Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.

com

REVIEW

Revolution in acute ischaemic stroke

care: a practical guide to mechanical
thrombectomy
Matthew Evans,1 Phil White,2 Peter Cowley,1,3 David Werring1

1
Stroke Research Centre, ABSTRACT occlusion due to either artery-to-artery
Department of Brain repair and Rapid, safe and effective arterial recanalisation to embolism or cardiac embolism. Early
Rehabilitation, University College
London Institute of Neurology,
restore blood flow and improve functional treatment is critical to rescue potentially
London, UK outcome remains the primary goal of hyperacute salvageable tissue (‘time is brain’)10 11:
2 ischaemic stroke management. The benefit of
Stroke Research Group, Institute safe, rapid and effective arterial recanali-
of Neuroscience and Newcastle intravenous thrombolysis with recombinant sation to restore blood flow and improve
University Institute for Ageing, tissue-type plasminogen activator for patients
Newcastle Upon Tyne, UK
functional outcome remains the primary
3 with severe stroke due to large artery occlusion is goal of hyperacute ischaemic stroke
Neuroradiological Academic
Unit, University College London limited; early recanalisation is generally less than management.12
Institute of Neurology, London, 30% for carotid, proximal middle cerebral artery Until recently, the only licensed treat-
UK or basilar artery occlusion. Since November 2014,
ment for acute ischaemic stroke was
nine positive randomised controlled trials of
Correspondence to intravenous thrombolysis with recombi-
Dr David Werring, Stroke mechanical thrombectomy for large vessel
nant tissue-type plasminogen activator
Research Group, University occlusion in the anterior circulation have led to a
College London Institute of (IV r-tPA). However, since November
revolution in the care of patients with acute
Neurology, London, UK; 2014, there have been nine positive rand-
ischaemic stroke. Its efficacy is unmatched by any
[email protected] omised controlled trials of mechanical
previous therapy in stroke medicine, with a
number needed to treat of less than 3 for
thrombectomy published (table 1),
Accepted 8 May 2017
improved functional outcome. With effectiveness leading to a revolution in the care of
shown beyond any reasonable doubt, the key patients with acute ischaemic stroke due
challenge now is how to implement accessible, to large vessel occlusion in the anterior
safe and effective mechanical thrombectomy circulation. The efficacy of this treatment
services. This review aims to provide neurologists is unmatched by any previous therapy in
and other stroke physicians with a summary of stroke medicine, with a number needed
the evidence base, a discussion of practical to treat of less than 3 for improved func-
aspects of delivering the treatment and future tional outcome.13 With effectiveness
challenges. We aim to give guidance on some of shown beyond any reasonable doubt, the
the areas not clearly described in the clinical trials key challenge is how to implement safe
(based on evidence where available, but if not, and effective services accessible to the
on our own experience and practice) and patients who need it.
highlight areas of uncertainty requiring further
research.
BACKGROUND: THE EVIDENCE
Intravenous recombinant tissue-type plasmin-
INTRODUCTION ogen activator (alteplase) and its limitations
In the UK, stroke is the the most Intravenous recombinant tissue-type
" http://dx.doi.org/10.1136/
practneurol-2017-001692 common serious neurological disease plasminogen activator (IV r-tPA) 0.9 mg/
(incidence 115–150 per 100 000 popula- kg is licensed for use in the UK up to
tion)1 2 and a leading cause of 4.5 hours post symptom onset.14 15 In a
death;3 there are more than 1.2 million meta-analysis of 6756 patients in nine
stroke survivors,4–7 of whom more than randomised trials comparing alteplase
50% have a disability.8 Improving with placebo or open control, treatment
To cite: Evans M, White P, outcome from stroke is thus a key health- within 3 hours resulted in good
Cowley P, et al. Pract Neurol care priority. About 80% of acute strokes outcome for 259 (32.9%) of 787
2017;0:1–14.
are ischaemic,9 mainly from large vessel patients who received alteplase

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 1

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

Table 1 Details of the nine positive thrombectomy trials

Onset of
symptoms
Trial Trial dates Centres Participants Primary outcome measure Age (years) IV r-tPA MT NIHSS

MR CLEAN35 2010–14 16 502 mRS at 90 days 18 4.5 6 >1

36
REVASCAT * 2012–14 4 206 mRS at 90 days 18–80z 4.5 8 >5
37
EXTEND 1A † 2012–14 10 70 Reperfusion at 24 hours, 18 4.5 6 No restriction
NIHSS at day 3
SWIFT-prime38† 2012–14 39 196 mRS at 90 days 18–80 3.5 6 8–29
39
ESCAPE † 2013–14 22 316 mRS at 90 days 18 4.5 12 >5
40
THRACE † 2010–15 26 402 mRS2 at 90 days 18–80 4 5 10–25
41
THERAPY * 2012–14 36 108 mRS2 at 90 days 18–85 4.5x 8¶ >7
42
PISTE 2013–15 10 65 mRS2 at 90 days 18 4.5 6 No restriction
43
EASI * 2013–14 1 77 mRS2 at 3 months 18 <3 6 >7**
*Enrolment was halted early after positive results for thrombectomy were reported from other similar trials.
†Trial stopped early due to efficacy.
zAfter enrolling 160 patients, inclusion criteria were modified to include patients up to the age of 85 years with an ASPECTS >8.
xThree-hour limit if patient>80 with diabetes, previous stroke, previous anticoagulation and NIHSS>25.
¶Revised protocol reduced cut-off to 5 hours.
**Or the presence of clinical imaging mismatch, and suspected or proven occlusion of the M1 or M2 segments of the middle cerebral artery, supraclinoid
internal carotid artery or basilar artery.
IV r-tPA, intravenous thrombolysis with recombinant tissue-type plasminogen activator; mRS, modified Rankin Scale; MT, mechanical thrombectomy;
NIHSS, National Institutes of Health Stroke Scale.

compared with 176 (23.1%) of 762 who received Intra-arterial versus intravenous thrombolysis
control (OR 1.75, 95% CI 1.35 to 2.27).13 Rapid The PROACT II trial randomised 180 patients with
delivery of intravenous thrombolysis after stroke acute ischaemic stroke due to proven occlusion of the
onset is crucial: the number needed to treat for an middle cerebral artery and without haemorrhage or
excellent outcome roughly doubles from 5 (for major early infarction signs on CT scan to heparin
treatment within 90 min) to 9 (when treatment is and intra-arterial pro-urokinase or heparin alone;
given at 3.0–4.5 hours).13 However, the relative 40% in the intervention arm achieved a good
benefit of IV r-tPA appears to be consistent regard- outcome compared with 25% in the control arm.25
less of age or stroke severity.13 Stroke services This promising endovascular approach led to the
burgeoned around intravenous thrombolysis, with development of mechanical thrombectomy.
development of hyperacute stroke units, pathways
and protocols, emergency stroke teams and public
awareness campaigns, in order to allow populations Mechanical thrombectomy
to access this effective treatment as quickly as The introduction of mechanical intra-arterial clot
possible. retrieval into clinical practice heralds a new age in the
However, the benefit of IV r-tPA for patients with acute management of ischaemic stroke for patients
severe stroke due to large artery occlusion is vari- with acute large artery occlusive stroke. The Food and
able, due largely to failure of early recanalisation Drug Administration gave clearance to the first endo-
(generally less than 30% for carotid, proximal vascular device: Merci Retrieval System (MERCI), in
middle cerebral artery or basilar artery occlusion).16 August 2004.26 The MERCI trial27 demonstrated a
More importantly, there is a good clinical outcome recanalisation rate (including the basilar artery) of
in only ~25% of patients (at best) with proximal 46% by MERCI device alone and 60.8% when
anterior circulation or basilar artery occlusion.17 combined with adjuvant intra-arterial recombinant
Important independent risk factors predicting poor tissue-type plasminogen activator. Intracranial
outcome post intravenous thrombolysis are the haemorrhage occurred in 7.8%. The MultiMERCI
length18–22 and location23 24 of the arterial trial28 used a later-generation MERCI device and
thrombus. This lack of efficacy of the only licensed demonstrated 69.5% recanalisation after device and
treatment led to efforts to remove larger arterial adjunctive lytic (intra-arterial or intravenous) with
clots using intra-arterial techniques, initially using favourable clinical outcomes in 34%, but there was
lytic but then mechanical means. no control medical therapy group.

