Endovascular Device Choice and Tools for Recanalization of Medium Vessel Occlusions: Insights From the MeVO FRONTIERS International Survey

Background: Endovascular treatment (EVT) for stroke due to medium vessel occlusion (MeVO) can be technically challenging. Devices and tools are rapidly evolving. We aimed to gain insight into preferences and global perspectives on the usage of endovascular tools in treating MeVOs. Methods: We conducted an international survey with seven scenarios of patients presenting A3, M2/3, M3, M3/4, or P2/3 occlusions. Respondents were asked for their preferred first-line endovascular approach, and whether they felt that the appropriate endovascular tools were available to them. Answers were analyzed by occlusion location and geographical region of practice, using multinomial/binary logistic regression. Results: A total of 263 neurointerventionists provided 1836 responses. The first-line preferences of physicians were evenly distributed among stent-retrievers, combined approaches, and aspiration only (33.2, 29.8, and 26.8%, respectively). A3 occlusions were more often treated with stent-retrievers (RR 1.21, 95% CI: 1.07–1.36), while intra-arterial thrombolysis was more often preferred in M3 (RR 2.47, 95% CI: 1.53–3.98) and M3/4 occlusions (RR 7.71, 95% CI: 4.16–14.28) compared to M2/3 occlusions. Respondents who thought appropriate tools are currently not available more often chose stent retrievers alone (RR 2.07; 95% CI: 1.01–4.24) or intra-arterial thrombolysis (RR 3.35, 95% CI: 1.26–8.42). Physicians who stated that they do not have access to optimal tools opted more often not to treat at all (RR 3.41, 95% CI: 1.11–10.49). Stent-retrievers alone were chosen more often and contact aspiration alone less often as a first-line approach in Europe (RR 2.12, 95% CI: 1.38–3.24; and RR 0.49, 95% CI 0.34–0.70, respectively) compared to the United States and Canada. Conclusions: In EVT for MeVO strokes, neurointerventionalists choose a targeted vessel specific first-line approach depending on the occlusion location, region of practice, and availability of the appropriate tools.


INTRODUCTION
Given the high efficacy of endovascular treatment (EVT) for acute ischemic stroke due to large vessel occlusion (LVO) and recently recognized substantial morbidity associated with stroke due to medium vessel occlusions (MeVO; distal M2/3, A2/3, P2/3 vessel segments) (1), EVT is now increasingly considered as a treatment for MeVO stroke (2), despite the lack of highlevel evidence for MeVO EVT (3,4). The smaller caliber, more distal location, and longer and more tortuous course of the affected vessels of MeVO compared to LVO makes EVT for MeVO stroke more challenging. Thinner, more fragile arterial walls could increase the risk of dissection, perforation, and vasospasm-complications that could offset any benefit of EVT (3,4).
Currently, EVT tools and techniques are rapidly evolving, resulting in improved efficacy and safety of MeVO EVT. Several authors report promising results of primary aspiration as a firstline approach in MeVO stroke (5)(6)(7)(8)(9). Mini stent-retrievers are designed specifically for more distal occlusion locations, and novel approaches like the blind exchange mini-pinning (10,11) technique may lead to higher rates of first-pass recanalization and a lower incidence of symptomatic intracranial hemorrhage compared to the use of mini stent retrievers alone (10).
However, in light of these developments, and in the absence of guideline-based treatment recommendations, clinical practice with regard to EVT techniques for MeVO stroke may vary greatly between countries or individual physicians. Currently, there are little data on the variability in MeVO EVT approaches. Therefore, we sought to determine global patterns in preferences and utilization of EVT devices in MeVO stroke. In addition, we explored interventionalists' access to the appropriate EVT devices, and whether they thought that appropriate tools already exist and are available to them in their current practice.

METHODS
We conducted an international, cross-sectional, anonymous, invitation-only survey: MeVO-FRONTIERS (MeVO-Finding Rationales and Objectifying New Targets for IntervEntional Revascularization in Stroke). Approximately 1,400 stroke physicians from 44 countries were invited to participate in this survey through Qualtrics (www.qualtrics.com). There were no restrictions for respondents based on country, years of experience, career stage, or hospital setting. The current study analyzes the survey questions on EVT technique and includes responses from interventionalists who identified themselves as neuroradiologists, neurosurgeons performing endovascular procedures, and interventional neurologists.
Response data were obtained from November 12, 2020 to December 31, 2020. Data are available from the corresponding author upon request. Approval by the local research ethics board of the University of Calgary was obtained (REB20-2086).

