US of Neurovascular Occlusive Disease: Interpretive Pearls and Pitfalls 1

Transcription

1 EDUCATION EXHIBIT 1165 US of Neurovascular Occlusive Disease: Interpretive Pearls and Pitfalls 1 CME FEATURE See accompanying test at /education /rg_cme.html LEARNING OBJECTIVES FOR TEST 5 After reading this article and taking the test, the reader will be able to: Recognize abnormal spectral configuration patterns associated with various stenotic lesions or anatomic variations. Discuss the importance of detecting incongruities between PSV findings and B-mode images in cases of critical stenosis. Discuss the value of indirect tests (eg, transcranial and periorbital Doppler US) in evaluation of carotid artery disease. Javier M. Romero, MD Michael H. Lev, MD Suk-Tak Chan, PhD Molly M. Connelly, BA Ryan C. Curiel, SB Anna E. Jackson, AB R. Gilberto Gonzalez, MD, PhD Robert H. Ackerman, MD, MPH Ultrasonography (US) of the head and neck is a convenient but operator-dependent screening tool for detection and diagnosis of neurovascular occlusive disease. In US examination of the extracranial carotid arteries, stenosis is most commonly graded according to the peak systolic Doppler velocity in the region of maximal luminal narrowing rather than according to the percentage of atheromatous plaque occupying the lumen. However, the peak systolic velocity is not always reliable in estimation of the degree of stenosis. General diagnostic pitfalls include technical difficulties with scanning, failure to review the spectral waveform patterns, the presence of additional stenotic lesions, and anatomic variants. Specific examples of pitfalls include tandem lesions, differentiation of pseudo-occlusion from true total occlusion, pseudonormalization of velocities in cases of very severe stenosis, lesions of the carotid artery origin or aortic valve, progression of subclavian steal, underestimation of severe stenosis due to heavily calcified plaque, a persistent trigeminal artery, and contralateral carotid artery stenosis. Although conventional angiography remains the standard of reference for assessment of carotid artery disease, recognition of these common sources of error in US can improve the accuracy of this noninvasive test in diagnosis of carotid artery occlusion. RSNA, 2002 Abbreviations: CCA common carotid artery, ICA internal carotid artery, PSV peak systolic velocity Index terms: Carotid arteries, stenosis or obstruction, , Carotid arteries, US, , Subclavian steal syndrome, Ultrasound (US), Doppler studies, , Vertebral arteries, US, ; 22: From the Departments of Radiology (J.M.R., M.H.L., M.M.C., R.C.C., A.E.J., R.G.G., R.H.A.) and Neurology (R.H.A.), Massachusetts General Hospital and Harvard Medical School, 55 Fruit St, Boston, MA 02114; and the Department of Optometry and Radiography, Hong Kong Polytechnic University, Hunghom, Kowloon, Hong Kong (S.T.C.). Presented as an education exhibit at the 2001 RSNA scientific assembly. Received January 21, 2002; revision requested March 6 and received April 1; accepted April 1. Address correspondence to M.H.L. ( mlev@partners.org). RSNA, 2002

