Endovascular Therapy for Acute Stroke: New Evidence and Indications

Abstract

Endovascular therapy (EVT) has revolutionized the treatment of acute ischemic stroke. In the past few years, endovascular treatment indications have expanded to include patients being treated in the extended window, with large ischemic core infarction, basilar artery occlusion (BAO) thrombectomy, as demonstrated by several randomized clinical trials. Intravenous thrombolysis (IVT) bridging to mechanical thrombectomy has also been studied via several randomized clinical trials, with the overall results indicating that IVT should not be skipped in patients who are candidates for both IVT and EVT. Simplification of neuroimaging protocols in the extended window to permit non-contrast CT, CTA collaterals have also expanded access to mechanical thrombectomy, particularly in regions across the world where access to advanced imaging may not be available. Ongoing study of areas to develop include rescue stenting in patients with failed thrombectomy, medium vessel occlusion thrombectomy, and carotid tandem occlusions. In this narrative review, we summarize recent trials and key data in the treatment of patients with large ischemic core infarct, simplification of neuroimaging protocols for the treatment of patients presenting in the late window, bridging thrombolysis, and BAO EVT evidence. We also summarize areas of ongoing study including medium and distal vessel occlusion.

Introduction

Endovascular therapy (EVT) has revolutionized the treatment of acute ischemic stroke (AIS). EVT is very effective in reducing the long-term disability with a number needed to treat (NNT) ranging from 2.6 to 3.0.^1,2) However, the global proportion of patients who receive this treatment remains low.^3,4) Although barriers such as cost, resources, and systems of care remain obstacles in many parts of the world,^5,6) providers are working on broadening the indications for this highly effective disability-saving treatment.⁷⁾ Published guidelines remain universally conservative in their recommendations with only minor variations in indications between guidelines. The 2021 Japanese stroke guidelines provide Grade A recommendations with a high level of evidence for EVT in patients with an internal carotid artery (ICA) or middle cerebral artery (MCA) M1 occlusion with an Alberta stroke program early computed tomography score (ASPECTS) of 6 or greater and National Institutes of Health Stroke Scale (NIHSS) score of 6 or greater in adult patients with a prestroke modified Rankin scale (mRS) score of 0–1, and a last known well (LKW) within 16 hours. Intravenous (IV) alteplase is recommended if eligible.^8,9) Other endovascular interventions or deviations from the above parameters were given lower levels of evidence. The American Heart Association/American Stroke Association Acute Stroke Guidelines and the European Stroke Guidelines from 2018 also have a similar level of recommendations with the addition of recommending advanced imaging for the late window.^10,11) Since the above guidelines, numerous clinical trials, and observational studies have been performed to ascertain whether patients outside these indications would benefit if treated by EVT.

