LVO is an occlusion of the most proximal intracranial vessels, such as the ICA, the M1 or the proximal M2 segments of the MCA, the A1 segment of the anterior cerebral artery (ACA), the BA, the vertebral artery, or the P1 segment of the posterior cerebral artery (PCA). More distal occlusions (e.g., distal M2, M3, A2, and P2) can be classified as medium vessel occlusions. However, inconsistencies exist as some authors categorize M2, A1, and P1 as medium vessel occlusions [28].

Within the 6-hour time window, the selection of patients with anterior circulation stroke eligible for MT is based on non-contrast computed tomography (NCCT) or magnetic resonance imaging (MRI), including angiography (CT-A or MR-A). The Alberta Stroke Program Early CT Score (ASPECTS, based on NCCT) estimates the amount of infarcted brain parenchyma in the MCA territory (Fig. 1) [29]. Overall, a beneficial effect of MT can be expected in patients with ASPECTS $\ge$6 [30]. MRI can rule out intracranial hemorrhage as reliably as NCCT using gradient recalled echo sequences (GRE) [31]. Diffusion-weighted imaging (DWI) is used to visualize the infarct core. In patients with acute stroke, DWI-ASPECTS is an average of 1 point lower than ASPECTS based on NCCT [32]. A DWI-ASPECTS $\ge$5 is associated with good functional outcomes after MT [33].

Fig. 1.

ASPECTS. Visualization of the ASPECTS territories (A,B). The following areas are covered: M1 (anterior MCA cortex, frontal operculum), M2 (anterior temporal lobe, laterally to the insula), M3 (posterior temporal lobe, posterior MCA cortex), M4 (anterior MCA cortex superior to M1), M5 (lateral MCA cortex superior to M2), M6 (posterior MCA cortex superior to M3), insula (I), internal capsule (IC), caudate (C), and lentiforme nucleus (L). Each area accounts for 1 point. The maximum ASPECTS score is 10. Hypodensity in a described area leads to a deduction of one point. (C,D) show an example of CT ASPECTS. Hypodensity in the M2 and M6 areas is observed. Total ASPECTS: 8.

The use of additional perfusion imaging (CT-P or MR-P) in the early time window (within 6 hours of symptom onset) is controversial and is not recommended in routine clinical practice [34, 35, 36, 37]. Perfusion imaging can be used to estimate the infarct core and potential tissue at risk (penumbra). Fig. 2 (Ref. [38, 39, 40]) illustrates CT-P parameters and potential thresholds for both the infarct core and penumbra. These thresholds are debatable and not universally accepted, and may change with the duration of symptoms [38, 40]. In an early time-window, CT-P can overestimate the infarct core, possibly because of a lack of contrast arrival overall [41]. In a pooled analysis from the Highly Effective Reperfusion evaluated in Multiple Endovascular Stroke Trials (HERMES) collaboration, adding CT-P in an early time window has not been found to be associated with functional outcomes [37].

Fig. 2.

CT-Perfusion. NCCT ASPECTS 10 (A). CT-A with M2-occlusion (B). Interpretation of CT-P: the cerebral blood volume (CBV) is symmetrical without a regional decrease (C). Cerebral blood flow (CBF) is reduced in the posterior MCA territory on the left (D). The mean transit time (MTT) of the contrast agent (E) and Tmax (time to maximum; time delay between the contrast agent arrival in the proximal large vessel arterial circulation and the brain parenchyma perfusion [F]) are prolonged. The infarct core in CT-P shows a markedly reduced CBF (25 mL/100 g/min) and CBV (2 mL/100 g) together with an increase in MTT and Tmax [38, 39]. Penumbral tissue shows a delay in MTT ($>$145%) and Tmax ($>$6 sec) as well as a reduced CBF and a normal or slightly increased CBV. Beyond these absolute values, relative CT-P thresholds are mentioned (e.g., infarct core [CBF] defined as 30% of the contralateral CBF) [40]. CT-P parameters in Fig. 2 suggest a large penumbra with prolonged MTT and Tmax. There is a reduction in CBF with normal CBV. MT was performed in this patient.

In posterior circulation stroke, MT is currently recommended in carefully selected patients with BA occlusion [34, 35, 36]. Imaging tools such as the posterior circulation collateral score (PC-CS) or posterior-circulation ASPECTS (pc-ASPECTS; Fig. 3) might contribute to the decision-making process [42]. A pc-ASPECTS $\ge$5 appears to be a reasonable cut-off even in late-presenting patients [43].

