Review

Imaging of acute stroke

Keith W Muir, Alastair Buchan, Rudiger von Kummer, Joachim Rother, Jean-Claude Baron

Thrombolytic therapy has led to a higher proportion of patients presenting to hospital early, and this, with parallel developments in imaging technology, has greatly improved the understanding of acute stroke pathophysiology. Additionally, MRI, including diusion-weighted imaging (DWI) and gradient echo, or T2*, imaging is important in understanding basic structural informationsuch as distinguishing acute ischaemia from haemorrhage. It has also greatly increased sensitivity in the diagnosis of acute cerebral ischaemia. The pathophysiology of the ischaemic penumbra can now be assessed with CT or MRI-based perfusion imaging techniques, which are widely available and clinically applicable. Pathophysiological information from CT or MRI increasingly helps clinical trial design, may allow targeted therapy in individual patients, and may extend the time scale for reperfusion therapy.

Lancet Neurol 2006; 5: 75568 Division of Clinical Neurosciences, University of Glasgow, Institute of Neurological Sciences, Southern General Hospital, Glasgow (K W Muir FRCP); University of Oxford, Nueld Department of Medicine, University of Oxford, John Radclie Hospital, Oxford (A Buchan FRCP); University of Dresden, University Hospital, University of Technology, Dresden, Germany (R von Kummer MD); Department of Neurology University of Hanover, Klinikum Minden, 32427 Minden, Germany (J Rother MD); and Department of Clinical Neuroscience, Neurology Unit, University of Cambridge, Cambridge (J-C Baron FMedSci) Correspondence to: Prof Jean-Claude Baron, University of Cambridge, Department of Clinical Neurosciences, Neurology Unit, Addenbrookes Hospital, Box 83, Cambridge CB2 2QQ, UK jcb54@cam.ac.uk

Introduction
Recent developments in imaging have revolutionised our approach to acute stroke by allowing us to directly image the ischaemic process and giving us a better understanding of the pathophysiology. Translation of basic research concepts into clinical practice fuelled pivotal randomised controlled trials of intravenous thrombolysis, which now constitutes the basis of management. Imaging has documented substantial individual dierences in both the time course of infarct growth and the occurrence of spontaneous recanalisation, supporting the idea that management decisions be based on the individual patients pathophysiological diagnosis.1 Physiological imaging is now incorporated in the design of many randomised controlled trials, particularly at the dose-nding stage, and is the basis of the current shift towards small trials of highly homogeneous samples powered to address specic pathophysiological processes, with expected increases in trial eciency.2,3 In this review we will focus on the imaging of acute ischaemic stroke (about 85% of all strokes), particularly, middle cerebral artery (MCA) territory stroke (about 50% of all ischaemic strokes).4 MCA stroke is the syndrome most studied in terms of pathophysiology and therapy. Posterior circulation and small vessel strokes will be discussed separately. After a brief pathophysiological overview, we will look at the advances achieved with MRI, currently the most useful imaging technique to assess acute stroke; and the use of CT in combination with CT perfusion for decisionmaking in MCA stroke. This review is not intended to be a guideline but aims to discuss how imaging can help in decision-making for therapy, from a neurological and pathophysiological, rather than a radiological, perspective. Levels of evidence have not been included, but where relevant systematic reviews have been done these will be mentioned in this review.

proximal MCA occlusion reduces distal cerebral perfusion pressure. Cerebral perfusion pressure reductions in the cortical MCA territory are most severe in its centre (perisylvian region) and least in the watershed areas; the lentiform nucleus and some of the white matter, which have far fewer anastomoses, are the most severely aected.5 Cerebral perfusion pressure reduction is, in most areas, severe enough to override cerebrovascular autoregulation, resulting in hypoperfusion according to a similar gradient. Lost autoregulation means that local perfusion is aected by systemic blood pressure, proximal carotid disease, and intracranial pressure. Ischaemia, when severe, can result in tissue infarction (ie, pannecrosis); when less severe or prolonged, however, ischaemia can induce selective neuronal death, an intermediate state reputed to be undetectable on CT.6 The key pathophysiological concept is the distinction of hypoperfused tissue into three operational compartments;7 tissue that will inevitably die (core), tissue that will in principle survive (oligaemia), and tissue that may either die or survive (the ischaemic penumbra; gure 1). Reduction of cerebral blood ow from its normal mean, around 50 mL/100 g per min, to less than 20 mL/100 g per min, results in impaired neural function8 but preserved tissue integrity; this denes the penumbra. Changes in water homoeostasis, including shrinkage of
Oligaemia Penumbra

Core

Pathophysiology
General concepts
The pathophysiological model developed from animal studies, particularly in primates, forms the basis for our understanding of the acute phase of MCA stroke.5 Acute
http://neurology.thelancet.com Vol 5 September 2006 Figure 1: A diagram showing the three hypoperfused tissue compartments in acute MCA stroke A further compartment with normal perfusion but partially exhausted vascular reserve, not illustrated here, may surround the oligaemic compartment (see text).

755

Review

Panel: Currently accepted operational criteria dening the penumbra Hypoperfusion <20 mL/100 g per min Abnormal neuronal function documented by a correlation with acute clinical decit Physiological and/or biochemical characteristics consistent with cellular dysfunction but not death Uncertain fate Salvage of this tissue is correlated with better clinical recovery

the extracellular space9 and net uptake of water from plasma,10 also take place in the penumbra. Unless early reperfusion occurs, the penumbra is gradually recruited into the coreie, the volume of irreversibly damaged tissue grows and the amount of penumbra decreases.7 Tissue outcome depends on two factors: the severity of ow reduction and its durationie, the cerebral blood ow threshold that denes the core increases with time until it reaches the penumbra threshold, at which point all the penumbra has been recruited. Thus, within the penumbra, the lower the cerebral blood ow, the higher the risk of early infarction. Because both the core and the penumbra can contribute to neurological decits, it is impossible to clinically determine their relative eects. Rescue of the penumbra, either by restoration of blood supply or by interruption of the adverse metabolic or neurochemical cascade, is the basis of acute stroke therapy. Imaging studies, with various techniques, have established the clinical importance of penumbral salvage, showing a clear association between the volume of penumbra not progressing to infarction and the improvement in neurological scores.1115 To be accepted as penumbra, the aected tissue must full several welldened operational criteria (panel). Tissue with small reductions in cerebral blood ow 2050 mL/100 g per min, dened as oligaemia, probably maintains its function for a very long time and is unlikely to proceed to infarction. However, the oligaemia may be pushed into penumbral stateand hence potentially into the coreby secondary events that reduce the cerebral perfusion pressure, such as vasogenic oedema and systemic hypotension, or by factors that aggravate the owmetabolism mismatch such as hyperglycaemia and pyrexia.16 This may explain the benets from avoiding physiological complications and maintaining blood pressure early after MCA stroke.16

carotid occlusion. The volume of core is associated with both admission clinical scores and nal infarct volume.20 Substantial penumbra is present in up to 90% of patients within 6h of onset; this falls to about 50% within 9 h but is still about 30% 18 h after onset.7 Up to 52% of the ultimate infarct still shows penumbra 16 h after onset,21 suggesting delayed therapeutic opportunities in some patients. The penumbra progresses to, or escapes, infarction partly or in all, depending on subsequent events such as reperfusion.11,2224 Survival of the penumbra is the main determinant of clinical recovery11,13 and probably underpins peri-infarct reorganisation.25,26 Thus, although mapping the core provides a marker of the inevitable damage, mapping the penumbra shows the potential for recovery. Oligaemia displays mild misery perfusionie, a moderately reduced cerebral blood ow (above the penumbra threshold) and a high oxygen extraction fraction. A high oxygen extraction fraction in itself does not equate with penumbra as this fraction rises as soon as the cerebral blood ow decreases.27 The high oxygen extraction fraction area represents exclusively the penumbra in patients but virtually only the oligaemia in others.11 Early spontaneous hyperperfusion, present in about a third of patients within 18 h,22 almost invariably predicts preserved cerebral oxygen metabolism (CMRO2) and tissue integrity, suggesting this as a marker of prior recanalisation that salvaged the penumbra. After thrombolysis, however, hyperperfusion may signal poor tissue outcome,28 possibly because pressure is forced into a damaged vascular tree. An extensive core invariably predicts poor clinical outcome and malignant infarction, whereas extensive spontaneous hyperperfusion invariably predicts excellent outcome, independently of admission clinical scores.19,29 By contrast, when there is extensive penumbra, the clinical course is unpredictable.19,29 The occurrence of these three main patterns of changes highlights the substantial pathophysiological heterogeneity underlying acute MCA stroke.

Overview of imaging techniques

Following major technological advances over the past 15 years, imaging can now characterise brain structure and the pathological status of established lesions, brain perfusion, intracranial and extracranial vascular pathology (including direct visualisation of the clot), tissue viability, and metabolic state, bringing complex physiological concepts into everyday clinical practice. Structural imaging with either CT or conventional MRI sequences has only limited sensitivity to acute ischaemic changes (4070% in the rst 6 h across all subtypes),30,31 depending notably on anatomical location, background abnormalities, scan quality, and reader experience. Although CT remains the best method for detection of intracerebral haemorrhage, MRI, with appropriate sequences, can be equally sensitive.32,33
http://neurology.thelancet.com Vol 5 September 2006

Insights from PET

PET shaped the concepts underlying modern acute stroke imaging7 and remains the gold standard. Based on validated thresholds, aected tissue can be classied as core, penumbra, oligaemia, and hyperperfused.7,17 In early proximal MCA occlusion, the core typically involves the striatocapsular area,18 whereas the penumbra typically involves the cortical mantle. However, occasionally the core extends widely into cortical areas as early as 4 h after onset,19 probably due to inadequate pial collaterals or a
756

Review

Figure 2: Early CT signs in acute MCA stroke Left and middle: Hyperdense left MCA sign (yellow arrow), hypoattenuated left basal ganglia (red arrow), and cortical swelling (blue arrows) in the same patient. Right: Dot sign (yellow arrow) in the left sylvian ssure.

