Journal of the Formosan Medical Association (2018) 117, 806e813

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.jfma-online.com

Original Article

Endovascular thrombectomy for acute

ischemic stroke: A single-center experience
in Taiwan
Hai-Jui Chu a,c, Sung-Chun Tang a, Chung-Wei Lee b,
Jiann-Shing Jeng a,*, Hon-Man Liu b,**

a
Stroke Center and Department of Neurology, National Taiwan University Hospital, Taipei, Taiwan
b
Department of Medical Imaging and Radiology, National Taiwan University Hospital, Taipei, Taiwan
c
Department of Neurology, En Chu Kong Hospital, Taipei, Taiwan

Received 23 January 2017; received in revised form 12 August 2017; accepted 30 September 2017

KEYWORDS Background/purpose: Endovascular thrombectomy has been strongly recommended for treat-
Acute ischemic ment of acute ischemic stroke (AIS) with large vessel occlusion. This study aimed to evaluate
stroke; its efficacy and safety in an Asian population from a single center in Taiwan.
Outcome; Methods: Patients who experienced AIS and received endovascular thrombectomy during the
Thrombectomy; period of September 2014 to September 2016 at National Taiwan University Hospital were
Thrombolysis included. Factors related to favorable outcome, defined as modified Rankin scale 0e2 at 90
days after stroke, were analyzed.
Results: During the study period, 65 patients (mean age, 71.9
12.4 years; 44.6% females)
received endovascular thrombectomy, including 33 who received intravenous thrombolysis
before the endovascular treatment. A significant trend of increasing thrombectomy therapy
was observed. The median National Institutes of Health Stroke Scale (NIHSS) score on admis-
sion was 19 (interquartile range, 15e26). The sites of vessel occlusion were middle cerebral
artery in 47 (72.3%) patients, intracranial internal carotid artery in 8 (12.4%), anterior cerebral
artery in 1 (1.5%), and basilar artery in 9 (13.8%). The median times from stroke onset to groin
puncture and from groin puncture to recanalization time were 200 and 29.5 min, respectively.
Successful revascularization was achieved in 41 (63.1%) patients. Two (3.1%) patients had
symptomatic hemorrhagic transformation. At 90 days, 25 (38.5%) patients achieved favorable
outcome. A shorter time from onset to puncture, and successful recanalization were indepen-
dent predictors of favorable outcome.

* Corresponding author. Department of Neurology, National Taiwan University Hospital, No 7, Chung-Shan South Road, 100, Taipei, Taiwan.
Fax: þ886 2 23418395.
** Corresponding author. Department of Medical Imaging and Radiology, National Taiwan University Hospital, No 7, Chung-Shan South Road,
100, Taipei, Taiwan. Fax: þ886 2 23418395.
E-mail addresses: jsjeng@ntu.edu.tw (J.-S. Jeng), inr.liu@gmail.com (H.-M. Liu).

https://doi.org/10.1016/j.jfma.2017.09.016
0929-6646/Copyright ª 2017, Formosan Medical Association. Published by Elsevier Taiwan LLC. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
Thrombectomy for acute ischemic stroke 807

Conclusion: This study demonstrated the therapeutic promise of endovascular thrombectomy

for treatment of AIS with large vessel occlusion in a clinical setting.
Copyright ª 2017, Formosan Medical Association. Published by Elsevier Taiwan LLC. This is an
open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-
nc-nd/4.0/).

