Association Between Ipsilateral Stroke and Nonstenotic (<50%) Carotid Disease: Secondary Analysis From the AcT Trial

Abstract

Background

Symptomatic nonstenotic (<50% stenosis) carotid disease in the presence of high‐risk plaque features is a potential cause of ischemic stroke. We assessed stroke risk associated with symptomatic nonstenotic carotid disease.

Methods

This cross‐sectional secondary analysis of the AcT (Alteplase Compared to Tenecteplase) randomized controlled trial evaluated baseline computed tomography angiograms for degree of internal carotid artery stenosis, plaque features and the presence of intraluminal thrombi, webs, dissection, and rim sign. Stroke location was evaluated on 24‐hour follow‐up imaging. At a carotid level, mixed‐effects logistic regression models adjusted for age and sex, with patient identity as a random effect, examined associations between “concordant stroke” (ipsilateral acute stroke in the internal carotid artery territory) and symptomatic nonstenotic carotid disease.

Results

Of 1577 patients enrolled, 1407 (89.2%) with interpretable imaging were included: 329 (23.4%) had no carotid disease, 869 (61.8%) had nonstenotic carotid disease, and 209 (14.9%) had stenotic (≥50%) carotid disease in either the left or right internal carotid artery. Median age was 73 years (interquartile range, 63–83), with 48% female patients. Among 2519 (89.5%) internal carotid arteries with nonstenotic disease, 689 (27.4%) concordant strokes were identified. Intraluminal thrombi, carotid webs, carotid dissections, and carotid rim sign were significantly associated with concordant stroke (adjusted odds ratio, 8.11 [95% CI, 1.60–41.08]; adjusted odds ratio, 3.58 [95% CI, 1.53–8.35]; adjusted odds ratio, 6.77 [95% CI, 1.72–26.75]; and adjusted odds ratio, 3.17 [95% CI, 1.39–7.23], respectively). Results remained unchanged after excluding patients with atrial fibrillation and lacunar infarctions.

Conclusions

Features other than the degree of stenosis should be considered when evaluating patients with carotid disease.

Nonstandard Abbreviations and Acronyms

AcT

Alteplase Compared to Tenecteplase

ECST

European Carotid Surgery Trial

ICA

Internal Carotid Artery

INTERRSeCT

Identifying New Approaches to Optimize Thrombus Characterization for Predicting Early Recanalization and Reperfusion With IV Alteplase and Other Treatments Using Serial CT Angiography

NASCET

North American Symptomatic Carotid Endarterectomy Trial

OPTIMISE

Optimizing Patient Treatment in Major Ischemic Stroke

PARISK

Plaque At Risk study

QuICR

Quality Improvement and Clinical Research stroke registry

STRATIS

Systematic Evaluation of Patients Treated with Neurothrombectomy Devices for Acute Ischemic Stroke

TOAST

Trial of ORG 10172 in Acute Stroke Treatment

Clinical Perspective.

What Is New?

• This large secondary analysis from the AcT (Alteplase Compared to Tenecteplase) trial shows that in patients with nonstenotic (<50%) carotid disease, certain high‐risk plaque features, including intraluminal thrombi, carotid webs, dissections, and rim sign, are strongly associated with ipsilateral acute ischemic stroke.

• The degree of luminal narrowing alone fails to capture the full spectrum of carotid‐related stroke risk in this population.

What Are the Clinical Implications?

• Current stroke workups and treatment decisions often rely heavily on the degree of carotid stenosis; our findings highlight the need to integrate plaque morphology and vulnerability features into routine imaging assessment.

• Identifying and reporting these features on baseline computed tomography angiography could enhance risk stratification, guide secondary prevention strategies, and help select patients for future interventional trials.

• This study supports a shift toward a “beyond stenosis” paradigm in carotid disease evaluation, particularly for patients presenting with acute ischemic stroke.