Large Core

Early randomized clinical trials focused on patients with a large vessel occlusion (LVO) and small regions of an established infarct (“core”). A pooled meta-analysis of these trials suggested that patients with ASPECTS 3–5 could benefit from thrombectomy.¹²⁾ However clinical equipoise remained in the absence of dedicated randomized controlled trial (RCT) data in this group and other data suggested higher symptomatic intracranial hemorrhage (sICH) rates.¹³⁾ With the release of large core RCTs including RESCUE-JAPAN LIMIT, ANGEL ASPECT, and SELECT 2, the evidence favors treating these patients with EVT,^14–17) and the TESLA trial showed a direction of benefit although the results did not achieve a statistical significance.¹⁸⁾ The RESCUE JAPAN trial examined patients with an ICA or M1 occlusion and an ASPECTS of 3–5 who were within 6 hours of LKW or had a diffusion weighted imaging (DWI)–FLAIR mismatch in the late window. The primary outcome of a mRS score of 0–3 favored the EVT group with 31% of patients achieving this compared to 12.7% in the medical management group (relative risk [RR] 2.43, 95% confidence interval [CI] 1.35–4.37, P = 0.002). Secondary outcomes of mRS scores 0–1 and 0–2 favored EVT but were not significant. There was significantly more intracranial hemorrhage (ICH) and numerically more sICH in the EVT group. Largely MRI-based selection and a Japanese population limited the generalizability of this study.¹⁴⁾ Subsequently, the ANGEL ASPECT and SELECT 2 trials confirmed the results of RESCUE-JAPAN LIMIT. ANGEL ASPECT enrolled patients with an ICA or M1 occlusion in China. The imaging criteria included patients with (1) ASPECTS 3–5, (2) ASPECTS 0–2 with infarct volumes of 70–100 mL on CT perfusion (CTP), or (3) ASPECTS of 5 or greater in the late window if infarct volume was 70 to 100 mL on CTP. Infarct volumes were calculated with MRI DWI or CTP. The trial was stopped early due to the evidence of the efficacy of EVT. The primary endpoint of an ordinal shift on the mRS score was in favor of EVT (odds ratio [OR] 1.37, 95% CI 1.11–1.69, P = 0.004). There were 30% of patients in the EVT group and 11.6% of patients in the medical management group who reached a mRS score of 0–2.¹⁵⁾ The probability of achieving a mRS score of 0–2 by 90 days was less than what was reported by the HERMES collaboration (which included patients treated with EVT having highly favorable preintervention imaging profiles)¹⁾; however, the magnitude of benefit with EVT over medical management is similar. SELECT 2 evaluated patients from America, Europe, Australia, and New Zealand with an ASPECTS of 3–5 or a CTP or MRI DWI infarct greater than 50 mL. This trial was also stopped early as there was a mRS shift in favor of the EVT group (OR 1.51, 95% CI 1.20–1.89, P <0.001). There were 20.7% of patients who reached a mRS score of 0–2, whereas only 7% of medical therapy patients reached this endpoint. sICH rates were not significantly higher in the EVT groups in all trials although RESCUE-JAPAN LIMIT and ANGEL ASPECT had numerically higher rates of sICH. RESCUE-JAPAN LIMIT and ANGEL ASPECT found significantly higher rates of any ICH.

Still, further questions need to be answered. In SELECT 2, a subgroup analysis showed that the presence of a core/penumbra mismatch did not mediate the treatment efficacy. Considering that core/penumbra hypothesis has been at the center of stroke treatment, these findings prompt the question of whether we are accurately assessing if the core/penumbra or acute stroke pathophysiology is more complex than just core/penumbra. Similarly, ASPECTS regions are not equal in their contribution to functional outcome. Subtle loss of gray–white differentiation and a well-circumscribed infarct result in a given region being scored as abnormal, but these two imaging patterns may respond to treatments differently.¹⁹⁾ Broocks et al. looked at this phenomenon with net water uptake (NWU), a measure that quantifies the level of hypodensity in an infarcted tissue.²⁰⁾ They found that a higher NWU or more hypodensity was associated with a decreased likelihood of a mRS score of 0–3 and that a successful recanalization mediated a treatment response only when NWU was low. Although they did not report the hemorrhage rates, it would also be interesting to see if the hemorrhage rates vary as well given the trends in the abovementioned RCTs. Lastly, as an ASPECTS of 3 to 5 is now an established target for mechanical thrombectomy and is likely cost-effective,^17,21,22) the question remains if there is a floor for the ASPECTS score and a response to recanalization.

Rescue Stenting

Approximately 10%–20% of stroke patients do not achieve successful recanalization after mechanical thrombectomy.^23,24) A common etiology for failed recanalization is intracranial atherosclerotic disease (ICAD) with higher rates of reocclusion.^25,26) Rescue therapy in this patient population comprised of rescue stenting (RS), angioplasty, or use of IV antiplatelet therapy to achieve vessel patency.^27,28) Hesitation in adopting this approach is related to stent thrombosis leading to reocclusion, bleeding with dual antiplatelet therapy or IV antiplatelet medication, and risk of vessel perforation. The level of evidence remains low with retrospective data and smaller studies. Mohammaden et al. retrospectively examined multicenter data of patients with M1 and M2 occlusions with a modified thrombolysis in cerebral infarction (mTICI) score of 0 or 1 after multiple passes.²⁹⁾ They compared the outcomes in 499 patients and performed a propensity score matched analysis with 107 patients each in the RS and non-RS arm. In both analyses, the RS arms achieved a positive shift on the 90-day mRS as well as higher rates of functional independence at 90 days compared to non-RS arms. There was a lower 90-day mortality with similar rates of sICH in both groups.