Non-invasive time-of-ight (ie, non-contrsast) magnetic resonance angiography and slightly more invasive CT angiography or contrast-enhanced magnetic resonance angiography (with intravenous contrast administration) now allow routine vascular imaging. Extracranial vascular imaging or ultrasound can identify individuals at high risk of repeated stroke, or those who may benet from specic interventions. Real-time monitoring of MCA ow is possible with transcranial doppler, which in addition can promote early recanaliasation after alteplase therapy.34

Plain CT
Contrary to previous dogma, plain CT can be abnormal within 3 h of onset in 75% of patients with MCA stroke.35 Early ischaemic changes include hypoattenuation or focal tissue swelling (gure 2), which dier in pathophysiological basis. Tissue density declines with time from vessel occlusion in the core and indicates increased water content.36,37 Net uptake of water can, however, also take place in the penumbra.10 Hypodensity is associated with the most severe reductions in cerebral blood ow and volume on perfusion imaging,38 and the anatomical extent of early hypodensity within 36 h of stroke onset predicts nal infarction, with few exceptions.39 However, brain swelling, without hypoattenuation, indicates increased cerebral blood volume,36 associated with normal apparent diusion coecient, moderate hypoperfusion, and increased mean transit timeie, viable tissue.40 Thus, conventionally dened early ischaemic changes include elements of both core and penumbra. Interobserver agreement for early ischaemic changes is moderate4144 but this depends on the individuals experience. Systematic review of CT using the Alberta stroke programme early CT score (ASPECTS),35 particularly with some clinical information, has better inter-rater reliability, indicates (severityduration) preceding
http://neurology.thelancet.com Vol 5 September 2006

hypoperfusion,45 is predictive of response to thrombolysis, and can be done in real time.46,47 Because current ASPECTS scoring combines swelling and hypodensity, it may not be the best method to distinguish irreversibly damaged tissue from viable tissue. Review of data from both the National Institute of Neurological Disorders and Stroke (NINDS) and ECASS-II trials with alteplase suggest that early ischaemic changes (assessed by ASPECTS) are not independently associated with poor clinical outcome, but a very poor ASPECTS score (2) was associated with increased risk of intracerebral haemorrhage in the ECASSII cohort, and poor outcome in the NINDS cohort.48,49 The presence and extent of early ischaemic changes probably indicate stroke severity and their prevalence does not vary substantially over time.39 The recognition that the anatomical extent of parenchymal hypodensity in patients treated with alteplase predicts the risk of intracerebral haemorrhage has led to the introduction of the one-third MCA territory rule (ie, do not thrombolyse) for randomised controlled trials.50 More studies are needed to clarify the respective predictive value of hypodensity and swelling for dierent time windows. Increased attenuation of an arterial segment, the hyperdense MCA sign (gure 2), is highly specic for occlusion by a thrombus51 but has limited sensitivity (2734% in large series).5254 Because radiograph attenuation of clots depends on haematocrit,55 a possible explanation is that white, brin-poor thrombi are not identied on CT, explaining its poor sensitivity.51 In a general stroke population, the hyperdense MCA sign is associated with poor prognosis and risk of thrombolysisassociated intracerebral haemorrhage5658 and its resolution with thrombolytic therapy is associated with favourable outcome. In patients with acute MCA occlusion, however, the hyperdense MCA sign has no independent prognostic value and is not predictive of responsiveness to intravenous alteplase given within 3 h.51 The presence of distal occlusion of the MCA can be
757

Review

FLAIR day 1

DWI after 7h

Figure 3: Main patterns of DWI and PWI changes in acute MCA stroke Top: Small DWI lesion in deep MCA terrirtory with extensive area of prolonged time to peak (ie, mismatch) in the face of complete proximal MCA occlusion on MR angiography. Bottom: Extensive DWI lesion covering a large part of the MCA territory with only slightly larger time-to-peak lesion, in the face of a near-occlusion of the proximal MCA; note growth of the DWI lesion to completely match the initial time-to-peak lesion 7 h later.

identied by the sylvian dot sign59 (gure 2), which has reasonable specicity (3846%).53,60

CT and magnetic resonance angiography

Although complete coverage from the aortic arch to the circle of Willis is possible with CT angiography, it is still rarely implemented. Acute CT angiography improves localisation and syndromic classication compared with clinical features alone,61 and the source images from CT angiography are more sensitive than plain CT.62 Intracranial CT angiography can identify the presence and site of occlusion, which has substantial implications for management of patients. For instance, the potential for intravenous alteplase to eect recanalisation is reduced for sites such as terminal carotid T occlusions,63,64 therefore inuencing the therapeutic approacheg, with intraarterial thrombolysis or mechanical clot retrieval. However, the value of these alternative approaches has yet to be established in prospective trials. By visualising the extracranial vasculature, CT angiography of neck vessels is important for management. Carotid dissection or complex atherosclerotic plaque might prompt alternative early treatments (eg, aggressive antiplatelet therapy, highdose statins, anticoagulants, or even stenting). Time-of-ight magnetic resonance angiography is the preferred method when combined with DWI/PWI (diusion weighted imaging/perfusion weighted imaging) because it avoids the repeated administration of contrast (required for PWI), and reliably detects proximal MCA occlusion (gures 3 and 4), although there can be diculty in distinguishing MCA occlusion from stenosis. This type of angiography can be used to image the neck
758

vessels with the newest scanners. However, contrast magnetic resonance angiography, when available, has higher quality and accuracy than time-of-ight magnetic resonance angiography.65 The sensitivity and specicity of CT and magnetic resonance angiography have been assessed against conventional angiography in extracranial carotid disease,66 but not in acute stroke.

Physiological brain imaging

In this review we will focus on the most widely used imaging techniques, MRI and CT. The main physiological imaging techniques are summarised in table 1; however several imaging techniques are used less commonly.67 Combining DWI with PWI and time-of-ight magnetic resonance angiography in acute stroke has successfully widened the opportunities to apply pathophysiological insights to clinical management.

The DWI lesion

DWI has revolutionised the diagnostic sensitivity of imaging ischaemia. DWI can indicate the degree of free diusion of water moleculesapparent diusion coecient. If the cerebral blood ow goes below the penumbra threshold this causes failure of energydependent processes, resulting in intracellular (cytotoxic) oedema, shrinkage of the extracellular compartment, and consequently reduced apparent diusion coecient.6872 The latter translates into a high DWI signal, resulting in even very small ischaemic lesions being conspicuous.73,74 Furthermore, DWI is abnormal within minutes of stroke onset.75 Apparent diusion coecient values gradually return to normal in 510 days and then increase in
http://neurology.thelancet.com Vol 5 September 2006

Review

FLAIR

ADC

DWI

TTP

<3 h

1 day

7 days

Figure 4: Main patterns of DWI and PWI changes in acute MCA stroke Top: fuzzy DWI lesion in left MCA territory matching an area of diminished time to peak, indicating local hyperperfusion and suggesting spontaneous recanalisation had occurred prior to imaging at 3 h after onset (note the prolonged time to peak at the posterior edge of the DWI lesion, suggesting distal branch occlusion). Middle: the next day, perfusion has essentially normalised as well as the DWI lesion, save for a narrow posterior streak, suggesting the spontaneous recanalisation saved the at-risk tissue from progressing to infarction. Bottom: At day 7, there has been no return of the DWI lesion, indicating the tissue was eectively salvaged, without even minimal damage.

chronic lesions.7678 This gradual increase may indicate the development of vasogenic oedema and cellular necrosis.79 However, the DWI lesion persists for another week or so because it also detects prolonged T2 signal (T2 shine-through). Proper interpretation of DWIeg, to distinguish acute recurrencemust therefore consider the apparent diusion coecient map. DWI is more sensitive to acute ischaemia than plain CT or T2-weighted MRI, which show only tissue changes caused by severe and prolonged ischaemia, usually sucient to cause infarction.37,80,81 Interobserver agreement in stroke more than 6 h after onset is substantially better with DWI than with plain CT.43,80,82 Within 6 h of stroke onset, DWI has reported sensitivity of 95% and specicity of nearly 100%.83
http://neurology.thelancet.com Vol 5 September 2006

DWIPWI mismatch
Unless there is early reperfusion, DWI lesions expand over a period up to 24 h,84 largely from recruitment of part or all of the surrounding hypoperfused tissue. DWI lesion expansion occurs almost exclusively in those patients who initially have a perfusion defect larger than the DWI lesion, DWIPWI mismatch.85,86 Factors that may hasten or amplify the growth of DWI lesions include hyperglycaemia,87 high haematocrit,88 old age,89 and systemic hypoxia.90 The DWIPWI mismatch pattern is present in about 70% of all patients with anterior-circulation stroke scanned within 6 h of onset,91 is strongly associated with proximal MCA occlusion,91 and resolves on reperfusion.92,93 Resolution of hypoperfusion from early sustained revascularisation prevents DWI lesion expansion94,95 and a
759