Large‐artery atherosclerosis affecting the extracranial internal carotid artery (ICA) is an important cause of ischemic stroke. ¹ According to the TOAST (Trial of ORG 10172 in Acute Stroke Treatment) criteria, this pathogenesis is considered if the involved artery is ≥50% stenosed due to plaque or stenosis. ¹ Meanwhile, patients with stroke and ipsilateral carotid stenosis of <50% are classified as stroke of undetermined pathogenesis or cryptogenic stroke if no other stroke pathogeneses are identified. ¹ Moreover, carotid revascularization with either carotid endarterectomy or stenting is currently recommended for symptomatic carotid stenosis of ≥50%. ² These guidelines were based on data from 2 pivotal trials, the ECST (European Carotid Surgery Trial) and the NASCET (North America Symptomatic Carotid Endarterectomy Trial), which showed that revascularization in patients with symptomatic carotid stenosis led to reduced stroke risk in the ipsilateral territory in a 5‐year follow‐up period. ³ , ⁴

The approach to managing patients with symptomatic nonstenotic carotid disease exhibiting <50% stenosis, remains unclear. ⁵ , ⁶ Nonstenotic carotid disease encompasses not only atherosclerotic plaques but also carotid webs, dissections, and other high‐risk features. ⁷ A recent meta‐analysis of patients with nonstenotic carotid plaques identified certain high‐risk carotid plaque features associated with an increased risk of stroke. ⁸ These features include ulceration, intraplaque hemorrhage, thrombus, fibrous cap rupture, echolucency, and plaque thickness measuring ≥3 mm as identified on magnetic resonance imaging, computed tomography angiography (CTA), or ultrasound. ⁸ In patients with embolic stroke of undetermined source, these high‐risk plaque features were more prevalent in the ipsilateral ICA regardless of the percentage of stenosis. ⁸

Nonstenotic carotid plaques are increasingly recognized as an important stroke pathogenesis. There are a few studies of nonstenotic carotid plaques in the setting of acute ischemic stroke and their role in stroke pathogenesis remains underrecognized. Such strokes, herein referred to as concordant strokes, offer the advantage of using a definitive clinical event as a concurrent outcome for evaluation of simultaneously detected carotid pathology on acute neurovascular imaging. ⁹ We sought to determine the association of neuroimaging plaque features with concordant stroke among patients with nonstenotic carotid disease in a large, pragmatic, phase 3, registry‐linked randomized controlled trial of acute ischemic stroke. These analyses could help further interrogate the association of various nonstenotic carotid disease features with ischemic stroke.

Methodology

Study Population

The present study is a cross‐sectional analysis of the AcT (Alteplase Compared to Tenecteplase) trial (https://www.clinicaltrials.gov; unique identifier: NCT03889249), a pragmatic, registry‐linked, multicenter, prospective, randomized controlled trial that evaluated the noninferiority of tenecteplase compared with alteplase for thrombolysis in acute ischemic stroke. ¹⁰ The trial was regulated by Health Canada and approved by institutional review boards at each participating center and used deferred consent procedures at sites where this was permitted. Details of the trial are published elsewhere. ¹¹ In brief, patients included in the trial were aged ≥18 years, diagnosed with ischemic stroke causing disabling neurological deficits, and presented within 4.5 hours of symptom onset. ¹⁰ Standard contraindications to intravenous thrombolysis applied. ¹⁰ Baseline demographic data, comorbidities, and procedure codes were gathered from the Canadian Institute for Health Information administrative databases that were linked with the trial data and with the QuICR (Quality Improvement and Clinical Research) stroke registry and the OPTIMISE (Optimizing Patient Treatment in Major Ischemic Stroke With EVT) registry. ¹¹ QuICR collects data for all patients with stroke treated with intravenous thrombolysis and/or endovascular thrombectomy in Alberta, while OPTIMISE is a national registry with similar data fields for provinces in Canada other than Alberta. ¹¹ Imaging core laboratory personnel blinded to treatment allocation assessed available baseline and 24‐hour imaging. These included noncontrast computed tomography (CT) scans and magnetic resonance imaging scans of the head, head and neck CTA, magnetic resonance angiograms, and digital subtraction angiograms, where applicable. In this analysis, “concordant stroke” was defined as stroke in the ipsilateral hemisphere supplied by the ICA. Meanwhile, “discordant stroke” was defined as stroke in a territory that is not supplied by the ICA of interest or having no strokes in the territory of interest. Stroke in the contralateral hemisphere, ipsilateral posterior circulation stroke, and multiterritorial strokes were also considered discordant strokes.

