10
INTRODUCTION Stroke is the third-most-common cause of death in adults and is a leading cause of morbidity in Australia. 1 Conventional structural imaging typically comprises CT to exclude intra- cerebral haemorrhage. However, 85% of strokes are the result of ischaemic infarcts for which, in the majority of patients imaged in the first 6 h, CT is either normal or demonstrates only subtle abnormalities that are easy to misinterpret. Even conventional MR images are commonly normal in this early period. The development of thrombolytic and neuroprotective agents for the treatment of acute stroke has created an imperative for improved imaging techniques in the evaluation of acute stroke. The National Institute for Neurological Diseases and Stroke trial was the first study to demonstrate that thrombolysis in patients with no intracerebral haemorrhage leads to a one-third increase in clinical outcomes if administered within 3 hours of the onset of ictus. 2 However, for patients treated beyond the 3 h of onset, it carries a substantial risk of intracranial bleeding. This setback in the use of thrombolysis has prompted the realization that a more rigorous selection of patients might reduce the risk of intracranial bleeding, and that the use of thrombolysis might have a bene- fit in selected patients extending beyond the current 3-h therapeutic window. 3 There have been several studies to suggest that the level of residual blood flow in the ischaemic zone is a useful indicator of the risk of intracranial bleeding. 4–7 There are distinct thresholds of cerebral blood flow for various functions of the brain. It is possible to use these thresholds to determine whether a particular brain region is salvageable. Perfusion CT offers the ability to positively identify patients with non-haemorrhagic stroke, to select those cases where thrombolysis is appropriate, and to provide an indication as to prognosis. Pictorial Essay Computed tomography perfusion imaging in acute stroke CJ Keith, 1,2 M Griffiths, 2,3 B Petersen, 1 RJ Anderson 1 and KA Miles 1,3 1 Southern X-ray Clinics, 2 Wesley Research Institute, The Wesley Hospital, and 3 Centre for Medical Health and Environmental Physics, School of Physical Sciences, Queensland University of Technology, Brisbane, Queensland, Australia SUMMARY The development of thrombolytic and neuroprotective agents for the treatment of acute stroke has created an imperative for improved imaging techniques in the assessment of acute stroke. Five cases are presented to illustrate the value of perfusion CT in the evaluation of suspected acute stroke. To obtain the perfusion data, a rapid series of images was acquired without table movement following a bolus of contrast medium. Cerebral blood flow, cerebral blood volume and mean transit time were determined by mathematically modelling the temporal changes in contrast enhancement in the brain and vascular system. Pixel-by-pixel analysis allowed generation of perfusion maps. In two cases, CT-perfusion imaging usefully excluded acute stroke, including one patient in whom a low-density area on conventional CT was subsequently proven to be tumour. Cerebral ischaemia was confirmed in three cases, one with an old and a new infarction, one with a large conventional CT abnormality but only a small perfusion defect, and one demonstrating infarct and penumbra. Perfusion CT offers the ability to positively identify patients with non- haemorrhagic stroke in the presence of a normal conventional CT, to select those cases where thrombolysis is appropriate, and to provide an indication for prognosis. Key words: computed tomography, functional imaging, stroke. CJ Keith MB BS; M Griffiths BSc, MAppSci; B Petersen Radiographer; RJ Anderson MB, ChB, FRANZCR; KA Miles MB BS, FRCR, MSc, MD. Correspondence:Carolyn Keith, Southern X-ray Clinics, 2nd Floor, Day Centre, The Wesley Hospital, 45 Coronation Drive, Auchenflower, Queensland 4066, Australia. Email: [email protected] Submitted 24 August 2001; accepted 24 September 2001. Australasian Radiology (2002) 46, 221–230

Computed tomography perfusion imaging in acute stroke

Embed Size (px)

Citation preview

Page 1: Computed tomography perfusion imaging in acute stroke

INTRODUCTIONStroke is the third-most-common cause of death in adults

and is a leading cause of morbidity in Australia.1 Conventional

structural imaging typically comprises CT to exclude intra-

cerebral haemorrhage. However, 85% of strokes are the result

of ischaemic infarcts for which, in the majority of patients

imaged in the first 6 h, CT is either normal or demonstrates

only subtle abnormalities that are easy to misinterpret. Even

conventional MR images are commonly normal in this early

period.

