Article Text

Download PDFPDF

Original research
Institutional and provider variations for mechanical thrombectomy in the treatment of acute ischemic stroke: a survey analysis
  1. Tapan Mehta1,
  2. Shailesh Male1,
  3. Coridon Quinn1,
  4. David F Kallmes2,
  5. Adnan H Siddiqui3,
  6. Aquilla Turk4,
  7. Andrew Walker Grande1,
  8. Ramachandra Prasad Tummala1,
  9. Bharathi Dasan Jagadeesan1
  1. 1 Department of Neurology, Neurosurgery and Radiology, University of Minnesota, Minneapolis, MN, USA
  2. 2 Department of Radiology, Mayo Clinic, Rochester, Minnesota, USA
  3. 3 Department of Neurosurgery and Radiology, University at Buffalo, Buffalo, New York, USA
  4. 4 Department of Neurosurgery and Radiology, Medical University of South Carolina, Charleston, South Carolina, USA
  1. Correspondence to Dr Bharathi Dasan Jagadeesan, University of Minnesota, 420 Delaware StreetS.E., Minneapolis, MN 55455, USA; jagad002{at}umn.edu

Abstract

Introduction Stent retriever combined with aspiration, or the ‘Solumbra technique’, has recently emerged as one of the popular methods of mechanical thrombectomy (MT). However, the variations in understanding and implementation of the Solumbra technique have not been reported.

Methods An 18 part anonymous survey questionnaire was designed to extract information regarding technical variations of MT with a focus on the Solumbra technique. The survey link was posted on the Society of Neurointerventional Surgery (SNIS) website in ‘SNIS connect’.

Results 80 responses were obtained over 4 weeks that were included in the final analysis. Direct aspiration without a balloon guide catheter (BGC) was the most favored technique among respondents (41.12%) followed by the Solumbra technique without a BGC (32.4%). Among those using the Solumbra technique, 77.6% reported that they wait between 2 and 5 min to allow clot engagement, 55.2% always remove the microcatheter before aspiration, and 69.1% commence aspiration through the intermediate catheter only when retrieving the stent retriever. Operators who infrequently used or did not use BGCs reported a higher incidence of >80% Thrombolysis in Cerebral Infarction 2b/3 annual recanalization rates (OR 8.85, 95% CI 2.03 to 38.55, P=0.004) compared with those who consistently used BGCs.

Conclusion Our study documents the variations in MT techniques, and more specifically, attempts to quantify variations in the Solumbra technique. The impact of these variations on recanalization rates and eventually patient outcomes are unclear, especially given the self-reported outcomes contained in this study.

  • thrombectomy
  • stroke

Statistics from Altmetric.com

Request Permissions

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.

Introduction

Endovascular treatment of large vessel stroke has evolved significantly over the past decade. The recent thrombectomy trials, compared with previous studies, have highlighted the importance of proper patient selection and modified thrombectomy techniques.1–3 Subsequently, Fargen et al showed the impact of thrombectomy trials on practice patterns.4 Along with the availability of multiple stent retrieval devices (Solitaire, Embotrap, ERIC, pREset, 3D, Phenox CRC, Aperio),5–11 several variations in the techniques for mechanical thrombectomy (MT) have been described (ie, forced arterial suction technique (FAST), a direct aspiration first pass technique (ADAPT), push and fluff (PFT), stent retriever and proximal aspiration thrombectomy (Solumbra), manual aspiration technique (MAT), dual aspiration technique (DAT), stent retrieving into an aspiration catheter with proximal balloon (ASAP) technique, thrombectomy using ‘clamping embolus with semi-retrieval’ technique (TCET), advancing distal access catheter over stent retriever technique (ADVANCE), proximal balloon occlusion together with direct thrombus aspiration technique (PROTECT), stent retriever assisted vacuum locked extraction (SAVE), and bare wire thrombectomy (BWT).12–20

Stent retriever combined with aspiration, or the ‘Solumbra technique’, has recently emerged as one of the popular methods of MT. However, the variations in understanding and implementation of the Solumbra technique have not been reported. We attempted to evaluate such variations in practices when performing the Solumbra technique with a questionnaire based survey.

Methods

A questionnaire survey with 18 questions was designed using Qualtrics software available through our institution. The survey questions were aimed at extracting information regarding technical variations of MT with a focus on the Solumbra technique. The survey allowed individuals to respond once while maintaining complete anonymity. After obtaining institutional review board (IRB) approval, the survey link was posted on the Society of Neurointerventional Surgery (SNIS) website in ‘SNIS connect’, an exclusive member only forum of SNIS.