2 1166 September-October 2002 RG f Volume 22 Number 5 Introduction Although conventional angiography is the standard of reference for assessing carotid artery disease (1,2), neurovascular ultrasonography (US) is an important and widely accepted technique for screening patients with extracranial arterial stenoses due to atheromatous disease (3). Nevertheless, neurovascular US is subject to a number of interpretive pitfalls that may not be well appreciated by inexperienced operators. The degree of carotid and vertebral artery stenosis is estimated on the basis of direct Doppler measurements of flow velocities in the vascular lumen, although the precise peak systolic velocity (PSV) cutoff values and ratios used to define a hemodynamically significant lesion are controversial and can vary between centers (4,5). Reliance on direct visualization of vessel anatomy with B-mode US to grade stenosis is perilous and has even greater variability (3). For these reasons, transcranial or periorbital Doppler examinations (referred to in this article as indirect tests) are useful ancillary screening examinations for accurate diagnosis of significant extracranial carotid artery stenosis. These tests examine the distal flow effects caused by stenotic lesions and therefore are often useful for accurate determination of residual lumen diameter and percent stenosis (6,7). In our experience, careful review of all US data, including not only PSVs and end diastolic velocities but also spectral configuration, B-mode images, and indirect measurements, is typically required to avoid misinterpretation. The purpose of this pictorial essay is to review a spectrum of interpretive pitfalls in US evaluation of the carotid and vertebral arteries, which may result in incorrect assessment of the degree or location of arterial stenoses. Recognition of these common sources of error can improve the accuracy of US in diagnosis of neurovascular occlusive disease. Specific topics discussed are tandem lesions, pseudo-occlusion versus true total occlusion, critical internal carotid artery (ICA) stenosis with velocities falling off, lesions of the common carotid artery (CCA) origin or aortic valve, progression of subclavian steal, length of calcified atheromatous plaque, abnormally low resistance waveforms and persistent trigeminal artery, and contralateral carotid artery stenosis. Tandem Lesions Infrequently, despite the finding of a severe arterial stenosis at B-mode imaging, the velocities and spectral configuration of the poststenotic segment are normal or only slightly abnormal (Figs 1, 2). This can occur in cases where an additional severe stenosis, downstream from the imaged lesion, causes dampening of the PSV in the vascular segment between the two lesions. When such tandem lesions are suspected on the basis of discordance between B-mode findings and Doppler Figure 1. Tandem lesions in a 75-year-old woman who experienced a presyncopal episode. (a) Longitudinal B- mode color flow US image of the proximal right ICA shows heterogeneously calcified plaque (arrows) with apparent moderate to severe luminal narrowing. (b) Transverse color flow US image shows heavily calcified plaque with shadowing of the ICA lumen. ECA external carotid artery. (c) Duplex US image shows a normal pulsed wave Doppler PSV of less than 100 cm/sec at the right ICA plaque. This velocity is inconsistent with the severity of the stenosis suggested by B-mode color flow imaging, although the spectral broadening raises the question of a proximal lesion. Transcranial Doppler US demonstrated reversal of blood flow in the ipsilateral A1 segment of the anterior cerebral artery and the ophthalmic artery; this result was also inconsistent with the PSV. These findings prompted performance of computed tomographic (CT) angiography to confirm the suspicion of a more severe stenosis or a tandem lesion. (d) Axial CT angiographic source image shows a severe stenosis of the proximal right ICA (arrow), which is consistent with the B-mode color flow imaging findings but not with the Doppler velocity measurements. (e) Curved reformatted CT angiogram of the right ICA shows the severe proximal stenosis (arrowhead). The absence of an elevated PSV at Doppler US is explained by the presence of a severe downstream tandem stenosis of the cavernous segment of the right ICA (arrow).

3 RG f Volume 22 Number 5 Romero et al 1167

4 1168 September-October 2002 RG f Volume 22 Number 5 Figure 2. Tandem lesions in a 77-year-old man with an acute stroke in the territory of the right middle cerebral artery. (a) Longitudinal color flow US image shows severe atheromatous disease in the proximal right ICA. However, pulsed wave Doppler velocities from the same vascular segment are within the normal range. (b) Curved reformatted CT angiogram of the right ICA shows a severe proximal stenosis (arrow). (c) Curved reformatted CT angiogram shows occlusion of the proximal right middle cerebral artery (arrow). This tandem lesion, which is downstream from the extracranial stenosis, explains the discordance between the B-mode color flow imaging findings and the Doppler velocity findings in the proximal right ICA. velocity measurements, transcranial Doppler US, magnetic resonance (MR) imaging of the circle of Willis, CT angiography, or conventional angiography should be performed. Also, spectral broadening without elevated velocities can be an indication of a proximal stenosis (Fig 1c). Tandem lesions are defined as stenoses that occur at more than one level, at least 3 cm apart, along the course of the carotid artery (8,9). When two stenotic lesions are present, the one with greater narrowing is the critical determinant of hemodynamic compromise. Because such lesions commonly occur intracranially at the carotid siphon or cavernous carotid artery, they may be overlooked if the US evaluation is restricted to the cervical carotid artery, resulting in underestimation of the degree of proximal ICA stenosis.

