These guidelines are the result of a joint effort from writing groups of the Brazilian
Stroke Society, the Scientific Department of Cerebrovascular Diseases of the Brazilian
Academy of Neurology, the Brazilian Stroke Network and the Brazilian Society of Diagnostic
and Therapeutic Neuroradiology. Those societies have been involved in treatment opinions
and educational projects related to cerebrovascular diseases in Brazil. Members from
the groups above participated in web-based discussion forums with pre-defined themes,
followed by videoconference meetings in which controversies and position statements
were discussed, leading to a consensus. This guidelines focuses on the implications
of the recent clinical trials on endovascular therapy for acute ischemic stroke (AIS)
due to proximal arterial occlusions (PAO), and the final text aims to guide specialists
and non-specialists in stroke care in managing patients with this condition.
The final recommendations are classified based on recommendation grade and evidence
level. Recommendation grades are divided in three main categories, based on the level
of certainty that the treatment is beneficial – certainly beneficial, of uncertain
benefit, and certainly not beneficial or even harmful – and its entailing proposition
for clinical care. Levels of evidence indicate the quality of the scientific evidence
that supports the recommendation. High levels of recommendation refer to results from
good quality randomized clinical trials that have been reproduced, from meta-analysis
of good quality clinical trials, and from a single clinical trial of good quality.
Recommendations from observational studies are classified as weak. We specify a weak
level of evidence for subgroups of patients, which are under-represented in good quality
clinical trials, and which have not been resolved by appropriate subgroup analysis;
if such analysis has been performed, then the recommendation qualifies as high level.
It is worth mentioning that recommendations from guidelines – as well as data from
clinical trials – apply to qualifying groups of patients, but not necessarily to individual
patients. Therefore, the recommendations serve as guidance to bedside decision-making
and do not substitute for patient-centered clinical reasoning.
EVIDENCE LEVELS
A) (high): the recommendation is supported by more than one randomized clinical trial
of good quality, OR by a well-performed meta-analysis.
B) (moderate): the recommendation is supported by only one randomized clinical trial
of good quality.
C1) (weak): the recommendation applies to under-represented subgroups of one or more
randomized clinical trials that have not been adequately resolved in subgroup analysis,
and for which the panel believes that there is sufficient clinical and biological
reasoning to support a recommendation.
C2) (weak): the recommendation is supported by observational or non-randomized studies.
E) (expert opinion): the panel considers that there is sufficient clinical and biological
reasoning to support a recommendation, in spite of the absence of good quality clinical
trials and observational studies.
Endovascular reperfusion therapy
About one third of patients with AIS present with an occlusion of a large, proximal
artery of the brain circulation[1]. Recently, endovascular treatment has been shown to improve functional outcome in
five randomized clinical trials of selected patients with AIS associated with PAO
of the anterior circulation[2],[3],[4],[5],[6],[7]. A summary of core design features and results is presented in the [Table]. These trials had important differences in design, including eligibility criteria,
which poses some relevant issues for interpretation and implementation of practice.
However, as a common denominator, they all conveyed a clear message that early endovascular
treatment with high rates of successful reperfusion leads to better clinical outcome
with very large effect sizes. The number needed to treat (NNT) to achieve functional
independence at three months after the stroke varied from 7.4 in the Multicenter Randomized
Clinical Trial of Endovascular Treatment for Acute Ischemic Stroke in the Netherlands
(MR CLEAN) – a less selective, pragmatical trial – to as low as 3.2 in the Extending
the Time for Thrombolysis in Emergency Neurologic Deficits – Intra-arterial (EXTEND-IA),
a penumbral mismatch-based, highly selective trial. The combined analysis of these
trials shows that the NNT for reduced disability of one point on the modified Rankin
Scale was 2.6, and 5.1 for achieving functional independency[8]. Altogether, these trials should bring endovascular treatment–when performed under
selected circumstances – to the standard-of-care as recommended by a high level of
evidence. Moreover, the combined results of these trials suggest that the therapeutic
benefit is dependent on well-organized stroke care, rapid initiation of the endovascular
procedure and high rates of early and successful reperfusion.