Introduction Program Early CT Score (ASPECTS).14 It is a 0e10 score system

to quantify the early ischemic change in affected areas.
Stroke is the second leading cause of death and third leading Segmental assessment of the MCA vascular territory is made
cause of disability worldwide.1 Until recently, intravenous and 1 point is deducted from the initial score of 10 for every
recombinant tissue-type plasminogen activator (r-tPA) was region involved: lower scores correspond to larger infarct
the only therapy that was proven to be effective for acute areas. Patients who were eligible for intravenous r-tPA
ischemic stroke (AIS).2,3 In 2015, five randomized controlled received the treatment before further evaluation for possi-
trials of endovascular treatment for AIS were published.4e8 bility of endovascular treatment.
The significant positive results indicated an evolution For AIS patients who were possible candidates for endo-
of treatment for AIS in selected patients. Nevertheless, vascular treatment, CT angiography and perfusion exami-
because these five randomized trials recruited mainly Cau- nations were arranged to obtain information about location
casians as trial participants, little is known about racial of vessel occlusion and volume of brain tissue at risk of
differences in endovascular treatment. For example, Asians infarction (ischemic penumbra), and volume of brain irre-
may have a higher overall disease burden of atrial fibrilla- versibly infarcted (ischemic core). Large mismatch means
tion,9 and intracranial atherosclerotic stenosis may be more the patient had a relatively small ischemic core and large
common compared to Caucasians.10 Here, we present a ischemic penumbra. By the way, patients with significant
single-center experience with endovascular treatment for renal function impairment, which defined by a serum
AIS in Taiwan and analyze the factors related to favorable creatinine level >2.0 mg/dl were carefully evaluated for the
functional outcome. risk of contrast related acute kidney injury and the benefit
from acute stroke treatment. All considerations were
explained to the patient and accompanied family, informed
Methods consent was obtained before the examination.
In general, the criteria for patients selected for endovas-
Patients who experienced AIS during the period between cular thrombectomy were the following: (1) groin puncture
September 2014 and September 2016 and were treated at can be initiated within 6 h for anterior circulation stroke and
National Taiwan University Hospital (NTUH), a university- within 24 h for posterior circulation stroke; (2) large vessels
affiliated medical center in Taiwan, were analyzed. The occluded such as intracranial internal carotid artery (ICA),
primary data used in this study were obtained from the NTUH first or second segment of middle cerebral artery (M1 or M2,
Stroke Registry, which was started in 1995 to study the etio- MCA), first segment of anterior cerebral artery (A1, ACA),
logical factors, clinical course, prognosis, and complications basilar artery (BA) or vertebral artery (VA) attributed to acute
of stroke.11,12 Among all stroke patients, those who received stroke symptoms; (3) the existence of large ischemic
intravenous thrombolysis and endovascular treatment were mismatch/penumbra for anterior circulation stroke; and (4)
selected for detailed analysis, including clinical characteris- informed consent given.
tics, neuroimaging findings, and functional outcome. This The following clinical information was collected: age,
study was approved by the Institutional Ethics Committee of gender, away from hospital or in-hospital onset of stroke,
NTUH. premorbid activities daily living, NIHSS scores on admission,
treatment with intravenous r-tPA, and comorbid myocardial
Patient selection infarction. The mechanisms of stroke were determined
according to Trial of Org 10172 in Acute Stroke Treatment
(TOAST) classification.15 Medical history such as the car-
The acute stroke protocol has been employed at NTUH since
diovascular risk factors, previous medication history, and
August 2010.13 This protocol involves the cooperation and
laboratory profiles were recorded.
integration of the emergency, radiology, laboratory medi-
cine, and neurology staffs who conduct the initial assess-
ment, imaging, and evaluation of AIS patients to expedite Thrombectomy procedures
acute stroke treatment. Neurologists were consulted once
acute stroke was suspected. Time of stroke onset was The following parameters related to the procedure were
checked and stroke severity was evaluated based on National recorded: time of stroke onset, intravenous r-tPA bolus,
Institutes of Health Stroke Scale (NIHSS) scores, which range groin puncture, and recanalization. Successful revasculari-
from 0 to 42. Head non-contrast computed tomography zation at the end of the thrombectomy procedure was
(NCCT) was performed immediately to exclude intracranial defined as a score of 2b or 3 on the modified Thrombolysis in
hemorrhage (ICH). For those with suspected AIS in the middle Cerebral Infarction [mTICI] scoring system (grade 0: no
cerebral artery territory, NCCT was graded as Alberta Stroke perfusion; grade 1: minimal recanalization; grade 2a: partial
808 H.J. Chu et al.