Survey Design
The survey consisted of seven narrative MeVO cases with illustrative images and three to six clinical case vignettes per scenario. The case vignettes included patient demographics, clinical symptoms, radiological images, and imaging-derived information like CT-perfusion volumes or early ischemic changes on non-contrast CT. At the end of each case, physicians were asked what their preferred first-line EVT approach for that particular case would be. Answer options were (a) stentretriever alone, (b) aspiration alone, (c) combined stentretriever and aspiration, (d) intra-arterial thrombolysis, or (e) no treatment. Participants were then asked whether they thought that optimal tools for treating MeVOs with EVT currently exist (Yes/No/There is substantial scope for improvement) and whether the appropriate material for MeVO EVT is available in their institution (full survey questions shown in Supplementary Methods). Prior to accessing the case scenarios, physicians provided basic personal data (age group, gender, subspecialty, years of experience in stroke treatment, annual center stroke treatment volume, and geographic region).

Statistical Analysis
Respondents' baseline characteristics were described using appropriate descriptive statistics. Univariable multinomial regression clustered by participant was used to assess the effect of occlusion site and respondent characteristics on preferred first-line EVT approach. Incidence rate ratios (IRRs) with 95% confidence intervals (CI) were reported.
For the following analysis, first-line approach was dichotomized into dummy variables (e.g., stent-retriever vs. others, aspiration vs. others, etc.). Then, preferred first-line EVT approach by occlusion site (M2/3, M3, M3/4, A3, P2/3) and region of practice (USA and Canada, Europe, rest of the world) were analyzed using binary logistic regression clustered by participant to calculate risk ratios (RRs) with 95% confidence intervals. M2/3 occlusion and responses from practitioners from the USA and Canada were chosen as reference values.
Multinomial regression analysis was used to determine treatment approach preference based on the existence and availability of specific endovascular tools and whether the interventionists thought they had adequate access to them in their current practice. In the multinomial regression model, the combined technique (stent retriever together with the contact aspiration) was chosen as the reference value and responses "Yes, tools exist" and "Yes, tools are available" were taken as base outcome.
The significance level was set at p < 0.05. Data analyses were performed in Stata 16.1 (StataCorp, College Station, TX, USA).

RESULTS
In total, 263 of 366 stroke physicians who participated were neurointerventionists and were included in the current study. Seven clinical scenarios (two M2/3 and two A3 cases; one M3, M3/4, and P2/3 case) resulted in 1,836 responses on treatment approaches for MeVOs. Detailed respondent characteristics are listed in Supplementary Table 1.

First-Line Device Preference by Occlusion Site
First-line treatment approach for each occlusion site is shown in Figure 1 and Table 1. Preference for stent-retrievers alone did not differ between middle cerebral artery (MCA) occlusions (M2/3, M3, M3/4; Supplementary Table 2). In A3 occlusions, stent-retrievers were the most commonly preferred approach [39.4%, 207/525 responses; RR 1.22, 95% CI 1.08-1.37 (reference category: M2/3)]. For P2/3 occlusions, the combined approach (stent-retriever and aspiration) was frequently chosen (37% in Two-hundred-and-three (77.2%) neurointerventionists indicated that they had access to the best available tools, and 60 (22.8%) stated that they did not (always) have access to the ideal tools. Those without access to the optimal tools more frequently chose no EVT at all as a first-line approach [RR 3.41, 95% CI 1.11-10.49 (reference category: having the access to the best available tools)] ( Table 3).