Two meta-analyses evaluated the use of RS in patients with LVO.^30,31) In the larger analysis, the RS group was younger and had lower rates of IV alteplase. For studies that listed an ASPECTS, most included small core patients with median scores ranging from 8 to 9 except for one study. There was heterogeneity in the definition of EVT failure with 6 studies that defined failed EVT as a mTICI score of 0–2a, 4 studies as a mTICI score of 0–1, and 2 studies as reocclusion. The primary endpoint of functional independence was achieved more frequently in the RS group (41% vs. 21.1%, OR 3.27, 95% CI 2.08–5.16) with a substantial between-study heterogeneity. Including single-arm studies, the rate of successful reperfusion after RS was 87% (95% CI 82%–91%). sICH occurred at similar rates in both groups (8.5% vs. 11.7%, OR 0.85, 95% CI 0.59–1.20) with low-certainty evidence and very little between-study heterogeneity. Mortality at 90 days occurred less in the RS group (22.5% vs. 33.8%, OR 0.47, 95% CI 0.32–0.69). Subgroup analysis for anterior or posterior circulation stroke did not change the trend of the above results. However, the meta-analysis is at a high risk of bias considering mostly the retrospective data, missing outcome data, heterogeneity in regards to definitions of EVT failure, treatment approaches, and etiology of occlusions that limit the validity. Therefore, stent placement to achieve vessel patency is likely protective, but questions regarding optimal patient population, the definition of EVT failure, and when to proceed with stenting or balloon angioplasty, stent type, and type of antiplatelet therapy remain open.^32,33)

Bridging Thrombolysis

The question of the added benefit of intravenous thrombolysis (IVT) surfaced after the advent of the positive EVT trials. Potential advantages of skipping thrombolysis could be reduced cost, less bleeding complications, and hesitancy in interventions such as stenting, given the need for antiplatelet therapy after IVT. Between 2020 and 2021, six trials were published that evaluated the role of bridging IVT.

DIRECT-MT evaluated 654 patients with ICA, M1, and M2 occlusions from 41 centers in China.³⁴⁾ Among those, 326 patients did not receive bridging therapy and 328 received bridging therapy. The trial demonstrated non-inferiority with an adjusted common OR for a shift in the mRS score at 90 days of 1.07 (95% CI 0.81–1.40, P = 0.04) and a prespecified lower CI of 0.8 for noninferiority. A successful recanalization prior to thrombectomy occurred significantly more often in the alteplase group and numerically more often without a statistical significance after EVT in the alteplase group. Both sICH and procedural complications did not vary between groups. Limitations of the trial that biased against the bridging arm were a wide noninferiority margin, long door to IVT times resulting in short delays between IVT to groin puncture, possibly due to payment models and consent. The DEVT trial evaluated 234 patients across 33 centers in China with LVO in ICA and M1.³⁵⁾ The trial was stopped early as the prespecified efficacy boundary was met on an interim analysis: 54.3% of patients in the direct EVT group reached the primary outcome of functional independence compared to 46.6% in the bridging group (difference 7.7%, one-sided 97.5% CI −5.1% to ∞). This was within the prespecified noninferiority margin of 10%. No differences were noted in the safety measures of mortality and sICH. Like DIRECT MT, the presence of long door-to-IVT times with short IVT-to-EVT times and generous noninferiority margins were the limitations. The SKIP trial enrolled 204 patients in 23 stroke centers in Japan with ICA or M1 occlusions.³⁶⁾ As per Japanese guidelines, bridging patients received a lower dose of 0.6 mg/kg of alteplase in contrast to the other trials. There were 59.4% of patients in the direct EVT group who reached the primary endpoint of functional independence compared to 57.3% in the bridging group (OR 1.09, one-sided 97.5% CI 0.63–∞) who did not meet the prespecified noninferiority margin of 0.74. Other secondary outcomes and sICH also did not differ. Longer door-to-IV thrombolysis times, usage of low-dose alteplase, and the start of IVT after an arterial puncture in 21.4% limit generalizability. DIRECT-MT, DEVT, and SKIP were all conducted in east Asian countries that are known to have higher percentages of ICAD, which may be less responsive to alteplase.³⁷⁾