Review

Parameters CT Perfusion Xe MRI DWIPWI CBF, CBV, MTT, TTP, ADC CBF, CBV, MTT, TTP CBF

Denition of penumbra

Advantages

Limitations

Relative CBF <66%* CBV >25 mL/100g CBF 720 mL/100g per min

Combined with plain CT; available; fast Quantitative, combined with plain CT

Limited brain coverage; poorly sensitive in posterior circulation; indirect visualisation of core; iodinated contrast Not fully validated; provides CBF only; pharmacological eects of xenon

Relative TTP (or MTT) delay >4s* and normal DWI Not validated Not validated Elevated lactate and normal NAA CBF 722mL/100g per min and OEF >070 Relative binding >34. and CBF < 14 mL/100g per min Uptake ratio >13*

Fast; best sensitivity; no radiation Limited availability; mismatch concept not fully validated; CBF and involved, directly visualises core CBV values not accurate; patient cooperation required; frequent contraindications No contrast needed Non-invasive mapping of OEF and CMRO2 Biochemically characterises tissue Provides CBF only; poor sensitivity to low ows Validity unclear Not validated; poor resolution

Arterial spin labelling MRI-based OEF (and PWI) Spectroscopy PET Multi-tracer 15O C-FMZ (+H215O)
11 18

CBF CBF, OEF, CMRO2 NAA, lactate

CBF, CBV, MTT, CMRO2, OEF Tracer binding Tracer uptake

Quantitative, validated Based on physiological neuronal integrity Produces a direct positive image of viable hypoxic tissue Cheap and relatively available

Complex; time consuming; not widely available; expensive As above and only suitable for cortex; basis not validated As above and validation incomplete; long imaging time

F-FMISO

SPECT 99mTc-labelled HMPAO or ECD CBF Relative CBF <65%* Provides perfusion only so thresholds uncertain; limited spatial resolution; slow brain kinetics

ADC=apparent diusion coecient; CBF=cerebral blood ow; CBV=cerebral blood volume; CMRO2=cerebral metabolic rate of oxygen; OEF=oxygen extraction fraction; DWI=diusion-weighted imaging; FMISO=uoromisonidazole; FMZ=umazenil; HMPAO=hexamethylpropyleneamine oxime; ECD=ethyl-cysteinate dimer; MTT=mean transit time; NAA=N-acetylaspartate; PWI=perfusion weighted imaging; SPECT=single photon emission CT; TTP=time to peak. *Relative to mean contralateral hemisphere.Relative to contralateral healthy white matter.The DWI lesion may not be associated with the core.

Table 1: Main physiological imaging techniques, with the parameters they map, the criteria they use to identify the penumbra, and their main advantages and limitations

large mismatch volume increases the chance of a spectacular shrinking decit on early reperfusion.96 Finally, the volume of tissue salvaged is associated with neurological recovery and functional outcome.97 Thus, combined DWI and PWI might dene the core (the DWI lesion) and the penumbratissue that has low perfusion and normal DWI or apparent diusion coecient but that can be recruited into the DWI lesion.98,99 Table 2 summarises the evidence base for the DWIPWI mismatch hypothesis. Although this concept seems appealing, it is challenged by the possibility of DWI lesions reversible.100 Normalisation of the apparent diusion coecient decline on reperfusion is reported in animal studies101 and in as many as 20% of patients in the 6 h time window.102 Predictors of apparent diusion coecient normalisation include thrombolytic therapy, particularly within the 3 h time window, recanalisation, and peripheral branch occlusion.102 The severity of hypoperfusion is probably the best predictor of the risk of infarction rather than the apparent diusion coecient value.103 However, delayed DWI lesion reappearance on later imaging (rst reported in animal models) has been reported in man.104 This lesion reappearance occurs in regions of intermediate ischaemia (as judged by apparent diusion coecient values) and causal hypotheses include cell death (which can be selective)105 due to various factors.81
760

There are also uncertainties regarding MRI-based bolus-tracking perfusion imaging. Denition of the optimum perfusion variable is not uniform across studies.106 The dierent variables (typically time to peak, mean transit time, cerebral blood volume and ow) dier in their predictive value for dening nal infarction. Lesion volumes depend partly on the selection of the arterial input function.107 Interpretion of studies that describe the correlation of initial PWI lesion volume with later infarct volume is further complicated by the varied natural course of recanalisation and reperfusion. Some studies suggest optimum variables and thresholds.108110 The hypothesis that mismatch is associated with penumbra has also been challenged by combined PET and MRI studies showing that the DWI lesion contains core but also penumbra,111 whereas the mismatch contains penumbra but also oligaemia.112 However, a mean transit time delay of more than 4 s adequately approximates the penumbra.113 Thus, for clinical purposes, the mismatch is reasonably robust in representing the at-risk tissue, whereas the DWI lesion represents the tissue that will probably progress to infarction without any intervention.

Other DWIPWI patterns

Apart from the mismatch pattern, acute patients with anterior circulation stroke can have matched DWIPWI
http://neurology.thelancet.com Vol 5 September 2006

Review

Operational criteria Hypoperfusion <20 mL/100 g per min Abnormal neuronal function documented by a correlation with acute clinical decit Physiological and/or biochemical characteristics consistent with cellular dysfunction but not death

PWIDWI mismatch* Absolute CBF values unreliable with PWI; but TTP >4 s delay shown to correspond to PET-derived CBF <20 mL/100 g per min Volume of DWI lesion and of mismatch signicantly correlate with acute-stage NIHSS In humans, acutely the ADC starts to decline for CBF values corresponding to the penumbra threshold. ADC declines represent cytotoxic oedema and consistently correspond to reduced CMRO2 in humans. In the rat, ADC declines shown to correspond with impaired aerobic glycolysis and ATP production Without early reperfusion, tissue-at-risk progresses to full infarction Salvage of tissue-at-risk transfers to better outcome and correlates with clinical recovery

Uncertain fate Salvage of this tissue is correlated with better clinical recovery

CBF=cerebral blood ow; TTP=time to peak; NIHSS=National Institutes of Health stroke scale; ADC=apparent diusion coecient; CMRO2=cerebral metabolic rate of oxygen.*See panel.

Table 2: Validity of the DWI-PWI mismatch to identify the penumbra

lesions or a DWI lesion without hypoperfusion (gures 3 and 4). These magnetic resonance patterns are consistent with the above described PET patterns. In the absence of data from prospective randomised controlled trials, the matched DWIPWI lesion pattern causes a management dilemma because it is unclear if the entire DWI lesion is already irreversibly damaged or whether it still contains a sizeable penumbra.111 There may therefore be an argument for reperfusion therapy, but the size of both the DWI and the PWI lesions, the magnetic resonance angiography ndings, and the clinical presentation can aect decisions. For instance, a matched pattern with a large DWI lesion, a severe neurological decit, and a proximal vessel occlusion predicts the development of a malignant MCA infarction.114 The third pattern of normal (or increased) perfusion with a variable size DWI lesion is indicative of spontaneous recanalisation and hence is inappropriate for thrombolysis.91,115

Direct thrombus visualisation

Analogous to hyperdensity on CT, signal change due to thrombotic occlusion of a vessel can be seen on FLAIR or gradient-echo MRI. These changes are of diagnostic value but are of no independent prognostic signicance.117 However, recent evidence suggests this sign is predictive of brin-rich emobli, as opposed to platelet-rich, and is independently predictive of subsequent recanalisation.118

CT perfusion
Technical features
Bolus-tracking CT perfusion119 also generates maps of cerebral blood ow and volume and time to peak or mean transit time.120 Most multidetector scanners restrict anatomical coverage typically to 20 mm (24 slices), reducing sensitivity to stroke not caused by large artery occlusion and preventing the full extent of perfusion changes to be imaged in MCA occlusion.121123 The lack of direct visualisation of the tissue with acute cellular injury is another serious drawback of CT perfusion as compared with MRI. The technique is similar in execution and analysis to MRI-based PWI but it provides more robust physiological values.124

Use of acute stroke MRI in clinical trials

Several DWI-based paradigms have been incorporated into randomised controlled trials. Several neuroprotectant trials included an MRI substudy with DWI-lesion expansion as a surrogate marker. The citicholine trials included more than 100 patients and supported the ability of DWI lesion expansion to act as a surrogate.76 The hypothesis that mismatch is associated with at-risk tissue is supported by two trials of the thrombolytic drug desmoteplase. The drug was given 39 h after stroke onset (only to patients with a DWIPWI mismatch 20%) and was shown to be potentially eective in improving clinical outcomes.3,116 The RESCUE study uses acute stroke MRI to select patients into a mechanical thrombectomy versus standard therapy trial. The ongoing EPITHET trial randomised all patients to alteplase or placebo 36 h after stroke onset regardless of baseline MRI ndings, but with the underlying hypothesis that patients with mismatch will benet more than those without, and thus may provide the most rigorous assessment.
http://neurology.thelancet.com Vol 5 September 2006

Mapping the penumbra and core

CT-perfusion lesion maps have been assessed for prediction of later infarct and clinical outcome (gure 5).12,125,126 Relative cerebral blood ow volume thresholds (compared with the contralateral hemisphere) in hypoperfused tissue, destined to infarct or to survive, are similar to those described for single photon emission CT or PET123,124,126 and therefore might allow denition of the penumbra. The extent of decits on cerebral blood ow or volume lesion maps has poorer interobserver agreement than mean transit time.127129 Mean-transittime and time-to-peak maps are more sensitive to ischaemia than plain CT or cerebral blood ow or volume maps.129 Mean transit time126,127 and time-to-peak15 lesion volumes are closely associated with nal infarct volume in patients who do not recanalise,15 or who reperfuse.126 Time-to-peak and mean-transit-time lesions are larger
761

