Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Stroke is the leading cause of adult disability in North America and is the fifth most common cause of death.1 ,2 The natural history of patients with acute ischemic stroke and occlusion of a major intracranial vessel such as the internal carotid artery (ICA), middle cerebral artery (MCA), or basilar artery is dismal, with high rates of mortality and low rates of disability-free survival.3 ,4 We introduce the term ‘Emergent Large Vessel Occlusion (ELVO)’ to describe this clinical scenario.
Among acute ischemic stroke, ELVO accounts for the greatest proportion of patients with long-term disability. For the past two decades the use of endovascular therapy has been performed in many centers across the world. The therapies have spanned from infusion of thrombolytic agents5 ,6 to mechanical embolectomy with the introduction of first-generation devices,7 ,8 aspiration-based embolectomy techniques,9 ,10 and the use of stent-retriever based procedures.11 ,12 However, these embolectomy trials were single-arm trials demonstrating safety of the procedure and technique or superiority over another, without direct comparison with standard medical therapy alone.
In the past 3 years, several major trials have been published comparing endovascular therapy with standard medical therapy alone. The purpose of this document is to summarize the results of these trials and synthesize the level of evidence supporting the use of embolectomy in patients with ELVO.
Materials and methods
This document was prepared by the Standards and Guidelines Committee of the Society of NeuroInterventional Surgery, a multidisciplinary society representing the leaders in the field of endovascular therapy for neurovascular disease. The strength of the evidence supporting each recommendation was summarized using a scale previously described by the American Heart Association.
Discussion and recommendations
Role of intravenous thrombolysis
In 1996 the FDA approved the use of recombinant tissue plasminogen activator (tPA) for the treatment of acute ischemic stroke within 3 h of symptom onset, based on the landmark NINDS trial.13 This medication is delivered intravenously (IV), with a total dose of 0.9 mg/kg delivered 10% as a bolus and the remainder as an infusion over 60 min. In 2009, based on a pooled analysis of multiple randomized trials including the European Cooperative Acute Stroke Study III, the time window for IV tPA was expanded to 4.5 h from symptom onset.14 Additional relative exclusion criteria for patients in the 3–4.5 h time window included age >80, use of any anticoagulant regardless of International Normalized Ratio (INR), the combination of both prior stroke and diabetes, and National Institute of Health Stroke Scale (NIHSS) score ≥25. While the FDA approval for tPA has not been changed to reflect this time window, national guidelines have endorsed this usage.15
The primary limitation with IV tPA is the limited rates of recanalization for ELVO. In a study using continuous transcranial ultrasound insonnation, rates of complete recanalization with ELVO were only 29% for M1 segment of the MCA, 20.3% for combined M1 segment and distal ICA, and 6% for total ICA occlusion.16
Summary and recommendation: IV thrombolysis should be offered to all eligible patients presenting within 3.0 h of symptom onset (AHA Class I, level of evidence A). Patients who meet the additional criteria also benefit from IV thrombolysis between 3.0 and 4.5 h from symptom onset (AHA Class I, level of evidence B). Intravenous thrombolysis should not be withheld in eligible patients in whom subsequent endovascular therapy is planned.
Intra-arterial pharmacological thrombolysis using pro-urokinase infusion in the Prolyse in Acute Cerebral Thromboembolism Trial I (PROACT-I) and PROACT II trials was safe and more effective than placebo infusion for patients with acute MCA occlusion within 6 h of symptom onset.5 ,6 In PROACT II, the Thrombolysis In Cerebral Ischemia (TICI) grade 2–3 recanalization rate in the treatment arm was 66%, although TICI 3 rates were only 19%.17 A meta-analysis combined the results from PROACT and PROACT-II with results from the Japanese Middle Cerebral Artery Embolism Local Fibrinolytic Intervention Trial (MELT) and found higher rates of good and excellent clinical outcomes in patients treated with thrombolysis compared with placebo.18
While no randomized trials have compared intra-arterial thrombolysis (IAT) with embolectomy, it is difficult to recommend primary IAT over embolectomy. This is primarily because the rates of near complete and complete recanalization using the TICI score (TICI 2b and TICI 3, respectively) appear to be higher with mechanical embolectomy than with IAT alone, especially with current mechanical thrombectomy device technology.19 Regardless, there may still be a role for IAT in cases where anatomic reasons preclude direct embolectomy.
