1302-1308.
41. Mazighi M, Serfaty JM, Labreuche J, et al;
RECANALISE investigators. Comparison of
intravenous alteplase with a combined
intravenous-endovascular approach in patients
with stroke and confirmed arterial occlusion
(RECANALISE study): a prospective cohort study.
Lancet Neurol. 2009;8(9):802-809.
42. Yoo DS, Won YD, Huh PW, et al. Therapeutic
results of intra-arterial thrombolysis after full-dose
intravenous tissue plasminogen activator
administration. AJNR Am J Neuroradiol. 2010;31(8):
1536-1540.
43. Katz JM, Gobin YP. Merci Retriever in acute
stroke treatment. Expert Rev Med Devices. 2006;3
(3):273-280.
44. Saver JL, Jahan R, Levy EI, et al; SWIFT
Trialists. Solitaire flow restoration device versus the
Merci Retriever in patients with acute ischaemic
stroke (SWIFT): a randomised, parallel-group,
non-inferiority trial. Lancet. 2012;380(9849):1241-
1249.
45. Nogueira RG, Lutsep HL, Gupta R, et al; TREVO
2 Trialists. Trevo versus Merci retrievers for
thrombectomy revascularisation of large vessel
occlusions in acute ischaemic stroke (TREVO 2):
a randomised trial. Lancet. 2012;380(9849):1231-
1240.
46. Furlan A, Higashida R, Wechsler L, et al; Prolyse
in Acute Cerebral Thromboembolism. Intra-arterial
prourokinase for acute ischemic stroke. The
PROACT II study: a randomized controlled trial. JAMA.
1999;282(21):2003-2011.
47. Wechsler LR, Roberts R, Furlan AJ, et al;
PROACT II Investigators. Factors influencing
outcome and treatment effect in PROACT II. Stroke.
2003;34(5):1224-1229.
48. Zivin JA. Factors determining the therapeutic
window for stroke. Neurology. 1998;50(3):599-603.
1843