Clinical Outcomes of Extracranial Carotid Artery-Related Stroke Eligible for Mechanical Reperfusion on Top of Per-Guidelines Thrombolytic Therapy: Analysis from a 6-Month Consecutive Patient Sample in 2 Centers

Background

Carotid atherosclerotic disease is an important [1–4] yet underestimated [5] cause of ischemic stroke. Strokes of carotid bifurcation origin constitute at least 20% of ischemic strokes [1,6]; some of those present with a co-existing occlusion of the intracranial vessel(s) (tandem lesions) [7,8]. Acute ischemic stroke of extracranial carotid artery origin (AIS-CA) has an unfavorable clinical prognosis due to the large volume of affected brain tissue and the typically large thrombus load, with recanalization rates below 10% using systemic intravenous thrombolysis (IVT) [9–11]. IVT is an established part of stroke reperfusion therapy that, in the setting of large vessel occlusions (LVO), should be combined with mechanical reperfusion [10–16]. There is some evidence in intracranial LVO strokes that IVT administration prior to emergency mechanical reperfusion (EMR) may be associated with increased early recanalization rate [14–16]; but this may be less applicable to AIS-CA due to the typically large thrombus burden in AIS-CA [6,9,17]. Some authors suggest that the bleeding risk with mechanical recanalization may be increased if IVT precedes mechanical thrombectomy (MT) [13,18].

Pivotal trials of mechanical intervention in AIS have been focused on intracranial LVO; patients with AIS due to extracranial carotid artery (CA) occlusion were routinely excluded or severely underrepresented [17,19,20]. As a result of trials, MT of the intracranial CA occlusion causing AIS is today recommended in all eligible patients (class I, level of evidence A recommendation) [12,17]. This contrasts with the data gap regarding EMR of causative occlusions of the extracranial internal CA, including tandem lesions. Today, stroke guidelines from the American Heart Association/American Stroke Association [12] indicate that emergent interventional management of extracranial CA lesions causing stroke “may be considered” (class IIb recommendation), in the absence of clarification regarding which treatment method (ie, endovascular [7,14,21], or carotid surgery±intracranial thrombectomy [21,22]) should be used. Surprisingly, the current European Stroke Organization guideline on carotid stenosis management states that it does “not address carotid revascularization done as part of acute stroke therapy”, leaving European stroke physicians without a guideline on AIS-CA patient management [23]. The relevance of those gaps for everyday clinical practice decision-making remains unknown.

Recent technological advances allow thrombus extraction using aspiration and/or stentrievers, with a high recanalization rate [11,17], from the extracranial internal CA (similarly to the intracranial segments of internal CA [24]), as well as a full reconstruction of the CA lumen with lesion sequestration [25]. However, data on the implementation of these new technologies in acute stroke are scarce [24].

In a sample of consecutive EMR-eligible patients in 2 centers, we performed an assessment of the real-life impact of randomized evidence gaps (resulting in class IIb guideline recommendation level) on contemporary referral and treatment pathways, as well as neurologic outcomes, in AIS-CA.

Material and Methods

ETHICS APPROVAL:

The study was approved by the Regional Ethics Committee (OIL/KBL/75/2021). Individual patient consents were waived due to retrospective analysis of anonymized data.

STUDY CENTERS:

Prospective data collection involved all consecutive patients with AIS receiving cerebral and vascular imaging in 2 collaborating centers with acute stroke neurology. Study Center 1 is a high-volume major tertiary 24/7 cardiovascular center with carotid artery stenting (CAS) expertise with a case load of 350–400 CAS procedures per year [26–28] and MT service under World Federation for Stroke Treatment (WIST) certification and supervision (Level-2 stroke center – Thrombectomy-Capable Stroke Center, TCSC) [24,29]. In addition, vascular surgery treatments (including carotid endarterectomy of ≈200 cases per year) are performed [26–28]. A multispecialty local Stroke Endovascular Mechanical Reperfusion Team with interventional cardiology and angiology, stroke neurology, radiology, and anesthesia is operational within the Level-2 WIST stroke center framework [24,29].