Summary and recommendation: Intra-arterial thrombolysis is reasonable in patients presenting within a major stroke with occlusion of the middle cerebral artery in whom direct catheter access for embolectomy is precluded for anatomic reasons, as long as thrombolysis can be completed within 6 h from symptom onset (AHA Class IIa, level of evidence B).
Embolectomy for ELVO
Feasibility and safety trials
The first device specifically designed for intracranial embolectomy was the Merci retriever (Stryker Neurovascular, Fremont, California, USA). The initial Mechanical Embolus Removal in Cerebral Ischemia (MERCI) trial showed TICI 2–3 recanalization rates of 60% and reasonable clinical outcomes with good clinical outcome (modified Rankin Scale (mRS) score 0–2) of 43.5%.7 The follow-up Multi Merci trial, which also allowed patients who had received IV thrombolysis to be included, continued to show reasonable recanalization rates of 68%, although a good outcome was only seen in 34%.8 However, neither of these trials was randomized. Similarly, device safety trials for suction embolectomy using the Penumbra system (Penumbra, Alameda, California, USA) demonstrated high rates of Thrombolysis in Myocardial Infarction (TIMI) 2/3 recanalization (81.6% using this technique), although a favorable clinical outcome was only seen in 25%.10 More recently, stent-retriever technology has been evaluated. The SWIFT trial established safety and efficacy of recanalization for the Solitaire FR device (Covidien, Mansfield, Massachusetts, USA), with 69% of patients achieving TIMI 2/3 recanalization and 58% having a favorable outcome (defined as 0 mRS 0–2 or return if pre-stroke mRS is pre-stroke mRS >2, or improvement in NIHSS ≥10 points).11 That trial also showed the superiority of the Solitaire device over the Merci retriever, with only 30% of patients randomized to the Merci device achieving 30% TIMI 2/3 recanalization. The latest device to be approved by the FDA in the USA is the Trevo (Stryker Neurovascular). The Trevo 2 trial demonstrated 86% TICI 2/3 recanalization (versus 60% in the MERCI group), and 40% had a favorable clinical outcome (mRS 0–2).12 While all of these trials showed increasing rates of recanalization as experience with stent-retrievers grew, no direct comparison with best medical therapy was performed.
Recent randomized trials comparing endovascular therapy with best medical management: lessons learned
In the SYNTHESIS expansion trial, patients were randomized to receive either IV tPA or endovascular therapy.20 No specific treatment protocol for endovascular therapy was required, nor was documentation of ELVO required for randomization. This trial showed similar rates of favorable clinical outcomes at 90 days for patients in both groups, despite an approximately 60 min delay in the start of endovascular therapy. In addition, the recanalization rates with endovascular therapy were not published.
In the MR RESUCE trial, 127 patients were randomized to either best medical therapy or best medical therapy combined with embolectomy.21 Pre-treatment imaging was either MRI (using both diffusion and perfusion weighted sequences) or CT perfusion. Patients were stratified into either ‘penumbral’ or ‘non-penumbral’ patterns based on the size of infarct core and presumed volume of territory at risk. Embolectomy showed no benefit in this trial compared with best medical therapy. Embolectomy in MR RESCUE included the Merci or Penumbra device, and although the TICI 2/3 rate was 67%, only 27% had a TICI 2b/3 recanalization. This may be a major reason why this trial was not successful. Time was also likely to have been a contributing factor as median time from onset to enrollment was 5.5 h. Prior data from single-arm trials had suggested that recanalization beyond the 6 h window may not be beneficial.22
The IMS III trial was the largest randomized trial of endovascular stroke therapy.23 ,24 A total of 656 patients were randomized in a 2:1 fashion to a combination of IV tPA plus endovascular therapy (either additional tPA infusion or the use of an embolectomy device) or to IV tPA alone. Documentation of ELVO was not required for randomization. Patients were treated at a time of 208 min from symptom onset. Overall, the TICI 2/3 recanalization rate was 65% for ICA occlusions and 81% for occlusions involving the M1 segment of the MCA. TICI 2b/3 rates for these vessels were 38% and 44%, respectively. For the overall cohort, endovascular therapy showed no benefit over IV tPA alone. However, review of the subgroups as well as trial design does reveal some important subtleties.