The next 3 trials were conducted outside of Asia. MR CLEAN- NO IV enrolled 539 patients across 20 centers in the Netherlands, Belgium, and France with an occlusion of the ICA, M1, and proximal M2.³⁸⁾ The adjusted odds ratio (aOR) for a shift in an mRS score was 0.84 (95% CI 0.62–1.15, P = 0.28) which was not non-inferior with a prespecified lower boundary for the CI set at 0.8. The two groups did not differ with respect to the secondary outcomes, mortality, or sICH. SWIFT-DIRECT included 408 patients across 48 centers in Europe and Canada with an occlusion of the ICA or M1.³⁹⁾ The primary outcome of functional independence occurred in 57% of direct EVT patients compared to 65% of bridging patients (adjusted risk difference −7.3%, 95% CI −16.6 to 2.1). The lower limit of the one-sided 95% CI was −15.1%, which did not meet the prespecified noninferiority margin of 12%. Final recanalization occurred more often in the bridging group, and there was a trend toward more sICH and groin hematoma in bridging patients. DIRECT-SAFE evaluated 293 patients across 25 centers in Australia, New Zealand, China, and Vietnam with occlusions of the ICA, M1, and M2.⁴⁰⁾ Patients were also allowed to receive IV tenecteplase. The primary outcome of functional independence occurring in 55% of the direct EVT group and 61% of the bridging group (difference 5.1%, two-sided 95% CI −16% to 5.9%) did not satisfy the noninferiority margin of 10%. Secondary and safety outcomes did not differ. In contrast to the above Asian trials, patients enrolled in Asia had more benefit from bridging than patients in Australia and New Zealand in DIRECT-SAFE. These studies led to an ESO-ESMINT guideline recommendation and SVIN Brief Practice Update to keep the bridging therapy.^41,42) They also performed a study-level meta-analysis that did not show noninferiority of direct EVT across a range of noninferiority margins and significantly reduced the successful recanalization rates with direct EVT. Data from the Improving Reperfusion strategies in Ischemic Stroke (IRIS) collaborative, a patient level meta-analysis of the 6 trials also did not show non-inferiority of direct EVT.⁴³⁾

Considering that patients in these trials largely had successful thrombectomies, the role of IVT in patients who do not achieve recanalization is less well understood. Faizy et al. looked at patients with ICA, M1, or M2 occlusion who went for EVT with an unsuccessful final recanalization (TICI 0-2a) in the German Stroke Registry. They performed propensity score matching to include 746 patients. A higher portion of patients in the bridging group had the primary outcome of functional independence at 18.2% compared to 11.3% in the EVT-only group (aOR 2.63, 95% CI 1.41–5.11, P = 0.003). The bridging group also had better secondary outcomes of mRS 0–3 as well as a shift in the overall mRS scores. A subgroup analysis favored bridging patients across the predefined subgroups. Bridging patients benefited within 4.5 hours after onset but not after 4.5 hours or with an unknown time of onset. Bridging patients also had more benefit with partial reperfusion rather than failed reperfusion. sICH was not different in both groups, whereas mortality occurred less often in the bridging patients. Importantly, the effect of IVT varies by the patient population, of which many were not included in the above trials. These trials examined mostly the anterior circulation occlusions presenting to a thrombectomy-capable center, and it is possible that the benefit of bridging thrombolysis will be more apparent in transfer patients. Posterior circulation occlusions, distal occlusions, and smaller thrombi are also more responsive to thrombolysis and the first two of the three locations were not well represented in the above trials.^44–46)

Investigators of the CHOICE randomized clinical trial randomized 121 patients with a successful recanalization (mTICI 2b-3) of an anterior or posterior LVO to intra-arterial alteplase (0.225 mg/kg) or placebo over 15–30 minutes and found a higher rate of the primary outcome (90-day mRS score 0–1) among patients treated with alteplase (59% vs. 40%, P = 0.047).⁴⁷⁾ While most patients had achieved technically successful recanalization with mechanical thrombectomy, the added benefit of intra-arterial alteplase argues for the importance of microcirculatory failure that remains present despite successful large vessel recanalization.