DISCUSSION
This survey study found that the first-line device preferences of neurointerventionists for EVT in MeVO stroke vary based on the exact location of the occlusion, physician's region of practice, whether they think that adequate tools exist, and whether they have access to these tools in their current practice.
The optimal strategy for recanalizing MeVOs is currently not known, and the data on the efficacy of first-line aspiration vs. stent-retriever techniques are heterogenous and exclusively from non-randomized studies (12)(13)(14)(15). Use of stent-retrievers, either alone (16)(17)(18)(19) or combined with aspiration (10, 11), seems to be a widespread approach for MeVO EVT. Stent-retrievers alone or combined stent-retrievers and aspiration were the most commonly chosen first-line approaches, with the combined approach being preferred less often in more distal occlusions and stent-retriever alone more often in A3 occlusions.
EVT device choice in MeVO stroke is likely largely determined by the device's safety profile. Because the affected vessels in MeVO strokes are smaller, more distal, and more fragile, the risk of complications such as vasospasm, manipulation-related subarachnoid hemorrhage, and dissection is increased (3). These risks should be mitigated in order for EVT to result in a net benefit for the patients undergoing recanalization. Current  large bore aspiration catheters may increase the risk of vascular damage, which is reflected in the decreased preference for firstline aspiration in more distal MCA occlusions in our study. Indeed, most intermediate catheters are 5-6Fr in diameter, and thus, they may be too large for MeVOs, considering the average diameter of the M2 middle cerebral artery segments is around 1.4-2.3 mm (20). Furthermore, when navigating the aspiration catheter to distal occlusion sites, there is the possibility of the device getting stuck at a vessel branch point, e.g., the middle cerebral artery bifurcation, although this risk can be mitigated by the use of wedge-shaped microcatheters (21), or coaxial microcatheters better size-matched to the aspiration catheter so as to reduce the step in transition. When using a primary combined approach on the other hand, limited lengths of currently available intermediate and microcatheters may render distal occlusions unreachable. For stent-retrievers, the tortuosity and angle of the arteries may affect safety because of increased shearing at branch points during stent retrieval as well as and displacement of the arterial tree, both of which may result in subarachnoid hemorrhage. Use of stent-retrievers may hence be more desirable in straight arteries, such as the A3 where there is less tortuosity compared to the MCA branches (22).
Combined aspiration and stent-retrieval was shown to reduce the risk of distal embolization in large vessel occlusion stroke in some studies, with subsequent improved reperfusion quality and clinical outcomes (21). We found that a substantial number of physicians preferred this combined approach in the more proximal MeVO locations (M2/3 and P2/3). Recently proposed techniques such as blind exchange mini-pinning (11), in which the aspiration catheter is advanced introduced over the bare pusher wire once the stent-retriever is deployed, can circumvent problems related to catheter length and, at the same time, provide effective aspiration during the retrieval process. Techniques and specific tools for medium-sized arteries, such are mini stentretrievers, are under development (17,23); thus, it can be expected that more data on the safety and efficacy of these techniques will emerge soon.
Access to neurovascular tools and materials plays an important role in EVT decisions and first-line device choice for treating MeVOs as observed in this study. When practitioners thought that the appropriate tools do not exist, they more often chose to treat with stent-retrievers alone or opt for intra-arterial thrombolysis as a first-line approach. Those that felt they did not have access to the best available tools in their practice often chose not to treat at all.
Overall, interventionalists from Europe more often opted for stent-retrievers or combined stent-retrievers and aspiration as a first-line approach, whereas direct aspiration only was the more frequently preferred first-line approach in the USA and Canada. Availability of material and devices in different regions as well as local experience with these tools could potentially account for this variation in physicians' preferences as suggested in previous studies (24), in which the willingness to treat M2 occlusions increased under assumed ideal conditions in some regions. There is variability in the distribution and supply of stent retrievers across the world with some centers having access only to earlier generation devices, although device availability per country and center is hard to check and changing quickly. Physicians with no access to the optimal devices more often opted either to use stentretrievers alone or not treat with EVT at all forgoing endovascular treatment that could potentially benefit the patient.

LIMITATIONS
Our study has several limitations. First, decisions in endovascular treatment are highly dependent on details of patient anatomy and factors such as patient motion during EVT. Radiologic images were presented with all case scenarios to make them as realistic as possible; however, details in these images or cases may limit the generalizability of our study results to real clinical practice. Secondly, the landscape of EVT materials and tools changes fast, hence the results represent a snapshot in time and availability would differ as the tools continue to evolve in each region. In addition, the survey did not ask for the specific device brands that were available at respondents' institutions. As such, we do not know the exact EVT materials on which our results reflect, other than the devices that are currently approved in general for EVT. Our intention was to provide a general overview of the field. Thirdly, our respondent sample was collected through personal and professional networks of the study authors, which may introduce selection bias in the results (e.g., overrepresentation of teaching hospitals). Models of financial compensation for MeVO EVT may also have differed between respondents' practice settings.

CONCLUSION
In this study, neurointerventionalists chose a targeted vesselspecific approach when treating MeVOs. Stent-retrievers alone or combined stent-retriever and aspiration were the most commonly used first-line approach, with the combined approach being preferred less often in more distal occlusions and stentretriever alone more often in A3 occlusions as a first-line approach. Interventionalists from Europe used stent-retrievers and combined stent-retriever and aspiration more often as a first-line treatment, whereas direct aspiration only was more frequently preferred in North America. Physicians without access to the optimal devices more often used stent-retrievers alone or chose not to treat endovascularly at all, forgoing potential benefit.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

ETHICS STATEMENT
The Conjoint Health Research Ethics Board of the University of Calgary reviewed and approved this study (REB20-2086). Written informed consent to participate in this study was provided by the participants. Written informed consent was obtained from the individuals for the publication of any potentially identifiable images or data included in this article.

AUTHOR CONTRIBUTIONS
NK: study conception, data collection, graphical analysis, design of the work, interpretation of data, and manuscript drafting. PC, MG, and MK: study conception, data collection, design of the work, statistical analysis, interpretation of data, and manuscript drafting. JO: study conception, data collection, design of the work, interpretation of data, and manuscript editing. NSi: study conception, data collection, design of the work, statistical analysis, interpretation of data, and manuscript editing. MA: study conception, interpretation of data, and critical manuscript revision. JR: critical manuscript revision. JF and MC: study conception, data collection, design of the work, interpretation of data, and critical manuscript revision. NSa: study conception, data collection, and critical manuscript revision. RA: manuscript revisions. MH: manuscript drafting and critical manuscript revisions. All authors contributed to the article and approved the submitted version.