The documentation of ELVO was not required for randomization in IMS III. In fact, among the patients allocated to receive endovascular therapy, 16% did not have any thrombus or treatable thrombus (as defined as ICA, M1 or proximal M2 segment MCA occlusion). Among the IMS III group of patients with documented ELVO by CT angiography (CTA), Rankin shift analysis showed a trend towards benefit to treatment.25 In addition, patients with an admission NIHSS of ≥20 also had a strong trend towards benefit from IAT for ELVO, both in the acute phase24 and also when follow-up was carried out at 6 months.26
Another important relationship was the time from symptom onset to recanalization, a link which had already been shown for the precursor EMS and IMS II trials.22 In the IMS III trial, when recanalization could have been achieved before 347 min, patients benefited from the addition of endovascular therapy when evaluating quality-adjusted life years.27 However, when recanalization occurred later, there was no clinical benefit. A strong trend towards benefit was seen in patients who received IV tPA <2 h and who then had their groin puncture within 90 min of IV tPA administration.24 The likelihood of recanalization also played a role, with the trial spanning several generations of embolectomy devices and techniques. Of the patients who were actually administered IAT, the majority of patients were treated with IAT alone (138 patients, 41.3%). Most of the embolectomy procedures in IMS III used the Merci devices (95 patients, 28.1%), with a small portion treated with the Penumbra system (54 patients, 16.2%). Stent-retriever based techniques, which have been shown to have some of the highest recanalization rates, were only used in five patients (1.5%).
Summary: Lack of benefit for the overall cohort in MR RESCUE and IMS III overshadowed signals of benefit to selected subgroups. Patients with confirmed large vessel occlusion from the IMS III study as well as those treated early showed strong trends toward benefit from the addition of endovascular therapy. In addition, these trials did not make extensive use of modern embolectomy techniques, which have been shown to have higher recanalization rates than first and second generation techniques.
Current trials: MR CLEAN, ESCAPE, EXTEND-IA
Based upon the lessons learned from the aforementioned trials, several groups set out to design new trials to truly test the ability of the most recent techniques of endovascular therapy in patients with documented ELVO. A core component of these trials was the requirement of documented vessel occlusion with non-invasive imaging (ie, CTA or MR angiography (MRA)) prior to randomization. In addition, several of these trials placed strict time constraints in order to ensure rapid recanalization. Finally, the devices used in these trials were the third generation of embolectomy devices and have been shown to have some of the highest recanalization rates.19 Key inclusion criteria and differences between these trials are summarized in table 1.
MR CLEAN is a multicenter randomized clinical trial of best medical therapy versus best medical therapy with IAT, including intra-arterial thrombolytics, mechanical embolectomy or both, in patients with anterior circulation ELVO.28 Conducted in 16 centers in the Netherlands, 500 patients were randomized, with 233 patients assigned to the interventional arm and 267 assigned to the control arm. The groups were well balanced, with a median NIHSS score of 17 in the intervention group and 18 in the control group; 25.7% of the patients in the intervention group had ICA occlusions compared with 29.3% in the control arm. Nearly all patients received IV tPA (intervention arm 87.1%; control arm 90.6%) with excellent time to IV tPA treatment (intervention arm 85 min; control arm 87 min). Time to groin puncture in the intervention arm was 260 min.
In assessment of the primary outcome, there was an OR of 1.67 (95% CI 1.21 to 2.30) of better outcome as assessed by mRS 0–2 at 90 days in the intervention arm. This benefit also extended to the secondary outcome measures. Statistically significant benefit or trends towards benefit were also seen in nearly all predefined subgroups, with the exception of those with pre-treatment Alberta Stroke Program Early CT Score (ASPECTS) 0–4. TICI 2b/3 recanalization was seen in 58.7% of the patients treated in the intervention arm. Comparing CTA at 24 h, absence of residual occlusion was seen in 75.4% of patients in the intervention arm compared with only 32.9% in the control arm. Final infarct volume was also significantly less in the intervention arm (49 mL vs 79 mL).
No significant safety concerns were seen, with similar intracerebral hemorrahage rates (7.7% vs 6.4%) and mortality rates (18.9% vs 18.4% at 30 days). Procedural complications included embolization into new territory in 8.6%, vessel dissection in 1.7%, and vessel perforation in 0.9%.
This trial represented the first clinical trial showing benefit for IAT in patients with ELVO in the post-IV tPA era. Compared with prior trials, ELVO was confirmed with pretreatment vessel imaging and modern device technology in the form of stent-retrievers was used. There was also a higher proportion of ICA occlusion, which may have more relative benefit than other occlusion locations.29 This benefit was also seen despite only a modest TICI 2b/3 recanalization rate.
