Answers

Short answers

1 His ABCD2 score is 3; the cut-off score predicting high risk of very early stroke after "brain attack" is 4.¹

2 The upper part of fig 2 shows a longitudinal greyscale image of the internal carotid artery with superimposed colour flow (colour map); the lower part shows the recordings of flow velocity (spectral Doppler) over six cardiac cycles. In the upper image, calcified plaque causes posterior acoustic shadowing, which obscures the lumen at the site of the stenosis (white arrow) and mimics an occlusion (yellow arrow). There is turbulent flow with aliasing on both the colour map and on the spectral Doppler. In the lower image, the peak systolic velocity at the proximal internal carotid artery is 429 cm/s, and the end diastolic velocity is 123 cm/s. Although velocity critera are specific to the laboratory, many operators use a peak systolic velocity of >230 cm/s to indicate a 70% stenosis (by NASCET criteria).^{2 3} An end diastolic velocity of >100 cm/s also suggests a high grade stenosis.

View larger version (390K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 2 Carotid duplex images of the origin of the left internal carotid artery

3 The flinging movements of the right arm and leg indicate hemiballismus, which usually indicates a lesion of the subthalamic nucleus. In most cases the likely vascular territory is the posterior circulation, but in a few cases it could be attributed to the middle cerebral artery. Best medical therapy (antiplatelets, a statin, and an angiotensin converting enzyme (ACE) inhibitor if appropriate) is the mainstay of secondary prevention against further cardiovascular or cerebrovascular events. If the symptoms are thought to be attributable to the presumptive "index lesion" in the left internal carotid artery, the degree of stenosis is confirmed to be greater than 70% (using the NACSET method), and alternative sources of embolus are excluded, then carotid intervention in an experienced unit (endarterectomy or stenting) is indicated in a neurologically stable patient and for maximum benefit should be performed within two weeks if the ABCD2 score is 4.⁴

4 As duplex ultrasound is operator dependent, it is important to obtain accurate timely imaging to confirm the degree of carotid stenosis, assess the vertebral arteries, and exclude alternative lesions by non-invasive imaging such as contrast enhanced magnetic resonance angiography (CEMRA) or computed tomographic angiography (CTA). Imaging of the brain by magnetic resonance diffusion weighted imaging (DWI) may help confirm the vascular territory if there is any diagnostic uncertainty.

Figure 3 shows a lesion at the arch origin of the left common carotid artery (top) and a further, "tandem," lesion at the intrapetrous segment of the left internal carotid artery (bottom). It is important to evaluate the source images (not included here) to exclude artefact. The brachiocephalic artery origin was normal on the source images. Figure 4 was taken five days after onset of symptoms. It shows a new white lesion in the left cerebellar hemisphere, with corresponding signal "drop-out" on the apparent diffusion coefficient (ADC) image.

View larger version (129K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 3 Top: tandem lesions at the left carotid bifurcation (yellow arrow) and at the intrapetrous portion of the left internal carotid artery (white arrow). Bottom: stenosis at the origin of the left common carotid artery (white arrow)

View larger version (67K): [in this window] [in a new window] [PowerPoint Slide for Teaching]   Fig 4 The white arrow indicates a high signal (white) lesion on the diffusion weighted image in the left cerebellar hemisphere (white arrow), with corresponding signal "fall out" on the apparent diffusion coefficient image (yellow arrow)

• Age 60 years (1 point)
  
• Blood pressure: systolic >140 mm Hg or diastolic 90 mm Hg (1 point)
  
• Clinical features: unilateral weakness (2 points), speech impairment without weakness (1 point)
  
• Duration 60 min (2 points) or 10-59 min (1 point)
  
• Diabetes (1 point).
  

The National Institute for Health and Clinical Excellence guidelines on the diagnosis and initial management of acute stroke and transient ischaemic attack recommend urgent specialist assessment and investigation (within 24 hours) and carotid intervention (within two weeks) in patients with an ABCD2 score of 4, and assessment, investigation, and intervention within two weeks in all other neurologically stable patients.⁴

2 Carotid ultrasound images
Doppler ultrasound measures the direction and speed of blood as it moves through vessels. The movement of blood causes a change in pitch of the reflected sound waves (Doppler shift principle), and this is processed by computer to generate images and calculate velocity. Colour flow imaging provides a pictorial representation of velocity and direction of flow. Two dimensional grey scale and spectral Doppler information was termed "duplex" and two dimensional, colour flow, and spectral Doppler "triplex" scanning.

No agreement has been reached on assessment of degree of stenosis, but the peak systolic velocity in the internal carotid artery has historically been the primary diagnostic ultrasound criterion for carotid stenosis and is recommended as one of the important measurements by the recent Society for Vascular Technology/Vascular Society of Great Britain and Ireland joint recommendations document.² The table shows the increase in peak systolic velocity in the internal carotid artery with increasing percentage stenosis.

Discussions regarding carotid intervention should take place within a multidisciplinary team. Carotid intervention (endarterectomy or stenting) is appropriate within six months of presenting with symptoms if these are attributable to a carotid stenosis of >70% (by NACSET criteria).^{3 4 5} In patients with a >50% stenosis (NACSET), intervention may be considered within two weeks of presentation with symptoms attributable to the carotid lesion; for men with hemispheric symptoms, intervention may be considered up to two months after onset of symptoms.

Considering the ambiguity of the neurological symptoms and signs in this case, localisation of the lesion should be guided by clinico-radiological correlation. The presenting symptoms suggested transient hemiballismus, but when the patient was examined he had no involuntary movements and dysarthria and mild weakness in his arm. Hemiballismus is a rare variety of transient ischaemic attack mimicking one of the varieties of lacunar syndrome. In most cases this is due to a lesion in the subthalamic nucleus supplied by perforating branches of the posterior cerebral artery. However, hemiballismus can also be caused by lesions in caudate, striate, and other basal ganglia sites. These may be supplied by middle cerebral artery perforating branches. The territory of this patient’s symptoms is therefore uncertain, and DWI (diffusion weighted imaging) is helpful.

