Article Text
PDF
Abstract
Background Endovascular treatment has become the standard care for acute basilar artery occlusion (BAO). Uncertainty persists about the optimal thrombectomy technique.
Objective To compare aspiration thrombectomy with stent retriever thrombectomy in patients with BAO in a multicenter real-world patient population.
Methods We analyzed data from the German Stroke Registry-Endovascular Treatment (GSR-ET). Patients with isolated BAO who underwent either aspiration or stent retriever thrombectomy were compared, including propensity score matching (PSM). The primary outcome measure was the modified Rankin Scale shift analysis at 90 days. Secondary outcomes included symptomatic intracranial hemorrhage (sICH), procedure complications, and metrics.
Results Of 13 082 patients in the GSR-ET, 387 patients (mean age 72.0±13.1 years; 45.0% female) fulfilled the inclusion criteria. The thrombectomy technique was aspiration only in 195 (50.4%) and stent retriever only in 192 (49.6%) patients. Functional outcome did not differ between the groups, either before (common OR (cOR) 0.94; 95% CI 0.64 to 1.38) or after PSM (cOR=1.37; 95% CI 0.90 to 2.09). There was no significant difference in sICH (2.6 vs 5.5%; P=0.231; OR=0.46; 95% CI 0.14 to 1.47), but aspiration thrombectomy demonstrated fewer procedure-related complications (4.6% vs 12.5%; P=0.017), a shorter procedure duration (24 vs 48 min; P<0.001), and higher first pass recanalization rates (75.1% vs 44.8%; P<0.001).
Conclusions In this study both aspiration and stent retriever thrombectomy showed equal efficacy in terms of functional outcome in patients with BAO. However, procedure complications and metrics might favor aspiration over stent retriever thrombectomy.
- Stroke
- Thrombectomy
- Technique
Data availability statement
Data are available upon reasonable request. The data that support the results of this study are available from the corresponding author upon reasonable request.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
WHAT IS ALREADY KNOWN ON THIS TOPIC
Endovascular treatment has emerged as the standard care for acute basilar artery occlusions. However, uncertainty persists about the optimal revascularization technique. In this study, we compared aspiration thrombectomy with stent retriever thrombectomy in a real-world-setting.
WHAT THIS STUDY ADDS
We found that both thrombectomy techniques were equally efficient in terms of functional outcome. However, aspiration thrombectomy appears to be faster, has fewer procedure-related complications, and leads to higher first pass recanalization rates.
HOW THIS STUDY MIGHT AFFECT RESEARCH, PRACTICE OR POLICY
First-line aspiration might be favored over stent retriever thrombectomy in acute basilar artery occlusion.
Introduction
Endovascular treatment (ET) in patients with acute basilar artery occlusion (BAO) has recently been investigated in four randomized controlled trials.1–4 Current evidence indicates that ET is particularly effective in patients with BAO with moderate to high National Institutes of Health Stroke Scale (NIHSS) scores and in those who did not receive intravenous thrombolysis compared with secondary prevention treatment.5 In contrast to ET for anterior circulation strokes, where both aspiration catheters and stent retrievers exhibit similar safety and efficacy, uncertainty exists about the optimal revascularization technique in posterior circulation vessel occlusions.6 In the BAOCHE Trial the Solitaire stent retriever was primarily used as the thrombectomy device, while patients in the ATTENTION trial received a variety of thrombectomy approaches, with the majority undergoing a combined technique using both aspiration catheters and stent retrievers. Therefore, no randomized controlled data exist directly comparing aspiration and stent retriever ET in patients with a posterior circulation stroke.
