Internal Carotid Artery Tortuosity: Impact on Mechanical Thrombectomy

All patients with acute ischemic stroke admitted to our institute within 7 days from last known well were prospectively registered in the National Cerebral and Cardiovascular Center Stroke Registry.^14–17 For the present study, we retrospectively reviewed consecutive patients enrolled in this registry from January 2014 to June 2021, who met the following criteria: (1) occlusion of the ICA or M1 or M2 segment of the middle cerebral artery and (2) those in whom MT was attempted for the occlusion of the ICA or M1 or proximal M2 segment of the middle cerebral artery. Patients with tandem occlusion (concomitant extracranial and distal intracranial artery occlusion) were also eligible. Patients who underwent intracranial angioplasty or stenting were excluded. We also excluded patients in whom ICA tortuosity could not be angiographically evaluated during the MT procedure. Written informed consent for study registration was waived because the study was retrospective in nature and utilized anonymized data. Ethics approval was obtained from the local institutional review board (M23-073-9). The National Cerebral and Cardiovascular Center Stroke Registry is registered with https://www.clinicaltrials.gov (NCT02251665). The present study conforms to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cohort studies.¹⁸ A completed STROBE Statement is included in the Supplemental Material.

All endovascular procedures were performed by neurointerventionalists certified by the Japanese Society for Neuroendovascular Therapy, as recommended by the American Heart Association/American Stroke Association guidelines.¹ MT procedures included SR thrombectomy, CA, or combined SR and CA (retrieval of SR and aspiration catheter as a unit).¹⁹ Groin puncture or navigation of a microcatheter for MT without use of thrombectomy devices was considered as an MT attempt. Procedural device selection was at the treating physician’s discretion but limited to those available in Japan. The ones used are listed in Table S1. A balloon guide catheter was routinely used and navigated to the extracranial ICA as much as possible. All patients underwent MT under local anesthesia. Conscious sedation was added if necessary. Written informed consent for MT was obtained from each patient or a relative, if the patients had communication difficulties. Reperfusion status after MT was assessed according to the eTICI scale.²⁰ Follow-up brain computed tomography (CT) or magnetic resonance imaging was routinely performed after MT.

The following clinical data were collected: age, sex, premorbid modified Rankin Scale (mRS) score, baseline National Institutes of Health Stroke Scale (NIHSS) score, atrial fibrillation, cardiovascular risk factors (hypertension, diabetes, current smoking, dyslipidemia), ischemic heart disease (history of myocardial infarction, angina, or coronary revascularization treatment), previous stroke or transient ischemic attack, and systolic blood pressure on admission. The extent of ischemic change in the middle cerebral artery territory was graded using the Alberta Stroke Program Early CT Score on noncontrast CT or diffusion-weighted magnetic resonance imaging. Occlusion sites were determined using digital subtraction angiography on admission. All digital subtraction angiography studies were performed using biplane angiography units: Integris BV 3000 (Philips Medical Systems Nederland B.V., Best, the Netherlands), Allura Clarity FD 20/20 (Philips Medical System Nederland B.V.), and INX-8000 V/HT or V/JE (Canon Medical Systems, Tochigi, Japan). Intravenous thrombolysis was performed with alteplase at 0.6 mg/kg (the dose approved in Japan).²¹

Tortuosity was independently classified by 3 specialists certified by the Japanese Society for Neuroendovascular Therapy (J.K., K.T., and T.Y.) who were blinded to clinical data. Tortuosity of the extracranial ICA was classified based on a previously reported grading system as follows: straight (angle between the centerlines of the common carotid artery and the ICA was <15°), tortuous (angle between the common carotid artery and the ICA centerlines was >15° or S- or C-shaped course of the ICA), coiled (an exaggerated S-shaped curve or circular configuration of the ICA), and kinked (acute [<90°] angulation associated with stenosis).¹¹ The extracranial ICA was considered tortuous if it was coiled or kinked (Figure 1).

Cavernous ICA tortuosity was also classified based on a previously reported grading system.¹² Briefly, type I has open configurations/angles of the anterior and posterior genus (the posterior genu angle ≥90°). Type II is characterized by a closed configuration of the anterior genu (more acute angle of the anterior genu than type I). Type III is defined by posterior deflection of the posterior genu, which gives it a buckled appearance. Type IV is the most tortuous and has a shape characteristic of the Simmons-style angiography catheter where the posterior genu is buckled superiorly compared with the anterior genu. Types III and IV were considered tortuous cavernous ICA (Figure 1).