Data Availability

Data supporting the findings of this study are available from the corresponding author upon reasonable request.

Image Analysis

Imaging core laboratory members composed of radiologists and neurologists with at least 5 years of experience (F.B., N.S., and I.A.) read baseline neuroimaging. Further extracranial ICA morphology and concordant stroke were assessed in detail by 5 trained readers composed of 1 radiologist and 4 neurologists with 4 years of experience (K.I., S.N., I.A., C.K., and K.T.) who analyzed the CTA images for carotid features using OsiriX version 10.6. Readers were blinded to all clinical data. The training protocol included review of carotid imaging slide decks and an initial session where 50 CTAs were read together by the team followed by a feedback session to resolve discrepancies.

ICA morphology was assessed on all available planes. Measurements to determine the degree of stenosis and plaque thickness were done according to the NASCET criteria on the axial plane. ³ In the presence of stenosis, the ICA was measured at its narrowest portion. Distal ICA measurements were taken 1 to 2 cm beyond the ICA bulb where the vessel walls were parallel and no longer tapering. ⁶ , ¹² When the ICA was tortuous or dolichoectatic, measurements to determine stenosis were done on the sagittal or coronal planes. The degree of stenosis was then classified into 3 categories as no carotid stenosis, <50% or nonstenotic carotid disease, and ≥50% or stenotic carotid disease, which included near occlusion (tight stenosis with distal luminal collapse) and complete occlusion of the carotid artery.

Plaque thickness was measured at the thickest level on axial images. The interface between the vessel wall and surrounding tissue was set as the outer marker, while the interface between the plaque and vessel lumen was set as the inner marker for measurement. ⁶ Plaques were classified as purely hypodense, microcalcifications, predominantly noncalcified (<50% calcified), predominantly calcified (≥50% calcified) or purely calcific. Near‐occlusion of the carotid was defined as a narrowed distal lumen or distal luminal collapse, while occlusion was defined as the absence of distal flow.

The presence or absence of the following additional features were also assessed for: plaque irregularity, plaque ulceration, carotid web, carotid dissection, and intraluminal thrombus. Plaque irregularity was defined as surface irregularity in a plaque. ⁹ , ¹³ Plaque ulceration was defined as contrast within the plaque or beyond the vascular lumen of at least 1 mm in 2 planes. ¹³ A carotid web was identified on oblique sagittal images when a thin, smooth, membrane‐like filling defect was observed along the posterior wall of the ICA with a corresponding thin septum on axial images. ¹³ , ¹⁴ The Biffl scale was used to grade the severity of carotid dissection. ¹⁵ Grade 1 is defined as minor intimal irregularity, grade 2 as dissection with intramural hematoma causing >25% luminal narrowing, grade 3 as a pseudoaneurysm, grade 4 as arterial occlusion, and grade 5 as transection with extravasation. ¹⁵ An intraluminal thrombus was identified by a filling defect in the ICA lumen surrounded by contrast and visible on at least 2 contiguous axial images, also referred to as the donut sign. ¹⁶ The rim sign was defined as the presence of calcifications on the outer wall of the ICA with internal soft‐tissue plaque. ¹⁷ See Figure 1 for representative CT angiographic images of the different plaque features assessed. Ischemic stroke and stroke location were recorded as identified on head CT or magnetic resonance imaging.

Figure 1.

Representative computed tomography angiographic images for purely hypodense plaque (A, sagittal), mixed hypodense and calcified plaque (B, sagittal), irregular plaque (C, sagittal), carotid web (D, sagittal), carotid dissection and ulceration (E, sagittal) and intraluminal thrombus (F, axial).

Statistical Analysis

Statistical analysis was conducted at 2 levels: the patient level and the carotid level. At the patient level, patients were classified on the basis of the degree of the highest stenosis in either carotid artery. Meanwhile, at the carotid level, all nonstenotic carotids (left, right, or both) were included. For example, if a patient had 40% stenosis in the left ICA and 60% stenosis in the right ICA, they would be classified in the stenotic ICA group in the patient‐level analysis, while the left ICA would be included in the carotid‐level analysis.