Review

Cerebral blood ow 60

Cerebral blood volume 6

Time to peak 20

Day 1

60

20

Day 2

Outcome T1-weighted MRI

Figure 5: CT perfusion maps of the same patient with left MCA stroke as shown in gure 2 This gure shows extensive mismatch between small striato-capsular cerebral blood volume core (red arrows) and cortical hypoperfusion (cerebral blood ow: left) with prolonged time to peak (yellow arrows) in the acute stage. After successful reperfusion, there is full normalisation of cerebral blood ow, a residual area of reduced cerebral blood volume in the striatum, and shortened time to peak indicating hyperperfusion at day 2. Outcome MRI shows an infarct aecting only the area with the greatest reduction in cerebral blood volume at day 1. This patient made a complete clinical recovery (National Institutes of Health stroke scale=0 at 3 months).

than either cerebral blood ow or volume lesions in stroke within 36 h of onset that are associated with arterial occlusion.15,130 Wintermark and colleagues12,130 hypothesised, based on PET data and animal studies, that if cerebral blood volume is below a threshold of 25 mL/100 mg this signied the ischaemic core and a reduced relative cerebral blood ow below 64% would dene tissue at risk; the dierence between these two lesions therefore represents penumbra.12,130 These studies and others showed that the core dened by the cerebral blood
762

volume threshold closely matched DWI lesion volume and128,130 penumbral tissue plus core on CT perfusion matched the PWI mean-transit-time lesion.130 The proportion of penumbra not progressing to infarction was predictive of clinical recovery.12,15 Threshold values for mean transit time and cerebral blood ow were found to be reasonably robust in a moderately large series.131 Additional studies have conrmed good correlations of cerebral blood ow values on CT-perfusion maps compared with PET132 or xenon CT,133 although quantitative perfusion measurements remain problematic.124
http://neurology.thelancet.com Vol 5 September 2006

Review

CT angiography of source images

An alternative approach uses the source images of CT angiography to produce whole brain perfused cerebral blood volume-weighted images.61,134 Relative cerebral blood volume values obtained by this method can predict subsequent infarction within 6 h.135 This method is attractive because it only requires a single contrast bolus for both perfusion and CT angiography; however, full validation of this technique is lacking.

The 36 h interval
In the 36 h period after onset of stroke, treatment by revascularisation is supported by more limited evidence: randomised controlled trials of intravenous alteplase have not conrmed eectiveness, although a meta-analysis suggests substantial benet until 45 h,144 which forms the basis of the ECASS III trial. A single intra-arterial trial (of prourokinase) used catheter angiography to select patients with MCA occlusion and reported positive results.145 Analysis of the CT images of these patients showed the usefulness of reperfusion in patients who had a high ASPECTS score (<7) and little in the way of developed injury, whereas there was no benet if the injury had already developed.146 Therefore there is support showing the value of reperfusion in the 36 h window based on appropriate selection of patients. Although ideally, patients in this time window should be oered randomisation into a randomised controlled trial, not all centres are involved in trials; in this situation, physiological imaging may assist in treatment decisions. The body of evidence for MRI variables, particularly DWIPWI mismatch, is greater than for CT-perfusion maps at present. Two studies14,147 compared treatment with intravenous alteplase (within 36 h) based on the presence of DWIPWI mismatch with standard alteplase treatment in the rst 3 h with conventional plain CT criteria. In both studies, functional outcome was similar in the two groups; Ribo and colleagues14,147 also reported similar recanalisation rates and neurological improvement. Thomalla and colleagues,148 comparing their results with prior meta-analysis, found more favourable outcome and reduced rates of symptomatic intracerebral haemorrhage in their cohort of MRI-selected patients within 6 h. Further supporting the idea that MRI can eectively select patients that will probably benet up to 6 h after onset is a strong association between clinical improvement and the salvage of mismatch tissue.149,150 Because CT is not as sensitive as DWI for detection of lacunar or posterior fossa stroke, MRI has clear advantages over CT perfusion unless time is a constraint or the diagnosis of MCA stroke is clinically denite. CT perfusion can be used instead if the patient is unsuitable for MRI, and in future may be preferred if the penumbra dened by CT perfusion is validated in larger series.

Implementation of acute stroke imaging

Assuming equal access to DWIPWI and CT perfusion and expertise in their interpretation, decisions on optimum imaging use can be dictated by time elapsed since onset of stroke and severity of symptoms.

The 03 h interval
Within 3 h of onset, the most important decision is the eligibility for intravenous alteplase. All randomised controlled trials of patients being treated with alteplase use plain CT and this remains the typical method of acute MCA stroke decision-making in most centres. Based on earlier studies that associated anatomically extensive hypodensity with poor outcome, the concept that hypodensity greater than one-third of MCA territory should exclude patients from alteplase therapy as it represents extensive core or inaccurate time of onset has been used as a selection criterion in some randomised controlled trials (ECASS I and II). However, results from the 03 h group in the NINDS cohort do not support this exclusion from intravenous alteplase on the basis of the extent of early ischaemic changes alone;136 this issue is still the subject of debate. Although MRI oers equivalent detection of acute intracerebral haemorrhage,32,33 superior diagnostic conrmation of ischaemia, and sensitive detection of old haemorrhage, there are more restrictions regarding its application safety to acutely ill patients.137,138 It remains unclear if the risk of remote parenchymal haemorrhage after intravenous thrombolysis is associated with microbleeds (detectable on gradient-echo MRI).139,140 Screening for eligibility for intravenous thrombolysis within 3 h of stroke with MRI was feasible in an experienced centre, but at the expense of about a 20 min delay in treatment.141 If we believe that time is brain, then plain CT might be the preferred method. However, proponents of MRI argue that time lost may be compensated by improved exclusion of patients who may be harmed by treatment, a concept that is supported by lower haemorrhagic rates in MRI-selected series using a 06 h window,80 although not yet formally in a prospective randomised controlled trial. Furthermore, if it is conrmed that plain CT can be omitted, this delay may be reduced in the future, while current advances in MRI deliver a complete study within 10 min.142 Other studies report the safety and feasibility of MRI screening in acute stroke, and that shorter door-to-treatment times are achievable by increasing the familiarity of emergency physicians with its use.143
http://neurology.thelancet.com Vol 5 September 2006

Other scenarios
For stroke presenting more than 6 h after onset or minor symptoms, the sensitivity and specicity of DWI supports this as the optimum imaging modality. Imaging can also be of particular value in awakening stroke which represents about 25% of ischaemic strokes. These patients are usually excluded from clinical trials, yet if current thinking is correct, many occur shortly before waking and are therefore potentially within the time window for thrombolytic therapy.151 A recent report of two patients with unknown time of onset in whom decision to administer alteplase was guided by CT perfusion would support this idea.152
763

Review

Acute focal neurological decit

Plain CT (or T2* GRE/susceptibility-weighted MRI)

No evidence of acute haemorrhage

Acute haemorrhage

between clinical severity and small DWI lesion.157 This algorithm does accurately predict the presence of DWIPWI mismatch, but with low sensitivity.158 In a pilot study,159 the presence of a clinical-CT mismatch (ie, a disproportion between clinical scores and extent of abnormalities on plain CT) had no association with the likelihood of beneting from thrombolysis. However, no distinction was made between hypoattenuated and swollen tissue in this study.

DWI/PWI/MRA (or CT perfusion/CTA)

No thrombolysis

Lacunar stroke syndromes and posterior circulation strokes

Lacunar stroke syndromes
No signicant mismatch

Signicant PWIDWI mismatch (or CT perfusion mismatch)

MCA proximal/branch occlusion

Distal ICA T occlusion

Hypoperfusion present

No hypoperfusion

Consider IV/IA thrombolysis

Consider IA or combined IV/IA thrombolysis or mechanical thrombectomy

No thrombolysis

Extensive DWI lesion

No extensive DWI lesion

Consider brain decompression

Consider thrombolysis

Figure 6: Possible imaging-based reference framework for the management of acute anterior circulation stroke within 6 h of symptom onset

Although the clinical diagnosis of lacunar stroke within the rst 6 h is unreliable,160 it is improved if DWI shows a congruent deep small lesion suggestive of single perforator occlusion. Conversely, when several DWI lesions and a clinical source of emboli are present, the chance that the lacunar syndrome represents embolic stroke is high.161 A diagnosis of lacunar infarction in the acute stage has undecided implications for management. The NINDS study found no dierence in benet from intravenous alteplase in lacunar versus other stroke subtypes within 3 h of onset;162 however this was only based on clinical syndromes and CT. Because spontaneous recovery from lacunar infarcts is generally good,4 and the risk of intracerebral haemorrhage can increase due to associated small vessel disease,163 denitive conrmation of lacunar infarction may change the risk:benet ratio. This hypothesis remains to be tested.