antegrade reperfusion of <50% previously occluded target Statistical analysis

artery ischemic territory; grade 2b: antegrade reperfusion of
50e99% previously occluded target artery ischemic territory; Descriptive data were expressed as number of patients,
grade 3: complete perfusion).16 percentage, and mean value with standard deviation. NIHSS
and ASPECT scores were expressed as median (25the75th
Outcome measures percentile). Whether the continuous variables were in normal
distribution were tested. Prognostic predictors were deter-
mined using univariable and multivariable analyses. In the
The outcome was defined by modified Rankin scale (mRS)
univariable analysis, distributions in age, sex, NIHSS, and
scores. Favorable outcome was defined as score 0e2 at 90
other parameters between different outcome groups were
days after stroke.17 Other outcome parameters included rate
assessed using a chi-square, independent t-test or Mann
of successful recanalization; NIHSS at 24e48 h; rate of
eWhitney U test depending on the characteristics of data.
symptomatic ICH, defined as any apparently extravascular
Multivariable analysis was used to adjust for factors of age,
blood in the brain or within the cranium associated with
gender, initial NIHSS scores, IV rt-PA, time from onset to
an increase of4 or more points of NIHSS score in 36 h18;
puncture and successful recanalization. Statistical analyses
and all causes mortality at 90 days. We also analyzed the
were performed using SPSS software (IBM Corp. Released
trend of percentage of acute stroke patients that received
2013. IBM SPSS Statistics for Windows, Version 22.0. Armonk,
intravenous r-tPA and endovascular treatment per month
NY) (Chicago, IL, USA). A p value < 0.05 was regarded as
during the study period.

Table 1 Demographics in acute ischemic stroke patients receiving endovascular therapy.

Overall (n Z 65) mRS 2 (n Z 25) mRS 3 (n Z 40) p-value
Age, y 71.9
12.4 68.8
15.9 73.9
9.3 0.15
Female 29 (44.6) 9 (36.0) 20 (50.0) 0.27
In-hospital stroke 14 (21.5) 17 (68.0) 6 (15) 0.18
Pre-stroke mRS0-1 55 (84.6) 25 (100.0) 30 (75.0) 0.01
Co-mobidities and risk factors
Hypertension 40 (61.5) 11 (44.0) 29 (72.5) 0.02
Diabetes mellitus 21 (36.9) 5 (20.0) 16 (40.0) 0.09
Hyperlipidemia 24 (37.0) 9 (36.0) 15 (37.5) 0.90
Myocardial infarction 4 (6.2) 3 (12.0) 1 (2.5) 0.12
Atrial fibrillation 40 (61.5) 15 (60.0) 25 (62.5) 0.84
Previous stroke 20 (30.8) 6 (30.8) 14 (35.0) 0.35
Smoking 23 (35.4) 7 (28.0) 10 (25.0) 0.79
Pre-stroke antithrombotic agents
Previous anti-platelet use (n Z 61) 18 (29.5) 5 (20.0) 18 (50.0) 0.04
Previous anti-coagulant use (n Z 61) 15 (24.6) 9 (36.0) 6 (16.7) 0.13
Etiology of stroke
Cardioembolism 43 (66.2) 19 (76.0) 24 (60.0) 0.19
Large-artery atherosclerosis 15 (23.1) 3 (12.0) 12 (30.0) 0.09
Others 7 (10.8) 3 (12.0) 4 (10.0) 0.80
Initial status
NIHSS 19 (15e26) 19 (14e24) 19.5 (15e27) 0.48
Systolic BP, mmHg 150.0
31.3 141.2
29.0 155.5
31.8 0.07
Diastolic BP, mmHg 81.5
14.7 77.3
16.3 84.1
13.2 0.07
Serum glucose, mg/dL 132.4
43.0 126.8
31.9 139.1
51.2 0.29
Site of artery occlusion
ACA 1 (1.5) 0 1 (2.5) 0.43
MCA 47 (72.3) 22 (88.0) 25 (62.5) 0.03
M1 segment 39 (60.0) 17 (68.0) 22 (55.0) 0.30
M2 segment 8 (12.3) 5 (20.0) 3 (7.5) 0.14
Intracranial ICA 8 (12.4) 1 (4.0) 7 (17.5) 0.11
Ipsilateral extracranial ICA (n Z 56) 7 (12.5) 2 (8.7) 5 (15.2) 0.43
BA 9 (13.8) 2 (8.0) 7 (17.5) 0.28
ASPECTS on CT (n Z 56) 9 (8e10) 9 (8e9) 9 (8e10) 0.90
Target-mismatch profile (n Z 56) 53 (94.6) 23 (92.0) 30 (75.0) 0.18
The statistical exam was applied between mRS 2 and mRS 3.
Values are mean (standard deviation), median (interquartile range), or number (percentage).
mRS: modified Rankin scale; NIHSS: National Institutes of Health Stroke Scale; BP: blood pressure; MCA: middle cerebral artery, ICA:
internal carotid artery, ACA: anterior cerebral artery, BA: basilar artery; ASPECTS: The Alberta Stroke Program Early Computed To-
mography Score CT: computed tomography.
Thrombectomy for acute ischemic stroke 809