Most observational data, including two recent meta-analyses, have consistently shown similar functional outcome in both aspiration and stent retriever thrombectomy.5–8 One study, however, reported a superior functional outcome with aspiration catheter ET,7 and one meta-analysis, including nine studies showed a lower risk of mortality with aspiration thrombectomy.8 Moreover, some studies reported higher rates of symptomatic intracranial hemorrhage (sICH) with stent retriever thrombectomy.9–11 Current evidence presents mixed results for procedural parameters: aspiration thrombectomy was generally faster and associated with fewer recanalization attempts in most studies, with varying results regarding recanalization rates and procedural complications for both techniques.12–14
Here we compare aspiration with stent retriever thrombectomy in patients with BAO in a large multicenter and real-world patient population.
Methods
Patients
We used data obtained from the German Stroke Registry-Endovascular Treatment (GSR-ET; ClinicalTrials.gov Identifier NCT03356392). This ongoing academic-driven registry prospectively enrolls consecutive patients with acute ischemic stroke with a large vessel occlusion undergoing ET from 25 centers across Germany. Details have been described previously.15 16 Clinical decisions for treatment with ET and intravenous thrombolysis (IVT) were made in accordance with current national and international guidelines.17 18 IVT was administered using alteplase, whereas tenecteplase was not used in any participating center. Clinical and radiological assessments of all patients were conducted by local neurologists and neurointerventionalists at each center. The modified Rankin Scale (mRS) score at 90 days was assessed either by on-site visit or by telephone interview. All data were checked for plausibility, integrity, and completeness using a standardized computed protocol, and queries were sent to the respective centers in instances of inconsistent data.
The study was centrally approved by the ethics committee of the LMU Munich (protocol: 689–15), as the leading ethics committee. Further approval was obtained from local ethics committees according to local regulations. Data acquisition in the registry was approved under a waiver of informed consent by the local institutional review boards at each participating center. This study is reported in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines (see online supplemental material). The corresponding author had full access to all the data in the study and takes responsibility for its integrity and the data analysis. The data that support the results of this study are available from the corresponding author on reasonable request.
Supplemental material
We enrolled adult patients registered between June 2015 and December 2021, who underwent ET for isolated BAO, as determined by baseline CT angiography or magnetic resonance angiography and digital subtraction angiography (figure 1). Subjects with additional occlusion of a vertebral or posterior cerebral artery, or within the anterior circulation were excluded. Patients for whom both aspiration and stent retrievers were used in combination (cross-over) and patients with incomplete data (n=31) on the thrombectomy technique were excluded. An analysis was conducted between patients treated with aspiration, those treated with stent retriever, and those undergoing the combined (cross-over) technique, comparing baseline characteristics and outcome parameters (see online supplemental tables 10 and 11). In subjects with unknown stroke onset, we determined the median time between the time of last proof of good health and the time of recognition as stroke onset and used it for further calculations.
Our study included and compared a total of n=387 patients before propensity score matching, with n=195 and n=192 undergoing aspiration-only and stent retriever-only thrombectomy, respectively. Matching resulted in n=323 patients: 195 receiving aspiration and 128 receiving stent retriever thrombectomy. The choice of thrombectomy method was at the discretion of the neurointerventionalists. Aspiration thrombectomy was performed using the SOFIA aspiration catheter system in most of the subjects (n=198; 96.9%). Stent retriever thrombectomy was most frequently carried out with the Solitaire (n=80; 42.7%), followed by the pRESET (n=52; 27.1%) and the APERIO system (n=21; 10.9%). A total of 15 patients (7.8%) who underwent stent retriever thrombectomy were treated with a combination of more than one stent retriever. See online supplemental material for a comprehensive list of all thrombectomy devices used in each group. Peri-interventional intra-arterial medication included administration of alteplase and nimodipine.
Patient selection flowchart. BAO, basilar artery occlusion; CTA, CT angiography; DSA, digital subtraction angiography; PCA, posterior cerebral artery; VA, vertebral artery.