Table
Selected features of positive trials on endovascular reperfusion therapy for acute
ischemic stroke (AIS).
|
Variable
|
PAO actual location
|
Age, yo.
|
NIHSS eligibility
|
Neuroimaging eligibility (besides proof of a PAO)
|
IV TPA
|
Time window
|
Stent retrievers
|
|
MR CLEAN
|
IC ICA + M1 27%
M1 64%
M2 8%
Associated EC ICA 29%
|
≥1 8
|
≥ 2
|
None
|
89%
|
6h
|
97%
|
|
EXTEND-IA
|
IC ICA 31%
M1 57%
M2 11%
|
≥ 18
|
None
|
Perfusion mismatch on CTP
Tissue at risk as Tmax > 6s
Infarct core as CBF < 50% normal
Mismatch: mismatch ratio > 1.2 and mismatch volume > 10mL and ischemic core <7 0mL
|
Mandatory
|
6h
|
Solitaire FR mandatory
|
|
ESCAPE
|
ICA + M1 28%
M1/all M2 71%
M2 2%
Associated EC ICA 13%
|
≥ 18
|
> 5
|
ASPECTS ≥ 6 on CT
On CTA or CTP, excluded if:
Poor collaterals on > 50% of MCA territory
OR low CBV/CBF: ASPECTS < 6
OR low CBV/CBF > 1/3 of MCA territory
|
73%
|
12h
|
86%
|
|
SWIFT-PRIME
|
IC ICA 18%
M1 68%
M2 14%
|
18– 80
|
8–29
|
Eligible if ASPECTS ≥ 6 or ischemic signs on < 1/3 of MCA territory on CT or DWI
|
Mandatory
|
6h
|
Solitaire FR mandatory
|
|
REVASCAT
|
IC ICA + M1 25%
M1 65%
Single M2 10%
Associated EC ICA 19%
|
18–80
|
> 5
|
Eligible if ASPECTS ≥ 6 on CT
|
68%
|
8h
|
Solitaire FR mandatory
|
PAO: proximal arterial occlusions; NIHSS: National Institutes of Health Stroke Scale;
IV TPA: intravenous tissue plasminogen activator; MR CLEAN: multicenter randomized
clinical trial of endovascular treatment for acute ischemic stroke in the Netherlands;
EXTEND-IA: extending the time for thrombolysis in emergency neurological deficits
– intra-arterial; ESCAPE: endovascular treatment for small core and anterior circulation
proximal occlusion with emphasis on minimizing CT to recanalization times; SWIFT-PRIME:
solitaire with the Intention for thrombectomy as primary endovascular treatment; REVASCAT:
randomized trial of revascularization with solitaire FR device versus best medical
therapy in the treatment of acute stroke due to anterior circulation large vessel
occlusion presenting within eight hours of symptom onset; IC ICA: intracranial segment
of the internal carotid artery; M1 and M2: first and second segments of the MCA; EC
ICA: extracranial segment of the internal carotid artery; Tmax: time to maximum; CBF:
cerebral blood flow; ASPECTS: Alberta stroke program early CT score; CT: computed
tomography; MCA: middle cerebral artery; CTA: CT angiography; CTP: CT perfusion; CBV:
cerebral blood volume; DWI: diffusion weighted imaging; Solitaire FR: solitaire stent-retriever
device.
These guidelines aim to give recommendations addressing the question of whether endovascular
treatment can improve long-term functional outcome in patients with an AIS caused
by a PAO, when compared to standard clinical treatment. Three main issues arise when
translating these results into guideline recommendations: (1) there were subgroups
of patients that, despite being included in most of the protocols, were significantly
underrepresented in the final population; (2) due to differences in design, there
were subgroups that were not reproduced in more than one study; (3) and these results
reflect organized, optimal care of selected patients treated in specialized centers.