5 RG f Volume 22 Number 5 Romero et al 1169 Figure 3. Pseudo-occlusion in an 84-year-old man with syncope. (a) Longitudinal US image shows absence of color flow signal and Doppler flow within the right ICA, an appearance suggestive of a complete vascular occlusion. (b) Axial CT angiographic source image of the right ICA, obtained at the point of maximal stenosis, shows a patent hairline residual lumen (arrow). (c) Axial CT angiographic source image of a more distal extracranial segment of the same vessel shows the slim sign (arrow), thus confirming flow through the distal ICA. Pseudo-occlusion versus True Total Occlusion When color flow images fail to demonstrate signal or Doppler shift within the ICA, a complete vascular occlusion may be present (Fig 3) (12). However, because routine US may be insensitive for the detection of very slow flow in the presence of a critical stenosis, the possibility of a hairline residual lumen or pseudo-occlusion must also be considered. Although inappropriately low color gain settings can result in a pseudo-occlusion being incorrectly interpreted as a complete occlusion, the use of maximum gain settings can increase image noise, improving sensitivity but decreasing specificity (8,9). Gain should therefore be set to the maximum possible level that avoids image noise, with pulse repetition and filter settings lowered for the detection of slow flow. Newer techniques, such as power Doppler US which is less velocity dependent than color flow Doppler US may improve visualization of vascular flow in cases of high-grade stenosis (10).

6 1170 September-October 2002 RG f Volume 22 Number 5 Figure 4. Left ICA stenosis in a 64-year-old man with right hemiplegia and aphasia. (a) Longitudinal B-mode color flow US image shows hypoechoic atheromatous plaque in the proximal left ICA (arrow), an appearance suggestive of severe stenosis. (b) Pulsed wave Doppler spectrum of the proximal left ICA shows a delayed, curved systolic upstroke. The PSV of 87 cm/sec, which is within normal limits, is discordant with the findings on the B-mode color flow image (a). Results of subsequent transcranial Doppler examination together with the B-mode US and waveform findings suggested the presence of a very severe left ICA stenosis with velocities pseudonormalizing (falling off). (c) Two-dimensional maximum intensity projection image from a CT angiographic data set shows a very severe stenosis of the proximal left ICA (arrow) with a slim sign in the distal ICA. Despite the use of optimal US technique, however, CT or conventional angiography may be required to definitively distinguish a pseudo-occlusion from a true total occlusion (3). Although CT angiography is highly accurate for this determination (11), conventional arteriography is still considered the standard of reference for distinguishing a pseudo-occlusion from a true total occlusion (1,2). Critical ICA Stenosis with Pseudonormalizing Velocities PSVs in the normal range do not preclude a critical ICA stenosis. As a severe stenosis worsens and approaches a hairline residual lumen, previously elevated intraluminal flow velocities may begin to fall off or pseudonormalize. Normal or decreased peak velocities with a slow systolic upstroke may become apparent at, and distal to, the point of maximal stenosis (Fig 4), violating the principle that velocity is directly proportional to the degree of stenosis. The waveform changes associated with falling off (3) of velocities in the setting of a critical stenosis (typically a 0.7-mm residual lumen diameter) may be subtle or absent, and therefore a high index of suspicion for performing indirect tests in clinically symptomatic patients should be present (4). Typically, blood flow in the ipsilateral ophthalmic artery and across the anterior communicating artery is reversed in such cases. Thus, periorbital and transcranial Doppler US (or phase-contrast MR angiography of the circle of Willis, which also shows flow direction) may be of value in suggesting the presence of a critical proximal lesion. CT or conventional arteriography is required to definitively establish the degree of stenosis. Lesions of the CCA Origin or Aortic Valve Although the great vessel origins are typically too deep to be directly insonated by using standard neck US, origin stenoses can sometimes be inferred from their distal waveform patterns. Specifically, in a CCA with a severe proximal stenosis, PSVs may be dampened along the entire vessel length, with slow upstroke on Doppler waveforms relative to those of corresponding levels from the contralateral normal CCA and ICA (Fig 5a 5d). Origin stenoses are the second most