Outcome parameters
The primary outcome measure was the mRS shift after 90 days, assessed through ordinal regression analysis indicating the shift towards higher mRS scores (worse functional outcome). Secondary outcome parameters included dichotomized mRS (0–2 and 0–3) at 90 days, mortality, and neurological improvement indicated by the NIHSS at 24 hours and at discharge. Safety parameters comprised sICH (defined by the Safe Implementation of Thrombolysis in Stroke-Monitoring Study (SITS-MOST (criteria)) and any ICH assessed by intracranial imaging at 24 hours, as well as other procedure-related complications, including vessel dissection/perforation, vasospasm, and clot migration/embolization. Procedural metrics were compared, including successful reperfusion (defined as a modified Thrombolysis In Cerebral Infarction (mTICI) score of 2b to 3), first pass recanalization rate, number of recanalization attempts, and procedure duration (defined as the time from groin puncture to vessel reperfusion). Subgroups were defined according to qualifying stroke etiology (large artery atherosclerosis (LAA), cardioembolism (CE), and other etiologies, including dissection, embolic stroke of undetermined source, and other determined etiologies), NIHSS at admission (≤15 and >15 points) and stroke onset to groin puncture (≤6 hours and 6–24 hours). Effect estimates are depicted as the odds of association of aspiration thrombectomy with each respective parameter.
Statistics
Unmatched and matched cohorts were compared using the Mann-Whitney U test, the Χ2 test, Fisher’s exact test, and the Kruskal-Wallis test, where appropriate. Each variable is depicted with mean±SD, median±IQR, or counts and percentages, where applicable. Association with outcome and safety parameters as well as procedural metrics was assessed using ordinal, binary logistic, and multiple linear regression models, where applicable. Effect estimates are displayed as (adjusted) odds ratios (aOR) and common odds ratios (cOR) with 95% confidence intervals (95% CI). Propensity score matching was carried out, using logistic regression analysis based on unequal baseline parameters between the two groups, estimating the propensity score with a caliber of 0.2. One-to-one nearest-neighbor matching was used, matching patients who underwent aspiration and stent retriever thrombectomy. After matching, baseline characteristics were equally balanced between the two groups. Subgroups were assessed in the unmatched cohorts with effect estimates displayed as common OR adjusted for baseline confounders within the specific subgroup. Missing values of outcome parameters were not considered for analysis. A P value of <0.05 was considered statistically significant. The statistical analysis was performed using SPSS version 26 for Windows (IBM Corp, Armonk, New York, USA).
Results
Patient characteristics
Of 13 082 patients within the GSR-ET, 387 fulfilled the inclusion criteria (figure 1). Among these 387 patients, n=195 (50.4%) and n=192 (49.6%) received aspiration and stent retriever thrombectomy, respectively (table 1). Most major baseline, stroke, and treatment characteristics were similar between the groups, including age (72.1 vs 71.8 years), sex (46.2% vs 43.8% female), premodified Rankin Scale scores (0 vs 0), baseline NIHSS scores (16 vs 15), most time intervals and IVT treatment (38.7% vs 44.2%). However, some differences were observed: patients who underwent stent retriever thrombectomy exhibited higher rates of pre-existing dyslipidemia (46.8% vs 31.0%) and diabetes (31.6% vs 17.5%) and lower systolic blood pressure at admission (140 vs 150 mm Hg). They had a lower prevalence of atrial fibrillation (34.5% vs 53.5%) and consequently lower rates of oral anticoagulation (12.1% vs 20.0%). The time from admission to groin puncture was 9 min longer in patients who underwent aspiration thrombectomy (75 vs 66 min), while other time intervals were evenly balanced. During ET, 12.5% of stent retriever thrombectomy patients received basilar stenting compared with 4.6% in aspiration thrombectomy patients. The rate of peri-interventional intra-arterial administration of medication was twice as high with stent retriever compared with aspiration thrombectomy (20.4% vs 11.5%).