Therefore, this guideline will balance its recommendations favoring core features
shared by them, while noting subgroups that deserve more careful consideration. Fortunately,
two meta-analyses with patient-level data have addressed many of these questions regarding
specific subgroups[8],[9]. We will also emphasize the requirements for stroke care organization and performance
monitoring to improve the implementation of endovascular treatment. Some considerations
regarding eligibility, intervention and implementation are shown below. Future studies
are warranted to answer remaining issues regarding efficacy among subgroups, as well
as real-world clinical effectiveness and cost-effectiveness in different systems of
care. Finally, the efficacy of endovascular treatment has been demonstrated only in
the presence of high-level infrastructures (e.g. last-generation devices, fast clinical
and neuroimaging workflows, highly specialized personnel) and optimized systems of
care, thus its safety, efficacy, and cost-effectiveness in developing countries will
likely require further validation.
In Brazil, endovascular treatment carries some specific implications. Despite being
one of the leading causes of mortality in the country, stroke has been severely neglected,
with very poor stroke awareness in the population and very low rates of treatment
with intravenous thrombolysis[10],[11]. Unfortunately, Brazil is a country of great economic inequalities, with some stroke
units in the private and public hospitals, especially in the wealthier regions of
the country, presenting similar thrombolysis rates to those of tertiary stroke centers
of developed countries[11],[12],[13]. These services routinely perform endovascular treatment for stroke and were able
to develop a triage and quality control system, with some centers even being certified
by international institutions such as the Joint Commission to the Canadian Stroke
Network[14]. In more recent years, improvements in acute stroke care and prevention have led
to a decrease in stroke mortality and stroke incidence[15],[16],[17]. The emergence of endovascular treatment as a standard-of-care comes in the context
of an increasing number of stroke centers[18]. The necessity of making intravenous thrombolysis widely available has been a fundamental
reason for this increase. We believe that the implementation of endovascular treatment
in the Brazilian healthcare system could serve as an additional driving force to further
increase the national stroke network, both in size and in spectrum of complexity[19],[20].
Considerations on patient selection and endovascular intervention
The only core eligibility criterion equally present among all trials was the presence
of a target PAO – defined as modified Thrombolysis in Cerebral Ischemia (TICI) score
of 0-1 – of the anterior circulation assessed by computed tomography angiography (CTA)
or magnetic resonance imaging angiography (MRA) before endovascular treatment was indicated. This represents a major difference from the prior clinical trials on endovascular
treatment, which used clinical criteria, such as the National Institute of Health
Stroke Scale (NIHSS) score, and failed to demonstrate clinical benefit[21],[22],[23]. In the Interventional Management of Stroke III trial (IMS-3), patients were mainly
selected based on an NIHSS ≥10, and one third of them eventually did not have a target
occlusion on catheter angiography. Cut-off values of NIHSS scores with the highest
accuracy for detecting PAO (scores from 10 to 14) still present false-negative rates
>20%, and scores with sufficient sensitivity to detect >90% of patients with PAO (NIHSS
from 1 to 5) would result in sending almost every patient to catheter angiography[1],[24]. Therefore, vessel imaging by CTA or MRA is highly recommended for endovascular
treatment indication[23].
Transcranial Doppler (TCD) has been shown to have good accuracy for detecting PAO
and might be useful in patients with AIS who have contraindications to CTA and MRA[25],[26]. However, this method was not systematically used to select patients in a clinical
trial of endovascular treatment, and its use in the AIS setting demands a high level
of expertise.
Regarding the site of the target occlusion, the majority of patients included in the
recent trials had an M1 segment middle cerebral artery (MCA) occlusion (57–68%), followed
by top-of-carotid occlusions [intracranial internal carotid artery (ICA), with or
without M1 MCA; 18–31%] and M2 segment MCA occlusions (2–14%). The Endovascular Treatment
for Small Core and Anterior Circulation Proximal Occlusion with Emphasis on Minimizing
CT to Recanalization Times (ESCAPE) (13%), the Randomized Trial of Revascularization
with Solitaire FR Device versus Best Medical Therapy in the Treatment of Acute Stroke
Due to Anterior Circulation Large Vessel Occlusion Presenting within Eight Hours of
Symptom Onset (REVASCAT) (19%) and MR CLEAN (29%) allowed for randomization of patients
with extracranial ICA occlusions[2],[4],[6]. Two meta-analyses have confirmed benefit for patients with M1 and ICA occlusions,
but not for patients with M2 occlusions[8],[9]. However, the number of patients with M2 occlusions was small (98 among 1,287 enrolled),
and there was no interaction between treatment efficacy and PAO location. Although
more studies will be necessary to address this issue, we believe that it is likely
that early and successful reperfusion could be beneficial in patients with M2 occlusions.
