Article Text

Download PDFPDF

Original research
External carotid artery embolus with internal carotid artery occlusion with acute ischemic stroke: predictor of cardioembolic etiology
  1. Thomas Tomsick1,
  2. Senthur Thangasamy2,
  3. Thomas Stamatis3,
  4. Pooja Khatri4,
  5. Joseph Broderick4
  1. 1Department of Radiology, University of Cincinnati, Cincinnati, OH, USA
  2. 2Department of Radiology, University of Massachusetts, Worchester, MA, USA
  3. 3Department of Radiology, Ohio State University, Columbus, Ohio, USA
  4. 4Department of Neurology, University of Cincinnati Medical Center, Cincinnati, OH, USA
  1. Correspondence to T Tomsick, University of Cincinnati, 234 Goodman St, Cincinnati OH 45267, USA; tomsicta{at}


Arteriograms performed in the Interventional Management of Stroke studies were analyzed for external carotid artery embolus. Two cases were identified and diagnosed as thromboembolic due to a cardiac origin. This is an uncommon but useful finding on angiography which is helpful for further management. It is hypothesized that finding embolus within the external carotid artery on angiography in stroke patients with internal carotid artery occlusion allows confident ascription to a proximal, usually cardiac, source.

View Full Text

Statistics from


Acute thromboembolism is the primary cause of ischemic stroke accompanied by moderate or large neurologic deficits. The range of the arterial occlusive lesions in such patients has been outlined.1 Internal carotid artery (ICA) occlusions are identified in 25–30% of patients.2 The occlusion may be primary at the ICA origin with atherosclerotic thrombosis or occlusion, more distal in the cervical ICA with spontaneous or traumatic dissection or at any portion of the ICA with major thromboembolism. Distinguishing the etiology of such ICA occlusion is helpful in planning medical management in the subacute setting but is mandatory in decision making in the acute setting of ultra early stroke management by endovascular interventional techniques.

We have sporadically recognized external carotid artery (ECA) emboli in the setting of ICA occlusion during acute stroke management. Past reviews and reports of acute intracranial occlusion have overlooked this diagnostic finding. To more systematically study our observation, angiograms from the Interventional Management of Stroke (IMS) trials between 2001 and 2007 were also prospectively evaluated for such emboli, and herein we report the findings of their review and analysis. We hypothesize that identification of ECA emboli immediately allows recognition of a central thromboembolic stroke etiology mechanism, and facilitates conduct of subsequent diagnostic and therapeutic maneuvers. Identification of an imaging finding indicative of major central thromboembolism allows the interventional approach to be initiated not only without concern for encountering ICA atherosclerotic occlusion or dissection as the etiology but also with confidence that simple microcatheter techniques for thrombolysis or thrombus removal might be applied, and that preparation need not be initially made for balloon angioplasty and/or stent placement.

Methods and results

In the IMS studies, candidates aged >18 years but ≤81 years with a National Institutes of Health Stroke Scale (NIHSS) score of ≥10 undergo cerebral angiography as soon as possible after onset of stroke, typically within 4 h, after intravenous administration of reduced dose intravenous recombinant tissue plasminogen activator. In IMS I and II trials, cerebral angiography in 153 subjects demonstrated 21 proximal ICA stenoses >70%, 14 proximal or cervical ICA occlusions (eight atherosclerotic and six dissections), two distal extradural ICA occlusions and 21 distal ICA or carotid terminus occlusions. Arteriograms were retrospectively reviewed for ECA emboli for IMS I but prospectively analyzed for IMS II by the core laboratory neuroradiologist (TT). In two cases of intracranial ICA occlusion, no flow in the cervical ICA was accompanied by ECA emboli on baseline diagnostic arteriography. Both instances were subsequently attributed to major embolus of cardiac origin, one related to atrial fibrillation present at baseline, the other to mural thrombosis post myocardial infarction. No ECA embolus was identified in the other 14 cervical ICA occlusions, 21 ICA stenoses or in other subjects with intracranial carotid terminus or M1 occlusion.