Descriptive analysis of baseline, stroke and treatment characteristics comparing patients undergoing aspiration only and stent retriever only thrombectomy in the unmatched cohort
Clinical outcome, safety parameters, and procedural metrics
At 90 days, there was no difference in the shift of the mRS scores distribution before (cOR=0.94; 95% CI 0.64 to 1.38) and after propensity score matching (cOR=1.37; 95% CI 0.90 to 2.09) favoring neither one of the thrombectomy techniques (figure 2A). The rates of mRS 0–2 and 0–3 were similar in the unmatched (mRS 0–2: 36.2% vs 32.9%; P=0.561; mRS 0–3: 46.9% vs 47.0%; P=0.989) and matched cohorts (mRS 0–2: 36.2% vs 40.9%; P=0.454; OR 0.82; 95% CI 0.50 to 1.34; mRS 0–3: 46.9% vs 55.7%; P=0.224; OR 0.73; 95% CI 0.45 to 1.18; figure 3).
(A) Distribution of scores on the modified Rankin Scale (mRS) at 90 days in the unmatched cohorts. mRS scores at 90 days were available in n=163 (83.6%) and n=161 (83.9%) of the patients who underwent aspiration and stent retriever thrombectomy, respectively. Missing outcome values were not considered for shift analysis. (B) Subgroup analysis. * Adjusted common odds ratio (aOR) with 95% confidence interval (95% CI), indicating the odds that the patients undergoing aspiration thrombectomy would have a shift in the distribution of scores on the mRS towards higher scores compared with patients who underwent stent retriever thrombectomy. **Other stroke etiologies, including dissection, embolic stroke of undetermined source, and other determined etiologies. 1Adjusted for premodified Rankin Scale, primary admission at interventional hospital and time from stroke onset to groin puncture. 2Adjusted for diabetes, dyslipidemia, time from admission to groin puncture and angiosuite to groin puncture. 3Adjusted for witnessed stroke onset, diabetes, and basilar stenting. 4Adjusted for oral anticoagulation, diabetes, atrial fibrillation, dyslipidemia, smoking status, and time from arrival at angiosuite to groin puncture. 5Adjusted for systolic blood pressure at admission and diabetes. 6Adjusted for heart rate at admission, time from arrival at angiosuite to groin puncture, and basilar stenting. CE cardioembolism; LAA large artery atherosclerosis; NIHSS, National Institutes of health Stroke Scale.
Outcomes, safety parameters, and procedural metrics. *OR and β coefficient with 95% CI, indicating association of aspiration thrombectomy with each parameter. ɸPropensity score matching based on unequal baseline, stroke and treatment characteristics, including oral anticoagulation at admission, diabetes, atrial fibrillation, dyslipidemia, systolic blood pressure, time from admission to groin puncture, basilar stenting and peri-interventional intra-arterial medication. ¥Procedure-related complications including, vessel dissection/perforation, vasospasm, and clot migration/embolization. mRS, modified Rankin Scale; sICH symptomatic intracranial hemorrhage; mTICI modified Thrombolysis In Cerebral Infarction.
There were no significant differences in sICH rates in the unmatched (2.6% vs 4.7%; P=0.290) and matched patients (2.6% vs 5.5%; P=0.231; OR=0.46; 95% CI 0.14 to 1.47). The rate of procedure-related complications was lower with aspiration compared with stent retriever thrombectomy (4.6% vs 12.5%; P=0.017; OR=0.34; 95% CI 0.15 to 0.79). Procedure duration was shorter (24 vs 48 min; P<0.001; ß −14.43; 95% CI −27.62 to −1.24) and first pass recanalization was achieved more frequently with aspiration thrombectomy (75.1% vs 44.8%; P<0.001; OR=3.72; 95% CI 3.29 to 6.07). However, there were no differences in successful reperfusion between both groups before and after matching (95.3% vs 92.8%; P=0.457; OR=1.59; 95% CI 0.61 to 4.11). Other outcome parameters did not differ between the two groups and are presented in detail in the online supplemental material.
Subgroup analyses
The results of the subgroups analyses were consistent with the primary analysis. Neither thrombectomy approach was associated with a shift in the distribution of the mRS scores after 90 days according to stroke etiology, NIHSS scores, and time of stroke onset to groin puncture, as detailed in figure 2B. Descriptive analyses of each subgroup are depicted in detail in the online supplemental material.