The Endovascular treatment for Small Core and Anterior circulation Proximal occlusion with Emphasis on minimizing CT to recanalization times (ESCAPE) trial is a randomized trial primary performed in Canada focusing on patients with ELVO within 12 h of symptom onset with minimum NIHSS score >5 and with baseline CT showing ASPECTS >5.30 The study investigators have emphasized rapid imaging to treatment times, with a goal of 60 min from initial image on non-contrast CT (NCCT) scan to groin puncture, and a goal of recanalization within 90 min from first CT image.
Conducted in 22 centers in Canada, the USA and Europe, 315 patients were randomized, with 165 patients assigned to the intervention arm and 150 patients assigned to the control arm. The groups were well balanced, with a median NIHSS score of 16 in the intervention group and 17 in the control. In the intervention arm, 27.6% of the patients had intracranial ICA occlusions compared with 26.5% in the control arm. Most patients received IV tPA as well (intervention arm 72.7%; control arm 78.6%). Median time from imaging to groin puncture was 51 min.
When evaluating the primary endpoint, there was an OR of 1.7 for improved outcomes with the addition of embolectomy as opposed to best medical therapy alone, with 53% of the intervention arm achieving mRS 0–2 at 90 days versus 29% in the control arm. In addition, this benefit extended across all pre-specified subgroups, including patients below and above the age of 80 or those with and without cervical ICA occlusion. There were no safety concerns, with the rates of symptomatic intracranial hemorrhage of 2.7% in the intervention arm and 3.6% in the control arm. There was also a significant reduction in mortality, from 19.0% in the control arm to 10.4% in the intervention arm.
The Australian study EXTEND IA (Extending the time for Thrombolysis in Emergency Neurological Deficits with Intra-Arterial therapy) trial is a phase II randomized trial of ELVO patients with anterior circulation occlusion and a favorable imaging profile on CT or MRI advanced imaging including a mismatch ratio of >1:2, absolute mismatch >10 mL, and an ischemic core volume <70 mL.31 While only 70 patients were randomized, the study was halted by the Data and Safety Monitoring Board (DSMB) for reaching the pre-specified efficacy endpoint.32
All patients were treated with IV tPA and 35 patients received IAT in addition. As with the other trials discussed here, baseline characteristics were similar, with a mean NIHSS score of 13 in the control group and 17 in the intervention group. Also similar to the other trials, 31% of patients in both groups were ICA occlusions.
For the primary endpoint of mRs 0–2 at 90 days there was a significant treatment benefit, with 71% of patients in the intervention arm reaching independence compared with 40% in the control arm (p=0.01). There were no safety concerns, with symptomatic hemorrhage rates of 6% in the control arm and 0% in the intervention arm.
Soon to be published embolectomy trials: SWIFT-PRIME and REVASCAT
The Solitaire FR as Primary Treatment for Acute Ischemic Stroke (SWIFT-PRIME) trial is an international randomized controlled trial of IAT using the Solitaire stent-retriever in patients with anterior circulation ELVO.33 All patients were randomized to IV tPA alone or IV tPA with the addition of embolectomy. Patients were initially selected based on advanced imaging using target mismatch, but an amendment in February 2014 allowed enrollment if baseline imaging showed ASPECTS ≥6. The trial was also halted early, and preliminary analysis has shown a significant improvement in 90-day outcomes in the intervention arm compared with IV tPA alone.33
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 8 Hours of Symptom Onset (REVASCAT) trial is a trial conducted in the Catalonia region of Spain, with the intra-arterial procedures being performed at four comprehensive stroke centers. Similar in design to other recent trials, patients were randomized to either best medical therapy (including IV tPA when eligible) or best medical therapy combined with embolectomy using the Solitaire device. Patients were eligible for randomization within 8 h of symptom onset. On 12 December 2014, following the first pre-planned interim analysis, the DSMB concluded that equipoise no longer existed and the study was discontinued.
Choice of pretreatment imaging
A consistent theme across all of these studies was the use of pretreatment vessel imaging using CTA (or MRA) in order to confirm ELVO. Confirmation of vessel occlusion as quickly as possible is likely to be one of the reasons these studies were positive where others had failed. It will probably be important to have mechanisms in place to obtain rapid CTA (or MRA) on patients with a significant clinical deficit who may be candidates for embolectomy.
