Review

Haemodynamic stroke: clinical features, prognosis, and management

Catharina J M Klijn, L Jaap Kappelle
Lancet Neurol 2010; 9: 100817 Department of Neurology, Rudolf Magnus Institute of Neuroscience, University Medical Center Utrecht, Utrecht, Netherlands (C J M Klijn MD, L J Kappelle MD) Correspondence to: C J M Klijn, Department of Neurology (G03.228), University Medical Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, Netherlands c.j.m.klijn@umcutrecht.nl

Haemodynamic stroke is a type of ischaemic stroke that is caused by hypoperfusion rather than by embolism or local vasculopathy. It can be caused by systemic diseases such as heart failure or hypotension, but also by severe obstruction of the carotid or vertebral arteries. Patients with haemodynamic stroke or transient ischaemic attack might show specic clinical features that distinguish them from patients with embolism or local small-vessel disease. Ancillary investigations of cerebral perfusion can show whether blood ow to the brain is compromised and provide important prognostic information. Management of patients who have hypoperfusion as the major cause of ischaemic stroke or as a contributing factor is hampered by the lack of clinical trials. Treatment aimed at increasing cerebral blood ow might be considered in selected patients on the basis of information from case series. Further research is needed to dene criteria for the diagnosis of haemodynamic stroke and to investigate treatment options in controlled studies.

Introduction
The term haemodynamic stroke has traditionally been used for ischaemic stroke caused by hypoperfusion and not by emboli or by local vasculopathy as in lipohyalinosis. Hypoperfusion can result from a drop in blood pressure either regional, caused by severe obstruction of an artery, or systemic (eg, heart failure or systemic hypotension). Occlusion of arteries by emboli (eg, from the heart or atherosclerotic plaques in the internal carotid artery [ICA]), or secondary to lipohyalinosis as in lacunar stroke, also results in hypoperfusion distal to the blocked vessel and subsequently leads to a brain infarct. The feature that distinguishes haemodynamic transient ischaemic attack (TIA) and stroke from other stroke subtypes is that blood ow towards part of the brain is too low and results in ischaemia in an area of the brain or retina that is at a distance from the blocked vessel. When cerebral blood ow (CBF) is chronically marginal and no longer sucient to meet the oxygen requirement of the brain tissue, the amount of oxygen that is extracted from the blood (oxygen extraction fraction) can increase, a condition known as misery perfusion.1 Up to the mid-1980s, hypoperfusion was thought to be a major cause of ischaemic stroke, but in subsequent decades, thromboembolism became the main culprit in stroke aetiology. In recent years, the debate on hypoperfusion as a cause of stroke has intensied,2 although two issues that are crucial to the settling of this debate remain. First, there is no gold standard for the diagnosis of haemodynamic stroke. Second, in many patients, low perfusion pressure and embolism might interact to cause ischaemic stroke because both the location of the hazardous lodging of emboli in the brainsupplying arteries and their clearance might depend on the ow state of the brain.2 In recent years, important information has become available on clinical symptoms that might point to haemodynamic compromise and that aect prognosis, as well as on diagnostic tools to obtain information on the ow state of the brain. Recognition of hypoperfusion as a causal factor in ischaemic stroke, separately or as a contributing factor in embolic stroke,

could have important consequences for the care and management of patients,3 and should facilitate research directed at improving the ow state of the brain to improve outcome in patients with haemodynamic stroke. Here, we review haemodynamic stroke with respect to pathophysiology, clinical manifestations, diagnostic investigations, prognosis, and management.

Pathophysiology
Several factors might contribute to the haemodynamic cause of stroke, either alone or in combination. Both structural changes in the cerebropetal arteries and systemic abnormalities associated with arterial pressure or with blood volume might play a part.

Carotid artery obstruction

In hospital-based series, about 9% of patients with a TIA or ischaemic stroke have complete occlusion of the ICA.4,5 In 146 patients with symptomatic ICA occlusion, 124 (85%) had only transient symptoms or a partial anterior circulation infarct.4 In the rst days after the ICA occlusion occurs, propagation of the thrombus might be observed.6 Thromboembolism might be the main cause of an actual obstruction of the ICA and associated ischaemic stroke,7 but emboli from the heart or atherosclerotic lesions in proximal arteries can no longer reach the brain via the ICA when this artery is occluded. Nevertheless, some have suggested that emboli might arise from the distal or proximal stump of the occluded ICA even as long as 2 years after the ICA was found to be occluded.8 Plaques in the common or external carotid artery (ECA), or in the contralateral ICA,9,10 might be alternative sources of emboli that travel via collateral pathways to the hemisphere distal to the ICA occlusion. By contrast, others have favoured low perfusion pressure as the most common cause of cerebral ischaemia in patients with ICA occlusion.1113 This haemodynamic hypothesis is supported by clinical observations that symptoms can occur in circumstances in which blood pressure is temporarily low, such as after rising from a sitting or supine posture or after the introduction of
www.thelancet.com/neurology Vol 9 October 2010

1008

Review

antihypertensive drugs.14,15 Evidence is accumulating to suggest that patients with ICA occlusion with a TIA or minor disabling ischaemic stroke associated with compromised CBF have a higher risk of future stroke than if they have a normal haemodynamic brain state.16,17 Hypoperfusion might also contribute to signs or symptoms in patients with severe ICA stenosis. The ecacy of carotid endarterectomy to prevent recurrent stroke in these patients might be largely dependent on removal of the embolic source,18,19 but re-establishing normal ow through the vessel might also contribute. In patients with symptomatic severe stenosis who were randomly assigned to receive medical treatment in the European Carotid Surgery Trial or North American Symptomatic Carotid Endarterectomy Trial and who had a contralateral severe ICA stenosis or occlusion, low blood pressure at baseline was associated with an increased risk of stroke over time.20 A rare cause of haemodynamic stroke associated with progressive stenosis of the intracranial ICA and its proximal branches is moyamoya disease, which is reviewed elsewhere.21,22

patients, hypoperfusion and embolism might interact because poor contraction of the heart might also result in thromboembolism, and at autopsy, cholesterol crystal emboli have been found in the arterial borderzones in patients who died after cardiac surgery.29,30 A recent cohort study of observational data showed that not only mean arterial blood pressure but also visit-to-visit variability in blood pressure and maximum systolic blood pressure are