Discussion
We present a large multicenter real-world patient population, comparing aspiration with stent retriever thrombectomy in patients with isolated BAO. There was no impact of the thrombectomy approach on functional outcome, neurological improvement, or mortality both before and after propensity score matching. Moreover, the distribution of mRS scores at 90 days in our cohorts closely resembled those of patients who underwent ET in the BAOCHE and ATTENTION trials.1 2 Subgroup analysis with respect to stroke etiology, NIHSS, and time from stroke onset to groin puncture further supported the results from our primary analysis. Our findings contribute to the existing non-randomized evidence that in BAO, the thrombectomy approach does not appear to affect functional outcome.9–11 13 19 This finding corroborates data known for the ET technique in anterior circulation strokes.6
Consistent with previous studies, we noted significantly shorter procedure durations, fewer recanalization attempts, and higher first pass recanalization rates with aspiration thrombectomy.14 19–21 This could be attributed to the simplicity, in theory, of aspiration compared with stent retriever thrombectomy, which is, however, heavily influenced by individual patient factors, including vessel elongation, tortuosity, clot morphology, and the degree of intracranial macroangiopathy.19 22 The higher rate of general anesthesia in patients with BAO (approximately 90% in our cohort) compared with in patients with anterior circulation stroke might, however, suggest a safety advantage with aspiration thrombectomy, by potentially mitigating anesthesia-related intraprocedural and postprocedural complications.23 Furthermore, shorter procedure times are associated with more favorable functional outcome in patients with BAO, especially in those with well-established collaterals.24 25 Moreover, given that both thrombectomy approaches appear to be equally efficient, one might argue in favor of aspiration catheters from a pathophysiological point of view, as they might contribute less to endothelial damage and iatrogenic occlusion of thalamic perforators compared with stent retrievers.26
Although we observed numerically higher rates of sICH with stent retriever thrombectomy, statistical significance was not reached between the two groups. Rates resembled data from recent observational studies, but were slightly lower those reported by the BAOCHE and ATTENTION trials.1 2 11 12 This difference could be attributed to the substantially shorter time intervals from stroke onset to groin puncture and reperfusion in our registry. Two recent studies including one meta-analysis, reported higher rates of sICH with stent retriever thrombectomy, although statistical significance was reached only in one study.10 Notably, aspiration thrombectomy was associated with lower rates of procedure-related complications, which was also suggested by other studies.12 22 27 Stent retriever thrombectomy specifically exhibited higher rates of vessel perforation (7.0% vs 2.6% after matching), which might explain the elevated rates of sICH.
The higher first pass rates with aspiration thrombectomy did not translate into higher rates of successful reperfusion in our cohort, as we observed similar rates of mTICI scores of 2b to 3. This finding is consistent with recent data,10 although some other studies have reported higher recanalization rates with aspiration thrombectomy.11 21 28 In our cohort, mTICI 2b to 3 rates were, however, consistently high in the stent retriever group (92.6% before and 92.8% after matching). This might be attributed to the widespread use of new-generation devices, including the pRESET (27.1%) and APERIO (10.9%) systems in our study, probably contributing to these favorable angiographic outcomes.29 30
Our study has some limitations due to its observational nature, introducing inherent biases. There were some imbalances between the groups, which were adjusted for by matching, but nevertheless could have affected our findings. For instance, patients who received stent retriever ET exhibited higher rates of diabetes and dyslipidemia; conditions associated with macroangiopathy, along with lower rates of atrial fibrillation. LAA etiology was more frequent in patients who underwent retriever thrombectomy than among those who underwent aspiration thrombectomy. The risk factor profile might have influenced the preference for a stent retriever thrombectomy approach in these patients. Interventionalists might have been prompted to use stent retrievers due to their perceived effectiveness in handling potential challenges related to distal vessel access in the context of macroangiopathy. This hypothesis is supported by the elevated rates of intracranial stenting observed in the stent retriever group. Additionally, thrombus location might have influenced the choice of thrombectomy approach. Proximal occlusions are more frequently associated with atherosclerotic angiopathy and may be better accessed with stent retrievers. On the other hand, distal occlusions are often attributed to CE and may be more safely accessed with aspiration catheters, minimizing the risk of vessel harm.18 However, our subgroup analysis did not suggest an advantage of either thrombectomy technique in stroke attributed to both LAA and CE.