2 Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

Optimism about thrombectomy was diminished individual trials. HERMES showed that the propor-
when three early randomised control trials published tions of patients achieving a good (independent)
in 201329–31 failed to show improved efficacy of functional outcome (mRS 0–2 at 90 days) were 46.0%
endovascular clot retrieval compared with intravenous (mechanical thrombectomy) vs 26.5% (best medical
thrombolysis. However, the study designs were treatment). IV r-tPA was given to 83% of patients in
criticised because of the following: limitations in the thrombectomy population and 87% of those in
patient selection (in one of the studies,29 documented the control population. The number needed to treat
large vessel occlusion was not required), use of older for patients to achieve a reduction of 1 or more
technology (mainly first-generation clot retrieval points on mRS was 2.6. Reassuringly, mortality at
devices) and a long delay from stroke onset to inter- 90 days and risk of symptomatic intracerebral haemor-
vention. Nevertheless, in a post hoc subgroup analysis rhage did not differ between patients receiving IV r-
of those with CT angiogram-proven large vessel occlu- tPA and thrombectomy versus IV r-tPA alone. The
sion, there was a statistical benefit from endovascular benefit remained in subgroups of patients >80 years
treatment within 90 min of IV r-tPA.32 of age and those who did not receive IV r-tPA.
New-generation stent retriever devices (the Solitaire Thrombectomy led to consistent benefit across
FR Revascularisation Device and Trevo ProVue National Institutes of Health Stroke Scale (NIHSS)
Retriever) were studied in two small randomised scores, from milder to more severe strokes. Although
control trials,33 34 which showed significantly better there was no statistical heterogeneity of effect by the
recanalisation compared with the older MERCI degree of early brain ischaemia measured by the
device; indeed, the SWIFT study33 was stopped early Alberta Stroke Programme Early CT score
due to significantly better recanalisation with Solitaire (ASPECTS), there was clear benefit only for
(83% vs 48.1% with MERCI), as well as reduced ASPECTS >5. However, there were only a few
mortality at 3 months (17.2% vs 38.2%) and better patients with ASPECTS <5 included. Other recent
neurological outcome at 90 days. meta-analyses have confirmed the key findings from
Everything changed with the publication, in rapid HERMES.45–47
succession, of nine landmark randomised controlled Based on evidence from these trials, updated prac-
trials,35–43 testing new-generation stent retriever tice guidelines were rapidly published in the USA,4 48
devices (between December 2010 and February Canada,49 Europe50 and in the UK,51 52 recom-
mending that mechanical thrombectomy should be
2015), which showed the consistently clear superi-
provided to patients with occlusion of the internal
ority of endovascular clot retrieval over standard
carotid artery or proximal middle cerebral artery who
medical care alone in reducing disability at 90 days in
had received treatment with IV r-tPA within 4.5 hours
patients with ischaemic stroke due to anterior circula-
from of onset53 and who could undergo the proce-
tion large vessel occlusion, as measured by the
dure (arterial puncture) within 6 hours of symptom
modified Rankin Scale (mRS; the primary outcome
onset.
measure). The first study to report was the Multi-
A further meta-analysis of the five studies54 showed
center Randomised Clinical Trial of Endovascular
improved outcomes when thrombectomy was
Treatment for Acute Ischaemic Stroke in the Nether-
performed up to 7.3 hours after symptom onset, in
lands (MR CLEAN),3 35 with subsequent studies all
patients satisfying imaging criteria for the randomised
stopped early due to efficacy, loss of equipoise or
trials, but there was still clearly greater benefit with
both (and it should be noted that stopping early might
faster intervention (<2 hours). Patients with moderate
have caused the later trials to overestimate the effect infarct core volumes (ASPECTS 7–8) had a shallower
size of the treatment). tables 1–3 summarise some key decline in benefit with longer symptom onset to
features of these studies. Note that, unlike the reperfusion than patients with minor infarct core
previous neutral trials, these all selected patients with volumes (ASPECTS 9–10). The important message
proven large artery occlusion using CT angiography here is that, just as for IV r-tPA, speed of delivery of
and mostly randomised patients within 6 hours of mechanical thrombectomy is key to achieving the best
stroke onset (table 1). possible outcomes.
Powerful evidence for the safety and efficacy of
mechanical thrombectomy comes from the ‘Highly
Effective Reperfusion Evaluated in Multiple Endovas- HOW TO SELECT PATIENTS
cular Stroke Trials’ (HERMES) collaboration meta- The decision to proceed with mechanical thrombec-
analysis of the first five positive studies.44 By pooling tomy should be made by a physician trained in the
individual data from 1287 patients, the meta-analysis diagnosis and treatment of stroke, in conjunction
could also investigate the efficacy of thrombectomy in with a neurointerventionist who has the relevant
subgroups that were too small to investigate in the brain and arterial imaging available for review. It is

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 3

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

Table 2 Treatment details for participants in each cohort

Mechanical thrombectomy cohort IV r-tPA cohort
Treatment n Age Median Treatment n Age Median
(±SD) NIHSS (±SD) NIHSS
Trial (IQR) (IQR)

MR CLEAN35 IV r-tPA + MT  (IA r-tPA or intra- 233 65.8 17 (14– IV r-tPA 267 65.7 (55.5– 18 (14–
arterial urokinase) (54.5–76)z 21) 76.4)z 22)
REVASCAT36* IV r-tPA + M.T. 103 65.7 17 (14– IV r-tPA 103 67.2 17 (12–
(11.3)¶ 20) (9.5)¶ 19)
EXTEND IV r-tPA  M.T. 35 68.6 17 (13– IV r-tPA 35 70.2 13 (9–19)
1A37† (12.3)¶ 20) (11.8)¶
SWIFT- IV r-tPA  M.T. 98 65.0 17 (13– IV r-tPA 98 66.3 17 (13–
prime38† (12.5)¶ 20) (11.3)¶ 19)
ESCAPE39† M.T.  IV r-tPA 165 71 (60– 16 (13– IV r-tPA 150 70 (60–81)z 17 (12–
81)z 20) 20)
THRACE40† IV r-tPA  M.T. 200 66 (54– 18 (15– IV r-tPA 202 68 (54–75)z 17 (13–
74)z 21) 20)
THERAPY41* IV r-tPA  M.T. 55 67 (11)¶ 17 (13– IV r-tPA 53 70 (10)¶ 18 (14–
22) 22)
PISTE42 IV r-tPA  M.T. 33 67 (17)¶ 18 (6– IV r-tPA 32 64 (16)¶ 14 (6–
24)x 29)x
EASI43* IV r-tPA  M.T. 40 74 (62.7– 18 (13– IV r-tPA 37 71 (59–79)z 20 (12–
80)z 21.75) 23)
*Enrolment was halted early after positive results for thrombectomy were reported from other similar trials.
†Trial stopped early due to efficacy.
zMedian (IQR).
xMedian (range).
¶Mean (SD).
key: IV r-tPA , intravenous recombinant tissue-type plasminogen activator; IA r-tPA, intra-arterial recombinant tissue-type plasminogen activator;
MT, mechanical thrombectomy; NIHSS , National Institutes of Health Stroke Scale; IQR, interquartile range; SD, standard deviation

essential to have rapid, expert clinical assessment— for stroke diagnosis, localisation, severity

Table 3 Effect of mechanical thrombectomy compared with best medical therapy on good functional outcome (modified Rankin Score2* at 90 days)

Adjusted OR (95% CI)

Trial Mechanical thrombectomy Best medical therapy p value

MR CLEAN35 76 (32.6) 51 (19.1) 2.16 (1.39–3.38)

36
REVASCAT 45 (43.7) 29 (28.2) 2.1 (1.1–4.0)
37
EXTEND 1A 25 (71) 14 (40) 4.2 (1.4–12) p=0.01
38
SWIFT-prime 59 (60) 33 (35) 1.70 (1.23–2.33) p<0.001
39
ESCAPE 87 (53.0) 43 (29.3) 1.7 (1.3–2.2)
40 53
THRACE 106 85 (42) 1.55 (1.05–2.30) p=0.028†
THERAPY41 1938 1430 1.4 (0.60–3.3) p=0.55
42 35
PISTE 17 (57) 10 4.92 (1.23–19.69)p=0.021z
EASI43 20 (50)x 14 (38)¶ p=0.36
Figures are number of patients (%).
*This corresponds to slight or no residual disability as a result of the stroke.
†Value at 30 days.
zPer protocol population analysis.
x19/35 anterior circulation, 1/5 posterior circulation.
¶14/32 anterior circulation, 0/5 posterior circulation.
OR = odds ratio.