Massive MCA infarction (ie, malignant MCA stroke), carries a high risk of brain herniation153 and a mortality hazard as high as 80%.153 Decompressive craniectomy is commonly considered because early surgical intervention is associated with better outcome to deferred intervention in young patients.154 There are no good clinical predictors of progression to coma and death, but imaging seems very promising. Both PET29 and ethyl-cysteinate dimer SPECT155 accurately predict the advent of malignant MCA stroke. Signicant DWIPWI predictors include large apparent diusion coecient lesion (>82 mL), large time to peak delay (more than 4 s) volume (>162 mL), small mismatch, and National Institutes of Health Stroke Scale >19.114 A multicentre study is currently underway in Germany that is prospectively assessing these parameters. Recently, the rate of expansion of the DWI lesion (ie, the lesion volume divided by the time since stroke onset) was reported to have almost 100% accuracy in a preliminary report.156 Figure 6 shows a possible imaging-based reference framework for acute stroke in the 06 h time window.

Posterior circulation strokes

Sensitivity of DWI is better than CT in brainstem and cerebellar (and lacunar) strokes. Studies using physiological imaging have been limited in this scenario. Using multimodal MRI, Ostrem and colleagues164 studied ve patients with acute basilar artery occlusion and found substantial volumes of diusionperfusion mismatch, consistent with the apparent benet of intravenous alteplase in posterior circulation stroke up to 7 h after onset.165,166 CT perfusion is probably unhelpful in posterior circulation stroke due to bony artifacts. Vascular imaging by CT angiography or magnetic resonance angiography and fat-suppressed T1-weighted neck views are valuable in identifying, for example, basilar occlusion, large-vessel atherosclerotic disease or extracranial arterial dissection, which may alter management.

Conclusions
Structural, vascular, and physiological imaging of acute stroke increasingly informs both clinical trial design and individual patient management. It is likely that both CT and MRI-based techniques will be more widely applied in future, and the relative strengths and weaknesses of each imaging modality should be regarded as complementary rather than competing. Eectively, an ideal situation would be to have access to
http://neurology.thelancet.com Vol 5 September 2006

Alternative approaches
Recently, alternatives to the full use of physiological imaging have been proposed. In the clinical-DWI mismatch approach, which avoids the use of PWI, the presence of salvageable tissue is implied from a mismatch
764

Review

Search strategy and selection criteria References for this review were identied by searches of PubMed from 1980 to 20 May, 2006. References from relevant articles and numerous articles were also identied through searches of les of the authors. The search terms acute stroke, PET, Perfusion, SPECT, CT perfusion, Xenon CT, ADC, diusion weighted imaging and ASPECTS were used. Only papers published in English were reviewed. The nal reference list was generated based on originality and relevance to the topics covered in the review.

15

16 17

18

19

both imaging modalities to adjust to the various clinical situations and contraindications that present in the real world.
Contributors KM contributed to layout, literature search, wrote the rst draft, and contributed gures. AB and RvK contributed to the writing of the third draft and nal version. J R contributed to the writing of third draft and nal version and contributed gures. J-CB coordinated the project, drafted the layout, did the literature search, contributed to the writing of the rst and second drafts, contributed gures, edited the third draft and nal version. Conict of interest We have no conicts of interest. References 1 Baron JC, von Kummer R, del Zoppo GJ. Treatment of acute ischemic stroke: challenging the concept of a rigid and universal time window. Stroke 1995; 26: 221921. 2 Muir KW. Heterogeneity of stroke pathophysiology and neuroprotective clinical trial design. Stroke 2002; 33: 154550. 3 Hacke W, Albers G, Al-Rawi Y, et al. The Desmoteplase in Acute Ischemic Stroke Trial (DIAS): a phase II MRI-based 9-hour window acute stroke thrombolysis trial with intravenous desmoteplase. Stroke 2005; 36: 6673. 4 Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. Classication and natural history of clinically identiable subtypes of cerebral infarction. Lancet 1991; 337: 152126. 5 Astrup J, Siesjo BK, Symon L. Thresholds in cerebral ischemia: the ischemic penumbra. Stroke 1981; 12: 72325. 6 Lassen N, Olsen T, Hojgaard K, Skriver E. Incomplete Infarction: a CT-negative irreversible ischemic brain lesion. J Cereb Blood Flow Metab 1983 3: S60203. 7 Baron JC. Mapping the ischaemic penumbra with PET: implications for acute stroke treatment. Cerebrovasc Dis 1999; 9: 193201. 8 Heiss WD, Rosner G. Functional recovery of cortical neurons as related to degree and duration of ischemia. Ann Neurol 1983; 14: 294301. 9 Schuier FJ, Hossmann KA. Experimental brain infarcts in cats II: ischemic brain edema. Stroke 1980; 11: 593601. 10 Todd NV, Picozzi P, Crockard A, Russell RW. Duration of ischemia inuences the development and resolution of ischemic brain edema. Stroke 1986; 17: 46671. 11 Furlan M, Marchal G, Viader F, Derlon JM, Baron JC. Spontaneous neurological recovery after stroke and the fate of the ischemic penumbra. Ann Neurol 1996; 40: 21626. 12 Wintermark M, Reichhart M, Thiran JP, et al. Prognostic accuracy of cerebral blood ow measurement by perfusion computed tomography, at the time of emergency room admission, in acute stroke patients. Ann Neurol 2002; 51: 41732. 13 Markus R, Reutens DC, Kazui S, et al. Hypoxic tissue in ischaemic stroke: persistence and clinical consequences of spontaneous survival. Brain 2004; 127: 142736. 14 Rother J, Schellinger PD, Gass A, et al. Eect of intravenous thrombolysis on MRI parameters and functional outcome in acute stroke <6 hours. Stroke 2002; 33: 243845.

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

Muir KW, Halbert HM, Baird TA, McCormick M, Teasdale E. Visual evaluation of perfusion computed tomography in acute stroke accurately estimates infarct volume and tissue viability. J Neurol Neurosurg Psychiatr 2006; 77: 33439. Baron JC. Mapping the ischaemic penumbra with PET: a new approach. Brain 2001; 124: 24. Heiss WD, Kracht LW, Thiel A, Grond M, Pawlik G. Penumbral probability thresholds of cortical umazenil binding and blood ow predicting tissue outcome in patients with cerebral ischaemia. Brain 2001; 124: 2029. Wise RJ, Bernardi S, Frackowiak RS, Legg NJ, Jones T. Serial observations on the pathophysiology of acute stroke: the transition from ischaemia to infarction as reected in regional oxygen extraction. Brain 1983; 106: 197222. Marchal G, Serrati C, Rioux P, et al. PET imaging of cerebral perfusion and oxygen consumption in acute ischaemic stroke: relation to outcome. Lancet 1993; 341: 92527. Marchal G, Benali K, Iglesias S, Viader F, Derlon JM, Baron JC. Voxel-based mapping of irreversible ischaemic damage with PET in acute stroke. Brain 1999; 122: 2387400. Marchal G, Beaudouin V, Rioux P, et al. Prolonged persistence of substantial volumes of potentially viable brain tissue after stroke: a correlative PET-CT study with voxel-based data analysis. Stroke 1996; 27: 599606. Marchal G, Furlan M, Beaudouin V, et al. Early spontaneous hyperperfusion after stroke. A marker of favourable tissue outcome? Brain 1996; 119: 40919. Read SJ, Hirano T, Abbott DF, et al. The fate of hypoxic tissue on 18F-uoromisonidazole positron emission tomography after ischemic stroke. Ann Neurol 2000; 48: 22835. Read SJ, Hirano T, Abbott DF, et al. Identifying hypoxic tissue after acute ischemic stroke using PET and 18F-uoromisonidazole. Neurology 1998; 51: 161721. Jaillard A, Martin CD, Garambois K, Lebas JF, Hommel M. Vicarious function within the human primary motor cortex? A longitudinal fMRI stroke study. Brain 2005; 128: 112238. Cramer SC, Shah R, Juranek J, Crafton KR, Le V. Activity in the peri-infarct rim in relation to recovery from stroke. Stroke 2006; 37: 11115. Baron JC, Bousser MG, Rey A, Guillard A, Comar D, Castaigne P. Reversal of focal misery-perfusion syndrome by extra-intracranial arterial bypass in hemodynamic cerebral ischemia: a case study with 15O positron emission tomography. Stroke 1981; 12: 45459. Marchal G, Young AR, Baron JC. Early postischemic hyperperfusion: pathophysiologic insights from positron emission tomography. J Cereb Blood Flow Metab 1999; 19: 46782. Marchal G, Rioux P, Serrati C, et al. Value of acute-stage positron emission tomography in predicting neurological outcome after ischemic stroke: further assessment. Stroke 1995; 26: 52425. Gonzalez RG, Schaefer PW, Buonanno FS, et al. Diusion-weighted MR imaging: diagnostic accuracy in patients imaged within 6 hours of stroke symptom onset. Radiology 1999; 210: 15562. Mullins ME, Schaefer PW, Sorensen AG, et al. CT and conventional and diusion-weighted MR imaging in acute stroke: study in 691 patients at presentation to the emergency department. Radiology 2002; 224: 35360. Kidwell CS, Chalela JA, Saver JL, et al. Comparison of MRI and CT for detection of acute intracerebral hemorrhage. JAMA 2004; 292: 182330. Fiebach JB, Schellinger PD, Gass A, et al. Stroke magnetic resonance imaging is accurate in hyperacute intracerebral hemorrhage: a multicenter study on the validity of stroke imaging. Stroke 2004; 35: 50206. Alexandrov AV, Molina CA, Grotta JC, et al. Ultrasound-enhanced systemic thrombolysis for acute ischemic stroke. N Engl J Med 2004; 351: 217078. Barber PA, Demchuk AM, Zhang J, Buchan AM. Validity and reliability of a quantitative computed tomography score in predicting outcome of hyperacute stroke before thrombolytic therapy. ASPECTS Study Group. Alberta Stroke Programme Early CT Score. Lancet 2000; 355: 167074. Butcher K, Lee S, Parsons M, et al. Increased blood volume maintains viability in tissue and isolated focal swelling on CT in acute stroke. Stroke 2005; 36: 418 (abstr).