Clinical deficit requirements for eligibility in the trials were not homogeneous.
The majority of patients enrolled in all trials had high NIHSS scores, as should be
expected. Patients with minor clinical deficits who have a PAO have a higher likelihood
of suffering clinical deterioration, and could still benefit from reperfusion[27],[28]. Based on the current data and the divergent clinical deficit criteria used in the
trials, it is not possible to determine an unambiguous cut-off on the NIHSS for indicating
endovascular treatment. More importantly, waiting for persistence of clinical symptoms
after the full dose of intravenous tissue plasminogen activator (IV TPA) – “failure
of IV TPA” – was not a requirement in order to assess for thrombectomy eligibility
and randomization, and adoption of such criterion would delay treatment initiation.
The benefit of endovascular therapy was confirmed in patients older than 80 years
old[8],[9]. Although the presence of a modified Rankin Scale of ≤1 was required for enrollment
in all trials, it seems reasonable to consider that patients with prior mild to moderate
disability could still benefit from therapy.
Time to treatment was defined in the trials as the period from symptom onset (defined
as the time last seen well) to groin puncture. In three of the five trials, patients
were eligible if endovascular treatment was to be initiated no later than six hours[2],[3],[5]. The REVASCAT trial would treat patients up to eight hours, requiring the presence
of a small infarct on non-contrast computed tomography (CT)[6]. The ESCAPE trial would treat patients until 12 hours after symptom onset based
on the use of CT and CTA to determine the presence of a small infarct and good collateral
status[4]. However, in spite of these differences, the great majority of patients included
in all trials were treated within the early six-hour window [median onset-to-groin
in minutes: 185 (ESCAPE) to 269 (REVASCAT)]. In the SWIFT-PRIME trial, the rate of
functional independency decreased by 10% to 20% for every hour of treatment delay
in the intervention arm[29]. Both meta-analyses showed that the sub-group of patients randomized after five
hours benefited from therapy, but it is still unclear whether treatment initiated
after six hours is beneficial[8],[9].
Except for MR CLEAN, all trials explicitly required (at least) the presence of a small
infarct core on admission neuroimaging, most commonly defined by an Alberta Stroke
Program Early CT Score (ASPECTS) ≥6 on either CT or diffusion-weighted magnetic resonance
imaging (MRI). The MR CLEAN trial did not explicitly state any contra-indications
based on admission ischemic findings on CT. However, the median ASPECTS score of treated
patients in this trial was 9 (interquartile range: 7–10). Ninety-two subjects had
ASPECTS from 5–7 and only 28 from 0–4. In MR CLEAN, there was no interaction between
treatment effect and ASPECTS strata (0–4, 5–7 and 8–10), however the study was likely
underpowered to detect differences in the 0–4 sub-group[30]. In total, 121 patients had ASPECTS from 0–5 in the five trials, and the benefit
in this subgroup could not be confirmed in meta-analysis[8]. We believe that until more data is available on this matter, patients with large
infarct cores should not be treated with endovascular treatment.
Additional criteria were used in the other trials. All patients evaluated in EXTEND-IA
used the RAPID evaluation as an inclusion criterion to determine the presence of penumbral
mismatch on CT perfusion[31]. The SWIFT-PRIME also used the same strategy for the first 71 of 196 patients but
later adopted the ASPECTS core criteria to ensure quicker therapy and higher enrollment.
The ESCAPE trial adopted the presence of good collateral status or penumbral mismatch
as an additional criterion in addition to a small infarct core. Trials using these
additional criteria (e.g. collateral or perfusion imaging) might have selected patients
more likely to benefit from therapy, but might have excluded patients who could still
benefit from endovascular treatment. The presence of either penumbral mismatch (EXTEND-IA)
or good collateral flow (ESCAPE) should not be considered mandatory for treatment
indication, but might be used as ancillary data for individualized decisions.