Fig. 3.

pc-ASPECTS (posterior circulation ASPECTS). The pc-ASPECTS is a 10-point score evaluating the extent of ischemia in the posterior circulation. Scores of 10 points indicate no signs of ischemia in NCCT or MRI (diffusion weighted imaging, DWI). Each thalamus, occipital lobe and cerebellar hemisphere accounts for 1 point, and the mesencephalon and pons account for 2 points. Fig. 3 shows fluid attenuated inversion recovery (FLAIR) sequences because of better image quality.

Current guideline recommendations for patient selection and treatment indications for MT (notably the European Stroke Organisation [ESO]–European Society for Minimally Invasive Neurological Therapy [ESMINT], American Heart Association [AHA]/American Stroke Association [ASA], and Chinese Stroke Association [CSA]) are summarized in Table 1 [34, 35, 36].

After the publication of the DWI or CTP Assessment With Clinical Mismatch in the Triage of Wake-Up and Late Presenting Stroke Undergoing Neurointervention With TREVO (DAWN) and Endovascular Therapy Following Imaging Evaluation for Ischemic Stroke 3 (DEFUSE-III) trials in 2018, the indications for MT for anterior-circulation stroke (ICA or MCA [M1, M2]) were expanded to patients arriving up to 24 hours after symptom onset or in an unknown time window [23, 24]. The DAWN protocol required CT-P (with a CBF-threshold) or MRI (DWI) with a subsequent predefined age-dependent clinical-core mismatch (A: $>$80 years: NIHSS $>$10, core volume 21 mL; B: 80 years, NIHSS $>$10, core volume 31 mL; C: 80 years, NIHSS $>$20, core volume 31–51 mL) [24]. DEFUSE-III required a perfusion-core mismatch on CT-P or MR-P, defined as an infarct core 70 mL, a penumbra $>$15 mL (T-max delay $>$6 seconds), and a penumbra/core ratio $>$1.8 [23]. Both RCTs demonstrated the superiority of MT to standard care in eligible patients. The findings translated well into daily practice: large retrospective studies found MT to be effective and safe even with the application of less strict selection criteria [44, 45, 46, 47].

Multimodal imaging modalities such as CT-P or MRI might not be available at all times. Thus, focusing on those modalities alone could withhold a potentially beneficial therapy from patients. Hendrix et al. [48] have reported similar outcomes in patients in early and late time windows who were treated after NCCT and CT-A (criteria: ASPECTS $>$6; retrospective analysis). This finding has been supported by data suggesting that CT ASPECTS and CT-P-parameters were similar in predicting DWI lesions in patients with acute ischemic stroke [49]. In the CT for Late Endovascular Reperfusion (CLEAR) cohort, patients presenting 6 hours after symptom onset selected for therapy after NCCT/CT-A versus CT-P or MRI did not differ in outcomes [50]. The results from a post-hoc-analysis of DEFUSE-III data have suggested that an ASPECTS of 8–10 is associated with better functional outcomes [51]. The MT treatment effects remained stable regardless of baseline ASPECTS or infarct core volume (CT-P). This finding might suggest that patients with a mismatch in perfusion imaging (and meeting the DEFUSE-III inclusion criteria) could benefit from MT regardless of ASPECTS and infarct core volume [51]. However, because of the retrospective design and consecutively introduced biases, these results must be interpreted cautiously and require future validation.

For BA occlusions, preliminary data of The Basilar Artery Chinese Endovascular Trial (BAOCHE) have been presented at the European Stroke Organisation Conference (ESOC) 2022 in Lyon, France [52]. Patients with a BA (or bilateral V4) occlusion within 6 to 24 hours of symptom onset/last-seen-well (ineligible for IVT or IVT with futile recanalization), NIHSS 6 or higher, pc ASPECTS 6 and a pons-midbrain index of two or lower were eligible. Functional outcome (mRS 0–3) was significantly higher in the MT group (46.4%; compared to 24.3%; OR 2.92 [1.56–5.47]).

