Quantitative Assessment of the Ischemic Brain by Means of
Perfusion-Related Parameters Derived From Perfusion CT
Matthias Koenig, MD; Michael Kraus; Carmen Theek, MSc; Ernst Klotz, DPhys;
Walter Gehlen, MD; Lothar Heuser, MD
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Background and Purpose—Besides the delineation of hypoperfused brain tissue, the characterization of ischemia with
respect to severity is of major clinical relevance, because the degree of hypoperfusion is the most critical factor in
determining whether an ischemic lesion becomes an infarct or represents viable brain tissue. CT perfusion imaging
yields a set of perfusion related parameters which might be useful to describe the hemodynamic status of the ischemic
brain. Our objective was to determine whether measurements of the relative cerebral blood flow (rCBF), relative
cerebral blood volume (rCBV), and relative time to peak (rTP) can be used to differentiate areas undergoing infarction
from reversible ischemic tissue.
Methods—In 34 patients with acute hemispheric ischemic stroke ⬍6 hours after onset, perfusion CT was used to calculate
rCBF, rCBV, and rTP values from areas of ischemic cortical and subcortical gray matter. Results were obtained
separately from areas of infarction and noninfarction, according to the findings on follow-up imaging studies. The
efficiency of each parameter to predict tissue outcome was tested.
Results—There was a significant difference between infarct and peri-infarct tissue for both rCBF and rCBV but not for rTP.
Threshold values of 0.48 and 0.60 for rCBF and rCBV, respectively, were found to discriminate best between areas of
infarction and noninfarction, with the efficiency of the rCBV being slightly superior to that of rCBF. The prediction of
tissue outcome could not be increased by using a combination of various perfusion parameters.
Conclusions—The assessment of cerebral ischemia by means of perfusion parameters derived from perfusion CT provides
valuable information to predict tissue outcome. Quantitative analyses of the severity of ischemic lesions should be
implemented into the diagnostic management of stroke patients. (Stroke. 2001;32:431-437.)
Key Words: cerebral infarction 䡲 cerebral ischemia 䡲 perfusion 䡲 tomography, x-ray computed
F
ibrinolytic therapy has become an accepted modality for
the treatment of acute ischemic stroke in selected patients. Early recanalization of an occluded artery is necessary
to avoid the development of large infarctions and to achieve
functional recovery.
Fibrinolysis has shown to be able to recanalize cerebral
vessels within the short therapeutic time window. However,
restoration of blood flow does not necessarily indicate a
clinical benefit for the patient,1,2 because fibrinolytic therapy
will not rescue brain tissue with restricted blood supply from
infarction if the level of cerebral blood flow is below the
viability thresholds of ischemia. Moreover, reperfusion may
even cause additional damage to the ischemic brain by
inducing hemorrhagic complications if thrombolysis is applied to severely hypoperfused regions.3 According to recent
investigations with positron emission tomography (PET) and
single-photon emission CT (SPECT), the assessment of the
ischemic brain by means of quantitative blood flow parameters may be the clue to direct fibrinolysis to those patients
with potentially salvageable brain.4 – 6
Thresholds of irreversible brain damage in terms of absolute values of parameters of cerebral perfusion and metabolism have been defined with PET yielding a better understanding of the underlying pathophysiology7–11; because of its
limited availability, however, PET seems impractical for the
management of stroke patients in a clinical setting.
Quantitative measurements of various cerebral perfusion
parameters such as relative cerebral blood flow (rCBF),
relative cerebral blood volume (rCBV), and time to bolus
peak (TP) have been used for the characterization of ischemic
brain areas. Recently, their role as fundamental factors of the
cerebral hemodynamic status has been stressed by calculating
threshold values of each single functional parameter to
discriminate between infarction and “tissue at risk.”4,5,12
In another clinical study, a combination of rCBF and rCBV
values was also used to characterize the area of hypoperfusion with respect to tissue outcome; the parameter values,
however, originated from different imaging modalities.13
Perfusion CT has recently been introduced in the clinical
Received July 14, 2000; final revision received August 31, 2000; accepted September 26, 2000.
From the Departments of Radiology and Nuclear Medicine (M.K., L.H.) and Neurology (M.K., W.G.), Ruhr-University Bochum; the Department of
Statistics, University of Dortmund (C.T.); and Siemens Medical Engineering Group (Computed Tomography), Forchheim, Germany.
Correspondence and reprint requests to Matthias Koenig, MD, Department of Radiology and Nuclear Medicine, Ruhr-University Bochum,
Knappschaftskrankenhaus Langendreer, In der Schornau 23/25, D-44892 Bochum, Germany. E-mail [email protected]
© 2001 American Heart Association, Inc.
Stroke is available at http://www.strokeaha.org
431
432
Stroke
February 2001
Figure 1. Evaluation of relative perfusion
scores of ischemic tissue. An ROI mask
outlining areas of hypoperfusion as
shown on the CBF color map (a) was
created on an anatomic image (b). The
regions included areas of cortical and
subcortical gray matter infarction or noninfarction according to the follow-up
imaging study (d). The side-to-side ratios
were calculated from mirrored regions for
both CBF (black/white image in c), CBV,
and TP (not shown).