A large completed infarct core is a well-established exclusion criterion for endovascular treatment.34–36 The size of infarct, however, has varied across several studies, as has the choice of imaging in order to establish the size of the infarct core. Across these three studies, a variety of exclusions were used. MR CLEAN did not specifically exclude patients using any lower limit of NCCT ASPECTS value. ESCAPE excluded patients with NCCT ASPECTS value <6 and those with no collateral filling of the affected territory on CTA, as well as other exclusion criteria based on CT perfusion, if performed. SWIFT-PRIME also used perfusion imaging to identify a ‘target mismatch’, including having a core volume of <50 mL. Later, the trial also allowed inclusion of patients with ASPECTS ≥6. EXTEND IA also used CT perfusion imaging to select patients as described above. REVASCAT excluded those with NCCT ASPECTS <7 (<8 in those aged ≥85 in a protocol amendment in mid-2014).
We will further summarize the details of pretreatment imaging in a separate review. At the present time there is inadequate randomized literature to recommend one imaging modality over another.
Time frame for treatment
The aforementioned studies varied in their times from onset to treatment as well as from imaging to treatment. While MR CLEAN and SWIFT-PRIME restricted treatment to those in whom groin puncture could be performed within 6 h of onset, REVASCAT allowed enrollment up to 8 h and ESCAPE permitted treatment up to 12 h. In both ESCAPE and SWIFT-PRIME there were time criteria from imaging to puncture as well: 60 min from NCCT to puncture in ESCAPE and 90 min from CTA confirmed occlusion to groin puncture in SWIFT-PRIME. These studies highlight the need for rapid mobilization of the neurointerventional team and the need for adequate support staff to meet these time goals.
Summary and recommendation: For patients with anterior circulation stroke and documented ELVO affecting the ICA or M1 segment of the MCA and a corresponding clinical deficit, the addition of endovascular embolectomy results in superior clinical outcomes compared with best medical therapy alone. Embolectomy needs to be performed as rapidly as possible for the greatest clinical benefit, and is best when performed within 6 h from onset of symptoms (AHA Class I, level of evidence A).
Now that the efficacy of IAT for patients with ELVO has been established, it is important that research continues, both for procedural and peri-procedural care. There are many additional factors in developing successful systems of care for patients with ELVO. The ability to perform rapidly embolectomy is a requirement for comprehensive stroke centers, but not all patients are initially seen at those centers. As such, changes in current pre-hospital identification and transport mechanisms must be considered with an eye towards ensuring rapid access to embolectomy for all patients. Eligible patients evaluated at centers not capable of performing embolectomy need to be transported to comprehensive centers as quickly as possible.
In addition, we need to better elucidate the optimal imaging protocols for identifying patients most likely to benefit. Post-procedural care is also a vital component of good patient outcomes and should be emphasized. In addition, a quality improvement process within the institution which monitors outcomes and identifies opportunities to improve care on an ongoing basis will also likely improve the care we can deliver for these patients. Further optimization of embolectomy techniques needs to happen, as well as comparison of techniques (stent-retrievers vs direct aspiration; balloon guide catheters vs local aspiration).37–39 Refinement of techniques should also focus on minimizing emboli to additional territories, which occurred in 5.6% of the patients in the MR CLEAN trial.28
Another major area for further research is the role of embolectomy in patients presenting beyond the 6 h window, such as those who awaken with symptoms of stroke. Research should also focus on whether physiologic imaging using MR perfusion or CT perfusion techniques can help select patients in these time windows.
Twenty years ago the landmark NINDS trial demonstrated a benefit to IV tPA compared with conventional therapy for patients within 3 h from stroke onset and began a transformation in acute stroke therapy. Now we have multiple randomized multicenter prospective trials providing Level 1, Class A evidence that, in patients with anterior circulation stroke and an ELVO, the addition of embolectomy to best medical therapy is superior to best medical therapy alone. We feel this represents a watershed moment in the care of these patients with the deadliest forms of stroke. Centers performing embolectomy should strive to duplicate the excellent recanalization rates and rapid treatment times achieved in these trials. Future directions should focus on optimizing systems of care to maximize patient access to rapid embolectomy, as well as further research refining the techniques of embolectomy and patient selection.
Correction notice This article has been corrected since it published online first. The fifth author's name has been corrected.
Contributors All authors contributed.
Competing interests None.
Provenance and peer review Commissioned; internally peer reviewed.