Effect of treatment technique on radiation exposure in mechanical thrombectomy for acute ischaemic stroke: A matched-pair analysis

Abstract

Objective

This study aimed to compare radiation exposure (RE) in patients receiving mechanical thrombectomy (MT) for large-vessel occlusions in the anterior circulation using direct thrombo-aspiration (DT) versus stent-retriever thrombectomy under continuous distal aspiration (STA).

Methods

This was a retrospective single-centre analysis of an Institutional Review Board−approved stroke database of a comprehensive stroke centre focusing on RE per dose area product, procedure time (PT) and fluoroscopy time (FT) in patients receiving MT. Patients who received MT with DT were matched with patients treated using STA according to occlusion location, mode of anaesthesia, manoeuvre count and sex.

Results

Apart from patient age (DT: M = 74 years (standard deviation (SD)=13 years); STA: M = 79 years (SD = 11 years); p = 0.023), there was no difference in baseline patient characteristics (n = 68 per group). PT (DT: median = 26 minutes (interquartile range (IQR) = 21–38 minutes); STA: median = 49 minutes (IQR 37–77 minutes); p < 0.0001) and FT (DT: median = 12 minutes (IQR 7–18 minutes); STA: median = 26 minutes (IQR 14–43 minutes); p < 0.0001) were shorter in patients who received MT using DT. RE (DT: median = 62.6 Gy·cm² (IQR 41.7–89.4 Gy·cm²); STA: median = 89.8 Gy·cm² (IQR 53.7–131.7 Gy·cm²); p = 0.034) was significantly lower in patients who received MT using DT. This represents a relative increase of RE, FT and PT by 43.6%, 116.6% and 88.5%, respectively, in patients who received MT using STA.

Conclusion

MT using DT is associated with shorter FT and PT and lower RE compared to matched patients treated with STA.

Introduction

After mechanical thrombectomy (MT) was established as a first-line therapy for acute ischaemic stroke with large vessel occlusion in recent years,¹ a more detailed investigation of factors influencing radiation exposure (RE) associated with MT has been initiated.^2–4

When it comes to interventional techniques, there is still equipoise concerning the ‘right’ technical approach for successful re-canalisation.⁵ While the first multi-centre studies in 2015 proving the effectiveness of MT compared to intravenous (i.v.) recombinant tissue plasminogen activator (rtPA) mostly applied stent-retriever thrombectomy, evidence emerging in recent years that direct thrombo-aspiration (DT) – a more cost-effective and faster technique – shows equally high re-canalisation success rates.^6,7 However, a combined approach of stent-retriever thrombectomy under continuous aspiration (STA) as opposed to DT or stent-retriever thrombectomy alone results in a better reperfusion rate.⁵ The choice of technique for the first approach is known to be crucial for a good clinical outcome – also referred to as ‘first-pass effect’.⁸ Furthermore, successful re-canalisation of the target vessel becomes less likely with each re-canalisation attempt, mainly due to an increase in thrombus friction with the vessel wall and prolonged interaction time.⁹ With regards to reperfusion success, after five re-canalisation attempts, there seems to be no advantage of MT compared to i.v. rtPA alone.¹⁰ However, the number of passes is an independent predictor of good clinical outcome after stroke and mechanical MT.¹¹

RE on the other hand is of high relevance due to the increasing number of thrombectomy procedures worldwide.¹² Additional growth in intervention numbers is also foreseen in the face of the rising prevalence and incidence of stroke in Europe, North America, China and other regions, as well as the widened time window for MT after stroke onset.^12–15 Thus, ionising radiation’s stochastic and deterministic risks for the patient and the interventionalist are gaining relevance. Approximately 6% of procedures reach the threshold of potential deterministic risk for the patient undergoing MT.¹⁶ Additionally, cancer, premature cataracts, early cognitive impairment and premature vascular disease may affect staff working in a catheter laboratory.^17–19 As RE during MT depends on the interventionalist’s experience, this might also imply a change in teaching strategy for future interventionalists.³

In parallel to diagnostic dose reference levels (RL), which have already been shown to be effective in controlling RE and minimising its risks for diagnostic imaging in radiology departments across North America and Europe,^20,21 it is currently being discussed whether RL can contribute to the control of RE in MT.²

Purpose

As interventional techniques differ in complexity, intermediate steps and materials used, we hypothesised that the technical approach of STA versus DT has an influence on fluoroscopy time (FT), procedure time (PT) and RE during MT.