4 Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

stratification (NIHSS) and assessment of pre-stroke included in some of the trials and might also
functional status (modified Rankin score) and reasonably be offered treatment)
comorbidities—and adequate brain and vascular " thrombectomy can be performed within 6 hours
imaging acquisition (typically CT and CT angiog-
" good collateral circulation (though benefit in
patients with poor collaterals remains uncertain).
raphy) and interpretation. It is crucially important
to have interaction, discussion and teamwork Areas of remaining uncertainty include patients with
between neurologist and neurointerventionist to more distal occlusions (eg, M2); there was no statis-
make often complex and time-sensitive decisions. tical evidence of treatment effect heterogeneity in
Extracranial vessel imaging (easily obtained with the patients with M2 occlusion, but only 94 such patients
same CT angiogram) is essential to determine the were included; patients with substantial symptoms
feasibility of access to the target artery occlusion. and technically accessible occlusions in proximal M2
The selection criteria applied in practice should might thus be reasonable to treat, but we need more
parallel those of the successful trials, including the evidence. There is also still only limited evidence for
following: thrombectomy in basilar occlusion. The role of more
advanced imaging also remains to be defined (beyond
" documented large vessel anterior circulation occlu- the mandatory CT and CT angiography, eg, CT perfu-
sion (middle cerebral artery, M1 or carotid T) sion, MRI DWI and PWI, which can more accurately
" significant clinical deficit at the time of treatment
define ischaemic core volume as well as potentially
(this might be NIHSS>5 or a lower score that is
salvageable brain). Nevertheless, good outcomes have
functionally significant for the patient; note that
even mild deficit from proven large vessel occlusion been achieved in Netherlands and the UK using prag-
has a high risk of clinical deterioration) matic CT angiogram-based patient selection (MR
" lack of extensive early ischaemic change (those CLEAN and PISTE).
with ASPECTS more than 5 on plain CT clearly
benefit)
" pre-stroke functional status and lack of serious HOW IT IS DONE
comorbidities indicating potential to benefit from Devices, technique and clot types
treatment (note that age>80 years alone is NOT a After the positive randomised trials, the Solitaire FR
contraindication to treatment) stent retriever device became the benchmark for
" treatment with intravenous thrombolysis within mechanical thrombectomy. However, rapid and safe
4.5 hours (although patients ineligible for intrave-
recanalisation and reperfusion of brain is the key
nous thrombolysis due to bleeding risk were also
factor, rather than any specific device or technique;

Figure 1 A range of different clot types, which have different physical properties, potentially requiring a range of thrombectomy
techniques. These are experimental clot analogues, primarily from ovine blood. Image provided courtesy Neuravi.84

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 5

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

there are multiple options available. In addition to the clot, where it opens like a stent (but remains attached
primary device, many supplementary devices and to its pusher wire); once integrated with the clot, the
techniques are used, for example, balloon guide cathe- device is pulled back into the intermediate catheter, to
ters, intermediate catheters and suction pumps which suction is simultaneously applied. A balloon
versus manual aspiration, etc. Moreover, the variable guide, forming a cuff around the guide catheter, may
underlying mechanisms and anatomy of large vessel be used to stop forward flow and reduce the chance
occlusions may challenge the standard approach. of embolising fragments of the clot distally or into
There is a vast range of thrombus types (traditionally other territories; when using such a guide, the inter-
considered as ‘red’ or fibrin-rich (classically thought mediate catheter may be omitted. figure 2 shows
most likely to be cardioembolic) and ‘white’ or freshly removed clot in a stent retriever device.
platelet-rich (classically thought most likely to result An increasingly popular approach is to attempt to
from atherosclerotic plaques). This dichotomy is now aspirate the clot directly into an intermediate catheter,
recognised to be an oversimplification; figure 1 shows which has become feasible with the advent of large
a vast range of different potential clot types, which lumen catheters that can safely be navigated into the
have different physical properties potentially requiring M1 segment of the middle cerebral artery and
a range of thrombectomy techniques to optimise beyond. The key here is to choose a catheter with a
recanalisation; for example, friction properties (‘stick- lumen approaching the size of the vessel where the
iness’) might relate to the ratio of fibrin to red blood clot is lodged allowing the clot to be suctioned in its
cells. entirety; if this fails, a stent retriever may easily be
The issue of general anaesthetic versus awake deployed through the initial system.
thrombectomy remains controversial, with little Common challenges are tandem occlusions of
evidence to guide the decisions (see below). Our cervical internal carotid artery and intracranial vessels
current approach is to proceed with the patient awake and fixed intracranial stenosis, which can limit endo-
using local anaesthetic and analgesia and support vascular access; patients are often older, with tortuous
from an anaesthetist. However, in severe dominant and ectatic large vessels, often with difficult ‘unfolded’
hemisphere internal carotid artery and middle cere- aortic arch and redundant cervical carotid loops. The
bral artery occlusions, the patients may be very primary access point is the common femoral artery,
agitated and confused making the procedure very but the radial or brachial approach is an alternative
difficult and unsafe; under these circumstances, it is for those with aorto-ilial-femoral disease ; the direct
entirely appropriate to recourse to general carotid approach has also been proposed but remains
anaesthesia. unpopular due to safety concerns. When there is a
Normally, after femoral arterial puncture, a large tandem occlusion secondary to carotid disease in the
(8Fr) guide catheter is navigated into the internal neck, the interventionist has to decide which lesion to
carotid artery, within which are an intermediate (5– treat first and whether to deploy a carotid stent or
6Fr) catheter, which is directed to the circle of Willis, just angioplasty any stenotic lesions. There is currently
and a microcatheter, which must be navigated
no evidence or guideline to direct this decision. Simi-
through to the clot over a microguidewire. The
larly with fixed intracranial stenosis, when the clot
microwire is removed, allowing the stent retriever to
has been removed revealing a tight (and possibly
slide through the microcatheter to emerge inside the
unstable) stenotic plaque, the choice is between angio-
plasty and stenting. In both of these circumstances,
there is a need to maintain dual antiplatelet blockade
if a stent is left in situ, which might exacerbate any
haemorrhagic complications after stroke. A systematic
review of 32 studies included 1107 patients with
internal carotid artery occlusions found that acute
stenting of occlusions of the extracranial internal
carotid artery resulted in a higher recanalisation rate
(87% vs 48%, p=0.001) and better outcomes (68%
vs 15%, p<0.001) as well as lower mortality (18% vs
41%, p=0.048) when compared with intra-arterial
thrombolysis.55
Recent cohort studies suggest that tandem stenosis/
occlusions of the internal carotid/middle cerebral
Figure 2 Freshly removed clot enclosed in a stent retriever arteries can be treated with acute stenting of the
device. extracranial internal carotid and stent retriever

