Journal of Medical Ultrasound (2015) 23, 98e103

Available online at www.sciencedirect.com

ScienceDirect

journal homepage: www.jmu-onl ine.com

CASE REPORT

Discrepancy Between Duplex Sonographyand Digital Subtraction Angiography WhenInvestigating Extra- and IntracranialUlcerated Plaque

Fu-Yi Yang 1, Yu-Chin Su 1, Shinn-Kuang Lin 1,2*

1 Stroke Center and Department of Neurology, Taipei Tzu Chi Hospital, Buddhist Tzu ChiMedical Foundation, New Taipei City, and 2 School of Medicine, Tzu Chi University, Hualien, Taiwan

Received 21 November 2014; accepted 5 January 2015Available online 21 March 2015

KEY WORDScolor-coded carotidduplex sonography,

digital subtractionangiography,

transcranial color-coded sonography,

ulcerated plaque

Conflicts of interest: The authors* Correspondence to: Dr Shinn-Kuang

289, Jian Guo Road, 231, Xindian DistE-mail addresses: jy0428@totalbb.

http://dx.doi.org/10.1016/j.jmu.20150929-6441/ª 2015, Elsevier Taiwan LL

Noninvasive color-coded duplex sonography has become a good, convenient, and reproduciblescreening tool for the general population when studying cerebral hemodynamics and athero-sclerotic disease. Digital subtraction angiography (DSA) is still the gold standard for the diag-nosis of carotid stenosis, although other noninvasive imaging tools are also available. Atpresent, ultrasound scanning, followed by confirmatory DSA, is a cost-effective way to surveypatients suspected of suffering from cerebral arterial stenosis. We report two patients who hadcerebral ischemic symptoms due to high-grade stenosis of either the cervical internal carotidartery (ICA) or the middle cerebral artery (MCA), combined with an ulcerated plaque. Ultraso-nographic Doppler analysis identified high-grade stenotic lesions as marked elevations in theturbulent flow of the cervical ICA in one patient and of the middle cerebral artery in the otherpatient. Subsequently, huge plaque ulceration was found by color B-mode scanning of the pa-tient with cervical ICA stenosis. However, DSA was able to demonstrate only a mildemoderatedegree of stenosis associated with the lesions. High-grade stenotic lesions of the ICA and themiddle cerebral artery were reconfirmed by computed tomography angiography and magneticresonance angiography. An atheromatous plaque with ulceration is believed to be the cause ofthis discrepancy between ultrasonography and DSA.ª 2015, Elsevier Taiwan LLC and the Chinese Taipei Society of Ultrasound in Medicine.Open access under CC BY-NC-ND license.

have no conflicts of interest to declare.Lin, Department of Neurology, Taipei Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Numberrict, New Taipei City, Taiwan.net.tw, sk2022@tzuchi.com.tw (S.-K. Lin).

.01.002C and the Chinese Taipei Society of Ultrasound in Medicine. Open access under CC BY-NC-ND license.

Discrepancy Between Ultrasound and Angiography 99

Introduction

Carotid endarterectomy (CEA) is clinically beneficial whentreating symptomatic patients with severe carotid steno-sis, according to the North American Symptomatic CarotidEndarterectomy Trial and the European Carotid SurgeryTrial, both of which included angiographic studies [1,2].Carotid angioplasty with stenting provides an equivalentbenefit to CEA when protecting against ipsilateral stroke inpatients with symptomatic and asymptomatic carotidstenosis [3]. Conventional angiography or digital subtrac-tion angiography (DSA) has been suggested to be usefulwhen diagnosing and treating carotid stenosis [4]. Someauthors have proposed that carotid ultrasonography alone[5,6] or carotid ultrasonography with other noninvasiveimage studies [7,8], such as magnetic resonance angiog-raphy (MRA) or computed tomography angiography angi-ography (CTA), may be sufficient prior to CEA.Nevertheless, while endovascular stenting has become themain treatment of stenotic carotid disease in Taiwan, DSAhas remained the gold standard method for the diagnosisand stenting treatment of carotid stenosis. A lesioninvolving the presence of an ulcerated atheroscleroticplaque is believed to be a high risk factor in relation tocerebral ischemic events [9]. Plaque ulceration or intra-plaque hemorrhage is a common finding in patients withischemic symptoms. It is known that DSA has limitationswhen detecting plaque ulceration [10,11]. Some authorshave proposed that, because DSA presents a limitednumber of views, this might be one of the major problemsassociated with DSA [12]. B-mode ultrasonography, CTA,and enhanced MRA have all recently been reported todemonstrate ulceration more accurately than DSA [11]. Bycontrast, some studies have disagreed with the usefulnessof CTA as a detection method for plaque ulceration. Thelimitations of CTA are probably related to the slice thick-ness of the CT images [13].