significant. Odds ratio and the 95% confidence intervals were were the most common causes corresponding for those with
measured. the status of pre-stroke mRS 3.
With regard to site of occlusion, 47 (72.3%) patients had
MCA occlusion (39 M1 occlusion), 8 (12.4%) had intracranial
Results ICA occlusion, 1 (1.5%) had ACA occlusion, and 9 (13.8%) had
BA occlusion. The median ASPECTS score of the anterior
During the study period, there were 1763 AIS patients circulation infarct was 9 (8e10). The neuroimaging and
treated at our hospital. Among them, 140 (7.9%) received endovascular procedures for stroke at the anterior and
intravenous r-tPA and 65 (3.7%) received endovascular posterior circulation are demonstrated in Figs. 1 and 2,
treatment. Table 1 lists the baseline clinical and neuro- respectively.
imaging characteristics. Of the 65 patients who received Table 2 shows the parameters related to endovascular
endovascular treatment, 14 (21.5%) had in-hospital onset procedures. The median time from stroke onset to
stroke, 55 (84.6%) had a baseline mRS 0e2, and 33 (50.8%) receiving intravenous r-tPA, onset to groin puncture, and
received intravenous r-tPA before endovascular treatment. groin puncture to recanalization were 99 (77.5e144.5), 200
Among patients with in-hospital stroke, 10 (71.4%) were (148.5e273), and 29.5 (18e50) minutes, respectively.
from cardiovascular department while the others (n Z 4, Regarding the devices used in the procedure, for 30 (46.2%)
28.6%) were from oncology department. Besides, internal patients only a penumbra aspiration system was used, for
medical problems, orthopedic problems and old stroke

Figure 1 Images from a representative patient with acute ischemic stroke (AIS) of anterior circulation underwent intra-arterial
thrombectomy. Non-contrast computed tomography (NCCT) (A) and cerebral blood volume (CBV) (B) show no specific evidence of
AIS whereas prolonged mean transient time (MTT) of cerebral blood flow (C) is noted on the right hemisphere (arrowheads showed
the prolonged MTT areas). Computed tomography angiography (CTA) (D) and digital subtraction angiography (DSA) (E) show right
middle cerebral artery (MCA) occlusion (arrowhead indicated the site of occlusion). The occluded artery is approached by
Penumbra thrombectomy system (F, arrowhead indicates the clot aspiration device). After thrombectomy, DSA shows re-
canalized right MCA (G, arrowhead indicated recanalized site). Diffusion weighted imaging (DWI) from magnetic resonance im-
aging 24 h later shows abnormal diffusion restriction at a part of MCA territory (H, arrowheads), and MR angiography shows patent
right MCA (I).
810 H.J. Chu et al.