6 Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

mechanical thrombectomy in the middle cerebral comprehensive stroke centres with adequate neuroin-
artery with acceptable risk profile, but further terventional procedural volumes (>200 per annum),
research of the safety profile and benefit of this of which a reasonable proportion are mechanical
approach is needed.56–60 thrombectomy and undertaking regular assessment/
audit of technical and clinical results, process times
and complications. When complications do occur, the
Complications and how they are managed immediate availability of neurocritical care and (less
Complications of endovascular procedures can result frequently required) neurosurgical support are manda-
from direct device-related vascular injury, vascular tory and may be lifesaving. Figures 3–5 present three
access and the use of radiological contrast media. examples of thrombectomy procedures to demon-
Stent retriever devices are generally safe61 with lower strate some of the potential complexities of the
complication rates than first-generation devices. The procedure.
the most common complications include the
following: vessel perforation,62–64 which occurred in
How are patients cared for before, during and after the
1.6% patients in the five recent positive endovascular
procedure
trials (range 0.9%–4.9%); symptomatic intracranial
haemorrhage (3.6%–9.3%); subarachnoid haemor- Based on the published trial evidence, treatment
should ideally be undertaken in major neurointerven-
rhage (0.6%–4.9%); arterial dissection (0.6%–3.9%);
emboli to new territories (1.0%–8.6% in randomised tional centres with well-functioning hyperacute stroke
controlled trials); vasospasm; and vascular access site units and with rapid access to neurosurgical and
neurointensive care facilities. Currently intravenous
complications (including dissection, pseudoaneurysm,
retroperitoneal haematoma and infection). The thrombolysis is typically administered as soon as the
overall procedural complication rate from recent diagnosis of ischaemic stroke is made, if the patient is
within 4.5 hours and there are no contraindications.
randomised controlled trials is in the range of 15%,
but it must be emphasised that many do not adversely Evidence for ‘primary’ thrombectomy without intra-
affect clinical outcome. Stent retriever detachment65 venous thrombolysis remains limited but there are
66 trials both ongoing and proposed.
is an uncommon complication (about 2%–3% with
first-generation Solitaire FR device, but anecdotally
much lower with latest versions). Anaesthesia
The key strategy to minimise complications is The use of general versus local anaesthesia (conscious
obvious and simple: for thrombectomy to be only sedation) currently varies; each strategy has potential
performed in high-volume centres by trained physi- advantages. General anaesthesia reduces subject
cians competent in intracranial endovascular distress and movement, and it can make the technical
procedures and undertaking them regularly to main- aspects easier; on the other hand, conscious sedation
tain skills, as recommended by various allows continuous neurological monitoring for
multidisciplinary guidelines.51 52 Mechanical throm- complications, and it avoids any potential hazard of
bectomy should only be performed by a general anaesthetic agents. Retrospective data
multidisciplinary team operating within comparing general with local anaesthesia during the

Figure 3 Plain CT scan of head (a) and prethrombectomy (b) and post-thrombectomy (c) CT angiograms in a 49-year-old woman
with sudden onset left hemiparesis and confusion. Plain CT scan of head shows hyperdense clot in the right middle cerebral artery
(red arrow) and early perisylvian loss of grey–white matter differentiation. Prethrombectomy CT angiogram shows occluded right
proximal middle cerebral artery (blue arrow). The catheter is visible passing through the occlusion. Post-procedure imaging shows
good filling of all middle cerebral artery branches (yellow arrow). There was complete resolution of neurological signs and symptoms
following aspiration thrombectomy.

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 7

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

Figure 4 Plain CT scan of head (a) and prethrombectomy (b) and post-thrombectomy (c,e, f) digital subtraction angiograms in a 58-
year-old man with a short history of visual symptoms and vertigo followed by a rapid drop in conscious level. Plain CT scan of head (a)
shows thrombus in the basilar artery (red arrow) with complex plaque at the vertebral artery origin, confirmed on digital subtraction
angiography (b). Following successful thrombectomy (c), with removal of a large cast of thrombus (d) by aspiration, a stent was
deployed across the unstable stenotic plaque at the vertebral artery origin (blue arrows) (e and f). Basilar thrombi can often be
removed in bulk like this, possibly because of their physical composition.

procedure found that general anaesthesia, often asso- dissection, cardiac ischaemia, hypertensive encepha-
ciated with systolic blood pressure<140 mm Hg, was lopathy or pulmonary oedema).69 Following
associated with a poor functional outcome (mRS >2) thrombectomy, medical and nursing teams are often
at 90 days.67 However, a recent single-centre rando- uncertain how to manage blood pressure. However,
mised trial in 150 patients with acute ischaemic stroke there is limited evidence to guide how blood pressure
found similar early (24 hours) outcomes (measured as should be managed before, during and after throm-
NIHSS change) from general anaesthesia with intuba- bectomy. It has been suggested that the poorer
tion or conscious sedation without intubation during functional outcome (mRS>2) at 90 days associated
thrombectomy.68 Moreover, two studies presented at with general anaesthesia might relate to the generally
the 3rd European Stroke Organisation Conference lower blood pressure (usually <140 mm Hg systolic),
(ESOC) in 2017 (GOLIATH and ANSTROKE) both but this study could not account for confounding
suggested that general anesthaesia and conscious factors.68 Given the lack of evidence, we currently
sedation are equally safe. Thus, either approach recommend maintaining blood pressure within a
currently seems reasonable, and the choice should be physiological range (typically 110–160 mm Hg
guided by careful consideration of each individual systolic) in a high dependency (eg, neurocritical care)
patient (eg, agitation, neurological or haemodynamic setting following thrombectomy. Specific situations
stability, ease of vascular access to the target lesion, (eg, critical extracranial or intracranial stenosis with
etc). Ongoing trials of general anaesthesia versus haemodynamic failure or post-procedure intracranial
conscious sedation should provide a clearer answer. bleeding) may require different blood pressure targets.

Blood pressure Antithrombotic treatments

Based on several neutral randomised trials of blood There is little evidence on optimum antithrombotic
pressure lowering, guidelines suggest that lowering treatment during and after thrombectomy. Urgent
blood pressure in acute ischaemic stroke should be anticoagulation is not generally recommended in
postponed, at least for a day or two, unless it is acute ischaemic stroke due to the risk of intracranial
severely elevated (>220/120 mm Hg, or >200/ haemorrhage. Aspirin is not recommended within
100 mm Hg with acute kidney injury, aortic 24 hours of IV r-tPA but should be started orally (or

8 Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

Figure 5 Plain CT scan of head (a) and prethrombectomy (b, c), during thrombectomy (d, e, f) and post-thrombectomy (g, h) digital
subtraction angiogram images in a 61-year-old man who presented with a 10 min seizure, followed by left-sided weakness and
neglect. Plain CT scan of head shows hyperdense thrombus in the right middle cerebral artery (red arrow) (a) with angiogram
identifying a critical stenosis of the internal carotid artery origin (blue arrow) (b). We performed middle cerebral artery thrombectomy
using stent retriever technique (e and f). An internal carotid artery stent was inserted (green arrow, d) complicated by an iatrogenic
dissection (yellow arrow,e) necessitating stenting (purple arrow, h).

via nasogastric tube) within 24–48 hours after stroke generally switch to a single antiplatelet agent for long-
onset. Randomised trials and registries do not give term secondary prevention.
consistent data or recommendations regarding antith- In the hyperacute clinical setting of mechanical
rombotic use in mechanical thrombectomy. Some thrombectomy, it is easy to forget the important task,
units give a single procedural dose of heparin, as in all patients with acute stroke, of working out the
but they avoid antiplatelet medication or further anti- likely causative processes and mechanism(s) to opti-
coagulation for 24 hours from stroke symptom onset, mise preventive treatment. For example, this might
and they suggest follow-up brain imaging with CT or involve specific investigations for arterial dissection,
MRI to exclude haemorrhagic complications, but detailed cardiac structure and rhythm evaluation and
practice varies. If we do not deploy a stent, after investigations for thrombophilias or systemic disease.
24 hours and satisfactory clinical progress and follow-
up imaging to exclude significant haemorrhage, we
then give aspirin 300 mg for up to 2 weeks, followed CHALLENGES AND THE FUTURE
by long-term secondary prevention. This will depend Thrombectomy with reperfusion 6 hours after
on stroke mechanism: usually clopidogrel or aspirin symptom onset, alongside IV r-tPA, is clearly the new
for non-cardioembolic and oral anticoagulation for standard of care for the treatment of acute ischaemic
atrial fibrillation or other cardioembolic sources. If we stroke due to large vessel occlusion in the anterior
do deploy a stent, we recommend acutely starting circulation. The UK National Institute for Health and
treatment with aspirin and clopidogrel (or equivalent) Care Excellence (NICE) now approves the use of
dual antiplatelet therapy for at least 3–6 months. For mechanical thrombectomy in stroke.70 However,
stents in patients requiring anticoagulation we many challenges remain, including, crucially, its prac-
tical implementation.