Even though it is the gold standard, DSA might under-estimate the degree of stenosis, particularly when there isa coexisting large ulcerated plaque. Here, we report twopatients who had cerebral ischemic symptoms due to ste-notic arteries, with the extracranial internal carotid artery(ICA) being affected in one individual and the intracranialmiddle cerebral artery (MCA) being affected in the other.Discrepancies between DSA, color-coded carotid (CCD) so-nography, and transcranial duplex sonography (TCCS) weredetected when evaluating the degree of arterial stenosis.The stenotic lesions were finally reconfirmed with the helpof noninvasive MRA and CTA.

Case reports

Patient 1

A 70-year-old man was admitted to the ward owing to asudden onset of left limb weakness. On examination, hewas alert with left hemiparesis. A bruit was auscultated inthe area of the right upper neck. A brain CT showed arecent infarction affecting the right frontal lobe corre-sponding to the territory of the right MCA. A CCD studyfound a segmental heteroechogenic atheromatous plaque

with surface ulceration that had resulted in a 70% stenosisof the vessel diameter at the proximal portion of the rightICA. Doppler detection of the stenotic lesion revealedmarkedly elevated turbulent flow, with a peak systolicvelocity (PSV) of >300 cm/s and an end diastolic velocity(EDV) of > 120 cm/s (Fig. 1A). The PSV ratio of ICA to CCAon the right side was estimated to be 6 (> 300 cm/s:54 cm/s) and was suggestive of a � 70% stenotic lesion affectingthe ICA. There was also a slightly increased resistanceindex affecting the right common carotid artery (0.84) ascompared to the left common carotid artery (0.68). Theflow direction of the right ophthalmic artery was ante-grade. We performed DSA in order to further identify thepatient’s symptomatic tight stenosis of the ICA prior totreating the patient. However, DSA simply demonstratedthe presence of a narrowed lumen in the right proximal ICAwith diameter stenosis of < 50%, as well as a slit-like fillingdefect of contrast medium across the lumen (Fig. 1B). Afurther CTA with thin cut images and focal reconstructionwas able to document the presence of a tight stenoticatheromatous lesion in the right proximal ICA with hugeplaque ulceration (Fig. 1C). His hemiparesis improvedgradually after medical treatment, and he underwentsuccessful carotid stenting for the stenotic ICA 2 monthslater.

Patient 2

A 32-year-old man who had a history of hypertension wasadmitted to the ward due to frequent transient ischemicattack, which presented as episodic right upper limbweakness. There had been four attacks within 2 days, withthe duration of each attack being about 10 minutes. Theinitial brain CT did not show any abnormal lesions. A CCDstudy only showed the presence of a mild atheromatouslesion in the area of the left CCA bifurcation. TCCS found aturbulent flow with a markedly elevated flow velocity (PSV/EDV Z 373/208 cm/s) at a depth of 57 mm through lefttemporal window, which corresponds to the proximalportion of the left MCA and suggests a severe focal stenosisof the left MCA (Fig. 2A). MRA also demonstrated severestenosis at the proximal part of the left MCA (Fig. 2B). MRDiffusion-weighted imaging showed the presence of multi-ple tiny scattered acute embolic infarcts in the area of theleft MCA. An electroencephalogram revealed mild to mod-erate intermittent slow waves affecting the left hemi-sphere during a hyperventilation test, which is indicative ofrelative hypoperfusion of the left brain. To evaluate thepossibility of treatment using intracranial stenting, DSA wasperformed, but it showed only mild to moderate stenosis ofthe left proximal MCA (Fig. 2C). No more transient ischemicattacks occurred after medical treatment, and he wasdischarged with antiplatelet therapy. A follow-up TCCS 6months later showed a further increase in the flow veloc-ities of the left MCA (PSV/EDV Z 418/239 cm/s), whichsuggests focal tight stenosis. Further MRA also demon-strated severe stenosis in the left proximal MCA. Owing tothe young age of the patient and the progression of thestenosis, notwithstanding prompt medical treatment, hewas referred to a medical center and underwent successfulintracranial angioplasty with stenting.