At the patient level, baseline demographic and clinical characteristics were compared among patients with no carotid disease versus those with nonstenotic (<50%) carotid disease versus patients with stenotic (≥50%) carotid disease. χ² tests were used to assess differences between groups for categorical variables, while Wilcoxon rank‐sum tests were used for continuous variables. Frequencies and proportions were reported for categorical variables, while median and interquartile ranges were reported for continuous variables. Analyses were conducted using available data without imputation; missing values were reported in the tables where applicable. Further, the proportion of patients with ischemic stroke who had concordant nonstenotic carotid disease (<50% stenosis) on the same side as the infarct was assessed. The risk of ipsilateral ischemic stroke attributable to nonstenotic carotid disease and the population‐attributable risk were calculated using standard epidemiological formulas for attributable risk and population‐level impact. Risk estimates were reported with 95% CIs. A subgroup analysis was conducted excluding patients with atrial fibrillation and small‐vessel disease to isolate the effect of nonstenotic carotid disease in the absence of other major stroke mechanisms.

Aligning with the study aims, univariable analysis was performed at the carotid level to compare the prevalence of plaque features in nonstenotic carotid disease with concordant stroke compared with those with discordant or no strokes. At a carotid level, mixed‐effects logistic regression analysis was conducted to determine the association between each high‐risk plaque features and the presence of concordant stroke in this subpopulation. In carotid‐level analyses, each patient contributed up to 2 carotid arteries. To account for within‐patient correlation, we used a mixed‐effects logistic regression model with a random intercept for patient identification, fitted using unstructured covariance structure. The analysis was adjusted for age and sex, with patient identity as a random‐effects variable. Adjusted and unadjusted odds ratios, along with their corresponding 95% CIs, were reported. Logistic regression was used to plot predicted probability of concordant stroke for continuous measure of plaque thickness and degree of stenosis using the NASCET grade. A sensitivity analysis was performed excluding patients with concomitant atrial fibrillation and lacunar infarctions to further isolate the effect of nonstenotic carotid disease. A 2‐sided P<0.05 was considered statistically significant. Statistical analysis was conducted using STATA/MP 17.0 (StataCorp, College Station, TX).

Results

Of the 1600 patients enrolled in the AcT trial, 1407 (88%) patients were included in the current study. Reasons for exclusion can be found in the study flowchart (Figure 2).

Figure 2. Study flowchart.

AcT indicates Alteplase Compared to Tenecteplase; CTA, computed tomography angiogram; CT, computed tomography; ICA, internal carotid artery; and MRI, magnetic resonance imaging.

Overall, the median age of the study population was 73 (interquartile range, 20–100) years, 672 (48%) were women, median baseline National Institutes of Health Stroke Scale was 9 (interquartile range, 6–16), and baseline Alberta Stroke Program Early Computed Tomography Score was favorable (9 [interquartile range, 8–10]). A total of 725 patients (52%) received tenecteplase and 455 (33%) underwent endovascular thrombectomy. One hundred ninety‐two (14%) had atrial fibrillation, 646 (50%) had hypertension, and 251 (19%) had type 2 diabetes. Lacunar infarctions were noted in 92 patients (6.5%).

A total of 329 (23.4%) had no carotid disease, 869 (61.8%) had nonstenotic carotid disease, and 209 (14.9%) had stenotic carotid disease. Patients without carotid disease were younger compared with those with nonstenotic or stenotic carotid disease (median ages, 61 versus 77 and 74, respectively; P<0.001). Stroke severity was higher among patients with stenotic or occluded carotids compared with those with no carotid disease and nonstenotic carotid disease (median National Institutes of Health Stroke Scale, 16 versus 8 and 9, respectively; P<0.001). Among 869 patients with nonstenotic carotid disease, 114 patients (14.1%) had concomitant atrial fibrillation. A higher proportion of patients with stenotic carotid arteries underwent endovascular thrombectomy compared with those with no carotid disease and those with nonstenotic carotid disease (55.5% versus 28% and 26.7%, respectively; P<0.001) See Table 1 for baseline demographic, clinical, and radiologic characteristics.

Table 1.