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 9

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

Can mechanical thrombectomy be delivered in the UK? identified, there should be no delay in transferring
The PISTE trial (run in 10 English neuroscience them to an appropriate centre; however, administra-
centres) data confirm the generalisability of the tion of IV r-tPA should not be delayed, since this is
compelling earlier randomised trial results to UK still the cornerstone of initial treatment (and given in
neuroscience centres, confirming that mechanical around 90% of patients in the nine recently published
thrombectomy can be safely and effectively delivered clinical trials of thrombectomy).73 We need meticu-
within the NHS.42 At the time of writing, only lous organisation and robust, well-audited care
two UK centres offer 24 hour endovascular clot pathways to enable safe and rapid transfer. The two
retrieval, with most others providing the service only potential models for providing thrombectomy can be
during working hours. Modelling work based on described as ‘drip and ship’ (initial transfer to a local
SSANP data, randomised controlled trials and other stroke centre for diagnosis and intravenous thrombol-
high-quality evidence indicates that, based on the ysis, followed by rapid transfer to a specialist
current criteria and guidelines listed above, around thrombectomy centre) and ‘mothership’ (transfer
10% of all stroke admissions in the UK (around 9500 immediately to a specialist comprehensive stroke
patients) would be eligible for thrombectomy annu- centre able to undertake thrombectomy and other
ally. The great bulk of those would come from required neuroscience support services). The
patients presenting to hospital within 4.5 hours of optimum model will vary according to local geog-
stroke onset. However, ongoing randomised trials raphy including population density, transport
may well expand those eligibility criteria further over infrastructure and distance from specialist centres able
the next 3–5 years, for example, for strokes in the 6– to deliver the treatment safely and effectively. ‘Drip
12 hour window, in stroke of unknown time onset and ship’ might be the more appropriate solution for
and in mild strokes (NIHSS<6). more remote areas, while a ‘mothership’ model might
be a good solution for urban city populations.

What are the challenges in delivering thrombectomy as

standard clinical practice in the UK? How (and by whom) should thrombectomy be delivered?
Rapid access to appropriate imaging (non-contrast CT Endovascular clot retrieval is safe, but only in
scan of head and CT angiogram at a minimum) is experienced, appropriately trained, competent
mandatory in selecting patients appropriate for endo- hands. It is a complex procedure requiring an
vascular treatment. Although this has been recently experienced team to deliver, and it needs to be
recommended in UK national guidelines,71 it is not performed with great rapidity. Therefore, throm-
yet standard care in all acute stroke centres, and so bectomy delivery will need to be centralised so
routine availability of CT angiography for acute that centres and teams can develop expertise
stroke needs to be rapidly increased. Making CT angi- quickly and maintain 24/7 services robustly and so
ography a routine acute stroke investigation can bring that neurointerventionists can undertake a sufficient
major gains in speeding up diagnostic pathways, as number of cases to maintain expertise. In the UK,
stroke teams and radiographers develop expertise and published guidance on training and competencies
familiarity with the processes. Although there have for thrombectomy is helpful here, and it
been attempts to perform CT angiography only in indicates that a caseload of at least 40 intracranial
more severe strokes (who are more likely to have a endovascular interventions per year is required to
large vessel occlusion), patients with milder strokes maintain competency in neurointervention.74 We
commonly have large vessel occlusion (about 10% of need new ways of thinking about care delivery
those with NIHSS <6) with a high risk of clinical both before and after completion of training to
worsening.72 Once a potentially treatable large vessel expand the pool of skilled neurointerventionists.
occlusion is identified (which requires 24/7 access to Although there has been considerable interest in
trained neuroradiologists or stroke physicians), treat- whether thrombectomy could be safely delivered
ment must then be delivered quickly. In the positive by other (noninterventional neuroradiology) special-
trials detailed above, median time between symptom ties, the skills required to open cerebral arteries
onset to femoral artery puncture was less than quickly, safely and effectively might not be generic
4 hours; median time from symptom onset to recanali- across these other specialties; for example,
sation was 41/3 hours. This timeframe is currently a although coronary and cerebral arteries are of
challenge in certain parts of the UK, particularly similar calibre, brain arteries are more delicate
outside standard working hours, and will require with a thinner tunica media and adventitia, often
innovation in local imaging acquisition and interpreta- with proximal ectasia and tortuosity, making navi-
tion, as well as emergency transport services. Once a gation with a catheter both hazardous and
patient eligible for endovascular clot retrieval is challenging. Moreover, dealing with sudden

10 Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

neurological complications requires great skill and need randomised trials of thrombectomy, but it
expertise in navigating the complex cerebral vascu- might be challenging because of lack of clinical
lature. Indeed, the best results in clinical trials and equipoise (given the clear benefits of thrombec-
clinical practice were achieved by experienced tomy in the anterior circulation and the
neurointerventionists in high-volume centres.75 We devastating outcome from untreated basilar throm-
need international mechanical thrombectomy regis- bosis); one trial of treatment within 6 hours is
tries to identify whether the real-world experience underway.6 83
is commensurate with that seen in the positive The optimum form of imaging acquisition and proc-
clinical trials; SITS thrombectomy is one such essing (to determine the extent of ischaemic core,
registry. potentially salvageable tissue, collateral supply, etc.)
requires further study. Is MRI better than CT? Is
Is mechanical thrombectomy cost effective? perfusion imaging required, or will collateral assess-
One study76 modelled the hyperacute management ment and APSECTS suffice? This is a very important
of stroke using intravenous thrombolysis and question for thrombectomy implementation as
mechanical thrombectomy in the UK (compared imaging triage is likely to be critical in the ‘drip and
with intravenous thrombolysis alone) using Markov ship’ service model in particular. For many logistic
simulations of estimated lifetime costs and quality- reasons, it would be preferable to secondarily transfer
adjusted life years (QALYs), based on pooled for thrombectomy only those patients who are very
outcome data of five randomised controlled trials. likely to benefit. We also need to know whether all
This study found an incremental cost per QALY patients need advanced brain imaging or just some,
gained of mechanical thrombectomy over a 20 year and if some, who? The PISTEai (advanced imaging)
period of $11 651 (£7061). A more recent study trial is proposed in the UK to answer some of these
that modelled the intervention in a US setting questions.
found an incremental cost-effectiveness ratio for Another critical question is whether we should use
endovascular treatment (compared with standard advanced imaging in more delayed (including ‘wake-
care) of $3110/QALY (about £2500 per QALY) in up’ stroke) presentations. Trials are ongoing, including
all simulations, although cost effectiveness was POSITIVE (6–12 hour time window with appropriate
lower in more distal (M2) occlusion and with estab-
image selection). The DAWN trial (6–24 hour time
lished ischaemic injury (ASPECTS score 5). Both
window, including wake-up stroke) selected patients
of these studies show that the cost of mechanical
with substantial clinical deficit but a small ischaemic
thrombectomy is well below the frequently applied
£30 000 per QALY threshold used by NICE to eval-
Key points
uate new treatments.
" Thrombectomy for anterior circulation stroke due to proven
Future research questions proximal major vessel (carotid or M1) occlusion within 6
There are many remaining questions regarding hours of stroke onset is safe and highly effective, and it sets
the new standard of care
thrombectomy. We have few data on thrombec- " In a meta-analysis of randomised trials, the proportions of
tomy for basilar artery thrombosis; some registry patients achieving a good (independent) functional outcome
data suggest that a high proportion of patients (mRS 0–2 at 90 days) were 46.0% (mechanical thrombec-
(68%) have a poor outcome (mRS>3), with no tomy) vs 26.5% (best medical treatment); most patients also
difference according to the use of intravenous received intravenous thrombolysis
thrombolysis or mechanical thrombectomy.77 As in " Favourable outcome from mechanical thrombectomy is
the anterior circulation, recanalisation is a key strongly time dependent (’time is brain’), with the best
prognostic factor: a recent meta-analysis of 45 results achieved when there is no evidence of extensive early
ischaemic brain injury (e.g. ASPECTS score >5); if good
observational studies (n=2056) of reperfusion
recanalisation is achieved within 4.5 hours, the absolute rate
versus no reperfusion of acute basilar occlusion of good functional outcome is 61%
showed a number needed to treat to decrease " Complications of endovascular procedures can follow device-
death or dependency of 3.78 Small single-centre related vessel injury (perforation, dissection, subarachnoid
studies reported good functional outcomes haemorrhage), vascular access or radiological contrast media
following basilar thrombectomy, ranging from 30% " Thrombectomy must be delivered by appropriately trained
to 48%.79–82 The time window may be longer for interventionists
basilar thrombosis (possibly up to 12–24 hours), " The next challenge is in delivering the treatment across
perhaps relating to the tissue properties, clot healthcare systems; the optimal solution (eg, ’drip and ship’
versus ’mothership’) may differ according to geography and
composition and haemodynamics of the collateral
population density
vascular supply in the posterior circulation. We