Panel 1: Factors to be considered when treating patients with haemodynamic transient ischaemic attacks or ischaemic stroke associated with internal carotid artery occlusion When taking the history, specically check for the following: Limb shaking Precipitating circumstances of symptoms: rising from a supine position exercise transition from a cold to a warm environment having just consumed a meal coughing administration of antihypertensive drugs, or intensifying their dosage, or administration of other drugs that might lower blood pressure (eg, sildenal) bleeding anaemia Episodes of loss of consciousness Monocular loss of vision after looking into bright light (retinal claudication) Gradual deterioration of vision of one eye, with or without pain around the eye Signs of cognitive impairment Diagnostic studies Check for orthostatic hypotension On brain imaging, look for borderzone infarcts, particularly the internal borderzone type If occlusion of the ICA is found in the acute phase of stroke, conrm permanent occlusion several days later, because an initially occluded ICA might turn out to be a severe stenosis amenable to endarterectomy several days later Consider contrast angiography to assess collateral blood ow pathways and investigate presence of stenosis in arteries important for collateral blood supply Consider assessment of blood ow to identify patients at relatively high risk of recurrent ischaemic stroke Ophthalmological examination is valuable to check for venous stasis retinopathy and chronic ocular ischaemic syndrome, particularly in patients in whom the external carotid arteryophthalmic artery collateral pathway contributes to the collateral blood supply of the hemisphere ipsilateral to the ICA occlusion Prognostic factors Features associated with a relatively low risk of recurrent ischaemic stroke: Retinal symptoms only, no symptoms of ischaemia of the brain No recurrence of symptoms after documentation of the carotid artery occlusion Features associated with a relatively high risk of recurrent ischaemic stroke: Precipitating circumstances (rising, exercise, etc) of symptoms indicative of a haemodynamic cause of stroke Presence of leptomeningeal collateral blood supply from the posterior cerebral artery territory to the supply territory of the occluded ICA Compromised blood ow in the supply territory of the occluded ICA, most convincingly shown by an increased oxygen extraction fraction measured with PET
ICA=internal carotid artery.

Vertebral artery obstruction

By contrast with haemodynamic stroke related to the anterior circulation, information on haemodynamic stroke from stenosis or occlusion of the vertebral arteries is scarce. In a large registry of 407 patients with posterior circulation stroke, haemodynamic compromise was thought the most likely cause in 132 (32%) patients, but clear criteria for this diagnosis were not given.23 One study showed an impaired response of the basilar artery to a vasodilatory stimulus in patients with vertebrobasilar ischaemia, which was interpreted as exhausted autoregulation.24 Others studied patients with severe bilateral vertebral artery disease and found impaired autoregulation and an increased dependency on systemic blood pressure for the ow through the posterior cerebral artery after tilting.25 Whether a low ow state of the brain supplied by the posterior circulation is a predictor of recurrent stroke in patients with vertebral artery stenosis or occlusion has not yet been studied.

Low blood pressure

Another cause of haemodynamic stroke is a drop in blood pressure during hypovolaemia or during anaesthesia. A haemodynamic TIA or ischaemic stroke might also be induced by changes to upright posture, or by treatment with antihypertensive medication.15 In 98 patients who underwent diusion-weighted MRI after cardiac surgery, 47 (48%) had bilateral ischaemic lesions in the borderzone areas, and this was particularly likely in those who had a decrease in mean arterial pressure of at least 10mmHg during surgery.26 Symptoms of orthostatic hypotension have been shown to be an independent predictor of ischaemic stroke.27 The prevalence of ischaemic stroke is increased in patients with heart failure, particularly in patients with low systolic blood pressure.28 In these
www.thelancet.com/neurology Vol 9 October 2010

1009

Review

Figure 1: Signs of venous stasis retinopathy and chronic ocular ischaemic syndrome (A) Retinal signs of VSR in a 44-year-old man who presented with left-hemispheric stroke and episodes of transient monocular blindness of the left eye and left ICA occlusion. Marked peripheral venous tortuosity (*), microaneurysms (black arrow), and multiple small dot haemorrhages (green arrow) are shown on uorescein angiography of the left eye. Reproduced from Klijn et al,51 with permission from the American Heart Association. (B) Neovascularisation of the optic disc and (C) rubeosis iridis in advanced COIS in a 59-year-old man with COIS of the right eye and bilateral ICA occlusion. VSR=venous stasis retinopathy. COIS=chronic ocular ischaemic syndrome. ICA=internal carotid artery.

important predictors for recurrent vascular events.31 Patients with carotid or vertebral artery occlusion might be particularly vulnerable to a drop in blood pressure.14

Limb shaking
About 10% of patients with symptomatic ICA occlusion complain of spells of involuntary, coarse movements of an arm or leg, which can occur separately or simultaneously. This limb shaking usually lasts for less than 5 min, is often accompanied by paresis of the involved limb, and might be precipitated by rising or exercise.4345 Limb-shaking TIAs can resemble partial epileptic seizures, but electroencephalograms during attacks have not shown epileptiform activity.44,46 In patients with ICA occlusion, limb-shaking TIAs are associated with a poor haemodynamic state of the brain and with a relatively high risk of recurrent ischaemic stroke.45,47

Anaemia
Anaemia can precipitate TIA and ischaemic stroke, especially in patients with severe carotid artery disease.32,33 In addition to the increased thrombogenesis elicited by anaemia, hypoxia caused by reduced haemoglobin concentration is likely to be hazardous to those parts of the brain in which cerebral perfusion is already compromised, such as the borderzone regions.33

Clinical manifestations
In many patients with a TIA or ischaemic stroke, signs or symptoms do not allow distinction between haemodynamic failure and thromboembolism, but in some patients specic symptoms or precipitating factors are present that suggest a haemodynamic origin (panel 1).

Transient loss of consciousness without focal symptoms

In general, loss of consciousness without focal symptoms is not considered a sign of TIA or ischaemic stroke, but exceptionally it can be the only manifestation of compromised CBF in patients with severe stenosis or occlusion of multiple cerebropetal arteries.40,48,49 ICA occlusion should be suspected in patients with cough syncope without lung disease,40 and has also been reported after consuming a meal.49 Case reports have documented cessation of symptoms after operative revascularisation.40,49

Precipitating circumstances
Orthostatic hypotension,14 exercise,34 transition from a cold to a warm environment,35,36 consuming a meal,3739 and coughing36 can lower the blood ow to the brain and precipitate TIA or stroke. In these circumstances, the decrease in CBF is probably caused by diversion of blood to other parts of the body. After coughing, cerebral arterial vasoconstriction induced by hypocapnia contributes in addition to the decreased venous return and a decrease in cardiac output caused by Valsalvas manoeuvre.40 In patients with ICA occlusion, Hankey and Gubbay15 reported that administration of antihypertensive drugs can provoke recurrent TIAs and ischaemic stroke, which is consistent with our own experience (unpublished). We have treated patients with ICA occlusion who had had recurrent TIAs after use of sildenal. This phenomenon has also been described in the absence of stenosis of extracranial or intracranial vessels,41 and is thought to result from the hypotensive eect of sildenal.42
1010

Retinal claudication
Patients with ICA occlusion might complain of transient monocular blindness after exposure to bright light. This retinal claudication is thought to result from increased metabolic demand in the retina that cannot be met because of already marginal perfusion.35,50 Retinal claudication is a rare nding,47 but is highly suggestive of carotid artery occlusive disease.

