Multimodality imaging of carotid web: A case report and literature review

Case report

A 38 year-old Chinese man, who was a smoker, was admitted to emergency department with acute left hemispheric syndrome. One month ago, he suffered an episode of transient aphasia. He was under no medications and did not have any history of atrial fibrillation. Multimodal computed tomography imaging showed total occlusion of distal M1 segment of middle cerebral artery (MCA) and cerebral tissue hypoperfusion in left frontotemporal parietal lobe, radial crown and basal ganglion (Figure 1(a)). Additionally, carotid artery dissection was suspected in the initial CTA assessment. A subsequent DSA confirmed the finding of MCA occlusion and discovered a thin intraluminal filling defect along the posterior wall of the ipsilateral carotid bulb, compatible with CW (Figure 1(b)). It showed 3.1 mm long and approximately 10% stenosis according to the North American Symptomatic Carotid Endarterectomy Trial criteria. When retrospectively reviewing the CTA imaging, we found a shelf-like septum emanating from the posterior wall of the internal carotid bulb on a sagittal section and a corresponding axial section (Figure 1(c)). The patient accepted mechanical thrombectomy inside the therapeutic time window, with a good clinical and angiographic result. A red thrombus was retrieved with subsequent histological analysis by hematoxylin and eosin staining showing a mixed composition of fibrin/platelets and red blood cells, consisting of 50% fibrin/platelet ratio, 45% red blood cells, and 5% white blood cells (Figure 1(d)). On the third day, the patient with CW was considered normal by HRMRI in the initial report, but was identified in retrospect, which revealed protrusion with enhancement along the posterior wall of carotid bulb (Figure 1(e)). Because the complete reperfusion of left MCA was obtained, the patient was discharged with dual antiplatelet therapy and was recommended for endovascular or surgical treatment for CW. On the 36th day, CTA demonstrated an unchanged degree of stenosis of the left internal carotid bulb. The morphology of the intraluminal lesion was unchanged from 36 days prior. On the 56th day, he was transferred to neurosurgery department for surgery for CW. No recurrent stroke occurred during this period. Preoperative Doppler ultrasonography (DUS) revealed a hyperechogenic septum along the left carotid bulb associated with thrombus formation (Figure 1(f)). There was no new ischemic lesion on magnetic resonance imaging (MRI). Transthoracic echocardiography and transcranial Doppler with bubble test did not reveal any potential cardioembolic source of embolism. Transcranial Doppler study was negative for microembolic signals. He underwent carotid endarterectomy (CEA) to eliminate the web, with successful results and no complications. A thrombus was identified in the web pocket. Histology demonstrated smooth muscle cell–rich intima, increased elastic tissue, and myxoid degeneration in the resected shelf-like septum. The superimposed thrombus had a fewer fibrin/platelet ratio (20%), higher red blood cells (77%), and 3% white blood cells (Figure 1(g)). Postoperative CTA and DUS showed normal lumen without any flow or structural abnormalities (Figures 1(h) and (i)). After 11 months follow-up, the patient experienced no recurrent stroke/transient ischemic attack, and no restenosis was detected by DUS.OPEN IN VIEWER

Literature review

Data extraction

First, we manually extracted patient-level data for all reported CW patients, including demographics, conventional stroke risk factors, symptomatic or asymptomatic, recurrent stroke, multimodality imaging, and treatment strategy. Conventional stroke risk factors included hypertension, dyslipidemia, atrial fibrillation, diabetes mellitus, coronary artery disease, smoking, patent foramen ovale, and oral contraceptive use. Symptomatic CW was defined as stroke or transient ischemic attack ipsilateral to the CW, and other possible causes of stroke or transient ischemic attack were excluded. Then, the CW-level data for all reported CWs was extracted, including location, length, and degree of stenosis. The reported degree of carotid stenosis in the original article was used, and it was classified as normal, 50–69%, 70–99%, and occlusion. Data abstraction was completed independently by two authors (J.Z., Y.Y.) and any discrepancies were resolved through review and consensus.