The SWIFT-PRIME and EXTEND-IA only included patients who had received intravenous
thrombolysis according to standards-of-care. Patients not receiving intravenous thrombolysis
due to contraindications were 11%, 27% and 32% in MR CLEAN, ESCAPE and REVASCAT trials.
Subgroups analysis later confirmed that patients not receiving IV TPA also benefit
from endovascular treatment[8],[9].
Mechanical thrombectomy with stent retrievers was the mandatory endovascular treatment
technique in SWIFT-PRIME, EXTEND-IA and REVASCAT, and was also used for the vast majority
of patients in MR CLEAN (97%) and ESCAPE (86%). This choice was driven by the findings
of SWIFT and TREVO-2, two head-to-head randomized clinical trials comparing the reperfusion
performance of stent retrievers against the MERCI device, where stent retrievers yielded
a four-times higher chance of post-procedure reperfusion and higher rates of good
clinical outcome[32],[33]. Indeed, prior trials that did not show benefit of endovascular treatment used other
first generation mechanical thrombectomy devices (mainly the MERCI device) or intra-arterial
injection of TPA[21],[22],[34]. One assumed reason for the absence of clinical benefit in those earlier trials
was the low rate of successful reperfusion. Stent retrievers should be the first choice
of device whenever considered feasible by the treating interventionist. The angiographic
goal of endovascular radiation therapy should be complete or near-complete reperfusion,
defined as a grade of 2b or 3 in the modified Thrombolysis in Brain Ischemia (mTICI)
score.
The type of anesthesia used during endovascular treatment is still a matter of debate.
It has been described that the use of general anesthesia results in increased intensive
care unit stay, larger infarct volumes and worse clinical outcomes, when compared
to conscious sedation in observational studies[35]. In the MR CLEAN trial, patients who underwent general anesthesia had longer door-to-groin
times and worse functional outcomes, with no significant benefit compared to controls.
The SIESTA trial was a single-center, randomized, open-label trial with blinded endpoint
evaluation comparing conscious sedation with intubation with general anesthesia performed
by a neurointensivist in patients receiving endovascular treatment. There was no difference
between groups in the primary endpoint of early neurological improvement in the NIHSS
score after 24 hours. (Oral abstract at the European Stroke Organization conference,
2016) Given the limitations of this single-center trial, further clinical trials addressing
the use of general anesthesia or conscious sedation are still necessary.
Generalizability and institutional implementation
The contemporary trials on endovascular treatment were all conducted in comprehensive
stroke centers with experience in treating patients with AIS in developed nations.
It should be noted that the positive results also reflect very fast and early treatment
initiation (the median CT to groin puncture time in the ESCAPE trial was 51 minutes),
very high rates of successful reperfusion and low rates of complications. The generalizability
of these results requires that endovascular treatment should be performed in experienced
stroke centers. Also, quality improvement initiatives have been shown to improve important
metrics of stroke care. With the establishment of endovascular treatment as a standard
of care in anterior circulation ischemic stroke with PAO and its increased and widespread
use, it is important that institutional initiatives are held to monitor and improve
procedural metrics of safety and efficacy, including key metrics such as time of initiation
of therapy and rate of complete reperfusion. Moreover, the design of local protocols
should reflect local resources and expertise. The success of the implementation of
endovascular treatment on public and private healthcare systems relies on careful
planning and assessment of local resources, and the organization of a stroke network
that is able to identify and refer patients to the appropriate comprehensive stroke
center.
Recommendations
-
Endovascular treatment is recommended for adult patients with AIS due to a PAO in
the anterior circulation with a significant neurologic deficit measurable on the NIHSS
stroke scale and a relatively small infarct core on baseline neuroimaging (i.e., ASPECTS
≥6) who can have their treatment initiated within six hours of symptom onset and prior
treatment with IV TPA (Recommendation: 1-A).