Several studies have found MT plus standard care (IVT; bridging therapy) to be superior to IVT alone [18, 19, 20, 21, 22]. Most patients in the intervention arm (between 68% and 100%) were treated with IVT before MT. Subsequently, the question arose as to whether direct MT (without IVT) might lead to comparable results, thus sparing patients from potentially harmful IVT complications, such as intracerebral hemorrhage (ICH). In a HERMES collaboration analysis, MT has been found to be beneficial independently of IVT use [53].

In 2020 and 2021, three RCTs comparing direct MT and bridging therapy presented inconclusive results [54, 55, 56]. Different non-inferiority margins (NIMs) were defined in each study. In DIRECT-MT (NIM 0.8 [meaning that the lower boundary of the 95% confidence interval was 0.8 or higher]; odds ratio (OR) 1.07 [95% CI 0.81–1.40]), and Direct Endovascular Thrombectomy versus Combined IVT and Endovascular Thrombectomy for Patients With Acute Large Vessel Occlusion in the Anterior Circulation (DEVT; NIM –10% of the proportion of functional independent patients; –7.7%; OR 1.48 [0.81–2.74]) direct MT was non-inferior. In the Direct Mechanical Thrombectomy in Acute LVO Stroke (SKIP) study (NIM 0.74; 1.09 [97.5% CI 0.63 to $\mathrm{\infty }$]), non-inferiority was not demonstrated. Wide confidential intervals, the dosage of IVT (0.6 mg/kg versus 0.9 mg/kg), and an entirely Asian patient population limited generalizability. Subsequent meta-analyses including observational data provided inconclusive results regarding outcomes and treatment complications [57, 58, 59, 60, 61, 62]. However, in the case of multiple passages, IVT appears to be associated with less disability and a smaller overall stroke volume [63]. The Multicenter Randomized Clinical trial of Endovascular treatment for Acute ischemic stroke in the Netherlands (MR CLEAN)-NO IV study showed neither non-inferiority (NIM 0.8; 0.84 [0.62–1.15]) nor superiority of direct MT [64]. In 2021, preliminary results of the Solitaire™ With the Intention For Thrombectomy Plus Intravenous t-PA Versus DIRECT Solitaire™ Stent-retriever Thrombectomy in Acute Anterior Circulation Stroke (SWIFT-DIRECT) and A Randomized Controlled Trial of DIRECT Endovascular Clot Retrieval versus Standard Bridging Therapy (DIRECT-SAFE) trials were presented at the ESOC and the World Stroke Congress [65, 66]. Neither trial confirmed non-inferiority. To identify specific subgroups of patients who might benefit from one of the two therapeutic options, independent patient data meta-analyses are currently in preparation. Beyond an overall interpretation of these results, a debate is necessary regarding clinically acceptable overall non-inferiority margins. Other developments such as the effect of the neuroprotectant nerinetide (ESCPAE-NEXT; NCT04462536) in direct MT, or the use of tenecteplase rather than alteplase for bridging therapy, might further influence future decision-making [67, 68]. Current guidelines (e.g., the 2022 ESO/ESMINT recommendations on IVT before MT) do not suggest skipping IVT in eligible patients [69].

Patients in need of secondary transfer have not been included in these analyses. Longer transportation times lead to later recanalization. With the “drip-and-ship” concept, eligible patients receive IVT treatment during transfer for MT. Bridging therapy in this context is effective and safe, and has been found to be a strong independent predictor of early recanalization and favorable outcomes [70, 71, 72]. A potential fragmentation or distal translocation of the thrombus caused by IVT appears to be associated with better functional outcomes, possibly because of smaller final infarct size [73].

No RCT has investigated bridging therapy in BA occlusion. However, a recent cohort study has demonstrated the superiority of the bridging concept [74]. The findings must be confirmed in future trials.

Early recanalization in embolic LVO can be spontaneous or an effect of IVT. Analyses in the Endovascular treatment for Small Core and Anterior circulation Proximal occlusion with Emphasis on minimizing CT to recanalization times (ESCAPE) trial have indicated that early recanalization (demonstrated in an 8-hour follow-up CT-A) is crucial for functional outcomes [75]. Recanalization occurs in approximately 40% of patients treated with IVT. IVT recanalization has been found to depend on the clot length, collaterals and localization: 4.4% ICA, 32.3% M1, 30.8% M2 and 4% of BA occlusions [3]. In the case of bridging therapy, up to 10% of patients recanalized as detected in digital subtraction angiography [76]. In medium-size vessel occlusion, only 50% of patients experience early recanalization. This result translates into outcomes: only every second patient achieves an excellent outcome, defined by a modified Rankin scale (mRS) score of 0–1 [77]. If IVT leads to early recanalization, the outcome is favorable. However, the overall recanalization rates are lower than those with (additional) MT [18, 19, 20, 21, 22].