7 RG f Volume 22 Number 5 Romero et al 1171 Figure 5. (a d) Stenosis of the CCA origin in a 72-year-old man with a history of transient speech disturbance. (a) Pulsed wave Doppler spectrum of the left CCA shows that the velocities are within normal limits; however, there is a slow systolic upstroke with dampened waveforms. The peak velocities at every level were lower than those of the normal right CCA. (b) Pulsed wave Doppler spectrum of the left ICA also shows lower velocities and sluggish waveforms compared with those of the normal right ICA (c). (c) Pulsed wave Doppler spectrum of the right ICA, obtained at the same level as in b, shows normal waveforms and velocities. The relatively diminished velocities and dampened waveforms throughout the course of the left carotid artery suggest the presence of a more proximal severe stenosis. (d) Gadolinium-enhanced cervical MR angiogram shows a severe stenosis at the origin of the left CCA (arrow). (e) Duplex US image of a patient with aortic insufficiency. Pulsed wave Doppler spectrum of the CCA shows a characteristic pronounced diastolic notch (arrow).

8 1172 September-October 2002 RG f Volume 22 Number 5 Figure 6. Subclavian steal in a 75-year-old man who ultimately developed a 28 mm Hg difference in systolic blood pressure between the right and left arms (lower in the right arm than in the left). (a) Pulsed wave Doppler spectrum of the middle segment of the right cervical vertebral artery shows a midsystolic notch (arrowhead), which suggests a mild degree of subclavian steal. (b) Follow-up pulsed wave Doppler spectrum obtained 2 years later shows a more pronounced systolic notch, which indicates progression of stenosis. The result is increased subclavian steal. (c) Follow-up pulsed wave Doppler spectrum obtained 1 year after b shows partial reversal of vertebral artery flow. (d) Pulsed wave Doppler spectrum of the proximal right subclavian artery shows abnormally elevated velocities compared with those of more distal segments. frequent CCA lesions, after bifurcation atheromas, and can be an important source of occult stroke. Owing to dampening of the velocity waveforms downstream from these stenotic sites, review of both the amplitude and shape of the distal spectra can aid in detection of severe proximal stenoses of the CCAs. Note that aortic stenosis can produce a similar dampened waveform pattern, although this pattern would be present bilaterally throughout the neurovascular system. Aortic insufficiency, unlike aortic stenosis, results in a characteristic spectral configuration pattern with a pronounced diastolic notch (Fig 5e). Progression of Subclavian Steal Early subclavian steal may go undetected if the hemodynamic change is subtle and a significant difference in blood pressure between both arms (typically 20 mm Hg) is not present. In the early stages of subclavian steal, a mild decrease in systolic flow velocity and reduced luminal diameter can be observed in the vertebral artery ipsilateral to the stenotic subclavian artery (Fig 6), sometimes resulting in an incorrect diagnosis of hypoplastic or diseased vertebral artery (13). As stenosis progresses, the subclavian steal hemodynamics become increasingly abnormal. Clear biphasic waveforms with a pronounced systolic notch are present in the ipsilateral vertebral artery