The development of thrombolytic and neuroprotective

agents for the treatment of acute stroke has created an

imperative for improved imaging techniques in the evaluation of

acute stroke. The National Institute for Neurological Diseases

and Stroke trial was the first study to demonstrate that

thrombolysis in patients with no intracerebral haemorrhage

leads to a one-third increase in clinical outcomes if

administered within 3 hours of the onset of ictus.2 However, for

patients treated beyond the 3 h of onset, it carries a substantial

risk of intracranial bleeding. This setback in the use of

thrombolysis has prompted the realization that a more rigorous

selection of patients might reduce the risk of intracranial

bleeding, and that the use of thrombolysis might have a bene-

fit in selected patients extending beyond the current 3-h

therapeutic window.3 There have been several studies to

suggest that the level of residual blood flow in the ischaemic

zone is a useful indicator of the risk of intracranial bleeding.4–7

There are distinct thresholds of cerebral blood flow for various

functions of the brain. It is possible to use these thresholds to

determine whether a particular brain region is salvageable.

Perfusion CT offers the ability to positively identify patients

with non-haemorrhagic stroke, to select those cases where

thrombolysis is appropriate, and to provide an indication as to

prognosis.

Pictorial Essay

Computed tomography perfusion imaging in acute strokeCJ Keith,1,2 M Griffiths,2,3 B Petersen,1 RJ Anderson1 and KA Miles1,3

1Southern X-ray Clinics, 2Wesley Research Institute, The Wesley Hospital, and 3Centre for Medical Health and Environmental Physics,

School of Physical Sciences, Queensland University of Technology, Brisbane, Queensland, Australia

SUMMARY

The development of thrombolytic and neuroprotective agents for the treatment of acute stroke has created animperative for improved imaging techniques in the assessment of acute stroke. Five cases are presented to illustratethe value of perfusion CT in the evaluation of suspected acute stroke. To obtain the perfusion data, a rapid series ofimages was acquired without table movement following a bolus of contrast medium. Cerebral blood flow, cerebralblood volume and mean transit time were determined by mathematically modelling the temporal changes in contrastenhancement in the brain and vascular system. Pixel-by-pixel analysis allowed generation of perfusion maps. In twocases, CT-perfusion imaging usefully excluded acute stroke, including one patient in whom a low-density area onconventional CT was subsequently proven to be tumour. Cerebral ischaemia was confirmed in three cases, one with anold and a new infarction, one with a large conventional CT abnormality but only a small perfusion defect, and onedemonstrating infarct and penumbra. Perfusion CT offers the ability to positively identify patients with non-haemorrhagic stroke in the presence of a normal conventional CT, to select those cases where thrombolysis isappropriate, and to provide an indication for prognosis.

Key words: computed tomography, functional imaging, stroke.

CJ Keith MB BS; M Griffiths BSc, MAppSci; B Petersen Radiographer; RJ Anderson MB, ChB, FRANZCR; KA Miles MB BS, FRCR, MSc, MD.

Correspondence: Carolyn Keith, Southern X-ray Clinics, 2nd Floor, Day Centre, The Wesley Hospital, 45 Coronation Drive, Auchenflower, Queensland

4066, Australia. Email: [email protected]

Submitted 24 August 2001; accepted 24 September 2001.

Australasian Radiology (2002) 46, 221–230

Page 2: Computed tomography perfusion imaging in acute stroke

Measurements of cerebral perfusion can be made with

CT using conventional contrast agents as tracers. Perfusion

measurements using conventional contrast agents benefit

from wide availability and can be easily incorporated into the

conventional CT examination that is routinely performed to

exclude intracranial haemorrhage. By avoiding the need to

transfer the patient to another imaging device, perfusion CT can

save a significant amount of time in a situation where early

administration of thrombolysis is critical.

Five cases are presented that illustrate the application of

perfusion CT in suspected acute stroke.