In 2003, Higashida et al. [96] developed the Thrombolysis in Cerebral Infarction (TICI) grading system for evaluating the therapeutic success of IVT (Fig. 4 and Table 3, Ref. [96, 97, 98, 99]). The TICI score is derived from the Thrombolysis in Myocardial Infarction (TIMI) risk score. TICI scores of 0 or 1 indicate no or limited perfusion, respectively. TICI scores of 2a and 2b describe anterograde reperfusion of less or more than half of the occluded target artery previously ischemic territory, respectively. A TICI score of 3 indicates complete reperfusion without any visible distal vessel occlusion. The TICI system has been adapted and modified (mTICI, which is commonly used in both the literature and routine clinical practice) to include an additional TICI 2c category indicating near-complete perfusion except for slow flow or distal emboli in several distal cortical vessels [97, 98]. An excellent reperfusion outcome is defined by mTICI scores of 2c/3. The HERMES collaborators described the expanded TICI (eTICI) score in 2019 [99]. MRS-shift analyses at 90 days after MT have suggested differences in outcomes depending on the percentage of recanalized brain tissue (Table 3). The eTICI has been found to be an independent predictor of outcomes.

Fig. 4.

Modified Thrombolysis in Cerebral Infarction (mTICI) grading system for evaluating the therapeutic success of IVT. No perfusion of the right MCA - mTICI 0 (A). Antegrade reperfusion past the initial occlusion with only filling of a temporal branch of the right MCA - mTICI 1 (B). Antegrade reperfusion of only the superior division of the right MCA - mTICI 2a (C). Antegrade reperfusion of more than half of the previously occluded right MCA territory with persistent filling defect parieto-occipital - mTICI 2b (D). Near complete perfusion except for some distal emboli in several distal cortical vessels frontal and occipital - mTICI 2c (E). Complete antegrade reperfusion of the previously occluded right MCA - mTICI 3 (F).

First-pass reperfusion (FPR) describes the effect of excellent reperfusion following the first thrombectomy device pass (i.e., the first attempt to re-open the vessel). It was initially described with the Solitaire stent retriever in anterior circulation stroke. Many studies have found an association of FPR with outcomes [100, 101]. The time taken for each additional pass, and consequent continued growth of the infarction, makes an excellent functional outcome less likely, even in the case that excellent reperfusion is finally achieved [102, 103, 104]. In retrospective analyses, factors such as the site of the vessel occlusion (M1), door to groin time, and baseline ASPECTS have been found to influence FPR [105, 106, 107]. A meta-analysis by Abbasi et al. [106] has not detected differences in FPR’s independence of thrombectomy techniques (stent retriever, ADAPT, or combined use [e.g., Solumbra]).

Likewise, FPR is an independent predictor of functional outcomes in patients with posterior circulation stroke caused by an occlusion of the BA or the dominant vertebral artery [108, 109, 110]. Ultimately, independently of the device, technique, or site of the occluded vessel, FPR should be the therapeutic goal in acute ischemic stroke requiring MT.

In traditional stent retriever MT, the stent (attached to a wire) is introduced via a micro-catheter. At the side of the occlusion, the stent is released from the catheter and subsequently self-expands, pushing the thrombus against the wall. As the stent is pulled back into the catheter, the thrombus is removed and retracted. Widely used devices include the Solitaire stent retriever (Medtronic, Dublin, Ireland), the Trevo retriever (Stryker, Kalamazoo, Michigan, USA), EmboTrap (Neuravi, Galway, Ireland), and the pRESET thrombectomy device (phenox GmbH, Bochum, Germany). Newly invented smaller devices such as the Tigertriever 13 (Rapid Medical, Yoqneam, Israel) have shown promising recanalization rates in distal vessel occlusions (e.g., A2, M3, and P2) [111].