Imaging Modality

Although many of the landmark trials selected patients based on advanced imaging, particularly in the late window, this approach is not without its disadvantages: increased cost, increased time to acquire images, use of contrast, and decreased availability of advanced imaging capability.^6,48) As a result, there has been an influx of studies examining if simpler imaging paradigms may provide a benefit to our patients without harming them.⁴⁹⁾

The largest study was the CLEAR study, a multicenter, retrospective study that examined different imaging modalities to select patients for treatment with EVT in the late window.⁵⁰⁾ The study included 1530 patients across 15 sites in 5 countries. Patients were included who presented with an occlusion in the ICA or proximal MCA (M1/M2 segments) and NIHSS score of 6 or more in the extended window of 6 to 24 hours from LKW. Patients were selected based upon non-contrast CT head (NCCT) with CTA, CTP, or MRI. The NCCT group had higher presenting NIHSS, higher rates of hypertension, atrial fibrillation, presented more frequently as transfers, and higher rates of ICA occlusion. Rates of IVT were most common in MRI followed by NCCT and then CTP groups. The median ASPECTS was 8 across all 3 groups. Significantly shorter door-to-puncture times were observed for the NCCT group as compared to those for the CTP and MRI groups. Successful reperfusion occurred significantly more often in the NCCT and CTP groups compared to that in the MRI group. The primary endpoint of distribution of mRS at 90 days did not differ between the groups after adjusting for the prespecified variables. The probability of functional independence and safety measures of sICH and mortality were also similar across groups. A meta-analysis combining 5 studies also found similar results with similar rates of functional independence and sICH.⁵¹⁾ They also found that the CTP group had lower rates of mortality and higher rates of recanalization. Katsanos et al. noted that due to the lack of patient-level data and retrospective nature of the data analyzed, it is possible that unbalanced baseline characteristics accounted for this difference.⁵²⁾

MR CLEAN-LATE was a randomized trial that enrolled patients not meeting DAWN and DEFUSE-3 criteria using CTA collaterals as a selection method for mechanical thrombectomy in the Netherlands.⁵³⁾ Patients with ICA, M1, or M2 occlusions and presence of collateral flow on CTA divided into 3 grades were randomized into treatment with EVT or no treatment. Patients in the treatment arm had a better primary outcome with lower 90-day mRS scores (adjusted common odds ratio [cOR] 1.67, 95% CI 1.20–2.32). Dichotomized mRS trended in favor of EVT but was only significant for a mRS score of 0–3. sICH occurred more often in the EVT group, although in low numbers in both groups. Patients with weaker collaterals were associated with a stronger treatment effect.

Given the concerns of NCCT as a selection for EVT would lead to futile recanalization, reperfusion without functional independence,⁵⁴⁾ and/or sICH, the positive large core thrombectomy data presented above provide reassurance that these concerns may not be relevant. ANGEL ASPECT and SELECT 2 used NCCT and CTP as a selection tool and found benefit in large core stroke as adjudicated by low ASPECTS. The benefits of a simpler acute stroke imaging workflow are also paramount. Only mandating NCCT as a screening tool could allow greater access to this disability-saving treatment in low-resource settings.^55–57) There is also the argument for a faster workflow with potential neuronal and cost savings. The risk of over-selection by CTP parameters and delays incurred with advanced imaging may place some patients at risk of being excluded or delayed from a potentially disability-sparing treatment.

Medium/Distal Vessel Occlusions

Although the data for intervention in large vessel occlusion are robust, the data for medium vessel occlusion (MeVO) are incomplete.^58–65) There is heterogeneity in the sites of occlusion as anterior cerebral artery (ACA), MCA, and posterior cerebral artery (PCA) occlusions have different clinical–anatomical considerations and it may be worthwhile to examine each arterial site separately.