Study Center 2 is a major district hospital (MDH) incorporating regional stroke neurology and on-site 24/7 interventional cardiovascular facilities with neuroradiology procedures performed on an elective basis. Center 2 has developed a CAS program and is presently en route to serve as a stroke MT (Level-2) center for patients in the region. Study Center 2 has a well-established cardiovascular collaboration with Study Center 1. However, its LVO stroke patients are typically referred to a regional Comprehensive Stroke Center (CSC, Level-1 stroke center). Travel time from Study Center 2 to the CSC is ≈90 min, resulting in a minimum transfer time of 120 min.

There are 5 Level-3 stroke centers in the region; these collaborate with both the CSC and TCSC. The CSC is the only Level 1 stroke center in an administrative area of about 3.5 million inhabitants.

PATIENT ELIGIBILITY:

In a prospectively-run stroke imaging database in Center 1 and Center 2, 32 consecutive patients with AIS-CA were identified over a 6-month period. For the 2 patients accepted for EMR by the CSC, no final treatment status (ie, EMR delivered vs not performed) and no clinical outcome data could be obtained, resulting in 30 subjects in the present analysis.

A clinical committee consisting of a neurologist, diagnostic radiologist, and stroke management-certified interventionist reviewed the clinical and imaging data in all consecutive AIS-CA stroke patients. Eligibility for EMR [12] was defined as: (1) NIHSS (National Institutes of Health Stroke Scale) ≥6 or a significant neurologic deficit (eg, aphasia), (2) ASPECTS (Alberta Stroke Program Early Computed Tomography Score) ≥6, (3) pre-stroke modified Rankin scale (mRS) ≤2, and (4) evidence of internal carotid artery occlusion/sub-occlusion on computed tomography angiography (CTA). For each study patient, their EMR eligibility was verified by a senior interventional neuroradiologist with over 15 years of experience in acute stroke interventions [30]. Administration of IV -thrombolysis, referral for EMR (surgical or endovascular), and the treatment received were evaluated.

ENDOVASCULAR TREATMENT STRATEGY:

The endovascular EMR strategy involved internal CA revascularization under routine flow-reversal cerebral protection, which reduces cerebral embolism in elective CAS [31–34] and in stroke intracranial interventions [35,36]. Aspiration or carotid-dedicated large-diameter thrombectomy device [24] was used in case of CA large thrombus. Intracranial LVO MT [19] was performed in tandem lesions. The culprit CA lesion sequestration was performed using micronet-covered stents [4,25,28,37,38]. Functional status was routinely assessed at 90 days.

PERIPROCEDURAL IMAGING AND PHARMACOLOGICAL TREATMENT REGIMEN:

Guideline-recommended [12] study imaging involved: plain cerebral computed tomography (cCT), CTA, cerebral magnetic resonance imaging (MRI) with diffusion-weighted imaging (DWI), apparent diffusion coefficient (ADC); fluid-attenuated inversion recovery (FLAIR) sequences, carotid Doppler ultrasound (if performed), and (in patients with endovascular treatment) catheter angiography (Figures 1, 2). Perfusion imaging was performed as clinically indicated or as required to establish EMR eligibility.

Optimal (as per current criteria) medical therapy (OMT) prior to stroke onset was defined as pharmacological treatment involving at least 1 antiplatelet agent, statin dose titrated to achieve guideline-indicated LDL cholesterol level (or maximally tolerated dose), and angiotensin-converting enzyme inhibitor/receptor blocker [39].

At the index event, IVT was administered in eligible patients (ie, those without contraindications to thrombolysis and presenting within 4.5 h after stroke onset) [12]. In the group receiving endovascular treatment, peri-procedural pharmacotherapy included heparinization at a reduced or full dose (at operator’s discretion) and a single periprocedural antiplatelet agent (IV acetylsalicylic acid), followed by a second antiplatelet agent after the intervention [12].