Fig. 1 Patient 1. (A) Color-coded carotid duplex sonography of the right internal carotid artery shows a tight stenotic athero-matous lesion with an ulcerated plaque (arrow) and a markedly elevated flow velocity. (B) Digital subtraction angiography onlydemonstrates a moderate degree of atheromatous lesion (arrow) in the right internal carotid artery with a slit-like filling defect onthe opposite wall (small arrow). (C) Computed tomography angiography discloses a large amount of ulcerated plaques (small ar-rows) that is causing a tight stenosis (large arrow) of the right internal carotid artery (left: transverse views; right: longitudinalview).

100 F.-Y. Yang et al.

Discussion

Detection of tight stenotic carotid lesions for possible CEAor stenting is an important issue when attempting to pre-vent further stroke. For patients with ischemic stroke, CCDand TCCS are usually the primary studies carried out byphysicians when screening for carotid disease and evalu-ating cerebral hemodynamics. Once a stenotic

atherosclerotic lesion is detected by CCD or TCCS, variousmethods are available that allow further evaluation of theextra- and intracranial cerebral arteries, including nonin-vasive MRA and CTA, as well as the relatively invasive DSAapproach.

Ultrasonographic diagnosis of carotid stenosis is basedon both B-mode imaging and Doppler flow velocities. B-mode imaging with or without color Doppler is able to

Fig. 2 Patient 2. (A) Transcranial color-coded sonography shows a markedly elevated Doppler flow velocity in the left proximalmiddle cerebral artery, which is suggestive of a high-grade stenosis. (B) Magnetic resonance angiography discloses a high-gradestenosis (arrow) of the left proximal middle cerebral artery. (C) Digital subtraction angiography demonstrates only a mild tomoderate degree of stenosis (arrow) in the left proximal middle cerebral artery.

Discrepancy Between Ultrasound and Angiography 101

detect the stenosis directly using either a transverse or alongitudinal view. The degree of diameter stenosis, incomparison with the local vascular diameter, approachesthe measurements obtained by the European Carotid Sur-gery Trial. If the denominator is changed to the diameter ofdistal ICA, the results approach the measurements obtainedby the North American Symptomatic Carotid Endarterec-tomy Trial. As technology has advanced, the potential ofmodern ultrasonic instruments for detailed diagnosis ofcarotid stenosis and plaque morphology, including plaqueulceration, intraplaque hemorrhage, and dynamic imagingof a floating plaque or a moving thrombus, has improvedgreatly [14]. Doppler velocity provides another method bywhich carotid stenosis can be predicted [15]. Althoughsome studies reported CEA according to ultrasonic studyalone did not compromise operative outcome [6], otherstudies recommended at least two noninvasive studies(such as CCD with MRA) prior to CEA [4,7,16]. An ultrasonicstudy is considered to be an initial screening approachrather than a final diagnosis tool. However, limitations ofCCD exist and are mainly operator related. Furthermore,

identification difficulties that are caused by acousticshadowing still remain hard to overcome. MRA or CTA hasbecome a second choice for noninvasive study. At present,CEA has almost been replaced by carotid stenting whentreating patients with tight stenotic carotid lesions inTaiwan. Although noninvasive MRA/CTA can be used todetect stenotic lesions, all patients will still undergo a DSAstudy prior to or during stenting treatment.

An accurate analysis of extra- and intracranial ulceratedplaques should include the level of stenosis, degree of ste-nosis, and change in blood flow. DSA allows only a limitednumber of views, and this may lead to an underestimation ofthe degree of stenosis by as much as 40% when comparedwith histological analysis [11]. Although Comerota et al [17]have reported that diagnostic sensitivity when identifyingcarotid ulceration is not significantly different between B-mode carotid imaging and arteriography, AbuRahma et al[18] found carotid duplex ultrasound to be superior to ca-rotid arteriographywhen detecting an irregular or ulcerativeheterogeneous plaque associated with a mild degree ofstenosis. Ko et al [14] have reported that multiple embolic

102 F.-Y. Yang et al.

infarcts due to unstable floating or moving carotid plaquescan be detected in patients only by CCD.

In this report, the large-scale ulceration involving a ca-rotid plaque found in Patient 1 was detected by CCD as anintraplaque color flash on the near wall of the ICA. Thedegree of ICA stenosis was estimated by both B-modecalculation of diameter reduction (70% stenosis by the Eu-ropean Carotid Surgery Trial method) and the Dopplercriteria of an increased PSV ratio of ICA to CCA of up to 6.CTA also found that there was stenotic lesion with large-scale ulceration on both the axial and reformatted longi-tudinal views. Since the filling with contrast medium isbetter when DSA is used, blood flow inside the large ul-cerated space resulted in more coverage of the vascularlumen by the contrast. As a result, the ulcerated plaquewas indicated only by the presence of a slit-like fillingdefect when DSA was used, and as a result, the degree ofstenosis was underestimated.