Chronic ocular ischaemic syndrome

Carotid artery occlusion can cause progressive loss of vision of one eye that is sometimes accompanied by pain around the eye. This chronic ocular ischaemic syndrome
www.thelancet.com/neurology Vol 9 October 2010

Review

(COIS) is the result of long-term hypoperfusion of the retina. As a result, neovascularisation of the retina, the optic disc, or the iris (rubeosis iridis; gure 1) can develop. Cotton wool spots (oedema of the optic disc or the macula) can also be found. These signs and symptoms can develop into a uveitis-like syndrome with neovascular glaucoma and blindness. In particular, rubeosis iridis has been associated with a loss of vision.52 The early, often asymptomatic stage of COIS is called venous stasis retinopathy (VSR) and is characterised by dilatation, irregular calibre, and tortuosity of retinal veins, midperipheral microaneurysms, and small dot and blot intraretinal haemorrhages or nerve-bre-layer splinter haemorrhages (gure 1).53,54 VSR is present in around a third of patients with a TIA or minor disabling ischaemic stroke associated with ICA occlusion; progression to clinically manifest COIS is rare.51

Cognitive impairment
Chronic hypoperfusion of the brain has been recognised as a cause of cognitive impairment since 1951.55 Reversibility of cognitive impairment after extracranialintracranial bypass surgery has been reported in a 55-year-old patient with bilateral ICA and unilateral vertebral artery occlusion, severely reduced CBF, and impaired vascular reactivity.56 In a study of 39 patients with TIA associated with ICA occlusion, we found that about half of the patients had cognitive decits.57 Impairment was global rather than limited to certain cognitive domains, and was relatively mild. These cognitive decits were also found in patients who had presented with eye symptoms only and in patients with a cerebral TIA in whom no structural lesion was found on brain MRI.57

Figure 2: Rosary-like pattern of infarcts Left-hemisphere internal borderzone infarct between the deep and supercial arterial system of the middle cerebral artery in the white matter above the lateral ventricle in a rosary-like pattern on uid-attenuated inversion recovery MRI in a 57-year-old woman with left common and internal carotid artery occlusion, who had frequent (>10) recurrent transient ischaemic attacks of the left hemisphere (dysphasia, right sensory symptoms, and mild weakness of the right arm) and recurrent transient monocular blindness of the left eye.

Diagnostic investigations
Although there is no gold standard for the diagnosis of haemodynamic stroke, diagnostic tests could provide information on whether haemodynamic compromise is likely to play a part and on the risk of recurrent ischaemic stroke. Diagnostic studies to be considered in patients with a haemodynamic TIA or mildly disabling ischaemic stroke associated with ICA occlusion are summarised in panel 1.

hampered by the large variability in the supply territories of the major cerebral arteries.61,62 A recent selective arterial spin-labelling MRI study showed that this variability is largely dependent on anatomical variation in the circle of Willis.63 Of 11 patients with proven haemodynamic failure and recurrent stroke, only three had borderzone infarcts, whereas six had territorial infarcts (the type of infarct was undetermined in one patient, and another had a retinal infarct).64 This suggests that haemodynamic compromise and emboli might act together to cause infarction, even when the infarct appears to be caused by an embolus on CT or MRI.

Pattern of infarcts
Classically, infarcts in the borderzones between the vascular territories of the major cerebral arteries have been associated with a haemodynamic cause. In particular, internal borderzone infarcts (infarcts between the deep and supercial arterial system of the middle cerebral artery [MCA], or between the supply territories of the anterior cerebral artery and MCA, located in the white matter along or above the lateral ventricle in a so-called rosary-like pattern; gure 2) have been attributed to haemodynamic failure58 and severe carotid artery occlusive disease.59,60 In individual patients, diagnosis of a borderzone infarct according to the location of the lesion on brain imaging is
www.thelancet.com/neurology Vol 9 October 2010

Collateral blood ow
In patients with a TIA or minor disabling ischaemic stroke of suspected haemodynamic origin, study of collateral pathways can contribute to treatment decisions (see below). In patients with ICA occlusion, the relative contribution of the contralateral ICA, the ipsilateral ECA, and the vertebral arteries to the collateral blood supply should be assessed (gure 3). If the collateral blood supply is highly dependent on the ipsilateral ECA, particular attention should be paid to the development of VSR and COIS.51 A poor haemodynamic state of the brain has been associated
1011

Review

with the presence of collateral blood supply via the ECA and ophthalmic artery,65,66 or leptomeningeal pial branches from the posterior cerebral artery extending
A B

as far as the vascular territory of the MCA or anterior cerebral artery,67,68 although other studies did not show an association between collateral blood ow pattern and misery perfusion.69,70 In a prospective follow-up study of 117 patients with symptomatic ICA occlusion, the presence of leptomeningeal collateral blood supply from pial branches of the posterior cerebral artery on digital subtraction angiography independently predicted recurrent TIA or stroke (hazard ratio 41, 95% CI 13131).47 Contrast angiography might add valuable information in patients with ICA occlusion whose symptoms persist after documentation of the ICA occlusion. The less invasive procedures of CT angiography and magnetic resonance angiography can be used to determine the contribution of the anterior and posterior communicating arteries, but not the contribution of the ophthalmic artery and leptomeningeal collaterals. Stenosis in the contralateral ICA and ipsilateral ECA can be overestimated by duplex sonography and by magnetic resonance angiography in the presence of ICA occlusion.71 A promising new, non-invasive technique to assess the relative contribution of collateral blood ow pathways is selective arterial spin-labelling MRI, which has been successfully applied to determine the importance of the ophthalmic artery collateral pathway in patients with ICA occlusion.72

Haemodynamic measurements
Evidence of compromised CBF can be obtained with various techniques,73 including transcranial doppler, stable xenon or Xe CT, N-isopropyl-p-[I]iodoamphetamine single-photon emission computed tomography,74 PET,16,17,75 and more recently, CT and MRI perfusion techniques.76,77 Most of these are based on assessment of the autoregulatory response to a vasodilatory stimulus such as breath holding, inhalation of carbogen, or intravenous administration of acetazolamide. Haemodynamic compromise stage 1 is dened as failure of CBF or CBF velocity to increase in response to a stimulus, because vasodilatation has already reached its maximum. If cerebral perfusion decreases further, normal cerebral oxygen metabolism is maintained by increasing the amount of oxygen that is extracted from the blood (haemodynamic compromise stage 2), a phenomenon that can be measured only with PET. Direct comparisons of various techniques have shown that dierent modalities do not always identify the same patients at risk of recurrent stroke, and therefore methods are not interchangeable.75,78 Haemodynamic compromise has been associated with a high risk of recurrent stroke, the evidence being strongest for stage 2.79 The risk of recurrent stroke for patients with symptomatic ICA occlusion and an increased oxygen extraction fraction is about six to seven times higher than for patients in whom the oxygen extraction fraction is within the normal range.16,17
www.thelancet.com/neurology Vol 9 October 2010