Literature review

Flowchart of literature selection is reported in Figure 2. A total of 719 articles were identified. After reviewing the titles, abstracts, or full texts of each, we excluded articles in languages other than English, articles with unavailable full text, and irrelevant articles. The 80 remaining publications matched the inclusion criteria of our study (Supplemental Table 1). A total of 681 patients with CW were included; 680 from the 80 publications identified in our literature review and 1 case from our previous report. Four hundred and sixty-four (68.1%) patients were symptomatic, 39 (5.7%) were asymptomatic, and 178 (26.1%) could not be classified. Of these 80 publications, 64 (80.0%) were published after 2017, including 619 (90.9%) patients with CW and 415 (89.4%) patients with symptomatic CW.OPEN IN VIEWER

Demographics of patients with carotid web and morphological characteristics of Carotid web

The median age of CW patient population was 50.9 years (range 20–86 years), with a high prevalence of women (60.9%). The majority of patients (77.8%) were of African race. Forty-one percent of the patients had conventional stroke risk factors. The most commonly reported stroke risk factors were hypertension (33.1%), oral contraceptive use (31.2%), smoking (18.7%), dyslipidemia (15.1%), patent foramen ovale (13.3%), and diabetes mellitus (10.8%). Coronary artery disease and atrial fibrillation were uncommon. The laterality of the CW lesion was reported in 431 patients, of which 99 (23.0%) patients had bilateral CW (Table 1). In most patients, CW was located on the posterior wall (87.3%) of the carotid artery (98.7%). The mean length was 3.3 mm and mean stenosis rate was 20.9%. The degree of stenosis was reported in 258 patients, of which 244 (94.6%) patients had normal lumens (Table 2).

OPEN IN VIEWER

Table 1.

Baseline characteristics of patients with carotid web and symptomatic patients.

Characteristics	Patients with carotid web (n = 681)		Patients with symptomatic carotid web (n = 464)
	No. of positive patients	No. of patients with data reported	No. of positive patients (%)	No. of patients with data reported
Age, y (median, range)	50.9 (20–86)	647	48.3 (20–82)	410
Women, n (%)	378 (60.9)	621	248 (62.9)	394
Race, n (%)
Asian	23 (7.3)	315	23 (7.8)	294
African	245 (77.8)	315	236 (80.3)	294
White	47 (14.9)	315	35 (11.9)	294
Absence of vascular risk factors, n (%)	157 (59.0)	266	149 (59.6)	250
Stroke risk factors, n (%)
Hypertension	140 (33.1)	423	86 (25.9)	332
Diabetes mellitus	45 (10.8)	415	26 (8.0)	325
Dyslipidemia	59 (15.1)	392	35 (11.2)	313
Coronary artery disease	11 (6.9)	160	0 (0.0)	71
Smoking	70 (18.7)	374	58 (17.8)	326
Atrial fibrillation	8 (5.7)	141	2 (2.0)	101
Patent foramen ovale	6 (13.3)	45	6 (13.3)	45
Oral contraceptive use	5 (31.2)	16	5 (31.2)	16
Laterality, n (%)
Unilateral	332 (77.0)	431	183 (69.6)	263
Bilateral	99 (23.0)	431	80 (30.4)	263

OPEN IN VIEWER

Table 2.

Morphological characteristics of carotid web.

Characteristics	Carotid web (n = 780)		Symptomatic carotid web (n = 452)
	No. of positive patients	No. of patients with data reported	No. of positive patients (%)	No. of patients with data reported
Location, n (%)
Carotid artery	770 (98.7)	780	411 (98.8)	415
Vertebral artery	8 (1.0)	780	4 (1.0)	415
Subclavian artery	2 (0.3)	780	0 (0.0)	415
Wall of lumen, n (%)
Posterior	399 (87.3)	457	248 (90.5)	274
Lateral	10 (2.2)	457	0 (0.0)	274
Anterior	4 (0.9)	457	0 (0.0)	274
Postero-lateral	42 (9.2)	457	24 (8.8)	274
Postero-medial	1 (0.2)	457	1 (0.4)	274
Anterior-lateral	1 (0.2)	457	1 (0.4)	274
Length, mm (mean)	3.3	200	3.5	73
Stenosis rate, % (mean)	20.9	188	20.8	184
Degree of stenosis, n (%)
Normal	244 (94.6)	258	157 (96.3)	163
50–69%	9 (3.5)	258	4 (2.5)	163
70–99%	4 (1.6)	258	1 (0.6)	163
Occlusion	1 (0.4)	258	1 (0.6)	163