Methods

In this retrospective single-centre study, we report data from an Institutional Review Board−approved stroke database of a comprehensive stroke centre of 136 patients treated with MT for a large-vessel occlusion of the anterior circulation between January 2014 and 04/2019. The stroke centre performs more than 350 MTs annually and uses either DT or STA as manoeuvre techniques. The primary patient group (n = 68) received MT using a distal access catheter for DT.²² These patients were matched with patients (n = 68) with STA. Matching criteria in order of decreasing priority were location of intracranial occlusion, mode of anaesthesia, number of thrombectomy manoeuvres (defined as a planned and conducted attempt with the intention to re-canalise an occluded intracranial vessel) and sex.^4,23 The patients were individually matched according to these matching criteria.

Inclusion and exclusion criteria

Of all patients who received MT for a large-vessel occlusion in the anterior circulation between January 2014 and April 2019, we only included procedures in which a single interventional technique – either DT or STA – was consequently used for all endovascular stroke treatment (EST) manoeuvres. Furthermore we excluded patients treated (a) using a monoplane system, (b) when the interventionalist had performed 25 or fewer EST procedures³ and (c) when additional percutaneous balloon angioplasty or stenting of the parent or intracranial artery was needed (e.g. treatment of a tandem occlusion).

Performance of MT

Eligibility for MT in patients admitted to our hospital followed current guidelines and was therefore subject to adjustments over the time of the observation period.²⁴ After initial clinical examination and computed tomography or magnetic resonance imaging, the interdisciplinary decision for MT was made by a neurointerventionalist (primary operator) and a neurologist. Thrombectomy procedures were performed constantly (24 hours/365 days) with staff consisting of a neurointerventionalist, a resident or fellow in interventional neuroradiology (as scrub assistant) and a medical technical assistant (angio-nurse) to support the procedure.

For all procedures, femoral access was chosen. There was no change of access side in the investigated cases. As the standard approach, a triaxial system with a non-balloon-guide catheter (8F VistaBriteTip; Cordis, Santa Clara, CA), a distal access catheter (Sofia 5F or Sofia Plus 6F; MicroVention, Aliso Viejo, CA) and microcatheter (Rebar 18; Medtronic, Minneapolis, MN) was established. The choice of material (e.g. microwire, stent retriever) for intracranial vessel access and MT technique (DT vs. STA) depended on availability and was made at the discretion of the neurointerventionalist according to the patient’s anatomy, clot location and length. Negative pressure for continuous aspiration was established by using 60 cc vacuum pressure syringes (VacLok; Merit, Jordan, UT) attached to the distal access catheters and the guide catheter, respectively, in both patient groups.

Road mapping and contrast series were the main sources of RE during the procedure. There was no standardised protocol for additional imaging (e.g. contrast series in the territory of the contralateral internal carotid artery (ICA) to delineate collateral flow) before or after MT. Angiographic systems (Artis Zee Biplane and Artis Q; Siemens Healthineers, Erlangen, Germany) received approval for human use and underwent technical surveillance with repetitive constancy tests scheduled every three months and additionally after every maintenance work. Data acquisition of dose area product (DAP) and FT was automatically done by the angiographic systems and was calibrated regularly.

Primary target parameters

The analysis focuses on RE per DAP (in Gy·cm²), PT and FT (in minutes) as primary target parameters.

Data acquisition

All patient and procedure-related data were obtained from medical and interventional reports and entered in the stroke database using a standardised electronic data-entry sheet (U.N., S.F., C.W. and J.P.). All cases selected for the matched-pair analysis were re-evaluated for the correctly obtained data concerning matching criteria, with a special focus on the interventional technique used for MT (J.P. and C.W.).

Statistical analysis

Statistical analysis was performed using IBM SPSS Statistics for Windows v21.0 (IBM Corp., Armonk, NY). Data were analysed using the chi-square test or t-test and are shown as the median with interquartile range (IQR) or means with standard deviation (SD) as appropriate. For all analysis, two-tailed hypothesis testing was used, with p < 0.05 interpreted as statistical significance.

Results

According to the inclusion and exclusion criteria, 68 patients who received MT using DT were identified and could be matched according to the matching criteria with 68 patients who received MT using STA, thereby resulting in a total of 136 patients in this analysis. There was a significant difference in age between the two groups (DT: M = 74 years (SD = 13 years); STA: M = 79 years (SD = 11 years); p=0.023). Other than that, there were no differences in any of the included patients’ baseline characteristics.

In both groups, i.v. rtPA was administered in nearly half of the patients (DT: n = 37 (54.4%); STA: n = 33 (48.5%); p = 0.607). There was a trend towards more favourable reperfusion results in the DT group according to modified thrombolysis in cerebral infarction score (mTICI 2b–3: 94.1% and 91.2%; p = 0.068).