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 11

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

core on CT-perfusion imaging and randomised them adapt and build upon this work, for commercial use, provided the
to mechanical trhombectomy with the Trevo device, original work is properly cited. See: http://creativecommons.org/
licenses/by/4.0/
or to medical therapy alone. The DAWN trial was
© Article author(s) (or their employer(s) unless otherwise stated in
stopped early on 9/3/2017 after ~200/500 patients
the text of the article) 2017. All rights reserved. No commercial use
had been recruited. Data from the DAWN trial is permitted unless otherwise expressly granted.
presented at the European Stroke Organisation
Conference (ESOC) in May 2017 indicated that at 90 REFERENCE
days, 48.6% of patients in the intervention arm 1 Feigin VL, Forouzanfar MH, Krishnamurthi R, et al. Global and
achieved functional independence at , compared to regional burden of stroke during 1990–2010: findings from the
13.1% in the control medical therapy arm. Evidence Global Burden of Disease Study 2010. Lancet
of an extended time window for mechanical throm- 2014;383:245–55.
2 Wang Y, Rudd AG, Wolfe CD. Age and ethnic disparities in
bectomy potentially means that more stroke patients
incidence of stroke over time: the South London Stroke
might be eligible for endovascular
Register. Stroke 2013;44:3298–304.
treatment. However, these data should not detract
3 Mechanical clot retrieval for treating acute ischaemic stroke.
from the key message that the most rapid treatment National Institute for Health and Care Excellence (NICE).
possible remains the key aim to optimise outcomes Available: https://www.nice.org.uk/guidance/ipg548.
for all reperfusion therapies in acute stroke. 4 Health and Social Care Information Centre. Quality and
There are also trials proposed of thrombectomy in Outcomes Framework (QOF)—2014-15. 2015, Available:
patients with milder stroke. The role of direct throm- http://bit.ly/1PhPYmL.
bectomy (without intravenous thrombolysis) remains 5 NHS Wales Informatics Service Information and Statistics.
to be defined in randomised controlled trials. Hospital Admissions for Stroke. 2015.
Many of these pressing remaining questions about 6 Department of Health, Social Services and Public Safety. QOF
mechanical thrombectomy will probably be Achievement Data. 2014. Available: http://www.dhsspsni.gov.
uk/index/statistics/qof/qof- achievement.htm.
answered within the next 3–5 years. All patients
7 World Health Organisation. The top 10 causes of death. 2014.
undergoing mechanical thrombectomy should be
Available: http://www.who.int/mediacentre/factsheets/fs310/en/.
prospectively included in registries to obtain further 8 Adamson J, Beswick A, Ebrahim S. Is stroke the most common
evidence on effectiveness and safety in ‘real world’ cause of disability? J Stroke Cerebrovasc Dis 2004;13:171–7.
practice. 9 Luengo-Fernandez R, Gray AM, Bull L, et al. Quality of life
after TIA and stroke: ten-year results of the Oxford Vascular
Study. Neurology 2013;81:1588–95.
CONCLUSION 10 Saver JL. Time is brain—quantified. Stroke 2006;37:263–6.
Mechanical thrombectomy is a highly successful, 11 Khatri P, Yeatts SD, Mazighi M, et al. Time to angiographic
safe and cost-effective treatment for patients with reperfusion and clinical outcome after acute ischaemic stroke:
large artery occlusive stroke. It is therefore a ‘no an analysis of data from the Interventional Management of
brainer’ that the UK NHS and other healthcare Stroke (IMS III) phase 3 trial. Lancet Neurol 2014;13:567–74.
systems need to deliver it as soon as practicable. 12 del Zoppo GJ, Poeck K, Pessin MS, et al. Recombinant tissue
However, that will inevitably require reorganisation plasminogen activator in acute thrombotic and embolic stroke.
of UK stroke services and that will require some Ann Neurol 1992;32:78–86.
13 Emberson J, Lees KR, Lyden P, et al. Effect of treatment delay,
investment, great attention to care pathways and
age, and stroke severity on the effects of intravenous
extensive cooperation between services including
thrombolysis with alteplase for acute ischaemic stroke: a meta-
ambulance services and hospitals. analysis of individual patient data from randomised trials.
Lancet 2014;384:1929–35.
Contributors ME, PC, PW and DW wrote the manuscript and are
14 Alteplase for treating acute ischaemic stroke. London (UK):
responsible for the content.
National Institute for Health and Care Excellence (NICE).
Funding DW receives research funding support from the British 2012. Available: https://www.guideline.gov/summaries/
Heart Foundation and the Stroke Association (TSA BHF 2009/01;
summary/38443/alteplase-for-treating-acute-ischaemic-stroke-
PPA 2015/04). This work was undertaken at University College
London Hospitals NHS Foundation Trust/University College review-of-technology-appraisal-guidance-122.
London who received a proportion of funding from the 15 Hacke W, Kaste M, Bluhmki E, et al. Thrombolysis with
Department of Health’s National Institute for Health Research alteplase 3 to 4.5 hours after acute ischemic stroke. N Engl J
Biomedical Research Centers funding scheme. Med 2008;359:1317–29.
Competing interests None declared. 16 Bhatia R, Hill MD, Shobha N, et al. Low rates of acute
Provenance and peer review Commissioned; externally peer recanalization with intravenous recombinant tissue plasminogen
reviewed. This paper was reviewed by William Whiteley, Edinburgh, activator in ischemic stroke: real-world experience and a call for
UK. action. Stroke 2010;41:2254–8.
Open Access This is an Open Access article distributed in 17 Fischer U, Arnold M, Nedeltchev K, et al. NIHSS score and
accordance with the terms of the Creative Commons Attribution arteriographic findings in acute ischemic stroke. Stroke
(CC BY 4.0) license, which permits others to distribute, remix, 2005;36:2121–5.