Care was taken to provide a tight periprocedural blood pressure control with: (1) systolic blood pressure of 160–180 mmHg prior to and during the intervention (to enhance flow through the stroke-related CA and enhance collateral supply), followed by (2) systolic blood pressure reduction to 100–120 mmHg after reperfusion of the intracranial vessels to reduce the risk of cerebral bleed [40].

STATISTICAL ANALYSIS:

Continuous variables were reported as median (Q1–Q3) and categorical data were expressed as numbers and proportions. The Mann-Whitney U test or Wilcoxon matched pairs test was used for between-group and within-group comparisons. Bonferroni correction was applied for multiple comparisons. Statistical significance was defined as P<0.05. Statistica 10 (StatSoft GmBH, Hamburg, Germany) was used for computations.

Results

STUDY COHORT:

Table 1 shows clinical data in consecutive AIS-CA patients in the study (age 39–87 years, 26.7% women): 93.3% had ASPECTS ≥9; NIHSS was 12–22 (median 17); 83.3% had pre-stroke mRS-0, 16.7% mRS-1; all eligible patients received thrombolysis (33.3%); and 20 (66.7%) were referred for EMR (18 for emergency endovascular management, 2 for emergency carotid endarterectomy). Figures 1 and 2 present typical examples of the disease course in EMR-untreated and EMR-treated patients on top of IVT as per guidelines.

IMAGING AND CLINICAL PATHWAYS:

Seventy five percent EMR-treatment-eligible referrals were rejected by the CSC. Out of the 2 patients referred for emergency surgery (carotid endarterectomy, CEA), none were treated (1 rejected on referral, 1 initially accepted for CEA but surgery was not performed). Once rejected, an alternative center was not contacted or alternative reperfusion method referral was not attempted in any of the EMR-eligible patients rejected by CSC or vascular surgery.

Only 40% of all study participants (60% of all EMR referrals, n=12 patients) received EMR. The EMR-untreated group included EMR-eligible patients who were either not referred or were referred but rejected (n=18; 60% of the study cohort). There were no significant clinical, demographic, or lesion-related differences between the EMR-untreated group and EMR-treated group, including NIHSS and functional status prior to stroke onset (Table 1). Prior to stroke onset, 13 patients (43.3%) were on documented OMT.

Eighty percent of patients who (subsequently) did not receive EMR presented within 6 h of symptom onset and had a favorable ASPECTS ≥6. A further 20% of the EMR-untreated group were wake-up strokes with favorable DWI/FLAIR mismatch and an ASPECTS ≥6, consistent with EMR eligibility. Similarly, 75% of EMR-treated patients presented within the 6-h window. Cerebral perfusion imaging confirmed EMR eligibility in a further 12.5% who presented with stuttering symptoms. Another 12.5% of EMR-treated patients initially presented with a “low” NIHSS (NIHSS of 3 and 4) in peripheral hospitals, without an initial referral. These patients were referred for EMR when their NIHSS exceeded 6; imaging repeated within 1 h of EMR still showed ASPECTS ≥6.

In the patients who did not receive EMR despite eligibility, the reasons were (1) lack of referral (56%), (2) being referred but not accepted for EMR treatment by CSC (40.7%), or carotid endarterectomy treatment not being performed (3.3%, no patient referred for CEA or referred and accepted for CEA received CEA).

REPERFUSION STRATEGY:

IVT was administered in all eligible patients (10, 33.3%), of which only 2 were immediately transferred for EMR (Table 1).

Proximal cerebral protection with transient flow reversal was applied if feasible (91.7%; distal filter use in 8.3% of interventions). In patients with particularly large thrombus burden at the carotid bifurcation, thrombectomy was performed using a dedicated large-diameter (9-mm) stentriever on top of aspiration (Figure 2B) [24]. Tandem lesion presence mandated intracranial thrombectomy performed with aspiration as the technique of 1st choice (followed, if needed, by stentriever). Micronet-covered stents were routinely used for culprit lesion sequestration and lumen reconstruction (length 30 or 40 mm, diameters 9.0 or 10.0 mm). Stent post-dilatation optimization was routinely performed to ensure device embedding. Modified Thrombolysis in Cerebral Infarction (TICI) reperfusion grade 2b–c/3 was achieved in all EMR-treated cases.