PT (DT: median = 26 minutes (IQR 21–38 minutes); STA: median = 49 minutes (IQR 37–77 minutes); p < 0.0001) and FT (DT: median = 12 minutes (IQR 7–18 minutes); STA: median = 26 minutes (IQR 14–43 minutes); p < 0.0001) were shorter in patients who received MT using DT.

Additionally, RE (DT: median = 62.6 Gy·cm² (IQR 41.7–89.4 Gy·cm²); STA: median = 89.8 Gy·cm² (IQR 53.7–131.7 Gy·cm²); p = 0.034) was lower in patients who received MT using DT.

This represents a relative median increase of RE by 43.6%, of FT by 116.6% and of PT by 88.5% in patients who received MT using STA (for further details, see Table 1).

Table 1.

Baseline characteristics of matched patients who received mechanical thrombectomy with either DT or STA.

	DT (n = 68)	STA (n = 68)	p-Value
Age (years), M (SD)	74 (13)	79 (11)	0.023d
Male, n (%)a	24 (35.3)	24 (35.3)	1.000c
Pre mRS			0.134c
0, n (%)	28 (41.2)	19 (27.9)
1, n (%)	10 (14.7)	16 (23.5)
2, n (%)	11 (16.2)	18 (26.5)
3, n (%)	17 (25)	13 (19.4)
4, n (%)b	1 (1.5)	1 (1.5)
5, n (%)b	1 (1.5)	0
Initial NIHSS score, median (IQR)	16 (11–19)	19 (11–20)	0.348d
Intravenous rtPA, n (%)	37 (54.4)	33 (48.5)	0.607c
Unknown time of symptom onset, n (%)	21 (30.9)	26 (38.2)	0.471c
Time from stroke onset to groin puncture (minutes), median (IQR)	232 (146–383)	233 (168–492)	0.336d
Procedural aspects
Treatment in conscious sedation, n (%)a	60 (88.2)	60 (88.2)	1.000c
Location of intracranial occlusiona			0.736c
Carotid T, n (%)	12 (17.6)	12 (17.6)
M1, n (%)	51 (75)	51 (75)
M2, n (%)	4 (5.9)	4 (5.9)
M3, n (%)	0	1 (1.5)
ACA, n (%)	1 (1.5)	0
Number of thrombectomy manoeuvres, M (SD)a	1.26 (0.7)	1.32 (0.8)	0.641d
Procedure time (groin puncture to last image; minutes), median (IQR)	26 (21–38)	49 (37–77)	<0.0001 d
Dose area product (Gy·cm2), median (IQR)	62.6 (41.7–89.4)	89.8 (53.7–131.7)	0.034 d
Fluoroscopy time (minutes), median (IQR)	12 (7–18)	26 (14–43)	<0.0001 d
Final mTICI score			0.068c
0–2a, n (%)	4 (5.9)	6 (8.8)
2b, n (%)	11 (16.2)	18 (26.5)
2c, n (%)	9 (13.2)	15 (22.1)
3, n (%)	44 (64.7)	29 (42.6)

^aMatching criterion.

^bLow pretreatment mRS was secondary to a temporary, potentially curable illness, and mechanical thrombectomy was conducted.

^cChi-square test, two-sided.

^dTwo-sided t-test.

DT: direct thrombo-aspiration; STA: stent-retriever thrombectomy under continuous distal aspiration; SD: standard deviation; pre mRS: pre-stroke modified Rankin score; NIHSS: National Institutes of Health Stroke Scale; IQR: interquartile range; rtPA: recombinant tissue plasminogen activator; mTICI: modified thrombolysis in cerebral infarction.

Discussion

This study presents data of a matched-pair analysis focusing on FT, PT and RE, depending on the interventional technique used for MT in the anterior circulation. Our data show a lower RE, FT and PT associated with MT using DT when compared to stent-retriever thrombectomy. The very high percentage of successful re-canalisation of the target vessel occlusion (mTICI 2b–3: 94.1% for DT and 91.2% for STA) is due to the chosen selection criteria to concentrate on the RE by manoeuvre technique. All interventions where the technique was changed were excluded.