Baseline Demographic and Clinical Characteristics Comparing Patients With No Carotid Disease, Patients With Nonstenotic Carotid Disease, and Patients With Stenotic Carotid Disease

Baseline characteristics	Overall patient sample (N=1407)	No carotid disease (N=329)	Nonstenotic carotid disease or <50% stenosis (N=869)	Stenotic (≥50%) carotid arteries (N=209)	P value
Age, y, median (IQR)	73 (63–83)	61 (51–74)	77 (68–84)	74 (66–84)	<0.001
Females, n (%)	672 (47.76)	160 (48.63)	441 (50.75)	71 (33.97)	<0.001
Baseline NIHSS, median (IQR)	9 (6–16)	8 (5–14)	9 (6–16)	16 (9–20)	0.0001
ASPECTS, median (IQR)	9 (8–10)	9 (8–10)	9 (8–10)	8 (7–10)	<0.001
Comorbidities, n (%)
Atrial fibrillation, n (%)	192 (13.65)*	38 (11.55)†	128 (14.73)‡	26 (12.44)§	0.309
Diabetes	251 (19.28)*	38 (12.71)†	174 (21.48)‡	39 (20.21)§	0.004
Dyslipidemia	68 (5.22)*	18 (6.02)†	38 (4.69)‡	12 (6.22)§	0.540
Coronary artery disease	23 (1.77)*	5 (1.67)†	15 (1.85)‡	3 (1.55)§	0.952
Hypertension	646 (49.62)*	109 (36.45)†	435 (53.70)‡	102 (52.85)§	<0.001
Prior stroke/TIA	12 (0.92)*	1 (0.33)†	11 (1.36)†	0	0.100
Received Tenecteplase	725 (51.53)	167 (50.76)	442 (50.86)	116 (55.50)	0.460
Underwent EVT	455 (33.12)	92 (27.96)	258 (26.69)	116 (55.50)	<0.001
Baseline intracranial occlusion site, n (%)
Intracranial ICA	121 (8.60)	21 (6.38)	36 (4.14)	64 (30.62)	<0.001
M1 segment MCA	215 (15.28)	43 (13.07)	130 (14.96)	42 (20.10)
M2 segment MCA	287 (20.40)	52 (15.81)	195 (22.44)	40 (19.14)
Other distal occlusions	287 (20.40)	68 (20.67)	192 (22.09)	27 (12.92)
Vertebrobasilar arterial system	58 (4.12)	19 (5.78)	36 (4.14)	3 (1.44)
No visible occlusions	458 (32.55)	137 (41.64)	300 (34.52)	21 (10.05)
Lacunar infarctions, n (%)	92 (6.54)	23 (6.99)	63 (7.25)	6 (2.87)	0.066

ASPECTS indicates Alberta Stroke Program Early Computed Tomography Score; EVT, endovascular thrombectomy; ICA, internal carotid artery; IQR, interquartile range; MCA, middle cerebral artery; NIHSS, National Institutes of Health Stroke Scale; and TIA, transient ischemic attack.

^*105 missing observations.

^† 30 missing observations.

^‡ 59 missing observations.

^§ 16 missing observations.

Of the 1407 patients included in the study, 488 (34.7%) had concordant strokes associated with nonstenotic carotid disease. The risk of ipsilateral ischemic stroke attributable to nonstenotic carotid disease in the overall population was estimated to be 33.5% (95% CI, 24.7%–41.2%) and the population‐attributable risk was calculated as 23.7%. After excluding patients with atrial fibrillation and small‐vessel disease, the risk of ipsilateral ischemic stroke attributable to nonstenotic carotid disease was estimated to be 34.9% (95% CI, 25.0%–43.6%), and the population‐attributable risk was calculated as 24.5%.