Figure 3: Collateral pathways in a patient with haemodynamic transient ischaemic attacks and chronic ocular ischaemic syndrome associated with occlusion of the right internal carotid artery Angiogram of a 73-year-old man who presented with repeated episodes of transient monocular blindness and retinal claudication of his right eye for 1 year. He also had recent symptoms of chronic ocular ischaemic syndrome on the right side, and one episode of left-sided facial weakness with dysarthria consistent with a right-hemispheric transient ischaemic attack. His angiogram showed occlusion of the right internal carotid artery (green arrow), severe stenosis of the right ECA (black arrow, A), and only limited lling of ACA branches on the right (arrows) via the anterior communicating artery without collateral blood supply to the right MCA territory (B). Selective catheterisation of the right central carotid artery showed collateral blood supply via the stenosed ECA and ophthalmic artery pathway to the right MCA (arrow) and to a lesser extent the ACA territory (C, anteriorposterior view; D, lateral view). Selective catheterisation of the left vertebral artery shows extensive collateral blood supply via leptomeningeal pial branches from the posterior cerebral artery towards the vascular territory of the MCA and ACA (arrows) without collateral blood supply via the posterior communicating artery (E, anteriorposterior view; F, lateral view). ECA=external carotid artery. ACA=anterior cerebral artery. MCA=middle cerebral artery.

1012

Review

Ophthalmological examination
Measurement of visual acuity, intraocular pressure, slitlamp examination, and fundoscopy with emphasis on the early retinal signs indicative of VSR should be considered in all patients with symptomatic ICA occlusion, particularly if the ECAophthalmic artery collateral pathway contributes to the collateral blood supply of the hemisphere at risk. Distinction between VSR and diabetic retinopathy might be dicult, but asymmetry of retinal abnormalities and mid-peripheral haemorrhages along and outside rather than within the vascular arcades is a sign of severe carotid artery occlusive disease.53 In patients with COIS, progression to blindness might be prevented by early treatment with panretinal laser therapy, and if necessary with revascularisation procedures.

Panel 2: Treatment considerations in patients with haemodynamic transient ischaemic attacks or mildly disabling ischaemic stroke associated with internal carotid artery occlusion Patients with transient monocular blindness or retinal infarction who did not have signs or symptoms of cerebral ischaemia should probably not undergo invasive revascularisation procedures, but should be treated with antithrombotic medication and control of risk factors Antithrombotic medication and control of risk factors is also the treatment of choice in patients who have had a transient ischaemic attack or mildly disabling ischaemic stroke, but only before the ICA occlusion was documented, and not after Bed rest in the acute stage might be considered in patients with precipitation of symptoms by a drop in blood pressure on rising, but is not of proven benet Tapering of antihypertensive drugs might be considered in patients with low blood pressure, particularly if symptoms are precipitated by a drop in blood pressure, but is not of proven benet Operative revascularisation procedures are not of proven benet for prevention of recurrent ischaemic stroke, but might be considered in individual patients with compelling evidence for a haemodynamic cause of symptoms In patients with chronic ocular ischaemic syndrome, extracranialintracranial bypass surgery or endarterectomy of stenosis of the external carotid artery in the presence of ICA occlusion can be considered to prevent blindness Revascularisation is not of proven benefit for amelioration of cognitive dysfunction in patients with ICA occlusion, and should not be done solely for this reason
ICA=internal carotid artery.

Prognosis
Prognosis in patients with haemodynamic stroke might vary according to clinical characteristics and results of diagnostic tests (panel 1).

Carotid artery occlusion

In patients with a TIA or mildly disabling ischaemic stroke associated with ICA occlusion, the risk of recurrent ischaemic stroke is 56% per year; about two-thirds of strokes occur ipsilateral to the previously symptomatic ICA occlusion.80 If haemodynamic compromise can be shown, this risk is approximately two times higher,47,81 and is probably highest in the rst 12 years after the ischaemic event.17,74 The risk of recurrent ischaemic stroke is lower in patients with only symptoms of the eye and not of the brain.16 In a prospective follow-up study of patients with symptomatic ICA occlusion, we found that only one of the 24 patients who presented with eye symptoms alone had an ischaemic stroke during an average period of 10 years.82 The risk of recurrent infarction is relatively high in patients who have clinical features suggestive of haemodynamic compromise such as limb shaking or symptoms after rising or exercise, and in patients in whom TIAs continue after the occlusion of the ICA was shown.47

quantitative ow measurements with magnetic resonance angiography is predictive of recurrent ischaemic stroke.84

Management
To date, no randomised controlled trials (RCTs) of treatment options for patients with signs or symptoms of presumed haemodynamic origin are available to guide treatment. As with patients with TIA or ischaemic stroke in general, antithrombotic agents should be administered and hyperlipidaemia and diabetes mellitus should be rigorously controlled. Cessation of smoking should be actively pursued. Only the management of blood pressure might be dierent (panel 2).

Vertebral artery occlusive disease

Little is known about the prognosis of patients with cerebral ischaemia of presumed haemodynamic origin in association with vertebral artery stenosis or occlusion. A systematic review showed that the risk of subsequent stroke in patients with a vertebrobasilar TIA or mildly disabling ischaemic stroke was similar to that after carotid artery territory events.83 This comparison did not include information on the presence of stenosis or occlusion in either the carotid or vertebral artery. However, more recently, the results of a study of 50 patients with a vertebrobasilar TIA or mildly disabling ischaemic stroke suggested that low ow in the basilar artery and posterior cerebral arteries determined by
www.thelancet.com/neurology Vol 9 October 2010

Non-surgical treatment Blood pressure control

Because clinical manifestations of cerebral ischaemia can be precipitated by the lowering of blood pressure in patients with occlusive arterial disease,15,20 treatment of hypertension in patients with a TIA or ischaemic stroke of presumed haemodynamic origin should be done with caution, particularly in the acute phase. Sometimes hypertension should not be treated, and it might even be necessary to increase the blood pressure. In our experience and that of others,85 tapering of antihypertensive drugs can result in cessation of recurrent TIAs in patients with ICA occlusion and haemodynamic TIAs. In individual cases, we have sometimes had to accept systolic blood pressures as high as 200 mm Hg, at least temporarily.
1013

Review

In the chronic phase, we advise control of blood pressure in the same way as for patients with TIA and stroke in general. Because patients with haemodynamic TIAs and ischaemic stroke might be particularly vulnerable to changes in blood pressure, calcium-channel blockers might be preferred over other classes of antihypertensive drugs to avoid uctuations in blood pressure.31

Bed rest
In patients with recent and multiple ischaemic events precipitated by rising, we suggest bed rest for a few days. If the symptoms have not recurred in the meantime, we then generally advise the patient to gradually resume sitting positions, standing, and walking over several days. This approach has never been tested in an RCT.