Although there is a lack of definite criteria whenassessing the degree of intracranial stenosis by ultrasound,the presence of a markedly elevated flow velocity withinthe left MCA of Patient 2 could be detected by TCCS, andthis was indicative of a significant focal stenosis. Chimowitzet al [19] adopted criteria whereby a mean velocity of �100 cm/s indicated a � 50% stenosis reduction of the MCAdiameter. Felberg et al [20] reported that � 70% MCA ste-nosis is able to produce a stenotic/prestenotic mean ve-locity ratio of � 3/1. According to Lindegaard et al’s [21]report, an MCA mean velocity of � 200 cm/s is able topredict a residual lumen of � 1 mm. In Patient 2, the meanvelocity of the left proximal MCA was 268 cm/s, which issuggestive of a focal severe to tight stenosis. Further MRAscreening also demonstrated the presence of a severe totight stenotic lesion. Thus, the DSA findings, which showedonly a mild to moderate degree of stenosis, are not reallycompatible with the TCCS and MRA findings. One possiblecause of this mismatch between DSA and TCCS/MRA findingsmay be related to the presence of an ulcerated plaque inthe MCA, which is similar to the mechanism found to beassociated with the presence of an ulcerated plaque in theICA of Patient 1. Unlike CCD, direct visualization of a ste-notic atheromatous plaque with large-scale ulceration inthe MCA is not possible by TCCS. Thus, a markedlyincreased Doppler flow velocity still plays an important rolein the diagnosis of intracranial stenosis.

Impaired cerebral perfusion from tight stenosis andartery-to-artery emboli due to the presence of an ulceratedplaque, in the ICA or MCA, were the two major causes ofischemic stroke in these two patients. Embolic stroke mighthave contributed more to the events in these two patientsowing to the large scale of the ICA ulcerated plaque presentin Patient 1 and to the presence of multiple scattered tinyinfarcts within the MCA territory of Patient 2. The patho-genesis of artery-to-artery emboli is based more on plaquemorphology than on the severity of stenosis present in theaffected artery. Intraplaque hemorrhage and an ulceratedplaque with dislodged debris are believed to be responsiblefor the production of emboli. Plaque ulceration or intra-plaque hemorrhage is a common finding when patients aresuffering from ischemic symptoms [14]. In Patient 1, furtherangioplasty with stenting or CEA was the treatment ofchoice when treating the tight stenotic ICA with huge

plaque ulceration, rather than medical therapy. For Patient2, who had symptomatic stenotic MCA with ongoing pro-gression of the stenosis even though there had beenaggressive medical treatment, further intracranial angio-plasty with stenting was later performed successfully.Although it still remains controversial whether stenting oraggressive medical therapy is more effective in treatingintracranial arterial stenosis [22e24], Patient 2 chosefurther intracranial stenting and this had a good result.

In conclusion, an underestimate of the degree of arterialstenosis by DSA may occur in cases where there is an ul-cerated plaque. When there is a discrepancy between ul-trasonography findings and DSA findings, noninvasive MRA orCTA will help distinguish the presence of such lesionsfurther. Obviously, the identification of a tight stenoticlesion will influence patient care significantly.

Acknowledgments

The study was supported by a grant from Taipei Tzu ChiHospital, Buddhist Tzu Chi Medical Foundation (TCRD-TPE-103-RT-16).

References

[1] Barnett HJ, Taylor DW, Eliasziw M, et al. Benefit of carotidendarterectomy in patients with symptomatic moderate or se-vere stenosis. North American Symptomatic Carotid Endarter-ectomy Trial Collaborators. N Engl J Med 1998;339:1415e25.

[2] European Carotid Surgery Trialists’ Collaborative Group.Randomised trial of endarterectomy for recently symptomaticcarotid stenosis: final results of the MRC European CarotidSurgery Trial (ECST). Lancet 1998;351:1379e87.

[3] Brooks WH, Jones MR, Gisler P, et al. Carotid angioplasty withstenting versus endarterectomy: 10-year randomized trial in acommunity hospital. JACC Cardiovasc Interv 2014;7:163e8.

[4] Qureshi AI, Suri MF, Ali Z, et al. Role of conventional angiog-raphy in evaluation of patients with carotid artery stenosisdemonstrated by Doppler ultrasound in general practice.Stroke 2001;32:2287e91.