Depending on the vascular anatomy and potential underlying diseases, endovascular access to the occlusion site can be challenging. In approximately 10% of patients, reperfusion is not achieved (TICI 0/1) [112]. In one-third of patients, this outcome is due to a failure to reach the targeted occlusion site because of either the supra-aortic vessel anatomy, or cervical (e.g., kinking or coiling of the ICA) or intracerebral vessel tortuosity [112, 113]. Curved or angled intracerebral vessels may lead to stent retriever failure or escape of the blood clot from the stent retriever [114]. Potential alternative strategies include the use of ADAPT in cases of angled intracranial vessels, distal access-guiding/intermediate catheters in observing cervical or intracerebral vascular tortuosity, and a coaxial technique using a small-sized diagnostic catheter over a larger-scale BGC in cases with unfavorable anatomy of the aortic arch or transradial access (e.g., aortic disease, transfemoral access failure) [112, 113, 114].

A transradial approach as a first-line strategy does not appear to differ from transfemoral access in terms of duration, accessibility, and complications, according to large retrospective analyses [115, 116]. Further investigations are needed before this method can be recommended as an alternative first-line access strategy.

ADAPT was developed as an alternative approach to perform embolectomy in acute ischemic stroke. The thrombus is removed via first-pass direct aspiration with a large-bore aspiration catheter (e.g., 5MAX ACE [Penumbra]). Because of their higher aspiration capacity, catheters with larger diameters appear to be more effective [117]. The size of an aspiration catheter enables a stent retriever rescue strategy in cases of futile recanalization. Recently invented devices, such as the MIVI Q aspiration catheter system designed to maximize the lumen size, might serve as promising future tools for distal occlusions [118]. ADAPT has been suggested to be associated with less endothelial damage than the use of stent-retrievers [119]. Whether this effect has clinical significance is unknown.

ADAPT was initially investigated in M1 and intracranial ICA occlusions, in which it has been found to decrease the time to recanalization [120, 121]. It has also been found to be effective and safe in M2 and M3, as well as BA occlusions [122, 123, 124, 125, 126]. The Contact Aspiration versus Stent Retriever for Successful Revascularization (ASTER) trial showed neither superiority nor non-inferiority of ADAPT to stent retriever MT (because the study was underpowered) whereas the Comparison of Direct Aspiration versus Stent Retriever as a First Approach (COMPASS) study did confirm non-inferiority of ADAPT in ICA, M1, and M2 occlusions (mRS 0–2; aspiration: n = 69 [52%], stent retriever: n = 67 [50%]; p [non-inferiority] 0.0014) [122, 127, 128]. The main limitations were a high percentage of rescue therapy in the ADAPT group (ASTER: 32.8%; COMPASS: 21%) and an uneven distribution of the localization of LVO (ASTER: M2 27.6% in direct aspiration versus 17.6% in the stent retriever cohort). A recent meta-analysis suggested higher recanalization rates with ADAPT [129]. However, this finding was mainly driven by observational data and did not interfere with the outcome overall. Leading the way to individualized decision-making, Liao et al. [117] have found ADAPT to be superior in embolic vessel occlusion than in occlusions associated with intracranial atherosclerosis.

Large-scale balloon guide catheters (BGC) can be placed proximally to the occlusion site. When inflated, they generate blood-flow arrest while retrieving the thrombus. BGC can be used in combination with both stent retriever MT and ADAPT [130, 131, 132]. BGC appear to be associated with improved procedural and functional outcome parameters in observational data [130, 131, 132, 133]. Studies have reported a higher FPR, shorter time to recanalization, and fewer attempts to achieve excellent reperfusion [133, 134]. Anterograde flow arrest during the retrieval of the clot (via direct aspiration or stent retriever) leads to a decrease in distal embolization and embolization to new territories [135, 136]. Therefore, BGC appear to facilitate good functional outcomes [137, 138]. Further developments such as the Walrus BGC are under investigation [139]. Current AHA/ASA guidelines recommend using BGC [35].