As previously published, RE in MT depends on the manoeuvre count.^2,23 Farah et al. considered manoeuvre count as suitable to establish RLs for RE. However, the operational technique used for the respective manoeuvres was not investigated further in a larger subset of patients.² As there is still equipoise about the ‘right’ technical approach, especially for the first thrombectomy manoeuvre, both techniques – DT and STA – are in everyday use in centres performing EST. The choice of technique depends on the interventionalist’s preference, personal operating experience and other factors, including the availability and reimbursement of material used for MT. Additionally, the choice of MT technique also depends on the patient’s vessel anatomy. DT might come with reduced catheter stability compared to stent retriever thrombectomy. Manoeuvring a microwire and microcatheter for stent-retriever thrombectomy allows difficulties in anatomy such as tortuous vessels to overcome (e.g. by using the stent anchor technique). Therefore, based on experience and theoretical assumption, DT appears to be easier to perform in younger patients without tortious vessels. This might explain the difference in patient age between the study groups. The different level of RE linked with the technique might – and perhaps must not – influence the choice of technique for every manoeuvre, but should definitely be considered when studying RLs as a measure for quality control for RE in modern EST.

Our results concerning overall PTs are in line with the familiar shorter time for MT using DT.^6,25 As the number of performed thrombectomy manoeuvres was a matching criterion, a higher or lower first-pass effect of a technique, i.e. the need for additional manoeuvres, could not explain the higher RE observed in the STA group in this study. By excluding patients with ICA stenosis or occlusion (i.e. tandem occlusion), higher RE due to necessary PTA/stent implantation was omitted. However, clot aetiology (cardial or arterio-arterial embolus) and composition by itself can influence the MT intervention mainly by leading to a higher MT manoeuvre count for difficult interventions or tending towards STA (e.g. calcified thrombi). By excluding MTs where the technique changed during the intervention and matching the patients according to the manoeuvre count, the effect of thrombus origin and composition on RE during MT should be limited.

It appears that the relevant difference of RE depending on the manoeuvre technique is arguably due to the shorter FT associated with the DT technique. In DT, fluoroscopy is used for navigation and placement of the distal access catheter at the proximal surface of the clot and the removal of the distal access catheter. In contrast to this technique, fluoroscopy is required for navigation to the target vessel occlusion and passage of the thrombus with a microwire and microcatheter, deployment of a stent retriever, placement of the distal access catheter at the proximal clot surface and during removal of the devices in the STA technique. These additional steps in the STA technique require image guidance, leading to a longer FT and thus higher RE. When comparing previously proposed RLs for MT (e.g. 107 Gy·cm² for one thrombectomy manoeuvre, 148 Gy·cm² for any EST in the anterior circulation²) with the results of the current analysis, it becomes apparent that the different choice of manoeuvre technique strongly influences values of RLs, as they differ by approximately 44%.

Finally, it should be mentioned that the decision to select an operational technique for MT should not be made primarily depending on the expected RE, but rather on the expected treatment effect in a timely manner (i.e. a safe technique) with a high first-pass effect to re-canalise the occluded vessels as soon as possible.

Limitations

This study is limited because of its single-centre, retrospective character, and it is subject to a certain degree of bias due to the methodology and matching criteria. Additionally, while RE during MT might vary from one stroke centre to another – depending on the angiographic system, usage of a dose-reduction software and so on – the significant difference in RE between manoeuvre techniques appears to be transferable to other settings mostly because it has been shown before that the time necessary to perform DT is shorter and FT can be spared, as there is no need for imaging of a stent-retriever deployment and retrieval in DT. A further limitation of this study is the comparison of only two commonly used thrombectomy techniques. Thus, transferability of our data to other techniques, such as the BAlloon guiDe with large bore Distal Access catheter with dual aspiration with Stent-retriever as Standard approach, is limited.²⁶

Conclusion

RE during MT is dependent on the chosen thrombectomy technique. Because DT is associated with shorter procedure and FT compared to STA, RE is lower when using DT for MT. This finding should be considered when RE control and eligible RLs for MT are discussed.

Conflict of interest

Peter A. Ringleb reported personal fees from Boehringer Ingelheim, Bayer, BMS and Daiichi Sankyo outside the submitted work. Markus A. Möhlenbruch reported unrelated Board Membership of Codman; consultancy for Medtronic, MicroVention and Stryker; grants/grants pending from Balt and MicroVention (money paid to the institution); and payment for lectures, including service on speakers’ bureaus from Medtronic, MicroVention and Stryker. Martin Bendszus declared the following activities not related to the present article: grants and personal fees from Bayer, Codman, Guerbet, Medtronic and Novartis; grants from the German Research Council (DFG), European Union, Hopp Foundation, Siemens and Stryker; and personal fees from BBraun, Böhringer Ingelheim, Roche, Teva and Vascular Dynamics. Johannes Pfaff declared the following activities not related to the present article: travel and meeting expenses from Stryker and MicroVention. The remaining authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Charlotte S Weyland https://orcid.org/0000-0002-1374-7854

Johannes AR Pfaff https://orcid.org/0000-0003-0672-5718