12 Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

18 Mishra SM, Dykeman J, Sajobi TT, et al. Early reperfusion rates 35 Berkhemer OA, Fransen PS, Beumer D, et al. A randomized
with IV tPA are determined by CTA clot characteristics. AJNR trial of intraarterial treatment for acute ischemic stroke. N Engl
Am J Neuroradiol 2014;35:2265–72. J Med 2015;372:11–20.
19 Elofuke P, Reid JM, Rana A, et al. Disappearance of the 36 Jovin TG, Chamorro A, Cobo E, et al. Thrombectomy within 8
hyperdense MCA sign after stroke thrombolysis: implications hours after symptom onset in ischemic stroke. N Engl J Med
for prognosis and early patient selection for clot retrieval. J R 2015;372:2296–306.
Coll Physicians Edinb 2016;46:81–6. 37 Campbell BC, Mitchell PJ, Kleinig TJ, et al. Endovascular
20 Kamalian S, Morais LT, Pomerantz SR, et al. Clot length therapy for ischemic stroke with perfusion-imaging selection. N
distribution and predictors in anterior circulation stroke: Engl J Med 2015;372:1009–18.
implications for intra-arterial therapy. Stroke 2013;44:3553–6. 38 Saver JL, Goyal M, Bonafe A, et al. Stent-retriever
21 Riedel CH, Jensen U, Rohr A, et al. Assessment of thrombus in thrombectomy after intravenous t-PA vs. t-PA alone in stroke. N
acute middle cerebral artery occlusion using thin-slice Engl J Med 2015;372:2285–95.
nonenhanced Computed Tomography reconstructions. Stroke 39 Goyal M, Demchuk AM, Menon BK, et al. Randomized
2010;41:1659–64. assessment of rapid endovascular treatment of ischemic stroke.
22 Riedel CH, Zimmermann P, Jensen-Kondering U, et al. The N Engl J Med 2015;372:1019–30.
importance of size: successful recanalization by intravenous 40 Bracard S, Ducrocq X, Mas JL, et al. Mechanical thrombectomy
thrombolysis in acute anterior stroke depends on thrombus after intravenous alteplase versus alteplase alone after stroke
length. Stroke 2011;42:1775–7. (THRACE): a randomised controlled trial. Lancet Neurol
23 Hirano T, Sasaki M, Mori E, et al. Residual vessel length on 2016;15:1138–47.
magnetic resonance angiography identifies poor responders to 41 Mocco J, Zaidat OO, von Kummer R, et al. Aspiration
alteplase in acute middle cerebral artery occlusion patients: thrombectomy after intravenous alteplase versus intravenous
exploratory analysis of the Japan Alteplase Clinical Trial II. alteplase alone. Stroke 2016;47:2331–8.
Stroke 2010;41:2828–33. 42 Muir KW, Ford GA, Messow CM, et al. Endovascular therapy
24 Saarinen JT, Sillanp€a€a N, Rusanen H, et al. The mid-M1 for acute ischaemic stroke: the Pragmatic Ischaemic Stroke
segment of the middle cerebral artery is a cutoff clot location Thrombectomy Evaluation (PISTE) randomised, controlled
for good outcome in intravenous thrombolysis. Eur J Neurol trial. J Neurol Neurosurg Psychiatry 2017;88:38–44.
2012;19:1121–7. 43 Khoury NN, Darsaut TE, Ghostine J, et al. Endovascular
25 Furlan A, Higashida R, Wechsler L, et al. Intra-arterial thrombectomy and medical therapy versus medical therapy
prourokinase for acute ischemic stroke. The PROACT II study: alone in acute stroke: a randomized care trial. J Neuroradiol
a randomized controlled trial. Prolyse in Acute Cerebral 2017: S01509861pii):30120–1.
Thromboembolism. JAMA 1999;282:2003–11. 44 Goyal M, Menon BK, van Zwam WH, et al. Endovascular
26 Felten RP, Ogden NR, Peña C, et al. The Food and Drug thrombectomy after large-vessel ischaemic stroke: a meta-
Administration medical device review process: clearance of a analysis of individual patient data from five randomised trials.
clot retriever for use in ischemic stroke. Stroke 2005;36:404–6. Lancet 2016;387:1723–31.
27 Smith WS, Sung G, Starkman S, et al. Safety and efficacy of 45 Badhiwala JH, Nassiri F, Alhazzani W, et al. Endovascular
mechanical embolectomy in acute ischemic stroke: results of the thrombectomy for acute ischemic stroke: a meta-analysis. JAMA
MERCI trial. Stroke 2005;36:1432–8. 2015;314:1832–43.
28 Smith WS, Sung G, Saver J, et al. Mechanical thrombectomy 46 Bush CK, Kurimella D, Cross LJ, et al. Endovascular treatment
for acute ischemic stroke: final results of the Multi MERCI with stent-retriever devices for acute ischemic stroke: a meta-
trial. Stroke 2008;39:1205–12. analysis of randomized controlled trials. PLoS One
29 Broderick JP, Palesch YY, Demchuk AM, et al. Endovascular 2016;11:e0147287.
therapy after intravenous t-PA versus t-PA alone for stroke. N 47 Rodrigues FB, Neves JB, Caldeira D, et al. Endovascular
Engl J Med 2013;368:893–903. treatment versus medical care alone for ischaemic stroke:
30 Ciccone A, Valvassori L, Nichelatti M, et al. Endovascular systematic review and meta-analysis. BMJ 2016;353:i1754.
treatment for acute ischemic stroke. N Engl J Med 48 Powers WJ, Derdeyn CP, Biller J, et al. 2015 American Heart
2013;368:904–13. Association/American Stroke Association Focused Update of the
31 Kidwell CS, Jahan R. Endovascular treatment of acute ischemic 2013 Guidelines for the Early Management of Patients With
stroke. Neurol Clin 2015;33:401–20.
Acute Ischemic Stroke Regarding Endovascular Treatment: A
32 Demchuk AM, Goyal M, Yeatts SD, et al. Recanalization and
Guideline for Healthcare Professionals From the American
clinical outcome of occlusion sites at baseline CT angiography in
Heart Association/American Stroke Association. Stroke
the Interventional Management of Stroke III trial. Radiology
2015;46:3020–35.
2014;273:202–10.
49 Casaubon LK, Boulanger JM, Glasser E, Blacquiere D, et al.
33 Saver JL, Jahan R, Levy EI, et al. Solitaire flow restoration
Canadian Stroke Best Practice Recommendations: Acute
device versus the Merci Retriever in patients with acute
Inpatient Stroke Care Guidelines, Update 2015. Int J Stroke
ischaemic stroke (SWIFT): a randomised, parallel-group, non-
2016;11:239–52.
inferiority trial. Lancet 2012;380:1241–9.
50 Wahlgren N, Moreira T, Michel P, et al. Mechanical
34 Nogueira RG, Lutsep HL, Gupta R, et al. Trevo versus Merci
thrombectomy in acute ischemic stroke: consensus statement by
retrievers for thrombectomy revascularisation of large vessel
ESO-Karolinska Stroke Update 2014/2015, supported by ESO,
occlusions in acute ischaemic stroke (TREVO 2): a randomised
ESMINT, ESNR and EAN. Int J Stroke 2016;11:134–47.
trial. Lancet 2012;380:1231–40.