Intervention in isolated ACA occlusions was evaluated by Meyer et al. in A2 to A4 occlusions with 110 patients after propensity score matching.⁵⁹⁾ There were trends toward more large artery atherosclerosis as an etiology for the EVT group. There were no differences between the groups with respect to clinical or safety outcomes even when stratified by occlusion site, age, sex, occlusion site, NIHSS, and IVT, which may have been due to a low sample size. On multivariable analysis, improved reperfusion scores were associated with good outcomes. Importantly, mechanical thrombectomy was safe and feasible in this site of occlusion. There was only 2% sICH observed and low rates of embolization to the new territory and vessel perforation. Moreover, NIHSS and mRS scores may not detect the disability related to neuropsychological deficits as seen in ACA occlusions, limiting conclusions about the functional benefit. A1 occlusions were not included in this study, but a retrospective study by Filioglo et al. found similar results on the subgroup analysis by location.^59,66)

Similar to ACA occlusions, PCA occlusions also were not included in the mechanical thrombectomy trials.^{60,62,63,67,68)} The Posterior Cerebral Artery Occlusion (PLATO) study evaluated EVT versus medical management in 1023 patients with PCA occlusions of the P1–P4 segments including fetal PCAs within 24 hours of a symptom onset.⁶¹⁾ Patients in the EVT arm had higher NIHSS scores, more proximal PCA occlusion, were treated in a later chronological year, had higher rates of mRS 0 at baseline, more often transfers, less diabetes, hyperlipidemia, and prior stroke. They also had a higher posterior circulation acute stroke early prognosis CT score (pc-ASPECTS) and perfusion mismatch. The primary endpoint of an ordinal mRS shift at 90 days did not differ between the EVT and medical management groups, but the coprimary endpoint of an early decrease in NIHSS by greater than or equal to 2 occurred more often in the EVT group. Secondary endpoints of early NIHSS improvement and vision improvement also occurred more often in the EVT group. There was a significantly higher rate of sICH and mortality in the EVT group. In adjusted analyses, there was a higher likelihood of excellent outcome of mRS 0–1 at 90 days in the EVT group. An adjusted analysis of a subgroup of patients treated with IVT also showed the same results except that the primary endpoint of decrease in NIHSS of greater than or equal to 2 was not significant. Like the ACA occlusion discussion, mRS places a larger emphasis on motor and language abilities, which may miss the benefit regarding the vision seen by treating PCA occlusions. These improvements in the vision may explain both the benefit regarding mRS score 0–1 at 90 days as well as the NIHSS improvements. The increased sICH and mortality trends do warrant caution and should be considered during the clinical decision making and weighed against the clinical deficit and possible functional gains.

The highest quality data for medium and distal MCA occlusions are from the HERMES meta-analysis of the early window trials. In a cohort of M2 occlusions, functional independence was greater for patients treated with EVT.⁶⁴⁾ Most of the cohort had proximal and dominant or codominant MCA occlusions and the benefit was understandably larger for these subgroups. A study level meta-analysis by Loh et al. pooled 15 studies with 2252 patients with medium or distal vessel occlusions.⁶⁹⁾ Herein, 72.1% patients had M2 MCA occlusion, 23.7% had PCA occlusion, and 4.2% had ACA occlusion. They found no significant difference for functional independence between the EVT and MM groups. However, they noted a substantial heterogeneity and publication bias on their funnel plot in favor of medical management. No difference was noted in the excellent functional outcome but with a substantial heterogeneity and no significant publication bias. Safety outcomes of sICH and mortality also did not differ. A subgroup analysis of patients receiving EVT with intra-arterial thrombolysis primary or in addition to mechanical thrombectomy showed no difference in the rates of functional independence but significantly increased rates of the excellent functional outcome in comparison to medical management. The patients receiving mechanical thrombectomy alone did not differ in functional independence or the excellent functional outcome in comparison to medical management. There were no differences in the outcomes by an occlusion location. They also looked at the studies that focused on the patients presenting with mean or median NIHSS score <6 and found that the EVT group had a lower likelihood of functional independence and significantly higher rates of sICH compared to medical management. When looking at the studies with only mean or median NIHSS score >6, they found that the EVT group had a greater likelihood of excellent outcome with no differences in functional or safety outcomes.