Standard anteroposterior and lateral projection angiograms were analyzed for tortuosity classification. Baseline angiograms before MT were used for analysis when possible. For assessment of ICA tortuosity, the data set was divided into 3 subdatasets. In the first session, each of the 3 readers assessed tortuosity in 1 of the 3 subdatasets. In the second, each reader evaluated tortuosity in a sub-dataset different from the first session. The overall interobserver reliability for tortuosity (dichotomized as tortuous or nontortuous) in the extracranial and cavernous ICA was evaluated using the κ-coefficient between the interpretations of the first and second sessions. In cases with a discrepancy between the 2 sessions, disagreements were resolved by discussion among the 3 readers. For analyses, the data on the tortuosity after resolving the disagreements were used.

Patients were divided into 2 groups according to ICA tortuosity (tortuous or nontortuous). The tortuous group included patients with an extracranial or cavernous ICA classified as tortuous. Patients without tortuosity in either the cavernous or extracranial ICA were included in the nontortuous group.

Procedural outcomes were FPE (achievement of eTICI score 2c/3 after first pass), modified FPE (first pass eTICI score ≥2b), final eTICI 2c/3 reperfusion, final eTICI ≥2b reperfusion, time from puncture to eTICI ≥2b reperfusion, procedural time, procedural complications (arterial perforation, arterial dissection, and embolization in a new territory), and failure to reach the target occlusion (failure to navigate MT devices due to difficult access). FPE and modified FPE were assessed only among patients who underwent MT. Clinical outcomes were favorable outcome (mRS score of 0–2 at 3 months), excellent outcome (mRS score of 0–1 at 3 months), poor outcome (mRS score of 5–6 at 3 months), and neurological improvement (a ≥10-point decrease of the NIHSS score from baseline or the score of 0) at 7 days from onset. Safety outcomes were any intracranial hemorrhage (ICH) within 36 hours after the onset, the presence of subarachnoid hemorrhage, parenchymal hematoma (PH), and symptomatic ICH. ICH was assessed using CT or gradient echo magnetic resonance imaging. PH was defined as PH1 and PH2 as described in the hemorrhagic transformation classification system of the European Cooperative Acute Stroke Study.²² Symptomatic ICH was defined as ICH associated with ≥4-point increase in NIHSS score.²²

Continuous data are summarized as means with SD or medians with interquartile range. Categorical data are shown as frequencies with percentage. Groups were compared using the Wilcoxon rank-sum test, Kruskal-Wallis test, or Fisher exact test as appropriate. Baseline characteristics and outcomes were compared between groups using univariate analysis. Logistic regression models were then constructed for each binary outcome, and odds ratios with 95% CIs were calculated for the tortuous group using the nontortuous group as reference. The interaction between groups according to ICA tortuosity and the occlusion site regarding FPE was evaluated in the logistic regression model with calculation of P for interaction. Multivariate linear regression models were used to evaluate the association of ICA tortuosity with time from puncture to eTICI ≥2b reperfusion and procedural time. For the procedural and clinical outcomes, the following prespecified variables were included: age, sex, body weight, premorbid mRS score, baseline NIHSS score, occlusion site, first-line MT strategy, and onset-to-puncture time. For the safety outcomes, adjustments for age, sex, body weight, baseline NIHSS score, occlusion site, intravenous thrombolysis, first-line MT strategy, and onset-to-puncture time were performed. Subgroup outcome analyses of patients grouped according to location of ICA tortuosity (extracranial or cavernous) were performed using logistic regression models with the nontortuous group as the reference. Missing data were handled using the pairwise deletion. P<0.05 was considered significant. Statistical analyses were performed using the R Software, version 3.6.1 (R Foundation for Statistical Computing, Vienna, Austria), with the following packages: tidyverse, dplyr, sjPlot, magrittr, lmtest, margin, ordinal, and ggplot2.

The study flowchart is shown in Figure 2. A total of 370 patients (167 women [45%]; median age, 78 [interquartile range, 71–83] years; median NIHSS score, 19 [interquartile range, 13–24]) were analyzed. Of the 370 patients, 128 patients (35%) had occlusion in the ICA, 157 patients (42%) in the M1 segment, and remaining 85 patients (23%) in the M2 segment. One hundred and twenty patients (32%) were treated by SR thrombectomy, 132 patients (36%) by CA, and 96 patients (26%) by combined SR and CA as the first-line MT strategy. MT was only attempted in 18 patients (5%) due to difficult access in 4 patients, reperfusion before MT in 11 patients, and arterial perforation before MT in 3 patients. Successful reperfusion (final eTICI score ≥2b) was achieved in 315 patients (85%); median time from puncture to reperfusion was 45 minutes (interquartile range, 29–74).

Tortuosity in the extracranial ICA and cavernous ICA could be assessed by baseline angiography in 269 (73%) and 272 patients (74%), respectively. The overall κ-coefficient for interrater agreement of the dichotomized tortuosity status was 0.70 in the extracranial ICA and 0.79 in the cavernous ICA. For the extracranial ICA, discussion was required to resolve discrepancies in 30 cases; among these, 14 were classified as tortuous and 16 as nontortuous. For the cavernous ICA, discussion was required for 33 cases; 11 were classified as tortuous and 22 as nontortuous.