http://neurology.thelancet.com Vol 5 September 2006

765

Review

37

38

39

40

41

42

43

44

45

46 47

48

49

50

51

52

53

54

55

56

57

58

Kucinski T, Vaterlein O, Glauche V, et al. Correlation of apparent diusion coecient and computed tomography density in acute ischemic stroke. Stroke 2002; 33: 178691. Kucinski T, Majumder A, Knab R, et al. Cerebral perfusion impairment correlates with the decrease of CT density in acute ischaemic stroke. Neuroradiology 2004; 46: 71622. von Kummer R, Bourquain H, Bastianello S, et al. Early prediction of irreversible brain damage after ischemic stroke at CT. Radiology 2001; 219: 95100. Na DG, Kim EY, Ryoo JW, et al. CT sign of brain swelling without concomitant parenchymal hypoattenuation: comparison with diusion- and perfusion-weighted MR imaging. Radiology 2005; 235: 99248. von Kummer R, Holle R, Gizyska U, et al. Interobserver agreement in assessing early CT signs of middle cerebral artery infarction. Am J Neuroradiol 1996; 17: 174348. Marks MP, Holmgren EB, Fox AJ, Patel S, von Kummer R, Froehlich J. Evaluation of early computed tomographic ndings in acute ischemic stroke. Stroke 1999; 30: 38992. Saur D, Kucinski T, Grzyska U, et al. Sensitivity and interrater agreement of CT and diusion-weighted MR imaging in hyperacute stroke. Am J Neuroradiol 2003; 24: 87885. Grotta JC, Chiu D, Lu M, et al. Agreement and variability in the interpretation of early CT changes in stroke patients qualifying for intravenous rtPA therapy. Stroke 1999; 30: 152833. Hirano T, Yonehara T, Inatomi Y, Hashimoto Y, Uchino M. Presence of early ischemic changes on computed tomography depends on severity and the duration of hypoperfusion: a single photon emission-computed tomographic study. Stroke 2005; 36: 260108. Coutts SB, Demchuk AM, Barber PA, et al. Interobserver variation of ASPECTS in real time. Stroke 2004; 35: e10305. Coutts SB, Hill MD, Demchuk AM, Barber PA, Pexman JH, Buchan AM. ASPECTS reading requires training and experience. Stroke 2003; 34: e179. Patel SC, Levine SR, Tilley BC, et al. Lack of clinical signicance of early ischemic changes on computed tomography in acute stroke. JAMA 2001; 286: 283038. Dzialowski I, Hill MD, Coutts SB, et al. Extent of early ischemic changes on computed tomography (CT) before thrombolysis: prognostic value of the Alberta Stroke Program Early CT Score in ECASS II. Stroke 2006; 37: 97378. Hacke W, Kaste M, Fieschi C, et al. Randomised double-blind placebo-controlled trial of thrombolytic therapy with intravenous alteplase in acute ischaemic stroke (ECASS II). Second EuropeanAustralasian Acute Stroke Study Investigators. Lancet 1998; 352: 124551. von Kummer R, Meyding-Lamade U, Forsting M, et al. Sensitivity and prognostic value of early CT in occlusion of the middle cerebral artery trunk. AJNR Am J Neuroradiol 1994; 15: 915. Tomsick T, Brott T, Barsan W, Broderick J, Haley EC, Spilker J. Thrombus localization with emergency cerebral CT. Am J Neuroradiol 1992; 13: 25763. Barber PA, Demchuk AM, Hill MD, et al. The probability of middle cerebral artery MRA ow signal abnormality with quantied CT ischaemic change: targets for future therapeutic studies. J Neurol Neurosurg Psychiatr 2004; 75: 142630. Wolpert SM, Bruckmann H, Greenlee R, Wechsler L, Pessin MS, del Zoppo GJ. Neuroradiologic evaluation of patients with acute stroke treated with recombinant tissue plasminogen activator. The rtPA Acute Stroke Study Group. AJNR Am J Neuroradiol 1993; 14: 313. Kirchhof K, Welzel T, Mecke C, Zoubaa S, Sartor K. Dierentiation of white, mixed, and red thrombi: value of CT in estimation of the prognosis of thrombolysis phantom study. Radiology 2003; 228: 12630. Qureshi AI, Ezzeddine MA, Nasar A, et al. Is IV tissue plasminogen activator benecial in patients with hyperdense artery sign? Neurology 2006; 66: 117174. Derex L, Hermier M, Adeleine P, et al. Clinical and imaging predictors of intracerebral haemorrhage in stroke patients treated with intravenous tissue plasminogen activator. J Neurol Neurosurg Psychiatr 2005; 76: 7075. Tomsick TA, Brott TG, Olinger CP, et al. Hyperdense middle cerebral artery: incidence and quantitative signicance. Neuroradiology 1989; 31: 31215.

59

60

61

62

63

64

65

66

67 68

69

70

71

72

73

74

75

76

77

78

79

Barber PA, Demchuk AM, Hudon ME, Pexman JH, Hill MD, Buchan AM. Hyperdense sylvian ssure MCA dot sign: a CT marker of acute ischemia. Stroke 2001; 32: 8488. Leary MC, Kidwell CS, Villablanca JP, et al. Validation of computed tomographic middle cerebral artery dotsign: an angiographic correlation study. Stroke 2003; 34: 263640. Ezzeddine MA, Lev MH, McDonald CT, et al. CT angiography with whole brain perfused blood volume imaging: added clinical value in the assessment of acute stroke. Stroke 2002; 33: 95966. Schramm P, Schellinger PD, Klotz E, et al. Comparison of perfusion computed tomography and computed tomography angiography source images with perfusion-weighted imaging and diusion-weighted imaging in patients with acute stroke of less than 6 hours duration. Stroke 2004; 35: 165258. Wunderlich MT, Stolz E, Seidel G, et al. Conservative medical treatment and intravenous thrombolysis in acute stroke from carotid T occlusion. Cerebrovasc Dis 2005; 20: 35561. Jansen O, von Kummer R, Forsting M, Hacke W, Sartor K. Thrombolytic therapy in acute occlusion of the intracranial internal carotid artery bifurcation. Am J Neuroradiol 1995; 16: 197786. Laub G. Principles of contrast-enhanced MR angiography: basic and clinical applications. Magn Reson Imaging Clin N Am 1999; 7: 78395. Wardlaw JM, Chappell FM, Best JJ, Wartolowska K, Berry E. Noninvasive imaging compared with intra-arterial angiography in the diagnosis of symptomatic carotid stenosis: a meta-analysis. Lancet 2006; 367: 150312. Baron JC. Stroke: imaging and dierential diagnosis. J Neural Transm Suppl 2002; 63: 1936. Lin W, Lee JM, Lee YZ, Vo KD, Pilgram T, Hsu CY. Temporal relationship between apparent diusion coecient and absolute measurements of cerebral blood ow in acute stroke patients. Stroke 2003; 34: 6470. Wang Y, Hu W, Perez-Trepichio AD, et al. Brain tissue sodium is a ticking clock telling time after arterial occlusion in rat focal cerebral ischemia. Stroke 2000; 31: 138691. Hoehn-Berlage M, Norris DG, Kohno K, Mies G, Leibfritz D, Hossmann KA. Evolution of regional changes in apparent diusion coecient during focal ischemia of rat brain: the relationship of quantitative diusion NMR imaging to reduction in cerebral blood ow and metabolic disturbances. J Cereb Blood Flow Metab 1995; 15: 100211. Busza AL, Allen KL, King MD, van Bruggen N, Williams SR, Gadian DG. Diusion-weighted imaging studies of cerebral ischemia in gerbils: potential relevance to energy failure. Stroke 1992; 23: 160212. Guadagno JV, Jones PS, Fryer TD, et al. Local relationships between restricted water diusion and oxygen consumption in the ischemic human brain. Stroke 2006; 37: 174148. Moseley ME, Kucharczyk J, Mintorovitch J, et al. Diusion-weighted MR imaging of acute stroke: correlation with T2-weighted and magnetic susceptibility-enhanced MR imaging in cats. Am J Neuroradiol 1990; 11: 42329. Warach S, Gaa J, Siewert B, Wielopolski P, Edelman RR. Acute human stroke studied by whole brain echo planar diusionweighted magnetic resonance imaging. Ann Neurol 1995; 37: 23141. Hjort N, Christensen S, Solling C, et al. Ischemic injury detected by diusion imaging 11 minutes after stroke. Ann Neurol 2005; 58: 46265. Warach S, Pettigrew LC, Dashe JF, et al. Eect of citicoline on ischemic lesions as measured by diusion-weighted magnetic resonance imaging. Citicoline 010 Investigators. Ann Neurol 2000; 48: 71322. Schlaug G, Siewert B, Beneld A, Edelman RR, Warach S. Time course of the apparent diusion coecient (ADC) abnormality in human stroke. Neurology 1997; 49: 11319. Lansberg MG, Thijs VN, OBrien MW, et al. Evolution of apparent diusion coecient, diusion-weighted, and T2-weighted signal intensity of acute stroke. Am J Neuroradiol 2001; 22: 63744. Knight RA, Dereski MO, Helpern JA, Ordidge RJ, Chopp M. Magnetic resonance imaging assessment of evolving focal cerebral ischemia. Comparison with histopathology in rats. Stroke 1994; 25: 125261.