Case reports

Case No 1

A 74-year-old known hypertensive patient with a history of atrial fibrillation presented with manifestations of left posterior cerebral arterial territory ischemia. Her baseline NIHSS score was 20. On angiography, she demonstrated evidence of occlusion of the right ICA with associated embolus within the ECA (figure 1). Her final diagnosis was cardioembolic infarct due to atrial fibrillation. Although she had partially recovered from the initial episode, she died on day 64 secondary to new left middle cerebral artery infarct.

Figure 1

Left: Common carotid artery injection demonstrates emboli in the internal maxillary artery and superficial temporal artery occlusion. Layered contrast in the cervical internal carotid artery (ICA) is depicted. The finding would be compatible with ICA dissection. Right: ICA injection (anteroposterior view) demonstrates tapered, layered, slow, antegrade flow in the cervical ICA terminating at an embolus in the petrous segment.

Case No 2

A 75-year-old patient with a past medical history of hypertension, diabetes and past history of four angioplasties for coronary artery disease presented with acute right hemiparesis with aphasia, with a baseline NIHSS score of 23. On angiography, there was complete occlusion of the distal left ICA with embolus within the lumen of the left ECA. Her discharge diagnosis was infarct due to emboli origin from the left ventricle following myocardial infarct.


To the best of our knowledge, this is the first prospective study series review focusing on angiographically documented ECA embolus with ICA occlusion in acute ischemic stroke treatment studies. Previous acute stroke treatment studies have suggested an incidence of 15–20% for proximal ICA occlusion or severe stenosis, and 10% distal ICA or carotid terminus occlusion.2 The variability of appearances of ICA occlusion on acute angiographic evaluation caused by atherothrombotic occlusion, dissection and thromboembolism validate this analysis and report.

The etiology of an ICA occlusion may be suspected clinically. Early clinical diagnosis of probable or possible cardiac embolism depends on the history or identification of a medium or high risk source.3 The causes of cardioembolic stroke are: atrial fibrillation (50%), ventricular thrombus (20%), structural heart defects or tumor (15%) and valvular heart disease (15%).4 A large embolus from the ascending aorta is less likely. Both of our presented cases were in high risk groups (atrial fibrillation and post-myocardial infarction mural thrombus) where the angiographer might anticipate an embolic source. Where patients may be in atrial fibrillation intermittently, or where no recent myocardial infarction has been diagnosed, the early clinical diagnosis at baseline may not yet be clear.

With acutely symptomatic ICA disease, common carotid artery (CCA) injection may show marked irregularity with high grade stenosis of the ICA origin, with distal occlusion or high grade stenosis, with an elongate whisp or channel of contrast identified within the markedly reduced caliber of distal artery, also known as the ‘string sign’ or ‘slim sign’. This severe tight narrowing at the ICA origin with complete or near complete occlusion may begin immediately at the ICA origin, or originate in the distal bulb, with a tapered or triangular stump. The low blood flow distal to this pseudo-occlusion and thin caliber of lumen limit demonstration of contrast in the distal vessel. Delayed images may reveal minimal opacification of nearly collapsed post-stenotic ICA.5 Flow may be best demonstrated in a hanging head position.6 Flow may slowly advance cranially, terminating in the last patent branch, such as the ophthalmic or posterior communicating arteries below a tandem intracranial embolus. The string sign may be indicative of atherosclerotic pseudo-occlusion, dissection or radiation vasculopathy.

With ICA dissection, the ICA bulb and proximal ICA appear normal, then flow slowly with a narrow, straight, smooth or irregular course through the distal cervical segment, frequently to the skull base, occasionally with kinking. It may taper to complete occlusion. The narrowing may continue intracranially with or without branch opacification, depending again on presence or absence of a tandem intracranial embolus. A long irregular ‘string sign’ flow pattern is more typical of carotid dissection. Specific findings of intimal flap and double lumen are seen in less than 10% of cases. At times, it may be difficult to differentiate from atherosclerotic occlusion/stenosis, distinguished most readily by its distal cervical location.7–9 The ‘string sign’ with normal ICA origin may be seen with distal intradural occlusion, or occasionally congenital hypoplasia.

Alternatively, the ICA bulb and proximal ICA may appear normal, with an abrupt termination or cut-off, with meniscus sign, or distinct filling defect in the distal cervical ICA, or beyond (figure 1C). This is typical of major ICA embolism, and identification of the presenting face of the embolus is virtually diagnostic. A long, incompletely occluding or recanalizing embolus may also rarely mimic the appearance of the string sign of atherosclerosis or dissection.