Revascularisation surgery Extracranialintracranial bypass

Supercial temporal artery (STA)MCA bypass does not prevent recurrent ischaemic stroke in patients with TIAs or minor stroke associated with ICA occlusion or MCA stenosis, or occlusion in general.86,87 A new bypass trial, the Carotid Occlusion Surgery Study, began in 2002 and addresses the question of whether the STAMCA bypass can prevent recurrent ischaemic stroke, at least in a subgroup of patients with recently symptomatic ICA occlusion and documented misery perfusion, as determined by an increased oxygen extraction fraction on PET.88 On June 24, 2010, the trial was stopped early by the US National Institutes of Health; results have not yet been published. The results of a Japanese trial have provided evidence for a benecial eect of the STAMCA bypass in 196 patients with ICA or MCA occlusive disease and haemodynamic cerebral ischaemia according to quantitative CBF.89 The second interim analysis showed that ve surgically treated patients had a recurrent stroke versus 14 patients who were treated medically (p=0046).89 The nal results have not been published in the English literature. The complication rate for fatal or major stroke after STAMCA bypass surgery in the Extracranial Intercranial Bypass Study was 31% or 45%, excluding or including strokes that occurred while waiting for surgery, respectively.86 Some observational series showed that this proportion is probably higher in patients who are neurologically unstable (12%),90 or who have recurrent symptoms and a compromised CBF (14%).91 The Excimer laser-assisted non-occlusive anastomosis (ELANA) extracranialintracranial bypass is a laserassisted bypass procedure that enables the construction of an anastomosis with the distal ICA or proximal MCA or anterior cerebral artery, without the need to temporarily clamp the recipient vessel.92,93 This technique results in a bypass with a higher ow than can be achieved with the conventional STAMCA technique,94 and in theory might better protect against future ischaemic stroke. An RCT to determine the ecacy of the ELANA bypass in prevention of stroke has not yet been done.
1014

According to current clinical standards, STAMCA or ELANA bypass operations to prevent ischaemic stroke in occlusive arterial disease should ideally be performed only in the context of a clinical trial. Nevertheless, while we await the results of the Carotid Occlusion Surgery Study, we occasionally consider STAMCA or ELANA bypass surgery in patients with extracranial ICA occlusion and continuing ischaemic events in whom the arguments for a haemodynamic cause are compelling and in whom there are no other treatment options.95 These patients are informed that the benet of the operation has not yet been proven. In patients with COIS, the ecacy of bypass surgery in addition to medical therapy and panretinal photocoagulation to prevent progression to blindness is uncertain. Disappointing and good results have been reported.51,96 In patients with cognitive decits associated with ICA occlusion, there is too little evidence to recommend surgery to construct an extracranial intracranial bypass. In the absence of an appropriate donor vessel for extracranialintracranial bypass, as might be the case in patients with common carotid artery occlusion, a contralateral donor vessel can be used with tunnelling of the bypass over the skull (so-called bonnet bypass).9799 An intracranialintracranial bypass is also possible with the ELANA technique.100 Information about the ecacy and risk of complications of these procedures is limited to case reports. The same applies to extracranialintracranial bypass surgery for posterior circulation stroke.101 Its ecacy in stroke prevention has not been studied in an RCT, and the risk of complications is higher than for anterior circulation bypasses.101

Other revascularisation procedures

Treatment of stenosis in blood vessels that are important for the collateral blood supply to the hemisphere on the side of the ICA occlusion has not been studied in a large clinical trial, and precise estimates of the risks of these procedures in patients with symptomatic ICA occlusion are not available. However, endarterectomy of a contralateral ICA stenosis or ipsilateral ECA stenosis can be considered in patients with symptomatic ICA occlusion in whom these blood vessels are important for collateral blood supply. Similarly, treatment of stenosis in the brachiocephalic, subclavian, or vertebral arteries can be considered if the stenosis compromises the collateral blood supply. We showed a sustained benet, at least in terms of blood ow to both hemispheres, after endarterectomy for contralateral ICA stenosis in patients with symptomatic ICA occlusion.102 In patients with COIS associated with ICA occlusion, endarterectomy of a stenotic ECA might be considered because the eye might be dependent on the ECA. Safety and feasibility of angioplasty for vertebral artery stenosis is currently being studied in the Vertebral Artery Stenting Trial.103
www.thelancet.com/neurology Vol 9 October 2010

Review

Search strategy and selection criteria References for this Review were identied through a search of PubMed with the broad search terms (stroke [MeSH] AND hemodynamics [MeSH]) NOT hemorrhage [MeSH], from 1966 to June, 2010. CJMK screened 3202 titles, which yielded 142 articles that were further evaluated for relevance by reading the abstracts. In addition, the personal les of the authors were searched. Finally, the bibliographies of all relevant articles were checked. Only papers in English were reviewed.

Acknowledgments We gratefully acknowledge the critical comments of J van Gijn and H B van der Worp (University Medical Center Utrecht, Netherlands) on an earlier version of the manuscript. We thank J van Gijn and P J Koudstaal (Erasmus Medical Center, Rotterdam, Netherlands) for providing the pictures of neovascularisation of the optic disc and rubeosis iridis in gure 1. Conicts of interest We have no conicts of interest. References 1 Baron JC, Bousser MG, Comar D, Soussaline F, Castaigne P. Noninvasive tomographic study of cerebral blood ow and oxygen metabolism in vivo. Potentials, limitations, and clinical applications in cerebral ischemic disorders. Eur Neurol 1981; 20: 27384. 2 Caplan LR, Wong KS, Gao S, Hennerici MG. Is hypoperfusion an important cause of strokes? If so, how? Cerebrovasc Dis 2006; 21: 14553. 3 Klijn CJM, Kappelle LJ, Tulleken CAF, van Gijn J. Symptomatic carotid artery occlusion. A reappraisal of hemodynamic factors. Stroke 1997; 28: 208493. 4 Mead GE, Wardlaw JM, Lewis SC, Dennis MS, Lothian Stroke Registry Study Group. No evidence that severity of stroke in internal carotid occlusion is related to collateral arteries. J Neurol Neurosurg Psychiatry 2006; 77: 72933. 5 Hurwitz BJ, Heyman A, Wilkinson WE, Haynes CS, Utley CM. Comparison of amaurosis fugax and transient cerebral ischemia: a prospective clinical and arteriographic study. Ann Neurol 1985; 18: 698704. 6 Finklestein S, Kleinman GM, Cuneo R, Baringer JR. Delayed stroke following carotid occlusion. Neurology 1980; 30: 8488. 7 Castaigne P, Lhermitte F, Gautier JC, Escourolle R, Derouesne C. Internal carotid artery occlusion: a study of 61 instances in 50 patients with post-mortem data. Brain 1970; 93: 23158. 8 Barnett HJM, Peerless SJ, Kaufmann JCE. Stump of internal carotid arterya source for further cerebral embolic ischemia. Stroke 1978; 9: 44856. 9 Georgiadis D, Grosset DG, Lees KR. Transhemispheric passage of microemboli in patients with unilateral internal carotid artery occlusion. Stroke 1993; 24: 166466. 10 Barnett HJM. Delayed cerebral ischemic episodes distal to occlusion of major cerebral arteries. Neurology 1978; 28: 76974. 11 Ausman JI, Diaz FG. Critique of the extracranial-intracranial bypass study. Surg Neurol 1986; 26: 21821. 12 Sundt TM. Was the international randomized trial of extracranial-intracranial arterial bypass representative of the population at risk? N Engl J Med 1987; 316: 81416. 13 Caplan LR, Piepgras DG, Quest DO, et al. EC-IC bypass 10 years later: is it valuable? Surg Neurol 1996; 46: 41623. 14 Dobkin BH. Orthostatic hypotension as a risk factor for symptomatic occlusive cerebrovascular disease. Neurology 1989; 39: 3034. 15 Hankey GJ, Gubbay SS. Focal cerebral ischaemia and infarction due to antihypertensive therapy. Med J Aust 1987; 146: 41214. 16 Grubb RL Jr, Derdeyn CP, Fritsch SM, et al. Importance of hemodynamic factors in the prognosis of symptomatic carotid occlusion. JAMA 1998; 280: 105560. 17 Yamauchi H, Fukuyama H, Nagahama Y, et al. Signicance of increased oxygen extraction fraction in ve-year prognosis of major cerebral arterial occlusive diseases. J Nucl Med 1999; 40: 199298. 18 European Carotid Surgery Trialists Collaborative Group. Randomised trial of endarterectomy for recently symptomatic carotid stenosis: nal results of the MRC European Carotid Surgery Trial (ECST). Lancet 1998; 351: 137987. 19 North American Symptomatic Carotid Endarterectomy Trial Collaborators. Benecial eect of carotid endarterectomy in symptomatic patients with high grade stenosis. N Engl J Med 1991; 325: 44553. 20 Rothwell PM, Howard SC, Spence JD, Carotid Endarterectomy Trialists Collaboration. Relationship between blood pressure and stroke risk in patients with symptomatic carotid occlusive disease. Stroke 2003; 34: 258390. 21 Kuroda S, Houkin K. Moyamoya disease: current concepts and future perspectives. Lancet Neurol 2008; 7: 105666.