Pathophysiology of acute strokeBasic physiological functions, such as synaptic transmission,

the membrane ion pump and energy metabolism, are critically

dependent on blood flow, and will fail at distinct blood-flow

levels. Normal cerebral perfusion is in the range of 50–60

mL/min per 100 g.When cerebral perfusion pressure decreases,

the brain has an intricate system of control that attempts to

maintain cerebral blood flow. The mildest impairment in the

cerebral circulation is associated with an autoregulatory

dilatation of cerebral arterioles resulting in a reduction in

vascular resistance and an increase in cerebral blood volume

(CBV) that is able to maintain cerebral perfusion at normal

levels (Table 1). With worsening ischaemia, vasodilatation is

insufficient to maintain perfusion, and perfusion then falls

below normal levels. At approximately 20 mL/min per 100 g,

electrical activity and water homeostasis are disrupted, which

is associated with abolition of somatosensory evoked potentials

and electroencephalogram (EEG).8,9 At this threshold, ischae-

mic impairment of tissue function is reversible and there is an

increased CBV and a prolonged mean transit time (MTT).

At 10–15 mL/min per 100 g, synthesis of adenosine triphos-

phate (ATP) is outstripped by demand resulting in disruption

of the membrane ion pump. Failure of the membrane ion

pump leads to an efflux of potassium from, and an influx of

calcium, sodium and water into cells resulting in membrane

depolarization and cytotoxic oedema.9 Later, the capillaries

become leaky, resulting in an accumulation of extracellular

water (vasogenic oedema). Both cytotoxic and vasogenic

oedema cause further compression of the microcirculation,

worsening the level of ischaemia.

Failure of the integrity of the cell membrane precedes the

irreversible destruction of the cell (infarction), and it is therefore

justified to associate the threshold of irreversible damage with

the threshold for ion-pump failure. However, it has been shown

that disturbed energy metabolism and ion-pump failure can fully

recover,10,11 which suggests that membrane failure might trigger

processes causing infarction without being the direct cause

of these processes.12 The cerebral blood-flow threshold for

membrane failure is therefore close to that of infarction, but the

development of necrosis and infarction is not only dependent

upon the level of perfusion but also the time for which the tissue

has been ischaemic. Heiss and Rosner13 found that the duration

of ischaemia required to induce permanent loss of neuronal

activity became progressively shorter as blood flow decreased.

In summary, for ischaemic non-viable tissue, both the

cerebral blood flow (CBF) and CBV are reduced, but MTT might

remain normal or slightly elevated. It is the mismatch between

CBF and CBV that discriminates between salvageable and

infarcted tissue. An area of infarction is frequently surrounded

by brain tissue that is ischaemic but viable. This surrounding

tissue is known as the penumbra and represents tissue at risk of

infarction that is potentially recoverable on re-establishment of

the circulation.The size of the perfusion defect has been shown

to correlate closely with the clinical outcome, and follow-up

studies have shown that recoverable tissue exhibits relatively

preserved blood volume.14

METHODSPerfusion computed tomographyTo obtain the perfusion data, a rapid series of images is

acquired without table movement following a bolus of contrast

medium. The protocol adopted for this study consisted of

60 acquisitions of 1-s duration with a 1-s interval following the

injection of 40 mL of non-ionic iodinated contrast material

(Isovue 370, 370 mg iodine/mL; Bracco, Milan, Italy) at a rate of

4 mL/s. The section chosen for study was at the level of the

basal ganglia as this level includes those vascular territories of

the brain that are frequently affected by acute stroke in the

carotid arterial territory. Cerebral perfusion, cerebral blood

volume and mean transit time are determined by mathemat-

ically modelling the temporal changes in contrast enhancement

in the brain and vascular system, and pixel-by-pixel analysis

222 CJ KEITH ET AL.

Table 1. Summary of the changes in cerebral vascular physiology with worsening circulatory impairment. (Note that transit time is proportional to

blood volume/perfusion)

Perfusion Blood volume Transit time

Autoregulatory range N ↑ ↑

Oligaemia (misery perfusion) ↓ ↑↑ ↑↑

Ischaemia (metabolic impairment) ↓ ↑ ↑↑

Irreversible damage (necrosis) ↓↓ ↓ ↓↑

Page 3: Computed tomography perfusion imaging in acute stroke

allows generation of perfusion maps. A number of commercial

perfusion CT software packages are now available. In this study,

data were transferred to a workstation and then evaluated with

‘CT Perfusion 2’ that uses a deconvolution-based method.