9 RG f Volume 22 Number 5 Romero et al 1173 Figure 7. Calcified atheromatous plaque in an asymptomatic, neurologically intact 85-year-old woman who had undergone right carotid endarterectomy. (a) Longitudinal color flow US image shows heavily calcified plaque in the wall of the left ICA. The plaque measures 1.5 cm in the long-axis dimension. (b) Duplex US image shows that neither a pulsed wave Doppler spectrum nor a color flow signal could be obtained in the region of calcified plaque owing to severe shadowing. (c) Duplex US image obtained distal to the calcified region shows normal pulsed wave Doppler measurements for velocity and waveform. Transcranial Doppler examination showed antegrade blood flow in the left anterior cerebral artery and left ophthalmic artery, a finding inconsistent with a more severe proximal stenosis. (d) Two-dimensional time-of-flight cervical MR angiogram shows a moderate stenosis of the proximal left ICA (arrows) with narrowing but no signal dropout. when significant steal develops. In cases of severe subclavian steal, blood flow in the ipsilateral vertebral artery may completely reverse. Because precise velocity thresholds for what is considered significant subclavian artery stenosis are not well established, recognition of the Doppler waveform changes in the vertebral artery associated with subclavian steal is important for detection and diagnosis of this entity. Conventional arteriography may be necessary to determine the degree of stenosis once the presence of steal is established. Length of Calcified Atheromatous Plaque In regions of marked acoustic shadowing by calcified atheromatous plaque, insonation of either Doppler waveforms or B-mode color flow images is often impossible (Fig 7). In such cases, if the calcified plaque covers only a short vascular segment (typically 1 cm in length), the presence of normal velocities and waveforms just proximal

10 1174 September-October 2002 RG f Volume 22 Number 5 Figure 8. Persistent trigeminal artery in a 76-year-old woman with recurrent episodes of rightsided weakness. (a) Pulsed wave Doppler spectrum of the left ICA shows abnormally low resistance waveforms with a resistive index of (Resistive index [PSV end diastolic velocity]/ PSV; normal range, approximately ) CT angiography was performed in an attempt to explain this finding. (b) Two-dimensional maximum intensity projection image shows a persistent trigeminal artery (arrow) arising near the top of the basilar artery with anastomosis to the cavernous segment of the left ICA. (c) Posterior volume-rendered three-dimensional view of the basilar artery shows the persistent trigeminal artery (arrow) coursing to the left and entering the lateral margin of the cavernous sinus. The left ICA also supplies the majority of flow to the left posterior cerebral artery via a large posterior communicating artery. These findings explain the low resistive index of the more proximal extracranial ICA. and distal to the shadowed region is sometimes sufficient to exclude a hemodynamically significant stenosis (3). However, when calcific plaque obscures US assessment of a large vascular segment (Fig 7), other imaging modalities are required to adequately assess the degree of stenosis. Abnormally Low Resistance Waveforms and Persistent Trigeminal Artery Abnormally low flow resistance with increased PSV is sometimes noted in the ICA, although no lesion is identified along the course of the artery on B-mode images (Fig 8). Anatomic variants, arteriovenous malformations, or high-flow tumors