NEUROLOGIC OUTCOMES:

Good neurologic outcomes (mRS 0–2, functional independence) at 90 days occurred exclusively in the EMR-treated group (91.7% vs 0%, P<0.001, Figure 3B). Poor functional outcome (mRS 3–5) prevalence was 8.3% vs 88.9% (P<0.001, EMR-treated vs untreated, Figure 3B). Intracerebral bleeding occurred in 1 patient (EMR-untreated group, thrombolysis given). There were 2 deaths by 90 days after the intervention (11.1%, mRS 6), and both occurred in the EMR-untreated group. The neurologic outcome difference between the study groups (Figure 3) was driven by patients who reached mRS 0–2 (P<0.001).

Discussion

LIMITATIONS:

Limitations include our retrospective analysis of prospectively collected data and the sample size. In the era of MT, obtaining further randomized evidence with regard to AIS-CA management in patients who qualify for EMR on top of thrombolysis is unlikely, as this would be unethical today. While the sample size may be considered moderate, it is similar to other AIS-CA cohort studies published [7,41,46,47]. An important element of our present work is the comparison of contemporary outcomes in the treated and untreated AIS-CA patients (Figure 3). With the moderate study size, however, the present comparison needs to be regarded as a pilot study with respect to the size of the treatment effect. As it would be unethical today to randomize AIS-CA patients to emergency intervention vs no intervention (or delayed intervention) [12], it is crucial to expand the body of knowledge by analysis of real-life datasets such as ours. Next, larger datasets, such as data from multiple centers and/or regional or national databases, are needed. However, it must be noted that detailed patient-level verification of criteria for emergency mechanical recanalization (which we achieved through detailed analysis using access to full imaging and clinical data) would be difficult for large patient cohorts with limited access to source data.

Regrettably, we could not obtain any specific information regarding reasons for non-referral/non-acceptance for EMR. Likely contributing factors include: (1) “symptomatic” carotid disease being historically considered a primarily surgical disease, to be addressed within 14 days from symptom onset [23]; (2) CSCs primary focus on intracranial LVO management [8,19,20]; (3) unfamiliarity with the treatment of carotid origin disease under flow reversal [7,21,45]; (4) shortage of stroke MT centers[50,51]; and (5) lack of established pathways for emergency AIS-CA stroke referrals that would involve centers with expertise in endovascular carotid revascularization [49]. Furthermore, clinical practice suggests that rejection of a particular patient type will impact further referrals of similar patients. Thus, “symptomatic” (elective) CA stenosis guidelines should not be applied blindly to AIS-CA, but should rather be used together with the acute stroke guidelines that demand emergency reperfusion not only in intracranial LVO but also in carotid LVO [12].

Finally, it is important to note that in many patients with large carotid-related strokes, once they reach the post-acute stroke stage, any potential intervention “within 14 days” [39] (whether surgical or endovascular) becomes futile due to the large irreversible loss of brain tissue and an unacceptable risk of intracranial hemorrhage. Such patients are not included in most statistics on symptomatic carotid stenosis and its revascularization.

Conclusions

An acute stroke due to an underlying carotid lesion represents a cerebral emergency, irrespective of the presence/absence of co-existing intracranial LVO and irrespective of any thrombolysis administration. Our analysis suggests that patients receiving emergency mechanical reperfusion have overwhelmingly better clinical outcomes. To increase the proportion of AIS-CA patients who receive mechanical reperfusion in acute stroke, cardioangiology centers with expertise in endovascular carotid revascularization and with knowledge of proximal cerebral protection and stenting should actively participate in delivering mechanical reperfusion for patients with acute stroke [55].