On carotid‐level analysis, 689 (27.4%) patients with nonstenotic carotid disease had concordant stroke, while 1830 (72.6%) had discordant stroke, of which 688 (37.5%) had no infarcts. Univariate analysis was performed at the carotid level to compare nonstenotic carotid disease with concordant stroke versus nonstenotic carotid disease with discordant or no stroke. Results showed that the degree of carotid stenosis did not significantly differ between the 2 groups (median, 15.84% versus 15.87%; P=0.846). The maximum plaque thickness also did not differ between the 2 groups (2.3 versus 2.2; odds ratio [OR], 1.02 [95% CI, 0.90–1.15]). The association between presence of plaque ulceration (1.83 versus 2.32%; OR, 0.78 [95% CI, 0.35–1.73]), plaque thickness of ≥3 mm (20.10 versus 18.49%; OR, 1.11 [95% CI, 0.83–1.49]), plaque irregularity (17.31 versus 17.51%; OR, 0.99 [95% CI, 0.74–1.32]), and predominantly hypodense plaques (44.98 versus 42.34%; OR, 1.11 [95% CI, 0.89–1.39]) did not significantly differ between the 2 groups (nonstenotic carotid disease with concordant strokes versus those with discordant or no strokes, respectively). These findings remained consistent after adjusting for age, sex, and clustering by patient identity. It was also found that every 1‐mm increase in plaque thickness was associated with a rise in the predicted probability of concordant stroke by 0.32%, although this was nonsignificant (OR, 1.02 [95% CI, 0.90–1.15]; Figure 3).

Figure 3. Graph of regression analysis showing relationship between maximum plaque thickness in mm and predicted probability of concordant stroke. Model accounts for patient identity as random effect, with CIs displayed to indicate the precision of the estimates.

Meanwhile, the presence of intraluminal thrombi (0.87 versus 2.00%; OR, 8.03 [95% CI, (1.62–39.88]), carotid webs (1.74 versus 0.49%; OR, 3.58 [95% CI, 1.50–8.55]), carotid dissections (1.02 versus 0.16%; OR, 6.25 [95% CI, 1.61–24.24]), and carotid rim sign (3.19 versus 1.03%; OR, 3.16 [95% CI, 1.44–6.89]) were significantly associated with a higher odds of concordant versus discordant stroke. These findings were consistent in adjusted analysis (adjusted OR [aOR], 6.98 [95% CI, 1.33–36.55 for intraluminal thrombus; aOR, 5.02 [95% CI, 1.93–13.10] for carotid webs; aOR, 6.77 [95% CI, 1.72–26.75] for carotid dissections; and aOR, 3.17 [95% CI, 1.39–7.23] for carotid rim sign). See Table 2 for carotid‐level analysis with adjusted and unadjusted analyses comparing high‐risk plaque features among patients with nonstenotic carotid disease with concordant stroke.

Table 2.

Carotid‐Level Analysis Comparing High‐Risk Plaque Features Among Patients With Nonstenotic Carotid Disease With Concordant Stroke and With Discordant Stroke

Carotid plaque features	Nonstenotic carotid disease with concordant stroke (N=689)	Nonstenotic carotid disease with discordant stroke (N=1830)	Unadjusted OR (95% CI)	Adjusted OR (95% CI)‡
Degree of carotid stenosis, %, median (IQR)	15.84 (8.92–24.27)	15.87 (9.44–24.39)	1.001 (0.99–1.01)	1.001 (0.99–1.01)
Maximum plaque thickness, mm, median (IQR)	2.3 (1.72–2.86)* N=388	2.2 (1.8–2.81) N=1022	1.02 (0.90–1.15)	1.03 (0.091–1.16)
Ulceration, % (n/total observations)	1.83 (8/438)	2.32 (27/1164)	0.783 (0.35–1.73)	0.80 (0.36–1.79)
Plaque thickness ≥3 mm, % (n/total observations)	20.10 (78/388)	18.49 (189/1022)	1.11 (0.83–1.49)	1.13 (0.85–1.50)
Plaque irregularity, % (n/total observations)	17.31 (76/439)	17.51 (204/1165)	0.99 (0.74–1.32)	0.99 (0.74–1.32)
Hypodense plaque ≥50% of plaque area, % (n/total observations)	44.98 (197/438)	42.34 (492/1162)	1.11 (0.89–1.39)	1.14 (0.92–1.42)
Presence of intraluminal thrombus, % (n/total observations)	0.87 (6/689)	0.11 (2/1830)	8.03 (1.62–39.88)†	8.11 (1.60–41.08)†
Carotid web, % (n/total observations)	1.74 (12/689)	0.49 (9/1829)	3.58 (1.50–8.55)†	3.58 (1.53–8.35)†
Dissection, % (n/total observations)	1.02 (7/689)	0.16 (3/1830)	6.25 (1.61–24.24)†	6.77 (1.72–26.75)†
Rim sign, % (n/total observations)	3.19 (14/439)	1.03 (12/1163)	3.16 (1.44–6.89)†	3.17 (1.39–7.23)†

IQR indicates interquartile range; and OR, odds ratio.