Symptomatic carotid web and secondary stroke prevention

The patients with symptomatic CW were relatively young (median age 48.3 years, range 20–82 years). The most commonly reported stroke risk factor was oral contraceptive use (31.2%). The length of CW was relatively long (mean 3.5 mm). Of the 464 patients with symptomatic CWs, 380 patients have available treatment data (Table 3). The stroke recurrence rate was 36.4% in patients without any treatment. Two hundred and one patients were managed medically alone, and the stroke recurrence rate was 27.9%. Of these patients, follow-up data were available in 97. One hundred and sixty-eight (44.2%) patients were treated with carotid revascularization. Of the 49 patients with symptomatic CW extracted from articles published before 2017, 23 (46.9%) were treated with carotid revascularization. Of the 415 patients with CW extracted from articles published after 2017, 145 (34.9%) were treated with carotid revascularization. Of all the patients treated with carotid revascularization, 5 (3.0%) patients with periprocedural complications were reported, including one stenting case with external iliac artery rupture, three stenting cases with hypotension/bradycardia, and one surgery case with mild postoperative tongue deviation. One hundred and twenty-six patients treated with carotid revascularization had available follow-up data. All the patients remained stroke free during this period. The medication after carotid revascularization was reported in 63 patients, of which 4 with mono antiplatelet therapy, 3 with dual antiplatelet therapy, 32 with dual antiplatelet therapy followed by monotherapy, 1 with anticoagulant therapy, and 23 with antiplatelet and anticoagulant therapy.

OPEN IN VIEWER

Table 3.

Secondary stroke prevention and stroke recurrence.

Treatment	Patients with symptomatic carotid web, n (%)	Recurrent stroke, n (%)	Months to recurrent stroke (median, range)	Months stroke free if No stroke recurrence (median, range)
None	11/380 (2.9)	4/11 (36.4)	0.2 (0.2–0.2)	72.0 (12.0–102.0)
Medical management alone	201/380 (52.9)	51/183 (27.9)	8.3 (0.0–97.0)	27.1 (2.0–100.0)
Antiplatelet	142/163 (87.1)	43/139 (30.9)	10.3 (0.3–97.0)	15.5 (2.0–100.0)
Anticoagulation	21/163 (12.9)	5/20 (25.0)	3.73 (0.2–9.0)	20.7 (6.0–48.0)
Carotid revascularization	168/380 (44.2)	0/127 (0.0)	—	24.9 (0.0–144.0)
Stenting	94/168 (56.0)	0/83 (0.0)	—	16.9 (2.4–144.0)
Surgery	74/168 (44.0)	0/44 (0.0)	—	18.7 (0.9–120.0)

Discussion

CW is a proposed stroke mechanism that may underlie cryptogenic stroke, particularly in younger patients without vascular risk factors. The current literature review is the largest and most comprehensive collection of CW patients published to date.

The prevalence of CW in patients who undergo carotid CTA varies from 0.5 to 2.3%.^8–10 The prevalence of CW in patients with ischemic stroke/TIA varies from 0.45 to 9.6%.^11–15 The prevalence of symptomatic CW in patients with cryptogenic stroke varies from 1.6 to 21.2% and 4.5–23.1% in young patients with cryptogenic stroke.^{1, 2, 11–13, 16, 17} The prevalence of symptomatic CW in patients with large vessel occlusion stroke varies from 0.7 to 2.5%.^{3, 4, 16, 18} The prevalence of symptomatic CW in patients who were operated on for a carotid stenosis varies from 3.0 to 15.6%.^{7, 19} After our literature review, the prevalence of symptomatic CW was 1.2% in patients with transient ischemic attack/stroke, 6.4% in patients with cryptogenic ischemic stroke, 1.1% in patients with large vessel occlusion stroke, and 4.4% in patients who operated on for carotid stenosis.

Similar to the systematic review published by Zhang et al. in 2018, ²⁰ our study found that patients with CW were relatively young, with a higher prevalence of women. The majority of these patients had few conventional stroke risk factors. The most commonly reported stroke risk factors were hypertension and oral contraceptive use. We also found that CW was most common among African population and rare in Asian population; however, there are several publications focused on CW in Asian population in recent years.^{8, 12, 14, 15, 21, 22} They were not included in the race data analysis of current literature review, since these studies did not explicitly state the race composition. In addition, we also updated some information about morphological features of CW. Bilateral CW was not uncommon, and 30.4% symptomatic CW had contralateral CW. CWs were located in carotid artery, vertebral artery, or subclavian artery, but it was most commonly located in the posterior wall of the carotid bulb. It may be partly due to the definition of CW. In most publications, CW was defined as a shelf-like septum in the posterior wall of the carotid bulb.^{1, 2, 4, 5, 7, 9-11, 13, 16–19, 23-28} On the other hand, the greater development and thickness of the intima and the greater amount of elastic fibers in carotid vessel are probably linked to the presence of a higher frequency of CW on the carotid artery. ²⁹ The mean length of symptomatic CW was relatively longer than asymptomatic CW. CW could cause moderate stenosis, severe stenosis or even occlusion of vascular lumen, but most CW still did not cause obvious lumen stenosis.