Conclusions
Prospective cohort studies have shown that when low blood ow in the brain plays a major part in causing TIAs or strokes, patients might have specic clinical features, and ancillary investigations can indicate compromised CBF. The same studies have provided evidence that these clinical features and ancillary investigations can be used to identify patients who are at high risk of recurrent ischaemic stroke. A low threshold for ancillary investigations aimed at assessment of the CBF should be used for patients in whom haemodynamic compromise is considered. Nevertheless, this diagnosis remains dicult because a gold standard for the diagnosis of haemodynamic stroke has not been established. Moreover, in many patients, embolism and hypoperfusion might act together. Despite these drawbacks, recognition of hypoperfusion as a contributing mechanism in TIAs and ischaemic stroke might guide treatment. To date, treatment of patients with haemodynamic stroke or TIA cannot be based on evidence from RCTs. Observational non-randomised studies support therapies aimed at increasing CBF, including bed rest in the acute stage, increasing blood pressure, and revascularisation procedures such as endarterectomy or stenting of collateral pathways, or construction of an extracranial intracranial bypass. Because none of these treatments has yet been proven eective in an RCT, risks and benets should be carefully assessed on an individual basis until further evidence becomes available. Future research should be directed at dening the best methods to determine the relative contribution of haemodynamic compromise to the risk of stroke, and diagnostic criteria should be established. Subsequently, treatment aimed at improving the haemodynamic state of the brain needs to be tested in RCTs in patients who are at high risk of ischaemic stroke. For posterior circulation stroke in particular, more information is needed on the natural history in patients in whom haemodynamic compromise seems to play a part.
Contributors Both authors were responsible for the concept of the Review. CJMK selected the papers to be included and prepared the rst draft. Both authors were involved in the writing and editing of the nal draft of the manuscript.

www.thelancet.com/neurology Vol 9 October 2010

1015

Review

22 23 24

25

26

27

28

29

30

31

32 33 34

35 36

37

38

39 40

41

42

43 44 45

46

47

Scott RM, Smith ER. Moyamoya disease and moyamoya syndrome. N Engl J Med 2009; 360: 122637. Caplan LR, Wityk RJ, Glass TA, et al. New England Medical Center Posterior Circulation registry. Ann Neurol 2004; 56: 38998. Kessler C, von Maravic M, Albrecht DM, et al. Doppler CO2-test in patients with vertebrobasilar ischemia. Acta Neurol Scand 1991; 84: 51922. Haubrich C, Kohnke A, Diehl RR, Moller-Hartmann W, Kltzsch C. Impact of vertebral artery disease on dynamic cerebral autoregulation. Acta Neurol Scand 2005; 112: 30916. Gottesman RF, Sherman PM, Grega MA, et al. Watershed strokes after cardiac surgery: diagnosis, etiology, and outcome. Stroke 2006; 37: 230611. Eigenbrodt ML, Rose KM, Couper DJ, Arnett DK, Smith R, Jones D. Orthostatic hypotension as a risk factor for stroke: the atherosclerosis risk in communities (ARIC) study, 19871996. Stroke 2000; 31: 230713. Pullicino PM, McClure LA, Wadley VG, et al. Blood pressure and stroke in heart failure in the REasons for Geographic And Racial Dierences in Stroke (REGARDS) study. Stroke 2009; 40: 370610. Price DL, Harris J. Cholesterol emboli in cerebral arteries as a complication of retrograde aortic perfusion during cardiac surgery. Neurology 1970; 20: 120914. Masuda J, Yutani C, Ogata J, Kuriyama Y, Yamaguchi T. Atheromatous embolism in the brain: a clinicopathologic analysis of 15 autopsy cases. Neurology 1994; 44: 123137. Rothwell PM, Howard SC, Dolan E, et al. Prognostic signicance of visit-to-visit variability, maximum systolic blood pressure, and episodic hypertension. Lancet 2010; 375: 895905. Shahar A, Sadeh M. Severe anemia associated with transient neurological decits. Stroke 1991; 22: 120102. Kim JS, Kang SY. Bleeding and subsequent anemia: a precipitant for cerebral infarction. Eur Neurol 2000; 43: 20108. Kiechl S, Furtner M, Knoach M, Werner P, Willeit J. Kaleidoscopic vision and a jerking leg on the ski slope. Lancet 2007; 370: 1878. Ross Russell RW, Page NGR. Critical perfusion of brain and retina. Brain 1983; 106: 41934. Niehaus L, Neuhauser H, Meyer BU. Transient visual blurring, retro-orbital pain and repetitive involuntary movements in unilateral carotid artery occlusion. Clin Neurol Neurosurg 1998; 100: 3132. Cheshire WP Jr, Meschia JF. Postprandial limb-shaking: an unusual presentation of transient cerebral ischemia. Clin Auton Res 2006; 16: 24346. Kamata T, Yokota T, Furukawa T, Tsukagoshi H. Cerebral ischemic attack caused by post-prandial hypotension. Stroke 1994; 25: 51113. Farid K, Renou P, Late M, et al. Postprandial recurrent hemiplegia. Clin Nucl Med 2010; 35: 8082. Linzer M, McFarland TA, Belkin M, Caplan L. Critical carotid and vertebral arterial occlusive disease and cough syncope. Stroke 1992; 23: 101720. Morgan JC, Alhatou M, Oberlies J, Johnston KC. Transient ischemic attack and stroke associated with sildenal (Viagra) use. Neurology 2001; 57: 173031. Lorberboym M, Mena I, Wainstein J, Boaz M, Lampl Y. The eect of sildenal citrate (Viagra) on cerebral blood ow in patients with cerebrovascular risk factors. Acta Neurol Scand 2009; 121: 37076. Fisher CM. Concerning recurrent transient cerebral ischemic attacks. Can Med J 1962; 86: 109199. Baquis GD, Pessin MS, Scott RM. Limb shaking a carotid TIA. Stroke 1985; 16: 44448. Persoon S, Kappelle LJ, Klijn CJM. Limb-shaking transient ischaemic attacks in patients with internal carotid artery occlusion: a case-control study. Brain 2010; 133: 91522. Yanagihara T, Piepgras DG, Klass DW. Repetitive involuntary movement associated with episodic cerebral ischemia. Ann Neurol 1985; 18: 24450. Klijn CJM, Kappelle LJ, van Huelen AC, et al. Recurrent ischemia in symptomatic carotid occlusion: prognostic value of hemodynamic factors. Neurology 2000; 56: 180612.