Discussion

The present study demonstrated the therapeutic promise of

endovascular thrombectomy for patients with AIS in Taiwan.
The results of our data were comparable to those of seven
previously published major international randomized endo-
vascular trials4e8,19,20 in terms of rates for successful recan-
alization, symptomatic ICH, and mortality at 90 days, though
our patients were older, had higher initial NIHSS score, and a
higher percentage had cardioembolism (Supplementary
Table). Regarding the functional outcome, ours has a
similar or lower percentage of good functional outcome at 3
months after stroke, whether there is an association between
atrial fibrillation related stroke or post procedures outcome
deserves further investigation. In the aspect of sICH rate,
ours was 3.1% and ranked in 5th low among the 8 studies
(Supplement Table). We used the criteria of second Euro-
peaneAustralasian Acute Stroke Study (ECASS II)18 as most
trials did. Notably, the definitions of sICH were different in
the Extending the Time for Thrombolysis in Emergency
Neurological Deficits-Intra-Arterial (EXTEND-IA), the Endo-
vascular Treatment for Small Core and Anterior Circulation
Proximal Occlusion with Emphasis on Minimizing CT to
Recanalization Times (ESCAPE), and the Solitaire with the
Intention for Thrombectomy as Primary Endovascular Treat-
ment (SWIFT-PRIME)trials. It may need to be carefully inter-
preted in comparison. Further study is suggested to clarify
the causes of sICH.
In addition, we have identified two independent prog-
nostic factors for stroke patients receiving endovascular
treatment as time from stroke onset to groin puncture and
successful recanalization. These two prognostic factors
emphasize the importance of “time” and “reperfusion” in AIS
Figure 2 Images from a representative patient with acute patients receiving endovascular treatment. For decades,
ischemic stroke (AIS) at posterior circulation received throm- “Time is Brain” has been established as central dogma
bectomy. Non-contrast head computed tomography (CT) shows regarding acute stroke treatment, especially after the era
no specific area of hypodensity yet hyperdensity of the basilar of intravenous r-tPA treatment. At our hospital in 2010 we
artery (BA) (A, arrowhead). CT angiography (B) and digital sub- implemented the stroke code protocol to improve the quality
traction angiography (DSA) (C) both show BA occlusion (arrow- of acute stroke treatment.13 After introduction of the acute
heads). After intra-arterial thrombectomy, DSA (D) shows stroke protocol, there was a significant increase in the per-
recanalized BA. Diffusion weighted imaging (DWI) of magnetic centage of acute stroke patients receiving intravenous r-tPA
resonance imaging 24 h later shows fluid restriction at pons (E, treatment and higher percentage of good functional outcome
arrowhead). There is no fluid restriction over further distal after intravenous r-tPA treatment.
territory. MR angiography shows patent BA (F). Previously we also demonstrated that the delivery of
patients via an emergency medical service system and pre-
10 (15.4%) only a stent retriever, and for 23 (35.4%) both hospital acute stroke notification,21,22 and improved Get
device systems were used. With The Guideline (GWTG)-Stroke program guideline
Successful recanalization was achieved in 41 (63.1%) pa- adherence23 could also shorten the time from stroke onset
tients. There were 2 (3.1%) patients who had symptomatic to treatment with intravenous r-tPA. Nevertheless, in the
and 14 (21.5%) who had asymptomatic ICH on the follow-up era of endovascular treatment, the stroke code protocol
images. At 90 days after stroke, 25 (38.5%) patients had a may need to evolve to improve quality of treatment,
favorable outcome. Importantly, as shown in Table 3, especially with regard to focusing on how to shorten the
multivariable analysis including factors of age, gender, initial time from door to groin puncture.
NIHSS score, intravenous r-tPA used, time from onset to Successful recanalization was the other independent
puncture and successful recanalization showed that only prognostic factor in our study patients. Up to now, there have
factors of onset to puncture time (OR 0.99, 95% CI 0.98e1.00, been three classes of mechanical thrombectomy devices that
p Z 0.02) and successful recanalization (OR 15.32, 95% CI were approved by the Food and Drug Administration in the
2.62e89.45, p < 0.01) were independently associated with Unites States: coil retrievers in 2004, aspiration devices in
favorable outcome. Furthermore, there is a significant trend 2008, and stent retrievers 2012. Currently, most evidence
of increasing rate of AIS patients receiving endovascular supports the use of stent retrievers over other types of me-
treatment during the study period in our hospital (Fig. 3). chanical devices in endovascular treatment for patients with
Thrombectomy for acute ischemic stroke 811

