вход по аккаунту


Contrast computed tomography scan in acute stroke УYou can't always get what you want butЕyou get what you needФ.

код для вставкиСкачать
Contrast Computed
Tomography Scan in Acute
Stroke: “You Can’t Always Get
What You Want But…You
Get What You Need”
X-ray computed tomography (CT) revolutionized the
diagnosis and treatment of patients with cerebrovascular accidents during the early 1970s.1 The prominent
increase in signal intensity caused by intracranial hemorrhage made it an invaluable tool for the emergency
evaluation of patients with a stroke-like syndrome; the
diagnosis of acute ischemic stroke was made when the
CT scan was “normal.” However, as early as the mid1970s, it was proposed that contrast-enhanced CT
could be used to map brain perfusion,2 and recently we
have realized that in major stroke, even the unenhanced CT is usually not normal.
The article in this issue of the Journal by Wintermark and colleagues, from Lausanne, Switzerland, entitled “Prognostic Accuracy of Admission Cerebral
Blood Flow Measurement by Perfusion-CT in Acute
Stroke Patients” is an elegant demonstration of the performance of CT perfusion imaging in its most modern
form.3 Using contrast CT scan data from a single slice
of brain, the authors make reasonably accurate predictions in individual patients about which brain tissue
will be salvaged with reperfusion and die without it.
The ability to make this distinction has been a “Holy
Grail equivalent” for those interested in image-based,
as opposed to purely time-based, stroke therapy. Quite
an achievement for a machine that is the time-honored
workhorse in most emergency rooms!
CT perfusion imaging using standard, nonionic iodinated contrast relies on the speed of modern helical
scanners to image faster than one brain slice per second.
At this speed, CT scanners can trace the entry and washout of a bolus of dye injected at 4 to 10ml/sec into a
large arm vein. CT dye becomes an intravascular tracer
and the CT scanner a detector of brain blood flow. The
general method is based on the indicator-dilution principle of measuring tissue perfusion.4,5 The change in signal intensity in Hounsfield units after dye injection is
proportional to the concentration of the dye in the pixel
imaged. The dye-related change in signal intensity in
pixels located within an artery enables calculation of the
concentration of dye in the arterial blood.6 With these
data, a pixel analysis of the dye-related increase in signal
intensity during the first pass of the bolus through the
brain provides maps of cerebral blood volume (CBV)
and cerebral blood flow (CBF).7–9
Mean transit time (MTT) ⫽ CBV/CBF
The linear relationship between dye concentration and
signal intensity is a major advantage of CT perfusion
imaging over gadolinium-based magnetic resonance
(MR) perfusion imaging. In MRI, the relationship between dye concentration and signal intensity is nonlinear; this stands as a severe impediment to quantification of brain blood flow using MR bolus tracking
perfusion imaging.
Wintermark and colleagues10 used a standard
“deconvolution-based” method to make quantitative estimates of brain blood flow. It is this quantitative ability
that allows the stratification of tissue according to specific perfusion thresholds. In their article, brain tissue
with a CBV of ⬍2.5ml/100g was considered, for all
practical purposes, dead, and it acted as such. A less severe threshold, brain with a CBF of ⬍34% of the corresponding contralateral side, was considered to encompass the brain at risk of dying, termed the “penumbra
plus infarct.” There are two major findings. First, in patients in whom reperfusion occurred, the final infarct
size matched the acute CBV lesion (defined by a CBV
threshold of ⬍2.5ml/100g). We have seen a similarly
highly predictive relationship between the more crude
whole-brain-perfused blood volume lesion and final infarct in patients who were successfully reperfused.11 The
second, and most important, finding is that brain tissue
with CBF of ⬍34% of the opposite side died in those
patients without reperfusion. The penumbra, defined as
the tissue that is salvageable by reperfusion, is then identifiable as that tissue below the CBF threshold, but
above the CBV threshold. The authors actually map out
these regions in color overlays on the CT scan calling
them “prognostic maps.” This is another rendition of
the “mismatch” idea, analogous to the popular mismatch between the diffusion-weighted imaging (DWI)
lesion and the perfusion abnormality on MR perfusion
imaging. In place of the DWI abnormality, Wintermark
and colleagues insert a CBV abnormality as highly predictive of dead or dying tissue. We agree; in our hands,
the CBV abnormality measured by either whole-brain
CT or MR perfusion scanning correlates extremely well
with the DWI abnormality.12,13 A collapse in the tissue’s perfusable vascular volume may be a necessary biological event on the pathway to infarction and DWI
abnormality. The major problem in MR perfusion imaging has been the difficulty in establishing thresholds of
CBF because of the difficulty in quantification. The perfusion abnormality seen on MR is oversensitive and
nonspecific and, in general, includes viable tissue. The
specificity in identifying the presence or absence of brain
tissue that will be salvaged by reperfusion is the true
achievement of this report. Combined with some sense
of the risk of hemorrhage, it is the measurement that
should guide reperfusion therapy.
After touting its promise, it is only fair to point out
© 2002 Wiley-Liss, Inc.
the limitations. Radiation exposure is always a concern.
It is the same order of magnitude as that of a head CT
scan, but it is concentrated in a single slice of brain.
Perhaps the major problem is that quantitative flow
mapping by CT is limited to a single brain slab. The
new multislice helical CT scanners can currently acquire up to four channels of image data simultaneously, but this is all sandwiched into a 2cm array.
This improves scanner speed and sells machines, but it
is an inherent obstacle to whole-brain CBF measurements, something stroke physicians want badly. Wintermark and colleagues address this problem by using
two data acquisitions, but this doubles the required
contrast and radiation dosage. The second major limitation is that quantitation of CT perfusion parameters,
as with MR perfusion, is dependent on the specific
mathematical model applied for image processing.