-
Even if endovascular treatment is indicated, IV TPA is recommended prior to endovascular
treatment in eligible patients according to previously published guidelines. (Recommendation
1-A)
-
Endovascular treatment should neither preclude nor delay intravenous thrombolysis
in patients who are candidates for both treatments. (Recommendation: 1-A). Also, presumed
“failure” of IV TPA should not be waited for in order to indicate or initiate endovascular
treatment. (Recommendation: 1-A). When eligible, endovascular treatment is recommended
even in patients with contra-indication to IV TPA (Recommendation: 1-A).
-
Patients eligible for endovascular treatment should be treated as soon as possible,
and the groin puncture should be performed within six hours of symptom onset (Recommendation:
1-A). It is uncertain whether treatment initiated after six hours may be beneficial
for selected patients (Recommendation: 2b-C1).
-
Selection of patients for endovascular treatment should be based on a clinical diagnosis
of AIS due to a documented PAO confirmed by CTA, MRA or digital subtraction angiography.
(Recommendation: 1-A). Transcranial doppler performed by skilled examiners might be
useful for detecting PAO in this setting, in cases of contraindications for or unavailability
of CTA and/or MRA (Recommendation: 2b-C2).
-
Occlusion of the M1 segment of the MCA or the terminal ICA are considered eligible
PAO (Recommendation: 1-A). Patients with occlusions of the M2 MCA segment might be
considered for treatment (Recommendation: 2b-C1). Patients with an associated extracranial
ICA occlusion also benefit from therapy (Recommendation: 1-A), however it is not established
whether revascularization of the cervical ICA must be performed at the same time as
the target intracranial thrombectomy.
-
Age should not be used as an isolated criterion for excluding patients from therapy
(Recommendation: 1-A). The severity of the clinical deficit and the presence of prior
disability should not be used as absolute contraindications for therapy, and clinical
judgment is necessary when taking these issues into account (Recommendation: 3-E).
-
A non-contrast CT scan or diffusion-weighted MRI should be done to assure ischemic
type and that the baseline infarct core is limited (ASPECTS ≥6) (Recommendation: 1-A).
Penumbral mismatch assessment on perfusion imaging or collateral grading on CTA can
help select patients in individualized decision making (Recommendation: 2a-A), but
are not mandatory for treatment indication.
-
New generation devices like stent retrievers should be the primary choice of technique
for endovascular treatment (Recommendation: 1-A). The therapeutic goal of endovascular
treatment should be of complete or near-complete reperfusion (TICI 2b-3). (Recommendation:
1-A). If deemed useful in order to achieve successful reperfusion by the treating
interventionist, proximal balloon guiding catheters, large bore aspiration catheters
or alternative techniques (intra-arterial thrombolysis, alternative devices, etc.)
might be considered at the discretion of the interventionist (Recommendation: 2b-C1).
-
Conscious sedation, as opposed to general anesthesia, might be beneficial in patients
undergoing endovascular treatment in improving the chances of good clinical outcome
(Recommendation: 2b-C). When general anesthesia is deemed necessary, careful monitoring
of blood pressure, blood carbon dioxide and oxygen saturation levels is recommended
(Recommendation: 2b-B) by an anesthesiologist or neurointensivist who has experience
with AIS patients. (Recommendation: 2b-E)
-
Endovascular treatment should be performed in stroke centers with sufficient human
and material resources, and with experience with treating patients with AIS, including:
experience with intravenous thrombolysis; full-time, on-board personnel with a neurologist;
full-time CT scan or MRI scan with capability of performing angiographic studies;
full-time laboratory; full-time on-call board certified neurointerventionist personnel;
stroke unit or neurointensive care unit; full-time, on-call neurosurgical staff. (Recommendation:
1-E).
-
Hospitals performing endovascular treatment should establish institutional, multidisciplinary
protocols, and promote prospective, continuous reassessments of safety and procedural
efficacy for quality assessment and improvement, including: door to neuroimaging time,
door to initiation of endovascular treatment time, rate of procedural complications,
rate of reperfusion (TICI 2b-3), and rate of hemorrhagic transformation. (Recommendation:
1-E).
-
Stakeholders of institutional and regional health policies (public or private) should
assess local resources in order to plan for establishing a network capable of identifying
and referring potential candidates for endovascular treatment to comprehensive stroke
centers. (Recommendation: 1-E).