Imaging techniquesThere have been a number of methods developed for the

measurement of tissue perfusion using CT. These various

methods can be grouped under two classes: compartmental

analysis and deconvolution-based methods. The mathematics

for these methods has been described in detail elsewhere.15

Compartmental analysis-based methodsFor compartmental analysis, a single compartmental model is

used where arterial blood flows into the vascular compartment

and leaves via a single draining vein. Based on the Fick

principle, perfusion can be determined from the maximum rate

of accumulation of contrast agent within the tissue divided by

the peak arterial concentration of contrast agent. Blood-volume

measurements are obtained from the ratio of the areas under

the tissue and arterial time-attenuation curves.

Deconvolution-based methodsDeconvolution analysis mathematically calculates the tissue

time-attenuation curve that would have been obtained had

the arterial bolus arrived instantaneously. The height of this

idealized tissue curve (also known as the impulse response

function) is determined by perfusion whereas the area under

the idealized curve gives the relative blood volume (Fig. 1).

Both methods require time-attenuation data from the

vascular system to correct for interpatient variation in bolus

geometry. Although identification of some arterial vessels, such

as branches of the anterior and middle cerebral arteries, might

generally be possible on an axial scan through the basal

ganglia, arterial-enhancement values tend to be markedly

reduced due to partial volume effects.Therefore, measurement

of the maximum value in the superior sagittal sinus is used to

rescale the arterial enhancement curve. Cerebral CT perfusion

imaging has been validated against microspheres16 and H215O

PET17 with good reproducibility.

CASE REPORTSCase 1An 82-year-old man presented to the emergency department

having awoken from sleep with slurred speech and difficulty

walking. On examination, the patient’s speech was dysarthric

but there were no other cranial nerve abnormalities.There were

left cerebellar signs of incoordination and tremor, with marked

truncal and gait ataxia; however, power, tone and reflexes

were all normal, and sensation was also reported as normal.

An acute CT scan showed no abnormalities on conventional

scan, and there was no perfusion defect on CT-perfusion study

(Fig. 2). The patient’s condition improved throughout the

week. A follow-up conventional CT scan 1-week later was again

normal for age, with no evidence of a recent ischaemic event.

Case 2A 63-year-old man was admitted to another hospital with a

subacute bowel obstruction. On the third day of his stay, the

patient was found with a decreased level of consciousness and

dense right-sided weakness. On examination, there was both

receptive and expressive dysphasia. Muscular tone was normal

on the left and increased on both the right upper and lower

limbs, there was markedly reduced power on the right (1/5).

There were also abnormally brisk reflexes on examination of the

right lower limb, with a positive Babinski sign. A conventional CT

of the head performed on the same day was normal for age.The

patient was transferred to the hospital where this study was

conducted. A repeat CT in another 4 days showed a large low-

density area in the left cerebral hemisphere in the territory

supplied by the left middle cerebral artery. Perfusion CT

performed at the same time demonstrated a substantial

reduction in blood flow to a much smaller area, with a mild

reduction in blood volume and a greatly increased MTT in the

corresponding small area (Fig. 3). It was thought then that the

large lesion seen on conventional scan was in fact a small

infarct with surrounding oedema.Clinically, the patient improved

markedly. At the end of 2 weeks, he was speaking in full

sentences and understanding commands, power had returned

to his right upper limb and there was minimal return of power to

his right lower limb.

Case 3A 79-year-old man was brought into the emergency depart-

ment following a collapse while walking. On arrival, the patient

was alert and orientated to his surroundings. There were no

223PERFUSION CT IN STROKE

Fig. 1. Graphical representation of the deconvolution analysis

method, the height of this idealized tissue curve (also known as the

impulse response function) is determined by perfusion whereas the

area under the idealized curve gives the relative blood volume. AUC,

area under curve; CBF, cerebral blood flow; CBV, cerebral blood

volume; MTT, mean transit time.