^*Plaque thickness assessed only when measurable plaque present.

^† P<0.05.

^‡ Adjusted for age, sex, with patient as cluster variable.

A sensitivity analysis, after excluding patients with concomitant atrial fibrillation and lacunar infarctions, showed similar results. See Table S1 for carotid‐level analysis comparing high‐risk plaque features among nonstenotic carotid disease with concordant stroke and with discordant stroke excluding patients with atrial fibrillation.

Discussion

The present study is a secondary analysis of the AcT trial conducted to determine the association between concordant stroke and neuroimaging plaque features in patients with nonstenotic carotid disease, defined as <50% ICA stenosis. The analysis revealed that the presence of intraluminal thrombi, carotid webs, carotid dissections, and carotid rim signs were significantly associated with concordant stroke in this population. Meanwhile, other generally accepted high‐risk plaque features, including plaque ulceration, plaque thickness of ≥3 mm, plaque irregularity, and hypodense plaques, were not significantly associated with concordant stroke in nonstenotic carotids.

Recent studies suggest that nonstenotic carotid disease is a potential pathogenesis for stroke, especially among those with embolic stroke of unknown source. ⁸ Certain neuroimaging features of carotid plaques can contribute to stroke risk and may indicate the presence of vulnerable plaques, even in the absence of significant stenosis. ⁸ , ¹⁸ These features include intraplaque hemorrhage, fibrous cap rupture, plaque ulceration, larger plaque volume or increased plaque thickness, and plaque echolucency. ⁸ Other features that have been found to be associated with increased stroke risk include carotid dissection, the presence of intraluminal thrombi and carotid webs. ¹⁹ , ²⁰ , ²¹ Similarly, in our study, we found that the presence of intraluminal thrombi, dissection, and webs were associated with an increased odds of stroke. Intraluminal thrombi had a reported prevalence of 1.5% in a systematic review, and, although rare, this pathology is believed carry a high risk of ischemic events, with a 30‐day risk of transient ischemic attack, stroke, or death of 17%. ²² The management of intraluminal thrombi remains challenging since the optimal medical treatment and safety of surgical treatment are not yet established. ²⁰ The CTA rim sign has been identified as an imaging marker of intraplaque hemorrhage, which indicates plaque instability. ¹⁷ Its presence may therefore be associated with an increased risk of stroke, as was found in our analysis.

The present study adds to the limited but growing body of literature examining nonstenotic carotid disease in various patient registries. Analysis from the STRATIS (Systematic Evaluation of Patients Treated with Neurothrombectomy Devices for Acute Ischemic Stroke) registry compared nonstenotic carotid plaque characteristics between 226 patients with cardioembolic strokes and 141 patients with cryptogenic strokes. ¹³ The study found that plaque irregularity, plaque hypodensity, and increased plaque thickness were significantly associated with stroke among patients with embolic stroke of unknown source. ¹³ Similarly, analysis of the INTERRSeCT (Identifying New Approaches to Optimize Thrombus Characteristics for Predicting Early Recanalization and Reperfusion With IV Alteplase and Other Treatments Using Serial CT Angiography) cohort showed that in patients without large‐artery atherosclerosis, nonstenotic carotid disease was significantly associated with ipsilateral strokes (aOR, 1.83 [95% CI, 1.05–3.18]) in 446 patients. ²³ The PARISK (Plaque at Risk) study was a prospective multicenter cohort study of 244 patients with symptomatic carotid stenosis that found that intraplaque hemorrhage and total plaque volume were independent risk factors for recurrent ischemic stroke or transient ischemic attack. ²⁴