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 13

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

51 White PM, Bhalla A, Dinsmore J, et al. Standards for providing 68 Sch€onenberger S, Uhlmann L, Hacke W, et al. Effect of
safe acute ischaemic stroke thrombectomy services (September conscious sedation vs general anesthesia on early neurological
2015). Clin Radiol. The Royal College of Radiologists 2015;72, improvement among patients with ischemic stroke undergoing
e1-175–e9. endovascular thrombectomy: a randomized clinical trial. JAMA
52 Intercollegiate Stroke Working Party. National Clinical 2016;316:1986–96.
Guideline for Stroke. Royal College of Physicians 2012; 232. 69 Jauch EC, Saver JL, Adams HP, et al. Guidelines for the early
Available: http://www.rcplondon.ac.uk/sites/default/files/ management of patients with acute ischemic stroke: a guideline
national-clinical-guidelines-for-stroke-fourth-edition.pdf. for healthcare professionals from the American Heart
53 Demchuk AM, Coutts SB. Alberta Stroke Program Early CT Association/American Stroke Association. Stroke
Score in acute stroke triage. Neuroimaging Clin N Am 2013;44:870–947.
2005;15:409–19. 70 ISD Scotland. Quality and Outcomes Framework. 2014.
54 Saver JL, Goyal M, van der Lugt A, et al. Time to treatment Available: http://www.isdscotland. org/health-Topics/General-
with endovascular thrombectomy and outcomes from ischemic Practice/Quality-And-Outcomes- Framework/.
stroke: a meta-analysis. JAMA 2016;316:1279–88. 71 Royal College of Physicians. National clinical guideline for
55 Kappelhof M, Marquering HA, Berkhemer OA, et al. Intra- stroke, Fifth Edition, 2016. Available: https://www.strokeaudit.
arterial treatment of patients with acute ischemic stroke and org/SupportFiles/Documents/Guidelines/2016-National-Clinical-
internal carotid artery occlusion: a literature review. J Guideline-for-Stroke-5t-
Neurointerv Surg 2015;7:8–15. 72 Coutts SB, Modi J, Patel SK, et al. CT/CT angiography and
56 Cohen JE, Gomori JM, Rajz G, et al. Extracranial carotid artery MRI findings predict recurrent stroke after transient ischemic
stenting followed by intracranial stent-based thrombectomy for attack and minor stroke: results of the prospective CATCH
acute tandem occlusive disease. J Neurointerv Surg study. Stroke 2012;43:1013–7.
2015;7:412–7. 73 Meurer WJ, Barth BE, Gaddis G, et al. Rapid systematic review:
57 Stampfl S, Ringleb PA, M€ ohlenbruch M, et al. Emergency intra-arterial thrombectomy ("clot retrieval") for selected
cervical internal carotid artery stenting in combination with patients with acute ischemic stroke. J Emerg Med
intracranial thrombectomy in acute stroke. AJNR Am J 2017;52:255–61.
Neuroradiol 2014;35:741–6. 74 Lenthall R, McConachie N, White P, et al. BSNR training
58 Maurer CJ, Joachimski F, Berlis A. Two in one: endovascular guidance for mechanical thrombectomy. Clin Radiol 2017;72,
treatment of acute tandem occlusions in the anterior circulation. e11–175, e18.
Clin Neuroradiol 2015;25:397–402. 75 Gupta R, Horev A, Nguyen T, et al. Higher volume
59 Lescher S, Czeppan K, Porto L, et al. Acute stroke and endovascular stroke centers have faster times to treatment,
obstruction of the extracranial carotid artery combined with higher reperfusion rates and higher rates of good clinical
intracranial tandem occlusion: results of interventional outcomes. J Neurointerv Surg 2013;5:294–7.
revascularization. Cardiovasc Intervent Radiol 2015;38:304–13. 76 Ganesalingam J, Pizzo E, Morris S, et al. Cost-utility analysis of
60 Dababneh H, Bashir A, Hussain M, et al. Endovascular mechanical thrombectomy using stent retrievers in acute
treatment of tandem internal carotid and middle cerebral artery ischemic stroke. Stroke 2015;46:2591–8.
occlusions. J Vasc Interv Neurol 2014;7:26–31. 77 Schonewille WJ, Wijman CA, Michel P, et al. Treatment and
61 Campbell BC, Hill MD, Rubiera M, et al. Safety and Efficacy of outcomes of acute basilar artery occlusion in the Basilar Artery
Solitaire stent thrombectomy: individual patient data meta- International Cooperation Study (BASICS): a prospective
analysis of randomized trials. Stroke 2016;47:798–806. registry study. Lancet Neurol 2009;8:724–30.
62 Gascou G, Lobotesis K, Machi P, et al. Stent retrievers in acute 78 Kumar G, Shahripour RB, Alexandrov AV. Recanalization of
ischemic stroke: complications and failures during the acute basilar artery occlusion improves outcomes: a meta-
perioperative period. AJNR Am J Neuroradiol 2014;35:734–40. analysis. J Neurointerv Surg 2015;7:868–74.
63 Leishangthem L, Satti SR. Vessel perforation during withdrawal 79 Nagel S, Kellert L, M€ ohlenbruch M, et al. Improved clinical
of Trevo ProVue stent retriever during mechanical outcome after acute basilar artery occlusion since the
thrombectomy for acute ischemic stroke. J Neurosurg introduction of endovascular thrombectomy devices.
2014;121:995–8. Cerebrovasc Dis 2013;36(5-6):394–400.
64 Mokin M, Fargen KM, Primiani CT, et al. Vessel perforation 80 M€ ohlenbruch M, Stampfl S, Behrens L, et al. Mechanical
during stent retriever thrombectomy for acute ischemic stroke: thrombectomy with stent retrievers in acute basilar artery
occlusion. AJNR Am J Neuroradiol 2014;35:959–64.
technical details and clinical outcomes. J Neurointerv Surg
81 Baek JM, Yoon W, Kim SK, et al. Acute basilar artery occlusion:
2016:neurintsurg-2016-012707.
outcome of mechanical thrombectomy with Solitaire stent
65 Kim ST, Jin SC, Jeong HW, et al. Unexpected detachment of
within 8 hours of stroke onset. AJNR Am J Neuroradiol
Solitaire stents during mechanical thrombectomy. J Korean
2014;35:989–93.
Neurosurg Soc 2014;56:463–8.
82 Mourand I, Machi P, Nogue E, et al. Diffusion-weighted
66 Akpinar S, Yilmaz G. Spontaneous SolitaireÉ AB thrombectomy
imaging score of the brain stem: A predictor of outcome in
stent detachment during stroke treatment. Cardiovasc Intervent
acute basilar artery occlusion treated with the Solitaire FR
Radiol 2015;38:475–8.
device. AJNR Am J Neuroradiol 2014;35:1117–23.
67 Davis MJ, Menon BK, Baghirzada LB, et al. Anesthetic
83 van der Hoeven EJ, Schonewille WJ, Vos JA, et al. The Basilar
management and outcome in patients during endovascular
Artery International Cooperation Study (BASICS): study
therapy for acute stroke. Anesthesiology 2012;116:396–405.
protocol for a randomised controlled trial. Trials 2013;14:200.

14 Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

REVIEW

84 Duffy S, Farrell M, McArdle K, et al. Novel methodology to

replicate clot analogs with diverse composition in acute
ischemic stroke. J Neurointerv Surg 2017;9:486–91.

Evans M, et al. Pract Neurol 2017;0:1–14. doi:10.1136/practneurol-2017-001685 15

 Downloaded from http://pn.bmj.com/ on June 26, 2017 - Published by group.bmj.com

Revolution in acute ischaemic stroke care: a practical

guide to mechanical thrombectomy
Matthew Evans, Phil White, Peter Cowley and David Werring

Pract Neurol published online June 24, 2017

Updated information and services can be found at:

http://pn.bmj.com/content/early/2017/06/24/practneurol-2017-001685

These include:

References This article cites 74 articles, 30 of which you can access for free at:
http://pn.bmj.com/content/early/2017/06/24/practneurol-2017-001685#BIBL
Open Access This is an Open Access article distributed in accordance with the terms of the Creative
Commons Attribution (CC BY 4.0) license, which permits others to distribute, remix,
adapt and build upon this work, for commercial use, provided the original work is
properly cited. See: http://creativecommons.org/licenses/by/4.0/
Email alerting Receive free email alerts when new articles cite this article. Sign up in the box at the top
service right corner of the online article.

Topic Articles on similar topics can be found in the following collections

Collections Open access (30)

Notes

To request permissions go to:

http://group.bmj.com/group/rights-licensing/permissions

To order reprints go to:

http://journals.bmj.com/cgi/reprintform

To subscribe to BMJ go to:

http://group.bmj.com/subscribe/