Combined approaches using a stent retriever and distal aspiration aim to achieve the advantages of each technique. A combination of approved stent retriever devices and large-bore aspiration catheters is used (e.g., Solumbra; Solitaire stent, and ACE [Penumbra]) [140, 141]. The large-bore aspiration catheter is advanced via a microcatheter proximally to the thrombus. The stent retriever device is placed around the thrombus, as performed in stent retriever MT. Under continual aspiration the stent is retracted into the aspiration catheter and removed (together with the aspiration catheter if resistance is felt) [140]. In alternative approaches (e.g., Stent retriever Assisted Vacuum-locked Extraction [SAVE] or Continuous Aspiration Prior to Intracranial Vascular Embolectomy [CAPTIVE]), the thrombus is captured between the catheter tip and stent retriever while both are retracted as a unit (without the stent retriever being introduced into the aspiration catheter) [10, 142, 143]. The Balloon guide with large bore Distal access catheter with Dual Aspiration with Stent-retriever as Standard Approach (BADDASS) and EmboTrap Pinched In Catheter (EPIC) techniques also involve combined retraction of the aspiration catheter and stent retriever, with additional mandatory use of a BGC [144, 145].

Observational data have indicated higher FPR, shorter groin puncture to recanalization times without increased periprocedural complications, and advantages in functional outcomes in patients with occlusion of the intracranial ICA, or the M1 or M2-segment of the MCA [143, 144, 145, 146, 147, 148]. Yet, the ASTER2-trial, published in 2021, has not indicated differences in total or near-total reperfusion (eTICI 2c/3) between a combined approach and stent retriever MT (the use of BGCs in both groups was mandatory) [149]. As noted by the authors, the study was underpowered to detect smaller but potentially relevant differences between groups. In addition, novel technical developments such as very large-bore catheters could further increase aspiration capability [149]. Switching from either stent retriever MT or ADAPT to a combined approach as part of a rescue strategy after futile recanalization might improve recanalization rates [150, 151].

In approximately 10–20% of patients with anterior circulation LVO, stroke is caused by tandem occlusions [201]. Tandem occlusions are defined as complete or near-total occlusion of the extracranial ICA with an additional intracranial LVO. This definition is somewhat limited, because the causal link between the extra- and intracranial pathology is not highlighted. Beyond an LVO, the extracranial lesion might also trigger an intracranial medium or small vessel occlusion that is not suitable for endovascular therapy but might lead to severe symptoms. Overall, two mechanisms can be causal: local atherosclerotic disease or dissection of the extracranial ICA (discussed below) [202]. Atherosclerotic stroke appears to be more severe. Several treatment approaches have been proposed, and there is no consensus regarding which strategy to choose [203]. An anterograde approach (ICA stenting followed by MT), a retrograde approach (MT followed by ICA stenting), balloon angioplasty (without permanent stent placement) together with MT, and a conservative approach with MT only in the acute stroke setting have been proposed and investigated [201, 204, 205, 206, 207]. A meta-analysis from 2018 has not indicated differences in outcomes among these approaches [206]. Recent data suggest an overall benefit of emergent stenting (retrograde approach) in terms of recanalization and outcomes [201, 207]. Larger RCTs comparing the different endovascular options are currently recruiting participants (e.g., EASI-TOC [NCT04261478]) [208]. Prior IVT (in eligible patients) appears to be safe and to further improve recanalization rates [206, 209, 210]. Bracco et al. [211] have recently reported “hemodynamic” tandem occlusion with an acute total or sub-total occlusion of the extracranial ICA without sufficient collateral compensation. Emergent stenting restoring the antegrade blood flow might be crucial [211]. This topic requires further investigation.

The main complication of stent treatment is early stent thrombosis, which has been observed in up to 19% of patients in a retrospective cohort [212]. Most of these patients were treated with a single platelet aggregation inhibitor. This finding might underscore the need to strictly follow a dual platelet aggregation regime even in patients with acute stroke [212]. However, this is associated with an elevated risk of consecutive hemorrhagic complications, because an increase in symptomatic and asymptomatic intracerebral hemorrhage with dual anti-platelet treatment has been reported [213, 214]. Other (retrospective) data have indicated that the risk might be overestimated and that these hemorrhages do not interfere with functional outcomes [215]. Whether hemorrhagic complications differ between a “per protocol” stent placement (as in the treatment of tandem occlusions) and a rescue procedure (indicating a longer treatment duration and multiple thrombectomy passages) cannot be answered.

Data on vertebrobasilar tandem occlusion are scarce [216, 217]. However, this etiology might not be uncommon [216]. Endovascular therapy, including angioplasty or stent placement, might be feasible and safe in selected patients. Further studies are warranted to detect treatment effect complications and identify patients for whom therapy is suitable.