The rationale for the survey questions was as follows. Although several techniques are described for MT, for the survey purposes, we broadly categorized them as: (1) Solumbra technique without a balloon guide catheter (BGC); (2) Solumbra technique with a BGC; (3) stent retriever without aspiration with a BGC; (4) direct aspiration alone without a BGC; and (5) direct aspiration alone with a BGC. In addition, the variations in key steps while executing the Solumbra technique, such as aspiration duration, timing of stent retriever deployment, and the retrieval method of the stent, were surveyed. Finally, self-reported thrombectomy outcomes in the form of modified Thrombolysis in Cerebral Infarction (TICI) score were collected (appendix for the questions). We also tried to understand the distribution of various other factors, including operator experience, practice volume, and type of devices, in our sample.

Survey data were stratified based on the method of MT for description of frequencies of responses. Fisher’s exact test for frequency <5 and the χ2 test for frequency >5 were applied to determine significant differences (P<0.05) for univariate analysis (table 1). We calculated attributable fractions (figures 1 and 2) in the context of self-reported annual recanalization rates (dichotomized by modified TICI 2b/3 <80% or  ≥80%) to further understand the variations in responses for use of devices and techniques. We performed a multivariate binomial logistic regression analysis to further understand whether use of a BGC during the Solumbra technique is associated with better self-reported outcomes after adjusting for operator experience, volume of procedures, and background training. This analysis was inspired by recent evidence on higher success rates of recanalization with the use of a BGC in conjunction with the Solumbra technique. Similarly, multivariate regression analysis for predictors of BGC use and pure aspiration thrombectomy were also performed.

Figure 1

Relationship between device use during the Solumbra technique and self-reported Thrombolysis in Cerebral Infarction 2b/3 outcomes for >80% of cases.

Figure 2

Relationship between technical nuances during the Solumbra technique and self-reported Thrombolysis in Cerebral Infarction 2b/3 outcomes for >80% of cases. BCG, balloon guide catheter.

Table 1 Baseline characteristics of mechanical thrombectomy techniques

Results

Overview of variations in technique

Over a period of 4 weeks, 80 responses were obtained that were included in the final analysis. All survey responses were included in the final analysis. The survey results revealed that 60.3% of respondents were neuroradiologists, 17.6% neurologists, 16.1% neurosurgeons, and 5.9% from other or combined specialties. The majority of the respondents worked in an academic (40%) or semi-academic (25.7%) setting, and 34.3% worked in private settings. We found that 68% of respondents had been in practice for >5 years and 39.1% worked at a high volume center where at least 100 thrombectomies are performed every year. Direct aspiration without a BGC was the most favored technique among respondents (41.12%) followed by the Solumbra technique without a BGC (32.4%) and the Solumbra with a BGC (11.8%). ACE 68 (Penumbra Inc, Alameda, California) was the most preferred intermediate catheter (61.5%) while a wide variety of microcatheters were used during MT.

Variations in the Solumbra technique

Among those using the Solumbra technique, following stent retriever deployment, 77.6% of respondents reported that they wait between 2 and 5 min to allow clot engagement, during this time the microcatheter is always removed before aspiration by 55.2%, and left in place by the rest. With regards to aspiration, the majority of operators (69.1%) commence aspiration through the intermediate catheter only when retrieving the stent retriever, whereas others perform aspiration throughout and the overwhelming majority (95.5%) of respondents perform aspiration after positioning the intermediate catheter at the face of the clot. The relationships between the different devices and these technical nuances used during the Solumbra technique with perceived thrombectomy outcomes are depicted in figures 1 and 2. After adjusting for operator experience, practice type, and volume of procedures, practitioners not using a BGC with the Solumbra technique were more likely to self-report TICI 2b/3 rates >80% in their clinical practice compared with those using BGCs (table 2). The operators' background training, operator experience, thrombectomy volume, and academic practice setting did not predict the use of a BGC or pure aspiration thrombectomy in multivariate logistic regression analysis (table 3 and table 4),

Table 2

Multivariate logistic regression model for outcome of Thrombolysis in Cerebral Infarction 2b/3 for >80% of thrombectomies

Table 3

Multivariate logistic regression model for predictors of balloon guide catheter use

Table 4

Multivariate logistic regression model for predictors of pure aspiration versus stent thrombectomy

Discussion

MT for acute ischemic stroke has been shown to significantly improve patient outcomes compared with more conservative management, especially since the advent of the latest generation of devices, along with improved patient selection criteria.21 These benefits have now been shown to be sustained for up to 24 hours following symptom onset in carefully selected patients.22 However, the impact of variations in MT techniques on patient outcome in the presence of standardized patient selection criteria is less well known. This is of particular interest as a standardized technique for MT has yet to be identified and newer devices/techniques are rapidly being introduced. Currently, technical and device considerations for MT are not guided exclusively by the results of published studies with level I evidence but rather by anatomical factors such as vessel tortuosity, diameter of the occluded vessel, length of the clot, supposed composition of the clot with underlying etiology, cost, ease of use, as well as operator preference. Our survey was intended to document and if possible analyze the impact of technical variations in the conduct of MT in a real world situation on procedure success rates and perhaps indirectly on patient outcomes.