However, in less common instances, an ICA embolus may be totally occlusive and associated with essentially no flow through the post-bulbar segment, or with only layered contrast opacification and little or no forward flow, simulating a narrow cervical segment, with none of the other specific characteristics of embolism (figure 2). A kinked or looped cervical ICA, where contrast must flow against gravity around bends or curves, may contribute to failure of antegrade flow. The etiology in this latter appearance is less easily ascribed, and may be seen with atherosclerotic occlusion, dissection or embolism. It is in this setting that identification of an accompanying ECA embolus allows immediate, confident arteriographic diagnosis of stroke etiology.

Figure 2

(A) Left Common Carotid (LCC) arteriogram demonstrates immediate post-bulbar occlusion of the internal carotid artery (ICA), with embolus in the external carotid artery (ECA). (B) Delayed image demonstrates tapered, layered flow with no advance to the distal cervical ICA. (C) Right Common Carotid (RCC) arteriogram demonstrates a kinked segment of cervical right ICA, with arteriopathic dilatation. Kinked segments may be frequently bilateral and may predispose to dissection. (D) In this patient, pre-procedure CT angiogram demonstrated ICA occlusion as well as the ECA embolus.

When embolism of the formed thrombus enters the CCA, three immediate possibilities exist. ECA only embolism is not expected and would rarely create CNS symptoms. Most emboli will then bypass the high resistance ECA and enter the ICA, reaching the brain to cause an ischemic event. Finally, a CCA embolus may enter not only the ICA but also the ECA, as in our cases. A very large embolus may impact as a saddle embolus and obstruct flow in the distal CCA.10 A large embolus may fragment at the bifurcation, with portions entering each branch, leading to emboli that will typically be larger in the ICA than the ECA, occlusive at any level of either vessel, depending on clot volume and artery diameter and course. Very proximal embolic ICA occlusion may lead to the meniscus sign with outline of thrombus, as described above. Distal cervical, petrous or cavernous embolic occlusion may lead to layered, dependent flow that does not pass intracranially, and may simulate a string sign on initial common carotid injection (figure 1A). Embolic occlusion beyond the ophthalmic artery usually allows slow, layered contrast flow to the ophthalmic artery level but kinking or tortuosity of the ICA may not allow its demonstration. Only guide catheter advance with repeat injection or microcatheter contrast injection through the non-opacifying portion of the ICA may define the true level and type of occlusion. Identification of concomitant ECA emboli confirms the etiology of occlusion, if not the level.

Retrograde extension from the proximal ICA thrombus or embolus must be acknowledged as a possible occurrence but we have not identified it in the IMS studies or in our personal experience. Retrograde embolus into the ECA from the stump of chronic proximal ICA occlusion has been reported.11 12 However, these cases were found to be longstanding ICA occlusion, with a recent episode of minor deficit of transient monocular blindness. Stump emboli have been suggested as a potential etiology of ophthalmic artery or intracranial artery emboli but usually of small size and consequence.

Non-invasive imaging, such as duplex ultrasound, CT angiography and MR angiography, frequently performed prior to an intervention, still may not document the exact details of the etiology of an occlusion and ECA emboli, depending on the attention of the observer, size of the embolus and level of ECA occlusion (figure 2D). Conventional cerebral angiography may be the only way likely to confidently detect the thrombus within the ECA.

In the setting of ICA occlusion with no flow or layered, slow antegrade flow, identification of the ECA emboli allows immediate early ascription of the stroke mechanism. A therapeutic intervention can then begin not only with diminished concern for the nature of the obstruction to be encountered or for the difficulty in traversing it, but also without preparation for angioplasty or stent placement for atherosclerotic occlusion or dissection. Arterial sheaths and guide catheters may then be chosen that are more appropriate for simple thrombolytic or clot removal procedures, according to the preferences or biases of the interventionist.


Identification of thrombus within the ECA allows a definitive diagnosis of thromboembolism in the setting of ICA occlusion and appearance non-diagnostic or atypical for ICA dissection or atherosclerotic occlusion.


View Abstract


  • Competing interests None.

  • Patient Consent Obtained.

  • Provenance and peer review Not commissioned; externally peer reviewed.

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.