There are several useful tools to identify carotid webs, including DUS, CTA, MR, and DSA. The finding of CW closely depends on the experience of radiologists. At present, there is no consensus on which imaging method is more suitable. DUS is a convenient and non-invasive method, but the diagnosis can be challenging because of thrombus inside CW, which mimics a classic hypoechoic non-calcified plaque. Ben et al. found that the rate of misdiagnosis and missed diagnosis was 30.43% in the first ultrasonography. ²¹ Hassani et al. attempted to evaluate the suspected CWs using IVUS. In two of the three cases, IVUS showed an isoechoic-to-hyperechoic focal eccentric area at the posterior carotid bulb, consistent with CW. But the endoluminal protrusion was inconspicuous on IVUS due to the low resolution of ultrasound not allowing a clear differentiation between fibrosis, thrombosis, and atherosclerosis. ³⁰ CTA is the most widely used imaging tool for evaluation of CW. In the current review, only 0.9% patients with CW confirmed by DSA were misdiagnosed as normal or atherosclerotic vessels by CTA. Some scholars believed that the correlation between DUS and CTA showed a substantial agreement, but some believed not.^{31, 32} CTA and DSA shared very high strength of agreement. ³¹ DSA could show retrograde delayed filling of the carotid bulb web and stasis within the pouch. MR imaging enables an in-depth evaluation of patients with CWs: morphology (by MRA), composition (by multi-contrast FSE), and wall dynamics (by cineFSE). ²⁷ In the current literature review, CW was detected in seven patients by MRA and four CWs confirmed by DSA were not visible on MRA. The missed diagnosis rate of CW by MRA was higher than that by other imaging tools. MRA maybe not a good modality to investigate for CW. However, the limitation of this study is the small sample size. The imaging diagnosis of CW needs more large sample size research. Histopathological analysis is the gold standard for the diagnosis of CW, which revealed that the etiology of CW was a nonatheromatous intimal fibromuscular dysplasia of the extracranial carotid artery. However, few patients had histological analysis of thrombus superimposed in CW and embolic thrombus in LVO, which may be of significance to clarify the pathogenesis of CW. In the present case we reported, we found that the histological analysis of LVO stroke embolized thrombus by CW was similar to that of other stroke etiologies, similar to previous studies.^{18, 33, 34} But the superimposed thrombus had a fewer fibrin/platelet ratio and a higher red blood cells than Gao et al. reported, which was the only previous study on histological diagnosis of thrombus superimposed in CW, to the best of our knowledge. ³⁵

This literature review identified 43 patients with asymptomatic CW who did not take any treatment. No ischemic events were associated with these asymptomatic CW, with a median follow-up time of 44.0 months. Therefore, asymptomatic CW noted “on imaging only” is benign from a clinical perspective not requiring further treatment or surveillance. For patients with bilateral CWs, asymptomatic side also does not need intervention. Optimal management strategies for symptomatic CW include either medical treatment or carotid revascularization. Guglielmi et al. found that 1 of 6 patients with a symptomatic CW had a recurrent stroke within 2 years ⁴ Within the current review, we found that 36.4% untreated symptomatic CW patients had stroke recurrence, and more than a quarter of symptomatic CW patients treated with medical therapy alone had stroke recurrence with a median time to recurrent stroke of only 8.3 months. CW may be particularly amenable to revascularization, since these patients are relatively young and commonly lack of conventional stroke and cardiovascular risk factors. None of the patients treated with carotid revascularization had recurrent stroke, with a median follow-up of 24.9 months. Only 3.0% patients with periprocedural complications were reported, which is recovered in the first follow-up. The stroke recurrence rate in patients with CW who were treated with anticoagulant therapy (25.0%) was relatively lower than those in patients without any treatment (36.4%) or patients with antiplatelet therapy (30.9%). So anticoagulant therapy is recommended for those patients between diagnosis and definitive therapy. However, the stroke recurrence rate in patients with anticoagulant therapy is still higher than that in patients with carotid revascularization. Carotid revascularization is recommended for patients with symptomatic CW as early as possible. Besides, medication after carotid revascularization was reported in 63 patients. Some patients were treated with anticoagulant therapy, while others not. All the patients remained stroke free during follow-up. Therefore, postoperative anticoagulant therapy may not be necessary.

In the current literature review, most articles (80.0%) were published after 2017, which showed that both clinicians and radiologists had paid more and more attention to CW in recent years. After 2017, DSA (18.6%) was less used to detect CW, while non-invasive examination methods, such as CTA (82.2%), were more and more used to detect CW. However, only 34.9% patients with symptomatic CW extracted from articles published after 2017 were treated with carotid revascularization, which was relatively lower than that (46.9%) in symptomatic patients extracted from articles published before 2017. Clinicians should strength their learning and understanding of CW.