Page 4: Computed tomography perfusion imaging in acute stroke

difficulties with speech. The only finding was some mild left

lower-limb weakness, reflexes were all normal. The patient had

a past history of left-sided occipital stroke. A conventional CT

performed 3 days after onset of symptoms showed a large

low-attenuation area occupying the right occipital lobe,

consistent with an old cerebral infarct.The CT-perfusion images

showed a large area of greatly decreased CBF, CBV and

MTT corresponding to the old infarct demonstrated on the

conventional images. The perfusion images also show a small

area of reduced perfusion and blood volume with preserved

MTT in the right temporoparietal region just anterior to the old

infarct (Fig. 4). This perfusion abnormality is consistent with a

recent infarct.

Case 4A 56-year-old man presented with a sudden onset of expressive

dysphasia, followed in 5–10 min by a headache. While in the

emergency department, the patient was observed to have

some focal right-sided motor-seizure activity. The patient was

sent for a conventional CT scan that demonstrated a

hypodense area in the left hemisphere, which was reported

as likely to be an early infarct. There was no evidence of

224 CJ KEITH ET AL.

Fig. 2. Case 1. (a) Mean transit time. (b) Cerebral blood volume. (c) Cerebral blood flow. (d) Follow-up conventional CT image performed 1 week

later.

Page 5: Computed tomography perfusion imaging in acute stroke

intracranial haemorrhage. The CT perfusion images dem-

onstrated a slight increase in blood flow in the area

corresponding to the hypodense lesion in the left hemisphere

consistent with a low-grade tumour (Fig. 5). The MRI of the

head, including T1- and T2-weighted images, diffusion-

weighted images and a gradient-echo sequence, performed at

the same time and after 1 month demonstrated a focal

abnormality involving the left insula and adjacent frontal lobe.

There was apparent thickening of the grey matter with some

reactive oedema around the lesion. Post-gadolinium images

displayed no enhancement and no abnormal blood vessels.

The morphology on MR was consistent with a low-grade glioma

rather than an infarct.

Case 5A 78-year-old woman suffered a brief loss of consciousness

and was noted to have a right hemiplegia thereafter. She was

brought into the emergency department where examination

revealed aphasia with right-sided neglect and a complete right

hemiplegia. The right-sided reflexes were brisker and the

plantar response was extensor. A CT brain scan performed

within 4 h of the onset of ictus demonstrated reduced density

225PERFUSION CT IN STROKE

Fig. 3. Case 2. (a) Mean transit time. (b) Cerebral blood volume. (c) Cerebral blood flow. (d) Conventional CT image performed 4 days postictus.

Page 6: Computed tomography perfusion imaging in acute stroke

and loss of clarity of the basal ganglia, and there was also

subtle loss of clarity of the grey–white interface in the left

temporoparietal region. A CT-perfusion brain scan performed

at the same time demonstrated markedly decreased blood

flow to the left temporoparietal region and reduced blood

volume predominately in the basal ganglia with preservation

of blood volume in the cortex (i.e. there was a mismatch

between CBF and CBV). This indicated that although there

was some infarction in the left temporoparietal region in

the territory supplied by the left middle cerebral artery,

some of the ischaemic changes were likely to be reversible

(Fig. 6). The patient was transferred to another hospital.

Conventional CT imaging of the head performed the following

day to investigate a marked deterioration in consciousness

revealed an extensive intraparenchymal bleed with intra-

ventricular and subarachnoid extension. The patient died the

next day.

DISCUSSIONA range of physiological brain-imaging techniques have been

used in the assessment of acute stroke including perfusion

CT, perfusion and diffusion weighted MR, MR spectroscopy,

226 CJ KEITH ET AL.

Fig. 4. Case 3. (a) Mean transit time. (b) Cerebral blood volume. (c) Cerebral blood flow. (d) Conventional CT.

Page 7: Computed tomography perfusion imaging in acute stroke

positron emission tomography (PET), single photon emission

tomography (SPECT) and transcranial Doppler ultrasound.