The common theme in the above studies is that relevant endpoints should be reconsidered when looking at the efficacy of intervention in MeVOs or distal vessel occlusions that are specific to the site of arterial occlusion. Either a higher bar for an efficacy endpoint in mRS should be considered or possibly another endpoint for ACA or PCA occlusions where the mRS may not weigh the deficit accurately. Although intervention appears overall safe, there are some subgroups such as PCA or MCA occlusions with a low NIHSS where there are safety concerns and the presenting symptoms should be weighed against these risks. Further randomized trial data via the DISTAL, ESCAPE-MeVO, DISCOUNT, DISTALS, and DUSK trials will help further inform this space.

Basilar Artery Occlusion

Although the landmark positive EVT trials included anterior circulation LVO patients only, the natural course of basilar artery occlusions (BAOs) made this lesion a logical EVT target. The BEST and BASICS trials were the two neutral trials looking at EVT in BAO.^70,71) The trial results were met with skepticism due to concerns about the patient selection, high crossover rates and treatment outside the trial, as well as personal clinical experience with the natural course and treatment response of BAO.^72–78) Subsequently, two RCTs, ATTENTION and BAOCHE, confirmed the benefit of EVT in BAO.^79,80)

The ATTENTION trial was a multicenter RCT of EVT in patients with a BAO presenting with a NIHSS >10, an estimated time of BAO within 12 hours. Angioplasty and stenting were used in 40% of patients in addition to thrombectomy. The primary outcome of mRS score of 0–3 occurred more often in the EVT group (RR 2.1, 95% CI 1.5–3.0, P <0.001). All secondary clinical outcomes favored the EVT group. sICH occurred significantly more often in the EVT group but mortality was less in the EVT group. The BAOCHE trial was a RCT in China looking at EVT in patients presenting 6 to 24 hours from LKW with a presenting NIHSS ≥10 and pc-ASPECTS ≥6. Intracranial angioplasty or stenting after failed thrombectomy was performed in 55% of the EVT arm. The primary outcome of mRS score of 0–3 occurred more often in the EVT group (adjusted rate ratio 1.81, 95% CI 1.26–2.60, P <0.001). sICH was numerically higher in the EVT arm and mortality occurred less often in the EVT arm. For many, the ATTENTION and BAOCHE trials confirmed what many clinicians knew⁷⁴⁾ and provided a high quality evidence for EVT in BAO.⁴⁵⁾

In patients with BAO and mild deficit, the benefit is uncertain, where trial data were underpowered to show a difference as shown by a subgroup meta-analysis of patients with NIHSS <10.⁴⁵⁾ The Nationwide Inpatient Sample of patients admitted with BAO in the United States showed more discharges to home in the EVT group as compared to patients who were medically managed.⁸¹⁾

Conclusion

After the initial neutral mechanical thrombectomy trials in stroke, subsequent trials were restrictive to provide the proof of concept for mechanical clot retrieval in stroke. Therefore, guidelines are conservative in their recommendations, in keeping with the trial eligibility criteria.⁵⁶⁾ However, maximizing the number of patients who can be helped necessitates that we broaden our criteria for interventions and update our guidelines.^56,82) Given the low NNT of EVT in stroke, broadening our criteria may dilute the overall effect of EVT but reach a greater number of patients overall that will benefit. We must also search for optimizing first-pass effect^83–85) and adjunctive therapies to help maximize the benefit of EVT.^47,86–92) EVT has changed the stroke treatment landscape and the stroke literature is now flourishing with new studies. We provide a summary of select studies representing the newest evidence that may inform decisions outside of the typical indications for EVT in an AIS.

Disclosure Statement

Dr. Thanh Nguyen reports advisory board of Idorsia. All the other authors report no other conflict of interest.