766

http://neurology.thelancet.com Vol 5 September 2006

Review

Fiebach JB, Schellinger PD, Jansen O, et al. CT and diusionweighted MR imaging in randomized order: diusion-weighted imaging results in higher accuracy and lower interrater variability in the diagnosis of hyperacute ischemic stroke. Stroke 2002; 33: 220610. 81 Jaillard A, Hommel M, Baird AE, et al. Signicance of early CT signs in acute stroke. A CT scan-diusion MRI study. Cerebrovasc Dis 2002; 13: 4756. 82 Fiebach J, Jansen O, Schellinger P, et al. Comparison of CT with diusion-weighted MRI in patients with hyperacute stroke. Neuroradiology 2001; 43: 62832. 83 Lovblad KO, Laubach HJ, Baird AE, et al. Clinical experience with diusion-weighted MR in patients with acute stroke. Am J Neuroradiol 1998; 19: 106166. 84 Baird AE, Beneld A, Schlaug G, et al. Enlargement of human cerebral ischemic lesion volumes measured by diusion-weighted magnetic resonance imaging. Ann Neurol 1997; 41: 58189. 85 Barber PA, Darby DG, Desmond PM, et al. Prediction of stroke outcome with echoplanar perfusion- and diusion-weighted MRI. Neurology 1998; 51: 41826. 86 Beaulieu C, de Crespigny A, Tong DC, Moseley ME, Albers GW, Marks MP. Longitudinal magnetic resonance imaging study of perfusion and diusion in stroke: evolution of lesion volume and correlation with clinical outcome. Ann Neurol 1999; 46: 56878. 87 Baird TA, Parsons MW, Phanh T, et al. Persistent poststroke hyperglycemia is independently associated with infarct expansion and worse clinical outcome. Stroke 2003; 34: 220814. 88 Allport LE, Parsons MW, Butcher KS, et al. Elevated hematocrit is associated with reduced reperfusion and tissue survival in acute stroke. Neurology 2005; 65: 138287. 89 Ay H, Koroshetz WJ, Vangel M, et al. Conversion of ischemic brain tissue into infarction increases with age. Stroke 2005; 36: 263236. 90 Singhal AB, Benner T, Roccatagliata L, et al. A pilot study of normobaric oxygen therapy in acute ischemic stroke. Stroke 2005; 36: 797802. 91 Barber PA, Davis SM, Darby DG, et al. Absent middle cerebral artery ow predicts the presence and evolution of the ischemic penumbra. Neurology 1999; 52: 112532. 92 Staroselskaya IA, Chaves C, Silver B, et al. Relationship between magnetic resonance arterial patency and perfusion-diusion mismatch in acute ischemic stroke and its potential clinical use. Arch Neurol 2001; 58: 106974. 93 Seitz RJ, Meisel S, Moll M, Wittsack HJ, Junghans U, Siebler M. Partial rescue of the perfusion decit area by thrombolysis. J Magn Reson Imaging 2005; 22: 199205. 94 Schellinger PD, Fiebach JB, Jansen O, et al. Stroke magnetic resonance imaging within 6 hours after onset of hyperacute cerebral ischemia. Ann Neurol 2001; 49: 46069. 95 Jansen O, Schellinger P, Fiebach J, Hacke W, Sartor K. Early recanalisation in acute ischaemic stroke saves tissue at risk dened by MRI. Lancet 1999; 353: 203637. 96 Kraemer N, Thomalla G, Soennichsen J, et al. Magnetic resonance imaging and clinical patterns of patients with spectacular shrinking decit after acute middle cerebral artery stroke. Cerebrovasc Dis 2005; 20: 28590. 97 Baird AE, Lovblad KO, Dashe JF, et al. Clinical correlations of diusion and perfusion lesion volumes in acute ischemic stroke. Cerebrovasc Dis 2000; 10: 44148. 98 Schlaug G, Beneld A, Baird AE, et al. The ischemic penumbra: operationally dened by diusion and perfusion MRI. Neurology 1999; 53: 152837. 99 Rother J. CT and MRI in the diagnosis of acute stroke and their role in thrombolysis. Thromb Res 2001; 103 (suppl 1): S12533. 100 Kidwell CS, Saver JL, Mattiello J, et al. Thrombolytic reversal of acute human cerebral ischemic injury shown by diusion/ perfusion magnetic resonance imaging. Ann Neurol 2000; 47: 46269. 101 Rother J, de Crespigny AJ, DArceuil H, Iwai K, Moseley ME. Recovery of apparent diusion coecient after ischemia-induced spreading depression relates to cerebral perfusion gradient. Stroke 1996; 27: 98086. 102 Fiehler J, Knudsen K, Kucinski T, et al. Predictors of apparent diusion coecient normalization in stroke patients. Stroke 2004; 35: 51419.

80

103 Seitz RJ, Meisel S, Weller P, Junghans U, Wittsack HJ, Siebler M. Initial ischemic event: perfusion-weighted MR imaging and apparent diusion coecient for stroke evolution. Radiology 2005; 237: 102028. 104 Kidwell CS, Saver JL, Starkman S, et al. Late secondary ischemic injury in patients receiving intraarterial thrombolysis. Ann Neurol 2002; 52: 698703. 105 Baron JC. How healthy is the acutely reperfused ischemic penumbra? Cerebrovasc Dis 2005; 20 (Suppl 2): 2531. 106 Rivers CS, Wardlaw JM, Armitage PA, et al. Do acute diusion- and perfusion-weighted MRI lesions identify nal infarct volume in ischemic stroke? Stroke 2006; 37: 98104. 107 Thijs VN, Somford DM, Bammer R, Robberecht W, Moseley ME, Albers GW. Inuence of arterial input function on hypoperfusion volumes measured with perfusion-weighted imaging. Stroke 2004; 35: 9498. 108 Schellinger PD, Latour LL, Wu CS, Chalela JA, Warach S. The association between neurological decit in acute ischemic stroke and mean transit time Comparison of four dierent perfusion MRI algorithms. Neuroradiology 2006; 48: 6977. 109 Butcher KS, Parsons M, MacGregor L, et al. Rening the perfusiondiusion mismatch hypothesis. Stroke 2005; 36: 115359. 110 Shih LC, Saver JL, Alger JR, et al. Perfusion-weighted magnetic resonance imaging thresholds identifying core, irreversibly infarcted tissue. Stroke 2003; 34: 142530. 111 Guadagno JV, Warburton EA, Aigbirhio FI, et al. Does the acute diusion-weighted imaging lesion represent penumbra as well as core? A combined quantitative PET/MRI voxel-based study. J Cereb Blood Flow Metab 2004; 24: 124954. 112 Sobesky J, Zaro Weber O, Lehnhardt FG, et al. Does the mismatch match the penumbra? Magnetic resonance imaging and positron emission tomography in early ischemic stroke. Stroke 2005; 36: 98085. 113 Sobesky J, Zaro Weber O, Lehnhardt FG, et al. Which time-to-peak threshold best identies penumbral ow? A comparison of perfusion-weighted magnetic resonance imaging and positron emission tomography in acute ischemic stroke. Stroke 2004; 35: 284347. 114 Thomalla GJ, Kucinski T, Schoder V, et al. Prediction of malignant middle cerebral artery infarction by early perfusion- and diusionweighted magnetic resonance imaging. Stroke 2003; 34: 189299. 115 Kidwell CS, Saver JL, Mattiello J, et al. Diusion-perfusion MRI characterization of post-recanalization hyperperfusion in humans. Neurology 2001; 57: 201521. 116 Furlan AJ, Eyding D, Albers GW, et al. Dose Escalation of Desmoteplase for Acute Ischemic Stroke (DEDAS): evidence of safety and ecacy 3 to 9 hours after stroke onset. Stroke 2006; 37: 122731. 117 Schellinger PD, Chalela JA, Kang DW, Latour LL, Warach S. Diagnostic and prognostic value of early MR Imaging vessel signs in hyperacute stroke patients imaged <3 hours and treated with recombinant tissue plasminogen activator. Am J Neuroradiol 2005; 26: 61824. 118 Cho KH, Kim JS, Kwon SU, Cho AH, Kang DW. Signicance of susceptibility vessel sign on T2*-weighted gradient echo imaging for identication of stroke subtypes. Stroke 2005; 36: 237983. 119 Axel L. Cerebral blood ow determination by rapid-sequence computed tomography: theoretical analysis. Radiology 1980; 137: 67986. 120 Nagata K, Asano T. Functional image of dynamic computed tomography for the evaluation of cerebral hemodynamics. Stroke 1990; 21: 88289. 121 Koenig M, Klotz E, Luka B, Venderink DJ, Spittler JF, Heuser L. Perfusion CT of the brain: diagnostic approach for early detection of ischemic stroke. Radiology 1998; 209: 8593. 122 Maruya J, Yamamoto K, Ozawa T, et al. Simultaneous multi-section perfusion CT and CT angiography for the assessment of acute ischemic stroke. Acta Neurochir (Wien) 2005; 147: 38391. 123 Rother J, Jonetz-Mentzel L, Fiala A, et al. Hemodynamic assessment of acute stroke using dynamic single-slice computed tomographic perfusion imaging. Arch Neurol 2000; 57: 116166. 124 Klotz E, Koenig M. Perfusion measurements of the brain: using dynamic CT for the quantitative assessment of cerebral ischemia in acute stroke. Eur J Radiol 1999; 30: 17084.

