Adrien Guenego, MD and Matthew Leipzig, BS conducted all the statistical analyses.

Mechanical thrombectomy (MT) is an effective treatment for acute ischemic stroke patients (AIS due to large vessel occlusion (LVO). Rapid and successful reperfusion, defined as thrombolysis in cerebral infarction (TICI) 2b-3, increases the likelihood of a favorable outcome (1, 2). Nevertheless, MT does not result in successful reperfusion in up to 29% of patients (1), and biomarkers that identify patients at risk of failed reperfusion failure are needed.

Clot composition, length, and shape may impact MT success, and imaging predictors of clot response to MT may lead to tailored MT techniques, such as stent-retriever or contact-aspiration, to maximize the likelihood of successful treatment (3, 4). Magnetic resonance imaging (MRI) often demonstrates the thrombus on T2^* gradient-echo sequence (GRE) as a region of intravascular hypointense signal abnormality, which is termed the susceptibility vessel sign (SVS). SVS has been used as a measure of clot length to predict response to intravenous thrombolysis (5), to detect small distal occlusions (6), to assess multiplicity of intracranial thrombus fragments (7), and even predict clot composition or stroke etiology (8–13). However, whether SVS depiction of clot shape and extension into vessel branches impacts the likelihood of successful reperfusion has not been investigated. Thrombus that involves a bifurcation or accentuated angle may be more prone to fragmentation and may be more difficult to remove (14).

We hypothesized that SVS may be used to visualize the extent of the clot within the middle cerebral artery branches at the point of vessel occlusion and to determine whether the clot is located in a straight branch (S-SVS) or in an angulated/traversing a bifurcation segment (A-SVS). We determined SVS clot shape, branch occlusion patterns, and the impact of these factors on successful reperfusion and favorable clinical outcomes after thrombectomy for proximal middle cerebral artery occlusions.

The study protocol was approved by the institutional review board (IRB) and complied with the Health Insurance Portability and Accountability Act (HIPAA). Patient informed consent was waived by our review board for this single center retrospective analysis of anonymized data acquired prospectively. Adherence to the STROBE criteria (15) was enforced.

Sixty-two patients met inclusion criteria (Figure 2). Patients were dichotomized into S-SVS (27 patients; 44%) and A-SVS (35 patients; 56%) groups; reader agreement of SVS classification was substantial (Cohen's Kappa 0.711) (20).

There were no differences between S-SVS and A-SVS with respect to patient age [72 (IQR, 63–79) vs. 67 (IQR, 55–76), p = 0.960], female sex (52 vs. 54%, p = 1.000), baseline NIHSS [14 (IQR, 10–20) vs. 17 (IQR, 13–22), p = 0.12], or frequency of left sided occlusions (71 vs. 54%, p = 0.19), respectively (Table 1). A-SVS clots were significantly longer compared to S-SVS clots [19 mm (IQR, 15–24) vs. 8 mm (IQR, 10–15), p = 0.0002] (Table 2). There were no other significant differences between the two groups with regard to neuroimaging variables, which included core infarction volume, penumbra volume, and HIR collateral robustness (Table 2). Likewise, MT procedural outcomes were similar between S-SVS and A-SVS groups (Table 3). There were no differences in time from groin puncture to reperfusion [36 min (IQR, 25–62) vs. 44 min (IQR, 23–77), p = 0.70] or MT complication rates (4 vs. 11%, p = 0.27).

All other univariate analysis are described in Table 3.

In the multivariable binary logistic regression model (AUC = 0.893), A-SVS was an independent negative predictor of successful reperfusion [OR 0.04 (95% CI, 0.002–0.58), p = 0.02, Table 4].

There was no impact of A-SVS clot shape in the multivariable binary logistic regression models on good clinical outcome [OR 0.72 (95% CI, 0.21–2.46), p = 0.59], excellent clinical outcome [OR 0.35 (95% CI, 0.09–1.46), p = 0.15], or mortality [OR 3.22 (95% CI, 0.67–15.49), p = 0.14], respectively (Table 4).

In this study, we found that thrombus morphology measured by SVS-shape influences the likelihood of successful reperfusion after MT. However, SVS morphology did not affect the likelihood of achieving a favorable clinical outcome. Our findings that SVS thrombus morphology is a biomarker of reperfusion have important implications for MT.

Prior studies have used SVS identified on gradient-echo imaging (T2^*) (21) to detect (5), localize (22), and measure clot length (5, 22) without contrast administration. Whether SVS is a predictor of reperfusion after MT remains controversial (9, 23). Our findings support the hypothesis that SVS is a biomarker of reperfusion when thrombus morphology is considered. S-SVS are linear clots that do not extend into arterial branch vessels, and clots with this morphology were likely to undergo complete reperfusion compared to angulated clots and clots that extend into branch vessels (A-SVS). Whether a prospective change to MT technique results in higher rates of reperfusion of A-SVS clots requires further study. While clot morphology will not impact the decision to perform a thrombectomy procedure, we hypothesize that it could impact technical strategy, and the routine use of balloon guide sheaths, longer stentretrievers, or even double stentretriever techniques (placement of two devices into two branch points involved with a A-SVS clot) may increase the likelihood of complete reperfusion of A-SVS clots (24, 25).

Patients selection for MT depends on a fast identification of a LVO and on the evaluation of early ischemic changes, and computed-tomography (CT) and magnetic-resonance-imaging (MRI) are recommended for patient evaluation (26). However, MRI is superior to CT for detection of acute ischemia (27), is associated with better outcomes after thrombectomy treatment (28), whereas CT has been associated with an increased risk of futile reperfusion (29). While MRI duration is often reported to be longer in patient's screening for MT (30), MRI did not delay MT (30) nor impact patient's functional outcome in recent studies (30). Use of MRI in AIS screening may then depend on local protocols and optimal institutional workflows.

In contrast to few prior studies that evaluated the importance of clot length on reperfusion (22, 31) our study focused on clot morphology only. In contrast to intravenous thrombolysis stroke treatment, the impact of clot length on successful reperfusion after MT remains controversial (32, 33) and requires further study.

Successful reperfusion has been correlated to clinical outcome in multiple studies (34, 35) and, therefore, one would expect S-SVS to be correlated with a greater likelihood of a favorable clinical outcome. In our study, S-SVS patients had a lower NIHSS the day after thrombectomy and at discharge, but this early recovery did not translate to better outcomes at 90 days. We hypothesize that our study is under powered to detect an outcome difference between S-SVS and A-SVS patients.

SVS clot morphology appears to be an independent predictor of successful reperfusion after MT in our cohort.

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

The studies involving human participants were reviewed and approved by Stanford Medical Center Committee. Written informed consent for participation was not required for this study in accordance with the national legislation and the institutional requirements.

AG, RF, ES, ML, GA, BM, DM, GK, MM, ML, MW, and JH participated to study design, data collection, data analysis, and writing. All authors contributed to the article and approved the submitted version.

GA reports equity and consulting for iSchemaView and consulting from Medtronic. MM reports Ownership Interest in ThrombX Medical. JH reports Consultant or Advisory Board for Medtronic, Inc., MicroVention, Inc., and iSchemaView. The remaining authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.