Xenon CT requires specialized equipment that is not widely

available and the technique is associated with adverse side-

effects in a significant number of patients. Perfusion MR is

directly analogous to perfusion CT but MR has the advantage

of obtaining dynamic contrast-enhanced images over a larger

volume of brain tissue with no radiation burden. Further-

more, perfusion MR can be combined with diffusion-weighted

MR (DWI) to define more clearly the areas of infarction and

penumbra. At present, perfusion MR is relatively time-

consuming, and the availability is relatively limited, although

future advances in MR technology and availability can be

expected. The use of MR within the first few hours of stroke

also poses a problem regarding patient cooperation, and the

adequate monitoring of vital parameters in patients receiving

emergent care remains a challenge while patients are placed

inside the magnet. Magnetic resonance spectroscopy

estimates the concentration of normal and abnormal metab-

olites in brain tissue. Early cerebral ischaemia is associated

with increased levels of lactate, while levels of N-acetyl-

aspartate (NAA) fall in the later stages. However, as well as

227PERFUSION CT IN STROKE

Fig. 5. Case 4. (a) Mean transit time. (b) Cerebral blood volume. (c) Cerebral blood flow. (d) Conventional CT.

Page 8: Computed tomography perfusion imaging in acute stroke

the other drawbacks mentioned for perfusion MR, a relatively

large volume of tissue is required for analysis, thereby

constraining the spatial resolution of the biochemical data. By

using radiotracers such as 15O-water, 15O-carbon dioxide and15O-oxygen, PET can provide quantitative information about

blood flow, blood volume and oxygen extraction while 18F-MISO

can depict ischaemic tissue. However, PET is largely used as a

research tool, and its limited availability prevents routine clinical

use. Single photon emission tomography can provide an image

of relative cerebral perfusion and blood volume but is non-

quantitative. Transcranial Doppler can access patency and flow

within the middle cerebral artery but is unable to evaluate

perfusion at tissue level.

This pictorial essay demonstrates the use of perfusion CT in

the delineation of suspected cerebral ischaemia. The potential

benefits of perfusion CT in the evaluation of stroke include

the ability to rapidly demonstrate and aid in the correct

diagnosis of non-haemorrhagic stroke, to identify those patients

whom treatment will benefit and those to which it might be

detrimental, and to indicate the likely clinical outcome.

Perfusion CT is an effective imaging technique for the

evaluation of stroke because it can reliably demonstrate

228 CJ KEITH ET AL.

Fig. 6. Case 5. (a) Mean transit time. (b) Cerebral blood volume. (c) Cerebral blood flow. (d) Conventional CT.

Page 9: Computed tomography perfusion imaging in acute stroke

ischaemia within 1–2 h after symptom onset, thereby allowing

initiation of therapeutic strategies.14,18,19 This ability to aid in early

diagnosis is shown in several of the cases presented here.

Case 1, in which the CT-perfusion study was able to help

exclude stroke as a possible cause for the patient’s signs and

symptoms, was confirmed as accurate by the follow-up images

after 1 week. These results could have saved the patient

unnecessary treatment that carries with it a risk of increased

morbidity and mortality. In case 2, the conventional CT images

demonstrated a large area of hypoattenuation and were

reported as showing a large infarct. However, the CT-perfusion

images demonstrated a much smaller area of infarction and it

was thought the larger area seen on conventional imaging was

predominantly the result of surrounding oedema. In case 3,

conventional CT images revealed an old infarct whereas CT-

perfusion images are able to demonstrate both old and recent

infarcts. In case 4, perfusion CT changed the clinical diagnosis

of stroke to tumour. Perfusion CT has the additional benefit of

defining abnormal regions of blood flow in brain tumours.