In the Stenting versus Aggressive Medical Management for Preventing Recurrent Stroke in Intracranial Stenosis (SAMMPRIS) trial, primary stenting of intracranial stenosis has been found to be inferior to the best medical treatment [218]. In intracranial atherosclerotic LVO, stenting might be used as a rescue therapy (bail-out procedure) in the case of MT failure or early re-occlusion [219, 220]. In both anterior and posterior circulation stroke, permanent stent placement might be a strategy to secure revascularization in an otherwise unfavorable situation, in selected patients only [213, 221]. In a retrospective analysis, 44.8% of patients experienced a good functional outcome after the rescue procedure, whereas the frequency of symptomatic ICH was high (10.5%) [222].

According to current consensus, extracranial ICA dissection should be treated with either oral anticoagulation for 3–6 months or platelet aggregation inhibitors. Information on emergency extracranial ICA stenting due to dissection is limited to smaller observational studies and case reports. Most of the data have focused on endovascular strategies in tandem occlusions due to ICA dissection [223, 224, 225]. In a pooled analysis from registry data, emergent stenting in tandem occlusions has been suggested to be effective and safe, with a slight increase in mainly asymptomatic ICH [223, 224]. Although both the anterograde and the retrograde approaches might be feasible, some authors have suggested a more conservative approach with stent placement only in cases of insufficient collateralization [225, 226, 227].

Even though the outcomes in ICA dissection are favorable, outcomes deteriorate in cases of complete or near-total occlusions caused by ICA dissection [228, 229]. Smaller case series have suggested a beneficial effect of ICA stenting in those patients, with treatment aimed at reperfusion, consolidation of the hemodynamic status, and prevention of (future) emboli [228, 229].

Patient-specific characteristics such as age or pre-stroke disability should not exclude patients from endovascular therapy when imaging criteria support MT [230, 231, 232]. Although the overall outcomes in octo- and nonagenarians are poorer than those in patients below 80 years of age, these patients benefit from endovascular stroke therapy [230, 233]. Older patients and those with (unspecified) pre-stroke disability appear to show similar improvement rates to those in independently living patients, and the pre-stroke functional status can be attained [232]. In the case of active cancer, MT can have beneficial effects in selected patients, although the overall mortality is high, owing to the underlying disease [234].

Repeated MT is observed in approximately 1.5–6.6% of patients after initial endovascular therapy [235, 236, 237]. An early re-occlusion at the site of the initial MT must be distinguished from a recurrent LVO affecting the same or any other blood vessel after a period of weeks or months [236, 237, 238, 239, 240]. Risk factors associated with early re-occlusion are atherosclerotic etiology, residual thrombus material, and stenosis after MT [237, 238]. Because these patients might be at risk, prolonged and more intensive post-interventional monitoring should be evaluated [235]. Re-MT is feasible and safe, but the overall outcomes seem to be comparatively poor [237, 238, 240]. Nevertheless, endovascular therapy should not be withheld, because mRS 0–2 is observed in as many as 30–46% of patients in retrospective cohorts [241].

The main cause of a recurrent LVO after initial hospitalization is cardioembolic stroke, which is attributed mainly to a lack of anticoagulation [239, 242]. In retrospective cohorts, functional outcomes or improvements in patients with recurrent MT have been found to be similar to those after first-time MT [238, 239, 242].

RCT data are lacking regarding MT’s beneficial effects (or harms) in patients with large early infarction, defined by CT ASPECTS 6. Several studies are ongoing (e.g., TENSION [NCT03094715], SELECT-2 [NCT03876457], LASTE [NCT03811769], and TESLA [NCT03805308]) [243, 244, 245, 246]. The eligibility criteria differ in time to randomization, baseline NIHSS score, CT/DWI ASPECTS thresholds (between 0–5 and 3–5, depending on age), or imaging modalities (CT-P in SELECT-2), indicating uncertainty in patient selection. Retrospective data suggest a potential benefit of low-ASPECTS MT [247, 248, 249, 250]. This effect appears to be time-dependent [247]. In an early treatment cohort (a median 173 min from last seen well time to admission), the rate of symptomatic intracerebral hemorrhage did not increase [248]. In 2022, The Recovery by Endovascular Salvage for Cerebral Ultra-Acute Embolism–Japan Large Ischemic Core Trial (RESCUE-Japan LIMIT) was published [251]. N = 203 patients with ASPECTS 3-5 (detected on NCCT within 6 hours of symptom onset or MRI [6–24 hours of symptom onset; DWI ASPECTS without demarcation in FLAIR sequences]) were randomized to best medical treatment versus MT. IVT in a dose of 0.6 mg per kilogram body weight was allowed in eligible patients. Functional outcome was superior in the MT group (mRS 0–3; MT: 31%, control group: 12.7%; OR 2.43 [1.35–4.37]). There was an increase in overall intracranial hemorrhage but not in symptomatic intracranial hemorrhage. Generalizability is limited by an entirely Japanese population and a low percentage of patients (in both treatment groups; 28.4% in the medical treatment group, 26.7% in MT) getting IVT. Nevertheless, the study might point towards a further expansion of MT indications. Endovascular treatment in patients presenting with ASPECTS 6 requires individual and thorough decision-making, with a special focus on the duration of symptoms, patient age, and pre-stoke morbidity.