From the underlying data, we could not differentiate between patients undergoing an initial combined technique and those who transitioned to it later (cross-over patients). To prevent reporting bias, we excluded this cohort from our primary analysis. However, comparing all three cohorts (aspiration only with stent retriever only with combined/cross-over technique), we observed a significantly higher rate of mortality (45.2% vs 33.1% vs 34.8%) in combined/cross-over patients and numerically lower rates of mRS score 0–2 (29.8% vs 32.9% vs 36.2%) and mRS score 0–3 (37.8% vs 46.9% vs 47.0%). This might be partially attributed to the inclusion of patients undergoing cross-over thrombectomy, as this was probably done after a failed initial thrombectomy approach, indicating a difficult and complicated thrombectomy. This is supported by the fact, that the combined/cross-over cohort required basilar stenting and peri-interventional administration of IA alteplase, nimodipine, and heparin more often than the other two cohorts. Moreover, while almost all cases of aspiration thrombectomy were performed using the SOFIA system, the stent retriever group employed a variety of different systems and this heterogenicity potentially affected the study results.
Conclusion
In our study aspiration thrombectomy was as efficient in terms of functional outcome as stent retriever thrombectomy in patients with BAO. However, procedure complications and metrics might favor aspiration thrombectomy over stent retriever thrombectomy.
Data availability statement
Data are available upon reasonable request. The data that support the results of this study are available from the corresponding author upon reasonable request.
Ethics statements
Patient consent for publication
Ethics approval
This study involves human participants and was approved by the ethics committee of the LMU Munich (protocol: 689-15), as the leading ethics committee. Further approval was obtained from local ethics committees according to local regulations. Participants gave informed consent to participate in the study before taking part.
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
Collaborators GSR-ET investigators:Tobias BoeckhBehrens, Silke Wunderlich, Alexander Ludolph, Karl-Heinz Henn, Arno Reich, Omid Nikoubashm an, Ulrike Ernemann, Sven Poli, Christian H Nolte, Eberhard Siebert, Sarah Zweynert, Georg Bohner, Gabor Petzold, Fee Keil, Joachim Röther, Bernd Eckert, Jörg Berrouschot, Albrecht Bormann, Anna Alegiani, Jens Fiehler, Christian Gerloff, Götz Thomalla, Christoffer Kraemer, Jan Liman, Martina Petersen, Florian Stögbauer, Michael Braun, Gerhard F. Hamann, Klaus Gröschel, Timo Uphaus, Jan Borggrefe, Franziska Dorn, Marielle Ernst, Jörg Hattingen, Hannes Leischner, Jan Hendrik Schäfer, Maximilian Schell, Peter Schellinger, Christoph Trumm.
Contributors JW and LKel conceptualized the study. JW, HZ, and LKel collected the data. JW carried out the analysis and wrote the first draft of the manuscript. All authors reviewed and edited the manuscript and approved the final version of the manuscript. LKel acts as the guarantor of this study.
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 LKel has received funding for travel or speaker honoraria from Alexion, AstraZeneca, Bayer Vital, Boehringer Ingelheim, Bristol-Myers Squibb, Daiichi Sankyo, and Pfizer outside of this study. CHN has received funding for travel or speaker honoraria from Alexion, Astra-Zeneca, Bayer Vital, Bristol-Myers Squibb, Daiichi Sankyo, Pfizer, and Takeda.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.