Table 2 Comparison between favorable and unfavorable outcome in ischemic stroke patients receiving endovascular therapy.
Overall (n Z 65) mRS 2 (n Z 25) mRS 3 (n Z 40) p-value
IV-tPA therapy 33 (50.8) 12 (48.0) 21 (52.5) 0.72
Rate of receiving 0.9 mg/kg intravenous r-tPA 8 (24.2) 4 (33.3) 4 (19.0) 0.36
Onset to intravenous r-tPA time, min 99 (77.5e144.5) 87 (74.5e97) 123 (82e160.5) 0.04
Onset to groin puncture, min 200 (148.5e273) 159 (134e204) 223 (183.5e357) 0.00
Groin puncture to recanalization, min (n Z 58) 29.5 (18e50) 26 (15e53.5) 33 (21e47) 0.69
Onset to recanalization, min 231.5 (180e297) 187 (169e225) 265 (216e360) 0.00
Modes of thrombectomy
Penumbra system only 30 (46.2) 15 (60.0) 15 (37.5) 0.18
Stent retriever only 10 (15.4) 5 (20.0) 5 (12.5) 0.54
Both used 23 (35.4) 5 (20.0) 18 (45.0) 0.08
Number of retriever pass (n Z 35) 2 (1e3) 2 (1e3.5) 2 (1e3) 0.81
Outcome
Success of recanalizationa 41 (64.6) 23 (92.0) 19 (47.5) 0.00
Symptomatic ICHb 2 (3.1) 0 2 (5.0) 0.26
Asymptomatic ICH 14 (21.5) 2 (8.0) 12 (30.0) 0.26
NIHSS at 24e48 h 11 (4e20.5) 3 (0.5e5) 17 (12.5e24.5) 0.00
The statistical exam was applied between mRS 2 and mRS 3.
Values are number (percentage), median (interquartile range).
mRS: modified Rankin scale; r-tPA: recombinant tissue plasminogen activator; ICH: intracerebral hemorrhage; NIHSS: National Institutes
of Health Stroke Scale.
a
Defined as modified Thrombolysis in Cerebral Infarction score of 2b (50e99% reperfusion) or 3 (complete reperfusion).
b
Symptomatic ICH: any apparently extravascular blood in the brain or within the cranium associated with an increase of 4 or more
points in the score on the NIHSS that is identified as the predominant cause of neurologic deterioration in 36 h.

AIS. The neurointerventionists in our hospital applied aspi- randomized controlled trial to compare the efficacy of
ration devices more in AIS patients receiving endovascular endovascular treatment to previous standard intravenous r-
treatment (single use in 46.2% and combined with stent re- tPA including the superior third of the basilar artery as a
trievers in another 35.4%) and achieved a similar successful candidate for endovascular treatment.19 However, among
recanalization rate compared to five previous major trials. In the total 204 patients in the intervention group, basilar artery
summary, the statistical analysis did not reveal a significant occlusion accounted for only two (1%). Whether endovascular
impact from different types of mechanical devices on thrombectomy would provide a similar benefit for patients
outcome in our study. Further study with more cases and with posterior circulation stroke remains to be determined.
experience sharing will help to clarify the advantages and Our patients included those with both anterior (86.2%) and
disadvantages of these two mechanical devices in clinical posterior (13.8%) circulation stroke. Although we had a small
practice. case number and there was a relatively lower percentage of
Effect of endovascular treatment in patients with anterior favorable outcome in patients with posterior circulation
or posterior circulation strokes is another interesting issue. stroke compared to patients with anterior circulation stroke
Notably, the inclusion criteria in the aforementioned five (22.2% versus 41.1%, p Z 0.28), endovascular thrombectomy
major endovascular trials basically excluded patients with may have an acceptable chance to improve outcome in pa-
posterior circulation stroke. The THRACE trial is a tients with acute posterior circulation stroke. So far at least
two randomized trials are currently ongoing to prove the
Table 3 Factors related to favorable outcome in ischemic effect for endovascular treatment for posterior circulation
stroke patients receiving endovascular thrombectomy: a stroke.
multivariate analysis. Taiwan has 23 million people with 98% having Han Chinese
ethnicity. The incidence was around 80,000 of new or
Variable Odds ratio (95% p- recurrent strokes each year23 and 74% are ischemic sub-
confidence intervals) value type.24 Our data showed 3.7% of AIS patients received
Age, y 0.94 (0.88e1.01) 0.09 endovascular treatment with a significant trend of increasing
Female 0.66 (0.17e2.55) 0.55 percentage during the study period in our hospital. The rea-
Initial NIHSS score 0.96 (0.87e1.06) 0.45 sons of this trend may be multi-factorial and possibly related
Intravenous r-tPA use 2.59 (0.61e10.93) 0.20 to the strongly positive results of recent trials, knowledge
Time from onset to 0.99 (0.98e1.00) 0.02 improvement of physicians and the population, reimburse-
puncture, min ment of health insurance, and the aggressiveness and ability
Successful recanalizationa 15.32 (2.62e89.45) 0.00 of our hospital’s stroke team. Furthermore increasing fre-
r-tPA: recombinant tissue plasminogen activator; NIHSS: Na- quency of endovascular treatment in the coming years in
tional Institutes of Health Stroke Scale. Taiwan as well as in other countries is also expected. Based on
a
Defined as modified Thrombolysis in Cerebral Infarction score the above data, there would be over 2100 candidates for
of 2b (50e99% reperfusion) or 3 (complete reperfusion). endovascular treatment for AIS annually in Taiwan. The
812 H.J. Chu et al.