Neoplasms of the brain develop abnormal capillaries that are

fenestrated and allow the free passage of contrast agents. Also,

neovascularization within tumours results in an increased

vessel density and abnormally dilated venous channels, which

is reflected in an increase in the relative blood volume of the

tumours.20

Clinical outcome has been shown to correlate with the size

of the perfusion defect. Mayer et al., using CT perfusion, found

in their study of 70 patients for whom CBF maps predicted

the extent of the infarct with a sensitivity of 93% and a specificity

of 98%.14 Another study by a second group using the same

technique in 75 patients confirmed these findings.15 The

prognostic value of perfusion CT is demonstrated in this review

by case 2, in which the conventional CT images demonstrated a

large area of hypoattenuation and were reported as showing a

large infarct.However, the CT perfusion images demonstrated a

much smaller area of infarction and it was thought the larger

area seen on conventional imaging was predominantly the

result of surrounding oedema.This is supported by the fact that

the patient’s clinical status improved dramatically over the

following weeks with recovery of both speech and much of the

motor deficit he originally presented with.

A range of new therapy approaches directed at thromboly-

sis and reversing or minimizing ischaemic damage are cur-

rently undergoing investigation. Salvage of tissue at risk

(i.e. penumbra) is the target of many of these treatment options.

However, restoration of blood flow will not always result in a

clinical benefit because therapy will not rescue brain tissue if

the level of cerebral perfusion is below the viability thresholds

of ischaemia. An imaging technique that could identify and

characterize the potentially reversible ischaemic tissue would

be invaluable in the selection of patients for new therapies.

Presently, the best and most quantifiable technique for

determining the penumbra is xenon CT, although for the

reasons mentioned previously, this technique has limitations.

Recently, Wintermark et al.21 demonstrated a good correlation

of CBF measured by perfusion CT and xenon CT. Jansen et al.

showed that tissue at risk, defined by MR perfusion-weighted

imaging that exceeded tissue with diffusion abnormality,

could be salvaged by recanalization.22 Koenig et al.,23 using CT

perfusion, demonstrated significantly lower values for cerebral

blood flow in infarcted areas compared with those suffering

reversible ischaemia, and that a severe reduction of CBF is

followed by a reduction in CBV and indicates the core of the

infarction, while in the border zone with only moderate hypo-

perfusion the CBV was maintained or only slightly reduced.

Case 5, in which the CT perfusion images showed a mismatch

between CBF and CBV, demonstrated the potential for

perfusion CT to define the penumbra and identify patients

suitable for thrombolysis.

Perfusion CT adds only a few minutes to the duration of

a conventional CT scan, and it is easily performed using

current CT technology and readily available computer soft-

ware. The amount of contrast material used and the short

examination time easily allow repetition of the procedure to

obtain a second and, if required, a third CT section if the

standard section does not provide conclusive information.

Furthermore, future developments in CT technology, including

dynamic scanning with multisection data acquisition,24,25 might

further increase the value of this technique and provide

information about the 3-D extent of cerebral ischaemia for the

assessment of stroke.

In conclusion, perfusion CT can demonstrate abnormalities

in patients with acute stroke, even when conventional images

are normal. Perfusion CT in acute stroke can depict brain tissue

at-risk and provide prognostic information, and is a useful tool

for the selection of patients for thrombolytic therapy.

REFERENCES1. Anderson CS, Jamrozic KD, Burvill PW, Chakera TM, Johnson GA,

Stewart-Wynne EG. Ascertaining the true incidence of stroke:

experience from the Perth Community Stroke Study. Med. J. Aust.

1993; 158: 80–4.

2. National Institute for Neurological Disease and Stroke rt-PA Study

Group. Tissue plasminogen activator for acute ischemic stroke.

N. Engl. J. Med. 1995; 355: 1581–7.

3. 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

European-Australasian Acute Stroke Study. Lancet 1998; 352:

1245–51.

4. Lyden PD, Zivin JA. Hemorrhagic transformation after cerebral

ischemia: mechanisms and incidence. Cerebrovasc. Brain Met.

Rev. 1993; 5: 1–16.

5. Ueda T, Hatakeyama T, Kumon Y, Sakai S, Uraoka T. Evaluation

of risk of hemorrhagic transformation in local intra-arterial

thrombolysis in acute ischemic stroke by initial SPECT. Stroke

1994; 25: 298–303.