The tortuous group comprised 124 patients (34%). Among these, location of tortuosity was extracranial ICA in 35 patients (9%), cavernous ICA in 70 patients (20%), and both sites in 19 patients (5%). The remaining 246 patients (66%) were assigned to the nontortuous group. Patient characteristics stratified according to the group are shown in Table 1. The tortuous group was older and more female and had shorter time from onset to hospital arrival than the nontortuous group. The tortuous group had the marginally higher baseline NIHSS score. Atrial fibrillation was more frequently observed in the tortuous group than in the nontortuous group. No significant difference was found in the first-line MT strategy (Table 1). Patient characteristics among the 4 subgroups of patients with no tortuosity, with extracranial ICA tortuosity, with cavernous ICA tortuosity, and those with tortuosity at both sites are shown in Table S2.

Procedural outcomes according to the group are summarized in Table 2. The rate of FPE was significantly lower in the tortuous group than in the nontortuous group (adjusted odds ratio [aOR], 0.45 [95% CI, 0.26–0.77]). The rate of achieving final eTICI 2c/3 reperfusion was also significantly lower in the tortuous group than in the nontortuous group (aOR, 0.60 [95% CI, 0.38–0.95]). There were no significant differences in the rates of final eTICI ≥2b and procedural complications. Failure to reach the target occlusion was observed in 3 patients (2.5%) of the tortuous group and 1 patient of the nontortuous group (0.4%; P=0.11). The tortuous group had higher number of passes (2 [1–3] versus 1 [1–3]; P=0.04), longer time from puncture to eTICI ≥2b reperfusion (59 [37–89] versus 40 [28–63] minutes; P<0.001), and longer procedural time (80 [56–117] versus 68 [42–112] minutes; P=0.009) than the nontortuous group. On multivariate linear regression analyses, ICA tortuosity was associated with time from puncture to eTICI ≥2b reperfusion (β=23.19 [95% CI, 13.44–32.94]) but not with procedural time (β=9.54 [95% CI, −11.27 to 30.35]). The median time from stroke onset to eTICI ≥2b reperfusion was similar between groups (239 [166–429] versus 232 [171–445] minutes; P=0.90). Procedural outcomes among the 4 subgroups according to location of ICA tortuosity are shown in Figure 3A and 3B.

In the tortuous group, the rate of FPE was similar in patients who underwent combined SR and CA and in those who underwent SR thrombectomy or CA alone (22% versus 19%; P=0.80). However, in the nontortuous group, the FPE rate was significantly higher in patients who underwent combined SR and CA (52% versus 35%; P=0.02; Figure 4).

The rate of FPE was lower in the tortuous group than the nontortuous group in patients with ICA occlusion (aOR, 0.25 [95% CI, 0.10–0.67]) and M1 occlusion (aOR, 0.38 [95% CI, 0.16–0.93]) but not in those with M2 occlusion (aOR, 1.63 [95% CI, 0.49–5.44]; P for interaction, 0.04; Figure S1).

The rates of favorable outcome did not differ between groups (Table 2). The excellent outcomes were numerically lower in the tortuous group than in the nontortuous group, though the difference did not reach statistical significance. Neurological improvement at 7 days was less frequent in the tortuous group than in the nontortuous group (aOR, 0.56 [95% CI, 0.34–0.91]). The overall distribution of the mRS scores at 3 months in the tortuous and nontortuous groups is shown in Figure S2.

Regarding safety outcomes, the tortuous group had the higher frequency of any ICH than the nontortuous group (aOR, 1.61 [95% CI, 1.02–2.53]). PH was more frequent in the tortuous group (11%) than in the nontortuous group (6%; aOR, 2.41 [95% CI, 1.04–5.58]). However, frequency of symptomatic ICH did not significantly differ between groups. Frequency of ICH among the 4 subgroups is shown in Figure 3C.

Outcomes according to the location of ICA tortuosity are summarized in Table S3. The FPE rate was lower in patients with extracranial ICA tortuosity alone than in those without tortuosity (aOR, 0.37 [95% CI, 0.14–0.96]). The rate of achieving final eTICI 2c/3 reperfusion was lower in patients with cavernous ICA tortuosity alone than in those without tortuosity (aOR, 0.54 [95% CI, 0.31–0.94]). The frequency of subarachnoid hemorrhage (aOR, 1.96 [95% CI, 1.03–3.71]) was higher in patients with cavernous ICA tortuosity alone than in those without tortuosity. Patients with cavernous ICA tortuosity alone had numerically more PH than those without tortuosity (aOR, 2.65 [95% CI, 0.98–7.21]).

Tables S1–S3

Figures S1–S2

STROBE Statement