Had DWI shown a lesion in the territory of the left middle cerebral artery, the symptoms would likely be attributable to the high grade stenosis of the left internal carotid artery, and intervention would have been offered within two weeks. However, it showed an acute left posterior fossa lesion. The cerebellar infarct does not explain the patient’s symptoms and is likely to be asymptomatic. It is necessary to assess the posterior circulation and circle of Willis before deciding whether to intervene in this case, and all the necessary information was available on the CEMRA image. The vertebral arteries were normal. The symptoms may have been attributable to cardiac embolism, but the patient’s echocardiogram showed normal left ventricular function and no evidence of valvular heart disease. Under these circumstances, the relevance of the lesion of the left internal carotid artery is less clear.

Carotid intervention may still be justified in this relatively young man, and the decision to proceed might be influenced by the state of the remaining cerebral circulation, although there is no consensus on timing of intervention for asymptomatic stenoses.^{6 7} The management of tandem lesions (arch origin and carotid bifurcation) is not evidence based, but at our centre patients with symptoms attributable to such disease are offered combined carotid endarterectomy with retrograde stenting of the arch origin lesion via the surgical incision, on the basis that surgical treatment of the arch origin lesion would necessitate a sternal split. With respect to intracranial lesions, once other embolic sources have been treated or excluded, stenting could be considered for >70% stenoses in patients whose symptoms recur despite best medical therapy. NICE’s guidance recommends that randomised trials should compare interventional procedures against best medical therapy for intracranial lesions.⁸

4 Further imaging
Carotid imaging has two main objectives: accurate assessment of severity of carotid stenosis, plus exclusion of an additional or alternative embolic source at the arch origin or beyond the carotid bifurcation ("tandem" lesion). Accurate, timely carotid imaging avoids potentially hazardous and expensive carotid interventions in patients who do not require them and directs the provision of intervention to all those who may benefit.

Accurate assessment of degree of stenosis—Carotid duplex imaging is operator dependent, with the potential for "unacceptably wide interobserver variation" even within accredited vascular laboratories.⁹ Between-scanner and within-scanner differences could affect categorisation of carotid stenosis.¹⁰ The landmark surgical trials were based on selective catheter angiography,^{3 5} and subsequent non-invasive innovations are essentially surrogates of catheter angiography. A recent systematic review comparing duplex ultrasonography, time of flight magnetic resonance angiography (TOFMRA), contrast enhanced MRA (CEMRA), computed tomographic angiography (CTA), and catheter angiography concluded that CEMRA had the highest sensitivity (0.94) and specificity (0.93). Duplex had a sensitivity of 0.89 and specificity of 0.84.¹¹

A cost effectiveness analysis indicated that duplex as first or repeat test was more cost effective than intra-arterial angiography,¹¹ but the cost effectiveness model used is sensitive to accuracy of the test. Given that CEMRA has the highest diagnostic accuracy, a strategy of offering intervention on the basis of a single ultrasound scan alone will lead to unnecessary intervention in up to 16% of patients and deny intervention in up to 11% of those who might benefit. Such unnecessary interventions or "missed opportunities" may not be included in cost effectiveness modelling but are important considerations.

Exclusion of an alternative carotid embolic source—Duplex can directly insonate only the cervical portion of the extracranial carotid circulation and cannot evaluate the aortic arch or exclude distal lesions. In people with diabetes (a growing population), the incidence of tandem carotid lesions is 14.0-21.3% and of intracranial lesions is 17.3-24.0%.¹² Furthermore, those with carotid stenosis (>70%) are predisposed to aortic plaques >4 mm, which may predispose to stroke.¹³ Under these circumstances, the clinician cannot be sure of the "index" lesion. Although the immediate outcome of intervention for carotid bifurcation may not be influenced, the patient may still be exposed to risk of stroke from lesions remote from the bifurcation. The vertebral arteries should also be scrutinised, particularly in cases with evidence of a posterior fossa lesion. "Overview" imaging of the carotid and vertebral arteries (from the arch origins to the circle of Willis) can be provided by CEMRA or CTA.

Although fig 2 indicates a left carotid bifurcation lesion of the order of 50% on CEMRA, we selected a single orientation to optimally show the lesion at the carotid siphon; other views showed a bifurcation stenosis of the order of 70%. When duplex ultrasonography and CEMRA are discordant, catheter angiography should be used as an "adjudicating" imaging modality.

Emboli may have sources other than the carotid bifurcation, and an acute ischaemic focus in the left cerebellar hemisphere prompts a careful search for alternative sources of emboli. Diffusion weighted imaging (DWI) is extremely sensitive for detecting acute cerebral ischaemia from within minutes to around two weeks of the onset of symptoms, and shows a bright (white) lesion.¹⁴ Apparent diffusion coefficient (ADC) images (which are calculated from the diffusion data set) should be examined alongside the DWI images and note made of whether there is corresponding "signal fall out" (a dark area on the ADC image) indicating an area of restricted diffusion. These two findings in conjunction suggest an acute ischaemic lesion of up to 7-10 days old (at which point the ADC image tends to normalise). A white lesion on both the DWI and ADC images indicates a phenomenon known as T2 shine-through and is caused by a lesion with high T2 value that is not associated with restricted diffusion (some tumours).

DWI is advocated for excluding stroke mimics and when there is diagnostic uncertainty with respect to vascular territory.⁴

Cite this as: BMJ 2009:338:a3109