Wintermark and colleagues10 use a deconvolutionbased method to obtain quantitative information from
a bolus of dye that does not enter the brain as a tight
packet in time. The tissue perfusion will be a function
of the temporal profile of the dye entering and leaving
the artery that most directly supplies the tissue being
imaged. This ideal input function cannot be directly
measured; instead, an arterial input function is measured from a large artery present on the slice. Delay
and dispersion of the bolus that occur between this
measurable artery and the actual vessel supplying the
tissue will introduce errors (usually underestimation) in
the CBF.14 In addition, the model is very sensitive to
noise in the image and requires sophisticated algorithms to limit noise and to correct for dye extravasation, change in hematocrit in artery versus capillary,
and other factors. Xenon CT (in which Xenon is a diffusable tracer), arterial spin-labeling MRI (in which the
ideal arterial input function is measured), and positron
emission tomography may therefore be more suitable
for accurate quantitative blood flow measurements.
CT, however, wins on accessibility, speed, and low
cost. CT is also versatile. It reliably identifies hemorrhage. Contrast CT angiography (CTA) can provide
spectacular images of the vascular lesion.11 CT angiography also provides predictive images of whole-brain
perfused blood volume, which may be essential to
choose the most appropriate slice for CBF measurements.15 It is hoped that the future will see software
that can rapidly construct validated, color-coded, “infarcted (red) and penumbra (green) maps” on a patient’s brain slice, as shown in the authors’ “prognostic
maps.” Specific reperfusion therapies might make sense
in patients with significant penumbra (green for “go”),
but not in those in whom the entire brain has passed
on to infarction (red for “stop”). As opposed to MRI,
every stroke center in the country should be able to
complete this form of imaging within minutes of the
patient presenting to the emergency department.
Learning how to use this information reliably for the
Annals of Neurology
Vol 51
No 4
April 2002
benefit of the individual patient, or as a tool to search
for effective brain protective strategies, is the future.
Walter J. Koroshetz, MD
Massachusetts General Hospital
Michael H. Lev, MD
Massachusetts General Hospital
Boston, MA
1. Ambrose J, Hounsfield G. Computerized transverse axial tomography. Br J Radiol 1973;46:148 –149.
2. Zilkha E, Ladurner G, Iliff LD, et al. Computer subtraction in
regional cerebral blood-volume measurements using the EMIscanner. Br J Radiol 1976;49:330 –334.
3. Wintermark M, Reichhart M, Thiran JP, et al. Prognostic accuracy
of admission cerebral blood flow measurement by perfusion-CT in
acute stroke patients. Ann Neurol 2002;51:417– 432.
4. Meier P, Zieler K. On the theory of the indicator-dilution
method for measurement of blood flow and volume. J Appl
Physiol 1954;6:731–744.
5. Roberts G, Larson K. The interpretation of mean transit time
measurements for multi-phase tissue systems. J Theor Biol
1973;39:447– 475.
6. Yeung I, Lee T, Del Maestro R, et al. An absorptiometry
method for the determination of arterial blood concentration of
injected iodinated contrast agent. Phys Med Biol 1992;37:
7. Cenic A, Nabavi DG, Craen RA, et al. Dynamic CT measurement of cerebral blood flow: a validation study. AJNR Am J
Neuroradiol 1999;20:63–73.
8. Nabavi DG, Cenic A, Henderson S, et al. Perfusion mapping using
computed tomography allows accurate prediction of cerebral infarction in experimental brain ischemia. Stroke 2001;32:175–183.
9. Wintermark M, Thiran JP, Maeder P, et al. Simultaneous measurement of regional cerebral blood flow by perfusion CT and
stable xenon CT: a validation study. AJNR Am J Neuroradiol
10. Wintermark M, Maeder P, Thiran JP, et al. Quantitative assessment of regional cerebral blood flows by perfusion CT studies at low injection rates: a critical review of the underlying
theoretical models. Eur Radiol 2001;11:1220 –1230.
11. Lev MH, Segal AZ, Farkas J, et al. Utility of perfusion
weighted CT imaging in acute MCA stroke treated with intraarterial thrombolysis: prediction of final infarct volume and
clinical outcome. Stroke 2001;32:2021–2028.
12. Sorensen AG, Copen WC, Ostergaard L, et al. Hyperacute
stroke: simultaneous measurement of relative cerebral blood
volume, relative cerebral blood flow and mean tissue transit
time. Radiology 1999;210:519 –527.
13. Berzin TM, Lev MH , Goodman D, et al. CT perfusion imaging versus MR diffusion weighted imaging: prediction of final
infarct size in hyperacute stroke. In: Stroke: Proceedings of the
26th International Conference on Stroke and Cerebral Circulation, Ft Lauderdale, FL, 2001.
14. Ostergaard L, Chesler DA, Weisskoff RM, et al. Modeling cerebral blood flow and flow heterogeneity from magnetic resonance residue data. J Cereb Blood Flow Metab 1999;19:
690 – 699.
15. Hunter GJ, Hamburg LM, Ponzo JA, et al. Rapid assessment of
cerebral perfusion and arterial anatomy in hyperacute stroke
with 3D functional computed tomography: early clinical results. AJNR 1998;19:29 –37.
DOI 10.1002/ana.10167
Без категории
Размер файла
37 Кб
get, contrast, tomography, butеyou, scan, уyou, want, always, stroki, can, need, acute, computer
Пожаловаться на содержимое документа