http://neurology.thelancet.com Vol 5 September 2006

767

Review

125 Nabavi DG, Kloska SP, Nam EM, et al. MOSAIC: Multimodal Stroke Assessment Using Computed Tomography: novel diagnostic approach for the prediction of infarction size and clinical outcome. Stroke 2002; 33: 281926. 126 Parsons MW, Pepper EM, Chan V, et al. Perfusion computed tomography: prediction of nal infarct extent and stroke outcome. Ann Neurol 2005; 58: 67279. 127 Eastwood JD, Lev MH, Azhari T, et al. CT perfusion scanning with deconvolution analysis: pilot study in patients with acute middle cerebral artery stroke. Radiology 2002; 222: 22736. 128 Eastwood JD, Lev MH, Wintermark M, et al. Correlation of early dynamic CT perfusion imaging with whole-brain MR diusion and perfusion imaging in acute hemispheric stroke. Am J Neuroradiol 2003; 24: 186975. 129 Wintermark M, Fischbein NJ, Smith WS, Ko NU, Quist M, Dillon WP. Accuracy of dynamic perfusion CT with deconvolution in detecting acute hemispheric stroke. Am J Neuroradiol 2005; 26: 10412. 130 Wintermark M, Reichhart M, Cuisenaire O, et al. Comparison of admission perfusion computed tomography and qualitative diusion- and perfusion-weighted magnetic resonance imaging in acute stroke patients. Stroke 2002; 33: 202531. 131 Wintermark M, Flanders AE, Velthuis B, et al. Perfusion-CT assessment of infarct core and penumbra: receiver operating characteristic curve analysis in 130 patients suspected of acute hemispheric stroke. Stroke 2006; 37: 97985. 132 Kudo K, Terae S, Katoh C, et al. Quantitative cerebral blood ow measurement with dynamic perfusion CT using the vascular-pixel elimination method: comparison with H2(15)O positron emission tomography. AJNR Am J Neuroradiol 2003; 24: 41926. 133 Wintermark M, Thiran JP, Maeder P, Schnyder P, Meuli R. Simultaneous measurement of regional cerebral blood ow by perfusion CT and stable xenon CT: a validation study. AJNR Am J Neuroradiol 2001; 22: 90514. 134 Hunter GJ, Hamberg LM, Ponzo JA, et al. Assessment of cerebral perfusion and arterial anatomy in hyperacute stroke with three-dimensional functional CT: early clinical results. AJNR Am J Neuroradiol 1998; 19: 2937. 135 Hunter GJ, Silvennoinen HM, Hamberg LM, et al. Whole-brain CT perfusion measurement of perfused cerebral blood volume in acute ischemic stroke: probability curve for regional infarction. Radiology 2003; 227: 72530. 136 Demchuk AM, Hill MD, Barber PA, Silver B, Patel SC, Levine SR. Importance of early ischemic computed tomography changes using ASPECTS in NINDS rtPA Stroke Study. Stroke 2005; 36: 211015. 137 Hand PJ, Wardlaw JM, Rowat AM, Haisma JA, Lindley RI, Dennis MS. Magnetic resonance brain imaging in patients with acute stroke: feasibility and patient related diculties. J Neurol Neurosurg Psychiatr 2005; 76: 152527. 138 Barber PA, Hill MD, Eliasziw M, et al. Imaging of the brain in acute ischaemic stroke: comparison of computed tomography and magnetic resonance diusion-weighted imaging. J Neurol Neurosurg Psychiatr 2005; 76: 152833. 139 Kidwell CS, Saver JL, Villablanca JP, et al. Magnetic resonance imaging detection of microbleeds before thrombolysis: an emerging application. Stroke 2002; 33: 9598. 140 Kakuda W, Thijs VN, Lansberg MG, et al. Clinical importance of microbleeds in patients receiving IV thrombolysis. Neurology 2005; 65: 117578. 141 Kang DW, Chalela JA, Dunn W, Warach S. MRI screening before standard tissue plasminogen activator therapy is feasible and safe. Stroke 2005; 36: 193943. 142 U.-King-Im J, Trivedi RA, Graves MJ, et al. Utility of an ultrafast magnetic resonance imaging protocol in recent and semi-recent strokes. J Neurol Neurosurg Psychiatr 2005; 76: 100205. 143 Schellinger PD, Jansen O, Fiebach JB, et al. Feasibility and practicality of MR imaging of stroke in the management of hyperacute cerebral ischemia. AJNR Am J Neuroradiol 2000; 21: 118489. 144 Hacke W, Donnan G, Fieschi C, et al. Association of outcome with early stroke treatment: pooled analysis of ATLANTIS, ECASS, and NINDS rt-PA stroke trials. Lancet 2004; 363: 76874. 145 Furlan A, Higashida R, Wechsler L, et al. Intra-arterial prourokinase for acute ischemic stroke: the PROACT II study: a randomized controlled trial. Prolyse in Acute Cerebral Thromboembolism. JAMA 1999; 282: 200311.

146 Hill MD, Rowley HA, Adler F, et al. Selection of acute ischemic stroke patients for intra-arterial thrombolysis with pro-urokinase by using ASPECTS. Stroke 2003; 34: 192531. 147 Ribo M, Molina CA, Rovira A, et al. Safety and ecacy of intravenous tissue plasminogen activator stroke treatment in the 3 to 6 hour window using multimodal transcranial Doppler/MRI selection protocol. Stroke 2005; 36: 60206. 148 Thomalla G, Schwark C, Sobesky J, et al. Outcome and symptomatic bleeding complications of intravenous thrombolysis within 6 hours in MRI-selected stroke patients: comparison of a German multicenter study with the pooled data of ATLANTIS, ECASS, and NINDS tPA trials. Stroke 2006; 37: 85258. 149 Singer OC, Du Mesnil De Rochemont R, Foerch C, et al. Early functional recovery and the fate of the diusion/perfusion mismatch in patients with proximal middle cerebral artery occlusion. Cerebrovasc Dis 2004; 17: 1320. 150 Chalela JA, Kang DW, Luby M, et al. Early magnetic resonance imaging ndings in patients receiving tissue plasminogen activator predict outcome: insights into the pathophysiology of acute stroke in the thrombolysis era. Ann Neurol 2004; 55: 10512. 151 Todo K, Moriwaki H, Saito K, Tanaka M, Oe H, Naritomi H. Early CT Findings in Unknown-Onset and Wake-Up Strokes. Cerebrovasc Dis 2006; 21: 36771. 152 Hellier KD, Hampton JL, Guadagno JV, et al. Perfusion CT helps decision making for thrombolysis when there is no clear time of onset. J Neurol Neurosurg Psychiatr 2006; 77: 41719. 153 Hacke W, Schwab S, Horn M, Spranger M, De Georgia M, von Kummer R. Malignant middle cerebral artery territory infarction: clinical course and prognostic signs. Arch Neurol 1996; 53: 30915. 154 Schwab S, Steiner T, Ascho A, et al. Early hemicraniectomy in patients with complete middle cerebral artery infarction. Stroke 1998; 29: 188893. 155 Berrouschot J, Barthel H, von Kummer R, Knapp WH, Hesse S, Schneider D. 99m technetium-ethyl-cysteinate-dimer single-photon emission CT can predict fatal ischemic brain edema. Stroke 1998; 29: 255662. 156 Arenillas JF, Rovira A, Delgado-Mederos, et al. Speed of diusionweighted imaging lesion expansion within 3 hours from stroke onset predicts evolution to malignant MCA infarction. Stroke 2005; 36: 431 (abstr). 157 Davalos A, Blanco M, Pedraza S, et al. The clinical-DWI mismatch: a new diagnostic approach to the brain tissue at risk of infarction. Neurology 2004; 62: 218792. 158 Prosser J, Butcher K, Allport L, et al. Clinical-diusion mismatch predicts the putative penumbra with high specicity. Stroke 2005; 36: 170004. 159 Kent DM, Hill MD, Ruthazer R, et al. Clinical-CT mismatch and the response to systemic thrombolytic therapy in acute ischemic stroke. Stroke 2005; 36: 169599. 160 Toni D, Iweins F, von Kummer R, et al. Identication of lacunar infarcts before thrombolysis in the ECASS I study. Neurology 2000; 54: 68488. 161 Wessels T, Rottger C, Jauss M, Kaps M, Traupe H, Stolz E. Identication of embolic stroke patterns by diusion-weighted MRI in clinically dened lacunar stroke syndromes. Stroke 2005; 36: 75761. 162 Tissue plasminogen activator for acute ischemic stroke. The National Institute of Neurological Disorders and Stroke rt-PA Stroke Study Group. N Engl J Med 1995; 333: 158187. 163 Koennecke HC. Cerebral microbleeds on MRI: prevalence, associations, and potential clinical implications. Neurology 2006; 66: 16571. 164 Ostrem JL, Saver JL, Alger JR, et al. Acute basilar artery occlusion: diusion-perfusion MRI characterization of tissue salvage in patients receiving intra-arterial stroke therapies. Stroke 2004; 35: 3034. 165 Montavont A, Nighoghossian N, Derex L, et al. Intravenous r-TPA in vertebrobasilar acute infarcts. Neurology 2004; 62: 185456. 166 Lindsberg PJ, Soinne L, Tatlisumak T, et al. Long-term outcome after intravenous thrombolysis of basilar artery occlusion. JAMA 2004; 292: 186266.

768

http://neurology.thelancet.com Vol 5 September 2006