Figure 3 Trends of reperfusion therapy in acute ischemic stroke since October 2014 to September 2016. There is significant
increase in administering intra-arterial (IA) thrombectomy, in contrast to consistent performance rate of intravenous recombinant
tissue plasminogen activator (r-tPA) thrombolysis.

immediate problem would be the lack of interventionists and Appendix A. Supplementary data
a well-organized population-based stroke care system to deal
with the change. Supplementary data related to this article can be found at
The present study had some limitations. First, this study https://doi.org/10.1016/j.jfma.2017.09.016.
was conducted at a single medical center with a relatively
small case number. It may limit its generalizability to other
hospitals. Also, it was a clinical practice but not a clinical References
trial. So the comparison with previous studies of clinical trial
may need to interpret carefully. Second, the analysis was 1. Lozano R, Naghavi M, Foreman K, Lim S, Shibuya K, Aboyans V,
done retrospectively, although most cases were selected for et al. Global and regional mortality from 235 causes of death for
endovascular treatment following our pre-specified protocol. 20 age groups in 1990 and 2010: a systematic analysis for the
Global Burden of Disease Study 2010. Lancet 2012;380:2095e128.
Third, the number of qualified neurointerventionists is still
2. The National Institute of Neurological Disorders and Stroke rt-
insufficient at NTUH. The success of endovascular treatment
PA Stroke Study Group. Tissue plasminogen activator for acute
for acute ischemic stroke depends on the experience and ischemic stroke. N Engl J Med 1995;333:1581e7.
sufficient number of neurointerventionists. Soon, a pro- 3. Hacke W, Kaste M, Bluhmki E, Brozman M, Dávalos A,
spective, multi-center registry of endovascular treatment Guidetti D, et al. Thrombolysis with alteplase 3 to 4.5 hours
will be started in Taiwan, and a well-planned training pro- after acute ischemic stroke. N Engl J Med 2008;359:1317e29.
gram for neurointerventionists is ongoing. 4. Berkhemer OA, Fransen PS, Beumer D, van den Berg LA,
Lingsma HF, Yoo AJ, et al. A randomized trial of intraarterial treat-
ment for acute ischemic stroke. N Engl J Med 2015;372:11e20.
Conclusions 5. Campbell BC, Mitchell PJ, Kleinig TJ, Dewey HM, Churilov L,
Yassi N, et al. Endovascular therapy for ischemic stroke with
perfusion-imaging selection. N Engl J Med 2015;372:1009e18.
This study demonstrated a promising outcome for endo- 6. Goyal M, Demchuk AM, Menon BK, Eesa M, Rempel JL,
vascular thrombectomy in AIS with large vessel occlusion in Thornton J, et al. Randomized assessment of rapid endovascular
a clinical setting. Time from stroke onset to groin puncture treatment of ischemic stroke. N Engl J Med 2015;372:1019e30.
and successful recanalization were the two independent 7. Jovin TG, Chamorro A, Cobo E, de Miquel MA, Molina CA,
prognostic factors for AIS patients receiving endovascular Rovira A, et al. Thrombectomy within 8 hours after symptom
treatment. This could be promising for empirically applying onset in ischemic stroke. N Engl J Med 2015;372:2296e306.
endovascular treatment in patients with AIS in Taiwan and 8. Saver JL, Goyal M, Bonafe A, Diener HC, Levy EI, Pereira VM,
other Asian populations. et al. Stent-retriever thrombectomy after intravenous t-PA vs.
t-PA alone in stroke. N Engl J Med 2015;372:2285e95.
9. Tse HF, Wang YJ, Ahmed Ai-Abdullah M, Pizarro-Borromeo AB,
Chiang CE, et al. Stroke prevention in atrial fibrillationean
Conflicts of interest Asian stroke perspective. Heart rhythm 2013;10:1082e8.
10. Toyoda K, Koga M, Hayakawa M, Yamagami H. Acute reperfu-
The authors have no conflicts of interest relevant to this sion therapy and stroke care in Asia after successful endovas-
article. cular trials. Stroke 2015;46:1474e81.
Thrombectomy for acute ischemic stroke 813