11 RG f Volume 22 Number 5 Romero et al 1175 Figure 9. Left CCA stenosis in a 48-year-old woman with recurrent episodes of left-sided transient monocular blindness 6 months after stent placement in the left CCA and ICA. (a) Pulsed wave Doppler spectrum of the left CCA shows greatly elevated velocities (590 cm/sec), which represent severe stenosis. (b) Pulsed wave Doppler spectrum of the distal right CCA shows elevated velocities (196 cm/sec) in the absence of an evident stenosis on B-mode images. (c) Reformatted CT angiogram of the right carotid artery shows no evidence of stenosis of the distal CCA. PSVs alone can lead to overestimating the degree of stenosis in these cases. Careful evaluation of B-mode and color flow images in cases of severe contralateral stenosis or occlusion can be useful in preventing overestimation of the degree of carotid artery stenosis. may cause this finding. A persistent trigeminal artery is a remnant of the fetal cerebral circulation that bridges the carotid and basilar arterial territories. The artery functions as an additional branch of the intracranial ICA; MR imaging, CT, or conventional angiography can help establish its presence (Fig 8). Contralateral Carotid Artery Stenosis Velocities in the ICA and/or CCA can sometimes be falsely elevated in the absence of significant carotid artery stenosis (Fig 9). This can occur in cases where there is severe stenosis or occlusion of the contralateral ICA and/or CCA, resulting in increased (collateral) blood flow through the patent carotid artery (14). Thus, reliance on Conclusions Awareness of some of the pitfalls associated with neurovascular US examination can help improve the accuracy of this technique in detection and diagnosis of occlusive disease of the carotid arteries. Even in cases where PSVs appear normal, abnormalities in the spectral waveform patterns, B-mode images, color flow images, or indirect test results can suggest the presence of stenotic lesions or anatomic variants that otherwise would have been missed. A thorough examination of all US evidence available thus can prove valuable not only for diagnosis of stenosis at the carotid artery bifurcation, but also for detecting the presence of other stenotic lesions and anatomic variants.

12 1176 September-October 2002 RG f Volume 22 Number 5 References 1. Thiele BL, Young JV, Chikos PM, et al. Correlation of arteriographic findings and symptoms in cerebrovascular disease. Neurology 1980; 30: O Leary DH, Mattle H, Potter JE. Atheromatous pseudo-occlusion of the internal carotid artery. Stroke 1989; 20: Ackerman RH. Neurovascular non-invasive evaluation. In: Ferrucci T, ed. Radiology. Philadelphia, Pa: Lippincott, 1995; Grant EG, Duerinckx AJ, El Saden S, et al. Doppler sonographic parameters for detection of carotid stenosis: is there an optimum method for their selection? AJR Am J Roentgenol 1999; 172: Suwanwela N, Can U, Furie KL, et al. Carotid Doppler ultrasound criteria for internal carotid artery stenosis based on residual lumen diameter calculated from en bloc carotid endarterectomy specimens. Stroke 1996; 27: Krapf H, Widder B. The supratrochlear artery as an indicator of cerebral hemodynamics in carotid occlusion. Ultraschall Med 1998; 19: [German] 7. Can U, Furie KL, Suwanwela N, et al. Transcranial Doppler ultrasound criteria for hemodynamically significant internal carotid artery stenosis based on residual lumen diameter calculated from en bloc endarterectomy specimens. Stroke 1997; 28: Mohr JP, Gautier JC, Pessin MS. Internal carotid artery disease. In: Banks MA, ed. Stroke. New York, NY: Churchill Livingstone, 1986; Rouleau PA, Huston J 3rd, Gilbertson J, et al. Carotid artery tandem lesions: frequency of angiographic detection and consequences for endarterectomy. AJNR Am J Neuroradiol 1999; 20: Steinke W, Ries S, Artemis N, et al. Power Doppler imaging of the carotid artery stenosis: comparison with color Doppler flow imaging and angiography. Stroke 1997; 28: Lev MH, Ackerman RH, Rabinov J, et al. Hairline residual lumen or total occlusion of the internal carotid artery? The clinical utility of spiral CT angiography. In: Proceedings of the 35th annual meeting of the American Society of Neuroradiology. Oak Brook, Ill: American Society of Neuroradiology, Berman SS, Devine JJ, Erdoes LS, et al. Distinguishing carotid artery pseudo-occlusion with color-flow Doppler. Stroke 1995; 26: Kliewer MA, Hertzberg BS, Kim DH, et al. Vertebral artery Doppler waveform changes indicating subclavian steal physiology. AJR Am J Roentgenol 2000; 174: Henderson RD, Steinman DA, Eliasziw M, et al. Effect of contralateral carotid artery stenosis on carotid ultrasound velocity measurements. Stroke 2000; 31: This article meets the criteria for 1.0 credit hour in category 1 of the AMA Physician s Recognition Award. To obtain credit, see accompanying test at