48

49

50

51

52

53 54

55 56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

Yanagihara T, Klass DW, Piepgras DG, Houser OW. Brief loss of consciousness in bilateral carotid occlusive disease. Arch Neurol 1989; 46: 85861. AhChong AK, Law CB, Chiu KM. Carotid endarterectomy for nonhemispheric cerebrovascular symptoms: an unusual indication. Hong Kong Med J 1999; 5: 39193. Furlan AJ, Whisnant JP, Kearns TP. Unilateral visual loss in bright light. An unusual symptom of carotid artery occlusive disease. Arch Neurol 1979; 36: 67576. Klijn CJM, Kappelle LJ, van Schooneveld MJ, et al. Venous stasis retinopathy in symptomatic carotid artery occlusion: prevalence, cause, and outcome. Stroke 2002; 33: 695701. Sivalingam A, Brown GC, Magargal LE. The ocular ischemic syndrome. III. Visual prognosis and the eect of treatment. Int Ophthalmol 1991; 15: 1520. Carter JE. Chronic ocular ischemia and carotid vascular disease. Stroke 1985; 16: 72128. Kearns TP, Hollenhorst RW. Venous-stasis retinopathy of occlusive disease of the carotid artery. Proc Sta Meet Mayo Clin 1963; 38: 30412. Fisher CM. Senile dementiaa new explanation for its causation. Can Med Assoc J 1951; 65: 17. Tatemichi TK, Desmond DW, Prohovnik I, Eidelberg D. Dementia associated with bilateral carotid occlusions: neuropsychological and haemodynamic course after extracranial to intracranial bypass surgery. J Neurol Neurosurg Psychiatry 1995; 58: 63336. Bakker FC, Klijn CJ, Jennekens-Schinkel A, van der Tweel I, Tulleken CAF, Kappelle LJ. Cognitive impairment in patients with carotid artery occlusion and ipsilateral transient ischemic attacks. J Neurol 2003; 250: 134047. Momjian-Mayor I, Baron JC. The pathophysiology of watershed infarction in internal carotid artery disease: review of cerebral perfusion studies. Stroke 2005; 36: 56777. Del Sette M, Eliasziw M, Streier JY, et al. Internal borderzone infarction: a marker for severe stenosis in patients with symptomatic internal carotid artery disease. Stroke 2000; 31: 63136. Bladin CF, Chambers BR. Clinical features, pathogenesis, and computed tomographic characteristics of internal watershed infarction. Stroke 1993; 24: 192532. van der Zwan A, Hillen B, Tulleken CAF, Dujovny M, Dragovic L. Variability of the territories of the major cerebral arteries. J Neurosurg 1992; 77: 92740. Lang EW, Daertshofer M, Daertshofer A, Wirth SB, Chesnut RM, Hennerici M. Variability of vascular territory in stroke. Pitfalls and failure of stroke pattern interpretation. Stroke 1995; 26: 94245. van Laar PJ, Hendrikse J, Golay X, Lu H, van Osch MJP, van der Grond J. In vivo ow territory mapping of major brain feeding arteries. Neuroimage 2006; 29: 13644. Derdeyn CP, Carpenter DA, Videen TO, Grubb RL Jr, Powers WJ. Patterns of infarction in hemodynamic failure. Cerebrovasc Dis 2007; 24: 1119. Hofmeijer J, Klijn CJM, Kappelle LJ, van Huelen AC, van Gijn J. Collateral circulation via the ophthalmic artery or leptomeningeal vessels is associated with impaired cerebral vasoreactivity in patients with symptomatic carotid artery occlusion. Cerebrovasc Dis 2002; 14: 2226. Norrving B, Nilsson B, Risberg J. rCBF in patients with carotid occlusion. Resting and hypercapnic ow related to collateral pattern. Stroke 1982; 13: 15562. Muller M, Hermes M, Brckmann H, Schimrigk K. Transcranial Doppler ultrasound in the evaluation of collateral blood ow in patients with internal carotid artery occlusion: correlation with cerebral angiography. AJNR Am J Neuroradiol 1995; 16: 195202. Smith HA, Thompson-Dobkin J, Yonas H, Flint E. Correlation of xenon-enhanced computed tomography-dened cerebral blood ow reactivity and collateral ow patterns. Stroke 1994; 25: 178487. Derdeyn CP, Shaibani A, Moran CJ, Cross De Witte T, Grubb RL Jr, Powers WJ. Lack of correlation between pattern of collateralization and misery perfusion in patients with carotid occlusion. Stroke 1999; 30: 102532. Vernieri F, Pasqualetti P, Matteis M, et al. Eect of collateral blood ow and cerebral vasomotor reactivity on the outcome of carotid artery occlusion. Stroke 2001; 32: 155258.