11. Yip PK, Jeng JS, Lee TK, Chang YC, Huang ZS, Ng SK, et al. 18. Hacke W, Kaste M, Fieschi C, von Kummer R, Davalos A,
Subtypes of ischemic stroke. A hospital-based stroke registry in Meier D, et al. Randomised double-blind placebo-controlled
Taiwan (SCAN-IV). Stroke 1997;28:2507e12. trial of thrombolytic therapy with intravenous alteplase in
12. Lee HY, Hwang JS, Jeng JS, Wang JD. Quality-adjusted life acute ischaemic stroke (ECASS II). Second European-
expectancy (QALE) and loss of QALE for patients with ischemic Australasian Acute Stroke Study Investigators. Lancet 1998;
stroke and intracerebral hemorrhage: a 13-year follow-up. 352:1245e51.
Stroke 2010;41:739e44. 19. Bracard S, Ducrocq X, Mas JL, Soudant M, Oppenheim C,
13. Chen CH, Tang SC, Tsai LK, Hsieh MJ, Yeh SJ, Huang KY, et al. Moulin T, et al. Mechanical thrombectomy after intravenous
Stroke code improves intravenous thrombolysis administration alteplase versus alteplase alone after stroke (THRACE): a
in acute ischemic stroke. PLoS One 2014;9:e104862. randomised controlled trial. Lancet Neurol 2016;15:1138e47.
14. Barber PA, Demchuk AM, Zhang J, Buchan AM. Validity and 20. Mocco J, Zaidat OO, von Kummer R, Yoo AJ, Gupta R, Lopes D,
reliability of a quantitative computed tomography score in et al. Aspiration thrombectomy after intravenous alteplase
predicting outcome of hyperacute stroke before thrombolytic versus intravenous alteplase alone. Stroke 2016;47:2331e8.
therapy. Lancet 2000;355:1670e4. 21. Hsieh MJ, Tang SC, Chiang WC, Huang KY, Chang AM, Ko PC,
15. Adams Jr HP, Bendixen BH, Kappelle LJ, Biller J, Love BB, et al. Utilization of emergency medical service increases
Gordon DL, et al. Classification of subtype of acute ischemic chance of thrombolytic therapy in patients with acute ischemic
stroke. Definitions for use in a multicenter clinical trial. TOAST. stroke. J Formos Med Assoc 2014;113:813e9.
Trial of Org 10172 in Acute Stroke Treatment. Stroke 1993;24: 22. Hsieh MJ, Tang SC, Chiang WC, Tsai LK, Jeng JS, Ma MH. Effect
35e41. of prehospital notification on acute stroke care: a multicenter
16. Zaidat OO, Yoo AJ, Khatri P, Tomsick TA, von Kummer R, study. Scand J Trauma Resusc Emerg Med 2016;24:57.
Saver JL, et al. Recommendations on angiographic revascu- 23. Hsieh FI, Lien LM, Chen ST, Bai CH, Sun MC, Tseng HP, et al. Get
larization grading standards for acute ischemic stroke: a With the Guidelines-Stroke performance indicators: surveil-
consensus statement. Stroke 2013;44:2650e63. lance of stroke care in the Taiwan Stroke Registry: Get With the
17. van Swieten JC, Koudstaal PJ, Visser MC, Schouten HJ, van Guidelines-Stroke in Taiwan. Circulation 2010;122:1116e23.
Gijn J. Interobserver agreement for the assessment of hand- 24. Hsieh FI, Chiou HY. Stroke: morbidity, risk factors, and care in
icap in stroke patients. Stroke 1988;19:604e7. Taiwan. J Stroke 2014;16:59e64.