Patients presenting with a low NIHSS score at admission have been excluded in most MT trials. However, “mild” deficits, such as aphasia, hemianopia, or hemi-ataxia cumulating in an NIHSS score of 0–5, can be disabling and limit functional independence. In a retrospective French cohort study, 12% of patients with LVO and a baseline NIHSS score 5 experienced early neurological deterioration within 24 hours after IVT [252]. Deterioration was associated with poor outcomes. Successful reperfusion in low-NIHSS score cohorts with M1 and M2 occlusions has been found to lead to better short and long-term outcomes [253, 254]. A meta-analysis including data on anterior and posterior circulation stroke has demonstrated the potential superiority of MT to the best medical treatment (n = 581 patients; mRS 0–2: OR 1.68 [1.08–2.61]) [255]. Data indicating a potential benefit in specific patient populations (on the basis of the site of vessel occlusion, collateral status, or penumbral imaging) are lacking. RCTs are currently recruiting participants, such as MOSTE [NCT03796468] and ENDOLOW [NCT04167527] [256, 257].

Procedural complications during MT occur in up to 15–20% of patients [258, 259]. According to an Italian registry study, the complications include decreasing frequency distal clot embolization (7.6%), symptomatic intracerebral hemorrhage (7.4%), subarachnoid hemorrhage/arterial perforation (2.9%), dissection (1.7%), and access site complications (0.6%) [260].

Vessel perforation is one of the most feared complications and is observed in 0.9–4.9% of patients [18, 19, 20, 21, 22]. The risk increases during the occlusion site maneuver (particularly in situations with access difficulties), passing the occlusion site with a micro-wire or micro-catheter (particularly when observing resistance), and stent or aspiration catheter withdrawal [259, 260, 261]. Atherosclerosis and distal vessel occlusions have been discussed as additional risk factors, whereas MT in later time windows does not appear to increase perforation rates [262]. If a perforation occurs, the perforating device should be kept in situ, because it might at least partially occlude the vessel [258]. Further therapeutic strategies include lowering blood pressure, reversing anticoagulation, and inflating an intracranial balloon at the perforation site (for approximately 5–10 minutes) [258, 259, 260, 261]. In persistent bleeding, the perforated vessel must be sacrificed [258, 263].

A periprocedural dissection is observed in any vessel being manipulated. This complication is observed in 0.6–3.9% of MT cases [18, 19, 20, 21, 22]. Asymptomatic dissection without stenosis or a prominent dissection membrane might be treated with platelet aggregation. In the case of blood flow disturbances, stent placement is necessary [258, 264].

Embolization to new or previously non-affected vascular territories is seen in 1–8.6% of cases [18, 19, 20, 21, 22, 258, 259]. It is the most frequent procedural complication in MT. During retrieval of the thrombus parts of the clot can migrate to new vascular territories, or can break up and disseminate into tiny branches downstream, thus affecting parts of the same vascular territory that had previously been spared. Retrieval of a proximal clot increases the risk of embolization [265]. Embolization to new territories is associated with disability and mortality [266, 267]. MT in posterior circulation stroke appears to be associated with higher incidence of both distal embolization and embolization to new territories [268]. Proximal emboli might be removed immediately with the thrombectomy device in use or through aspiration. For distal embolization, intra-arterial thrombolysis might be an option [269]. BGCs appear to considerably decrease the risk of distal embolization [132, 133, 265].