[5] Dinkel HP, Moll R, Debus S. Colour flow Doppler ultrasound ofthe carotid bifurcation: can it replace routine angiographybefore carotid endarterectomy? Br J Radiol 2001;74:590e4.

[6] Loftus IM, McCarthy MJ, Pau H, et al. Carotid endarterectomywithout angiography does not compromise operativeoutcome. Eur J Vasc Endovasc Surg 1998;16:489e93.

[7] FrieseS Krapf H, Fetter M, et al. Ultrasonography and contrast-enhanced MRA in ICA-stenosis: is conventional angiographyobsolete? J Neurol 2001;248:506e13.

[8] Saouaf R, Grassi CJ, Hartnell GG, et al. Complete MR angi-ography and Doppler ultrasound as the sole imaging modalitiesprior to carotid endarterectomy. Clin Radiol 1998;53:579e86.

[9] Eliasziw M, Streifler JY, Fox AJ, et al. Significance of plaqueulceration in symptomatic patients with high-grade carotidstenosis. North American Symptomatic Carotid Endarterec-tomy Trial. Stroke 1994;25:304e8.

[10] Senkowsky J, Bell 3rd WH, Kerstein MD. Normal angiogramsand carotid pathology. Am Surg 1990;56:726e9.

[11] Randoux B, Marro B, Koskas F, et al. Carotid artery stenosis:prospective comparison of CT, three-dimensional gadolinium-enhanced MR, and conventional angiography. Radiology 2001;220:179e85.

[12] Streifler JY, Eliasziw M, Fox AJ, et al. Angiographic detection ofcarotid plaque ulceration. Comparison with surgical

Discrepancy Between Ultrasound and Angiography 103

observations in a multicenter study. North American Symptom-atic Carotid Endarterectomy Trial. Stroke 1994;25:1130e2.

[13] Walker LJ, Ismail A, McMeekin W, et al. Computed tomographyangiography for the evaluation of carotid atheroscleroticplaque, correlation with histopathology of endarterectomyspecimens. Stroke 2002;33:977e81.

[14] Ko PT, Lin SK, Chang YJ, et al. Carotid floating plaques asso-ciated with multiple cerebral embolic strokes. Case reports.Angiology 1997;48:255e61.

[15] Wong JK, Gibson RN, Mitchell PJ. Comparison of two Dopplerultrasound criteria for grading cervical internal carotid arterystenosis. Australas Radiol 1999;43:153e5.

[16] Johnston DC, Goldstein LB. Clinical carotid endarterectomydecision making: noninvasive vascular imaging versus angiog-raphy. Neurology 2001;56:1009e15.

[17] Comerota AJ, Katz ML, White JV, et al. The preoperativediagnosis of the ulcerated carotid atheroma. J Vasc Surg 1990;11:505e10.

[18] AbuRahma AF, White 3rd JF, Boland JP. Carotid endarterec-tomy for symptomatic carotid artery disease demonstrated byduplex ultrasound with minimal arteriographic findings. AnnVasc Surg 1996;10:385e9.

[19] Chimowitz MI, Kokkinos J, Strong J, et al. The Warfar-ineAspirin Symptomatic Intracranial Disease Study. Neurology1995;45:1488e93.

[20] Felberg RA, Christou I, Demchuk AM, et al. Screening forintracranial stenosis with transcranial Doppler: the accuracyof mean flow velocity thresholds. J Neuroimaging 2002;12:9e14.

[21] Lindegaard KF, Nornes H, Bakke SJ, et al. Cerebral vasospasmafter subarachnoid haemorrhage investigated by means oftranscranial Doppler ultrasound. Acta Neurochir Suppl (Wien)1988;42:81e4.

[22] Higashida RT, Meyers PM, Connors JJ, et al. Intracranial an-gioplasty & stenting for cerebral atherosclerosis: a positionstatement of the American Society of Interventional andTherapeutic Neuroradiology, Society of Interventional Radi-ology, and the American Society of Neuroradiology. J VascInterv Radiol 2005;16:1281e5.

[23] Chimowitz MI, Lynn MJ, Derdeyn CP, et al. Stenting versusaggressive medical therapy for intracranial arterial stenosis. NEngl J Med 2011;365:993e1003.

[24] Derdeyn CP, Fiorella D, Lynn MJ, et al. Intracranial stenting:SAMMPRIS. Stroke 2013;44(6 Suppl. 1):S41e4.