1016

www.thelancet.com/neurology Vol 9 October 2010

Review

71

72

73

74

75

76

77

78

79

80

81

82

83 84

85

86

van Everdingen KJ, van der Grond J, Kappelle LJ. Overestimation of a stenosis in the internal carotid artery by duplex sonography caused by an increase in volume ow. J Vasc Surg 1998; 27: 47985. van Laar PJ, van der Grond J, Bremmer JP, Klijn CJM, Hendrikse J. Assessment of the contribution of the external carotid artery to brain perfusion in patients with internal carotid artery occlusion. Stroke 2008; 39: 300308. Wintermark M, Sesay M, Barbier E, et al. Comparative overview of brain perfusion imaging techniques. J Neuroradiol 2005; 32: 294314. Ogasawara K, Ogawa A, Terasaki K, Shimizu H, Tominaga T, Yoshimoto T. Use of cerebrovascular reactivity in patients with symptomatic major cerebral artery occlusion to predict 5-year outcome: comparison of xenon-133 and iodine-123-IMP singlephoton emission computed tomography. J Cereb Blood Flow Metab 2002; 22: 114248. Nemoto EM, Yonas H, Kuwabara H, et al. Identication of hemodynamic compromise by cerebrovascular reserve and oxygen extraction fraction in occlusive vascular disease. J Cereb Blood Flow Metab 2004; 24: 108189. Smith LMcG, Elkins JS, Dillon WP, Schaeer S, Wintermark M. Perfusion-CT assessment of the cerebrovascular reserve: a revisit to the acetazolamide challenges. J Neuroradiol 2008; 35: 15764. Bokkers RP, Bremmer JP, van Berckel BN, et al. Arterial spin labeling perfusion MRI at multiple delay times: a correlative study with H215O positron emission tomography in patients with symptomatic carotid artery occlusion. J Cereb Blood Flow Metab 2009; 30: 22229. Sugimori H, Ibayashi S, Fujii K, Sadoshima S, Kuwabara Y, Fujishima M. Can transcranial Doppler really detect reduced cerebral perfusion states? Stroke 1995; 26: 205360. Derdeyn CP, Grubb RL Jr, Powers WJ. Cerebral hemodynamic impairment. Methods of measurement and association with stroke risk. Neurology 1999; 53: 25159. Klijn CJM, Kappelle LJ, Algra A, van Gijn J. Outcome in patients with symptomatic occlusion of the internal carotid artery or intracranial arterial lesions: a meta-analysis of the role of baseline characteristics and type of antithrombotic treatment. Cerebrovasc Dis 2001; 12: 22834. Adams HP Jr, Powers WJ, Grubb RL Jr, Clarke WR, Woolson RF. Preview of a new trial of extracranial-to-intracranial anastomosis. The carotid occlusion surgery study. Neurosurg Clin N Am 2001; 36: 61324. Persoon S, Luitse MJA, de Borst GJ, et al. Symptomatic internal carotid artery occlusion: a long-term follow-up study. J Neurol Neurosurg Psychiatry 2010 (in press). Flossmann E, Rothwell PM. Prognosis of vertebrobasilar transient ischaemic attack and minor stroke. Brain 2003; 126: 194054. Amin-Hanjani S, Du X, Zhao M, Walsh K, Malisch TW, Charbel FT. Use of quantitative magnetic resonance angiography to stratify stroke risk in symptomatic vertebrobasilar disease. Stroke 2005; 36: 114045. Leira EC, Ajax T, Adams HP. Limb-shaking carotid transient ischemic attacks successfully treated with modication of the antihypertensive regimen. Arch Neurol 1997; 7: 90405. EC/IC Bypass Study Group. Failure of extracranial-intracranial arterial bypass to reduce the risk of ischemic stroke. Results of an international randomized trial. N Engl J Med 1985; 313: 1191200.

87

88 89 90

91

92

93

94

95

96

97 98

99 100

101

102

103

Fluri F, Engelter S, Lyrer P. Extracranial-intracranial arterial bypass surgery for occlusive carotid artery disease. Cochrane Database Syst Rev 2010; 2: CD005953. Grubb RL Jr, Powers WJ, Derdeyn CP, Adams HP Jr, Clarke WR. The Carotid Occlusion Surgery Study. Neurosurg Focus 2003; 14: e9. JET Study Group. Japanese EC-IC Bypass Trial: the second interim analysis [in Japanese]. Surg Cereb Stroke 2002; 30: 43437. Sundt TM Jr, Whisnant JP, Fode NC, Piepgras DG, Houser OW. Results, complications, and follow-up of 415 bypass operations for occlusive disease of the carotid system. Mayo Clin Proc 1985; 60: 23040. Schmiedek P, Piepgras A, Leinsinger G, Kirsch C-M, Einhaupl K. Improvement of cerebrovascular reserve capacity by EC-IC arterial bypass surgery in patients with ICA occlusion and hemodynamic cerebral ischemia. J Neurosurg 1994; 81: 23644. Tulleken CAF, van der Zwan A, Verdaasdonk RM, Mansvelt Beck RJ, Moreira Pereira Ramos L, Kappelle LJ. High-ow Excimer laser-assisted extra-intracranial and intra-intracranial bypass. Oper Tech Neurosurg 1999; 2:14248. Langer DJ, Vajkoczy P. ELANA: Excimer laser-assisted nonocclusive anastomosis for extracranial-to-intracranial and intracranial-tointracranial bypass: a review. Skull Base 2005; 15: 191205. van der Zwan A, Tulleken CAF, Hillen B. Flow quantication of the non-occlusive Excimer laser-assisted EC-IC bypass. Acta Neurochir (Wien) 2001; 143: 64754. Klijn CJM, Kappelle LJ, van der Grond J, van Gijn J, Tulleken CAF. A new type of extracranial/intracranial bypass for recurrent haemodynamic transient ischaemic attacks. Cerebrovasc Dis 1998; 8: 18487. Kawaguchi S, Sakaki T, Morimoto T, Okuno S, Nishikawa N. Eects of bypass on ocular ischaemic syndrome caused by reversed ow in the ophthalmic artery. Lancet 1999; 354: 205253. Spetzler RF, Roski RA, Rhodes RS, Modic MT. The bonnet bypass. Case report. J Neurosurg 1980; 53: 70709. Zumofen D, Khan N, Roth P, Samma A, Yonekawa Y. Bonnet bypass in multiple cerebrovascular occlusive disease. Acta Neurochir Suppl 2008; 103: 10307. Garrido E, Freed MH. Fatal complication of the bonnet bypass: case report. Neurosurgery 1983; 13: 32021. Langer DJ, van der Zwan A, Vajkoczy P, Kivipelto L, van Doormaal TP, Tulleken CA. Excimer laser-assisted nonocclusive anastomosis. An emerging technology for use in the creation of intracranial-intracranial and extracranial-intracranial cerebral bypass. Neurosurg Focus 2008; 24: E6. Ausman JI, Diaz FG, Vacca DF, Sadasivan B. Supercial temporal and occipital artery bypass pedicles to superior, anterior inferior, and posterior inferior cerebellar arteries for vertebrobasilar insuciency. J Neurosurg 1990; 72: 55458. Rutgers DR, Klijn CJM, Kappelle LJ, Eikelboom BC, van Huelen AC, van der Grond J. Sustained bilateral hemodynamic benet of contralateral carotid endarterectomy in patients with symptomatic internal carotid artery occlusion. Stroke 2001; 32: 72734. Compter A, van der Worp HB, Schonewille W, et al. VAST: Vertebral Artery Stenting Trial. Protocol for a randomised safety and feasibility trial. Trials 2008; 9: 65.

www.thelancet.com/neurology Vol 9 October 2010

1017