229PERFUSION CT IN STROKE

Page 10: Computed tomography perfusion imaging in acute stroke

6. Ezura M, Takahashi A, Yoshimoto T. Evaluation of regional cerebral

blood flow using single photon emission tomography for the

selection of patients for local fibrinolytic therapy of single acute

cerebral embolism. Neurosurg. Rev. 1996; 19: 231–6.

7. Alexandrov AV, Black SE, Ehrlich LE, Caldwell CB, Norris JW.

Predictors of hemorrhagic transformation occurring spontan-

eously and on anticoagulants in patients with acute ischemic

stroke. Stroke 1997; 28: 1198–202.

8. Astrup J, Siesjo B, Symon L.Thresholds in cerebral ischemia – the

ischemic penumbra. Stroke 1981; 12: 723–5.

9. Hossmann KA. Viability thresholds and the penumbra of focal

ischemia. Ann. Neurol. 1994; 36: 557–65.

10. Ljunggren B, Ratcheson RA, Siesjo BK. Cerebral metabolic state

following complete compression ischemia. Brain Res. 1974; 73:

291–307.

11. Hossmann KA, Sakai S, Zimmerman V. Cation activities in

reversible ischemia of the cat brain. Stroke 1974; 8: 77–81.

12. Astrup J. Energy-requiring cell functions in the ischemic brain.

Their critical supply and possible inhibition in protective therapy.

J. Neurosurg. 1982; 56: 482–97.

13. Heiss WD, Rosner G. Functional recovery of cortical neurons as

related to the degree and duration of ischemia. Ann. Neurol. 1983;

14: 294.

14. Mayer TE, Hamann GF, Baranczyk J et al. Dynamic CT perfusion

imaging of acute stroke. AJNR 2000; 21: 1441–9.

15. Klotz E, Koenig M. Perfusion measurements of the brain, using

dynamic CT for the quantitative assessment of cerebral ischaemia

in acute stroke. Eur. J. Radiol. 1999; 30: 170–84.

16. Cenic A, Nabavi DG, Craen RA, Gelb AW, Lee T-Y. Dynamic CT

measurements of cerebral blood flow. A validation study. AJNR

1999; 20: 63–73.

17. Gillard JH, Minhas P, Hayball MP et al. Evaluation of quantitative

computed tomographic cerebral perfusion imaging with H215O

positron emission tomography. Neurol. Res. 2000; 22: 457–64.

18. Koenig M, Klotz E, Luka B, Vendernick DJ, Spittler JF, Heuser L.

Perfusion CT of the brain: diagnostic approach for early detection

of ischemic stroke. Radiology 1998; 209: 85–93.

19. Rother J, Jonetz-Mentzel L, Fiala A et al. Hemodynamic asses-

sment of acute stroke using dynamic single-slice computed

tomographic perfusion imaging. Arch. Neurol. 2000; 57: 1161–6.

20. Leggett DAC, Miles K, Kelley BB. Blood–brain barrier and blood

volume imaging of cerebral glioma using functional CT. A pictorial

review. Eur. J. Radiol. 1999; 30: 185–90.

21. Wintermark M, Thiran J, Maeder P, Schnyder P, Meuli R. Simul-

taneous measurement of regional cerebral blood flow by perfusion

CT and stable xenon CT. A validation study. AJNR 2001; 22:

905–14.

22. Jansen O, Schellinger P, Fiebach J, Hacke W, Sartor K. Early

recanalisation in acute ischaemic stroke saves tissue at risk

defined by MRI. Lancet 1999; 353: 2036–7.

23. Koenig M, Kraus M, Theek C, Klotz E, Gehlen W, Heuser L.

Quantitative assessment of the perfusion-related parameters

derived from perfusion CT. Stroke 2001; 32: 431–7.

24. Roberts HC, Roberts TP, Smith WS, Lee TJ, Fischbein NJ,

Dillon WP. Multisection Dynamic CT perfusion for acute cere-

bral ischemia. The ‘toggling-table’ technique. AJNR 2001; 22:

1077–80.

25. Jones TR, Kaplan RT, Lane B, Atlas SW, Rubin GD. Single versus

multi-detector row CT of the brain: quality assessment. Radiology

2001; 219: 750–5.

230 CJ KEITH ET AL.