Our results showed that there was significant variation in the broad classes of MT techniques (aspiration vs Solumbra vs stent retriever vs BGC) used by operators. Our results also showed (table 1) that factors such as operator experience or volume of thrombectomies performed annually were not associated with the use of one particular technique. More interestingly, there was wide variation in technique even within one of these classes (ie, the Solumbra technique) to the point where it is perhaps difficult to precisely define what the Solumbra technique is, except as a combination of a stent retriever and aspiration with or without a BGC, with any type of temporal intervals or device combinations being included in this definition. Such a variation makes the identification of best practices with this technique very difficult. We attempted to quantify outcomes with each variation in terms of self-reported annual per cent of TICI 2b/3 in order to identify specific technical nuances to which greater success rates can be attributed, and which could point towards a best practice technique.

Some of the factors to which successful outcomes could be attributed with the Solumbra technique seemed quite logical—for instance, 95.5% of respondents performed aspiration after positioning the intermediate catheter at the face of the clot and their attributable fraction in the context of self-reported annual recanalization rates (TICI 2b/3  ≥80%) was significantly higher than average compared with operators who kept the intermediate catheter between the stent retriever and guide catheter (figure 2). On the other hand, some of our results contradicted established wisdom with regards to successful technical nuances. In particular, after adjustment for potential confounders (operator experience, teaching status of the practice, and thrombectomy volumes), operators who infrequently used or did not use BGCs reported a higher incidence of >80% TICI 2b/3 annual recanalization rates (OR 8.85, 95% CI 2.03 to 38.55, P=0.004) compared with those who consistently used BGCs. This contradicts the results from a recent thoughtfully analyzed retrospective data set as well as the results of numerous other basic science and clinical studies which show a reduced incidence of distal secondary emboli and increased incidence of TICI 2b/3 recanalization with the use of BGCs.23–25 Another set of factors which could have been expected to have had an impact on TICI 2b/3 rates, such as duration of aspiration during the Solumbra technique, were found to have no influence on self-reported outcomes. A more nuanced analysis of this last described factor however showed that perhaps there is an optimal duration for aspiration (3–4 min), and that shorter or longer intervals are less likely to result in successful recanalization. Whether similar optimal quantitative parameters can be identified for other technical aspects, such as the magnitude of negative suction or the optimal relative sizing of the occluded vessel and the aspiration catheter, remains to be seen.

The presence of anomalous or counterintuitive results in our survey results can be at least partly explained by some of the limitations associated with a report of this nature. Perhaps the biggest limitation lies in the fact that the outcomes were self-reported by the operators themselves without the moderating and standardizing influence of a core lab. Recently, it has been shown that operators tend to significantly over estimate their TICI scores compared with a Core lab for the same set of images.26 A non-uniform distribution of this bias between the different operators could well be reflected in the results. Another limitation lies in the retrospective nature of the study, which makes it impossible to account for the impact of factors such as anatomical variations encountered in individual patients on the TICI 2b/3 rates. It is also very important to mention that the survey was only accessible to members of the SNIS and members of other societies or those from other parts of the world might have different practice patterns. However, our study sample is fairly representative of physicians from different backgrounds and practice settings.27 28 Even with these limitations, we believe that this survey gives a broad overview of the extremely wide variations in the techniques for the performance of MT in a real world setting. It also suggests that the impact of these variations on the outcomes of MT procedures can be difficult to predict based on extant literature. It will be extremely important to continue to more carefully study the impact of these technical variations, perhaps in a prospective manner given that recent data suggest that even the difference between TICI 2b and TICI 3recanalization can have significant impact on eventual patient outcomes.29

In conclusion, although limited by not comprehensively including all MT practitioners, our study reveals that there are wide variations in MT techniques, particularly in the Solumbra technique. The impact of these variations on recanalization rates and eventually patient outcomes are unclear, especially given the self-reported outcomes contained in this study. Therefore, it will be crucial for operators to objectively assess their outcomes and for practitioners in general to continue to evolve towards an optimal MT technique or set of techniques.

Acknowledgments

None

References

Footnotes

  • Contributors TM, SM, and BDJ: study design, data acquisition/analysis, and drafting and revising the article. CQ, DFK, AHS, AT, AWG, and RPT: study design, and review and revision of the article.

  • Funding The authors have not declared a specific grant for this research from any funding agency in the public, commercial, or not-for-profit sectors.

  • Competing interests None declared.

  • Patient consent Not required.

  • Ethics approval The study was approved by the institutional review board.

  • Provenance and peer review Not commissioned; externally peer reviewed.