Contrary to the above studies, in the present study comprising 1407 patients, no significant association was observed between concordant stroke and plaque ulceration, plaque thickness, plaque irregularity, or plaque hypodensity. While we performed a sensitivity analysis excluding patients with atrial fibrillation and lacunar infarctions, it is possible that other confounding factors and comorbidities were not accounted for. In addition, we only examined the association of the carotid disease features with the acute stroke itself; while this had the strength of examining a population that had experienced a definite clinical event, the absence of longer‐term follow‐up meant we could not prospectively relate these disease features to recurrent events. The strengths of the analysis was the inclusion of a large sample size from a pragmatic trial where patients across 22 stroke centers in Canada were included. ¹⁰

Although prior literature suggests an association between nonstenotic carotid disease and ischemic stroke among patients with high‐risk neuroimaging features, establishing causality remains challenging. Plaque irregularity, hypodensity, thickness, and ulceration are generally considered high‐risk imaging features, but the negative findings in this analysis highlight the complex relationship between plaque features and stroke risk. ²⁵ Several factors may have contributed to this negative finding. First, the study used CTA as the primary imaging modality, which, while widely available and effective for detecting luminal stenosis and some plaque characteristics, may have limitations in fully characterizing high‐risk plaque features compared with more advanced imaging techniques like MR vessel wall imaging. ⁷ Second, the definition of a “vulnerable plaque” in intracranial and extracranial circulation remains an evolving concept; we may not yet fully understand which plaque features truly confer an increased risk of recurrent stroke, especially in the absence of significant stenosis. Additionally, prior studies showing associations between plaque vulnerability and stroke often rely on longitudinal follow‐up, whereas our cross‐sectional design does not allow for assessment of future stroke risk based on plaque characteristics. Finally, given the sample size and event rate, the study may have been underpowered to detect smaller effect sizes.

The present study specifically examined high‐risk neuroimaging features in patients with nonstenotic carotid disease, addressing a gap in literature as few studies have focused on this subset of patients. ⁸ , ¹³ Nonstenotic plaques are often overlooked as a potential pathogenesis of stroke, as current management strategies primarily focus on degree of carotid stenosis. ⁸ , ²⁶ Prospective studies with larger sample sizes and longitudinal follow‐up are needed to establish causation between nonstenotic carotid disease and stroke.

Conclusions

Among patients with nonstenotic carotid disease, we found the presence of intraluminal thrombi, carotid webs, carotid dissections, and carotid rim sign to be significantly associated with an increased odds of concordant stroke. These findings underscore the importance of evaluating plaque characteristics beyond the degree of stenosis when assessing stroke risk in patients with carotid disease.

Sources of Funding

The authors thank the Canadian Stroke Consortium for their support of the trial recruitment efforts and the Canadian Institutes of Health Research (grant numbers 419 722 and 450 890), the Alberta Strategy for Patient Oriented Research Support Unit, Alberta Innovates, the Heart and Stroke Foundation, and the University of Calgary for funding support.

Disclosures

R.H.S. has stock options in FollowMD and receives salary support for research from the Heart and Stroke Foundation of Canada, Sandra Black Centre for Brain Resilience and Recovery, and Ontario Brain Institute. L.C. has received payments by Servier and consulting fees from Ischaemavie RAPID, Circle NV, and Canadian Medical Protective Association. J.S. has a grant from Medtronic to the University of Manitoba. B.K.M. has stock options in Circle NVI and has consulted for Biogen and Boehringer Ingelheim. A.G. reports consulting fees and honoraria from Alexion, Biogen Eisai, and Servier Canada. T.T.S. has received consulting fees from Circle NVI research support from Alberta Innovates, the Alzheimer Society of Canada, the Alzheimer Society of Alberta and Northwest Territories, Brain Canada, the Canadian Institutes of Health Research, Campus Alberta Neuroscience, the Government of Canada INOVAIT and New Frontiers in Research programs, the France–Canada Research Fund, the Heart and Stroke Foundation of Canada, Microvention, MSI Foundation, and Panmure House; and stock/stock options from SnapDx Inc and Collavidence Inc (Let’s Get Proof), all outside the scope of the published work. All other authors declare no competing interests.

Supporting information

Table S1