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
routine for the diagnostic workup of stroke patients. Based on
the indicator dilution principle, it provides various functional
maps of the cerebral perfusion status, which may be used for
the detection and characterization of the ischemic brain. The
aim of our study was to describe typical changes of the
perfusion-related parameters CBF, CBV, and TP in areas
undergoing infarction and in tissue compartments of reversible ischemia. We also intended to define threshold values of
ischemia for various parameters derived from perfusion CT
studies and to evaluate the quality of classification of hypoperfused tissue as infarction or as reversible ischemia.
Finally, we analyzed whether the potential of predicting the
tissue outcome could be increased by using a combination of
functional perfusion parameters.
Subjects and Methods
Patients
Thirty-four patients with acute nonlacunar ischemic stroke of the
supratentorial brain were enrolled in the study. There were 14 men
and 20 women (mean⫾SD age 65.4⫾13.5 years). A conventional
CT study, obtained within 6 hours of symptom onset to rule out
intracranial hemorrhage, was followed immediately by perfusion CT
that showed areas of cerebral hypoperfusion in all cases.
The study was approved by an institutional review committee, and
informed consent was obtained from all patients or from their close
relatives. Intravenous heparin was given to 20 patients in whom
stroke was suspected to be caused by cardiac emboli but who did not
fulfill the inclusion criteria for fibrinolytic therapy according to the
guidelines at our stroke center. One patient received intravenous
rtPA according to the NINDS Stroke Study protocol,14 and 11
patients were treated with local intra-arterial fibrinolysis (LIF). In an
additional 2 patients in whom LIF was initiated shortly after surgery,
treatment had to be stopped after injection of a small dose of rtPA (8
mg and 10 mg, respectively) because of bleeding complications. For
the remaining patients in the fibrinolysis group, the dosage and
duration of treatment ranged from 10 to 80 mg rtPA and from 30 to
180 minutes, respectively. For local intra-arterial fibrinolysis, rtPA
was superselectively administered into the thrombus by a microcatheter after demonstration of an embolic occlusion within the carotid
distribution. A bolus of 5000 IU heparin was given intravenously at
the beginning of the procedure in each patient. Arterial recanalization was rated by way of repeated angiography at the end of
intra-arterial fibrinolysis by using the classification of the Thrombolysis In Myocardial Infarction (TIMI) Trial,15 with scoring from 0
(persisting occlusion) to 3 (full recanalization). Follow-up CT or
MRI scans for the depiction of definite infarction were obtained at 2
to 10 days.
CT Perfusion Imaging
Our perfusion CT technique has been previously described in
detail.16,17 Briefly, axial, single-section dynamic CT at the level of
the basal ganglia was performed to encompass areas of the anterior,
posterior, and middle cerebral artery territory. The slice thickness
was 10 mm, and the matrix size for data acquisition was 512⫻512.
For each study, a bolus of 50 mL iodinated contrast agent was
injected into an antecubital vein with a power injector. A sequence of
32 to 40 images was then collected at a rate of 1 image/s to measure
the signal intensity change during the first passage of the bolus
through the vasculature of the brain. To achieve quantitative cerebral
perfusion data, we used the Perfusion CT software (Siemens) which
allows the calculation of CBF, CBV, and TP maps based on the CT
time attenuation curves for each pixel. Our previous investigations
indicated these parameters to be useful for the delineation of
hypoperfused brain tissue in acute stroke.18 Thus, in this study we
also focused on quantitative values of CBF, CBV, and TP to assess
the degree of ischemia, although other parameters (eg, time to bolus
start and maximum enhancement) are routinely provided by the
software. The calculation of CBF is based on the maximum slope
model, whereas for the CBV the normalized maximum enhancement
was used as an estimate. The TP is defined as the time lag between
the first arrival of the contrast agent within major arterial vessels
included in the section and the local bolus peak in the brain tissue.
Thus, while in the normally perfused brain the TP is expected to be
in the range of a few seconds because of the unimpaired antegrade
flow, in the ischemic hemisphere the TP may be prolonged,
reflecting the delayed perfusion caused by leptomeningeal pathways.
Image Analysis
For the quantitative assessment of reversible and irreversible ischemia, we compared the perfusion impairment shown on the CBF,
CBV, and TP maps with the findings on follow-up imaging studies.
The section of the follow-up CT or MRI study that best fit the chosen
scan level of perfusion CT was used to define those areas of
hypoperfusion that eventually became infarcted and those that did
not (Figure 1). In our experience, the results of these 2 imaging
techniques are generally quite similar as long as the phase of the
“fogging effect” is avoided for CT evaluation. Regions of interest
(ROIs) were manually drawn in the CBF map to outline areas of
infarction of the cortical and subcortical gray matter as shown on the
follow-up images. ROIs were also defined for those ischemic
compartments that did not undergo infarction. To reduce potential
errors from the differences of the physiological perfusion values
within the gray and white matter, care was taken not to involve
substantial parts of the cerebral white matter within a given ROI.
This was facilitated by an overlay of the ROI mask to a highresolution anatomic image provided by the software. To evaluate all
functional images, the ROI mask was then applied to the CBV and
TP maps in a similar manner. One to 4 ROIs were obtained in each
patient, yielding a total of 83 ROIs. From each ROI the mean values
of the CBF, CBV, and TP were calculated.
Previous investigations16,17 have shown that absolute measurements of the CBF and CBV do not provide accurate results because
of limitations of the underlying model. Similarly, the TP values
within the intracranial vascular system may be influenced by the
dispersion of the contrast bolus caused by the cardiovascular status
of the patient and the contrast injection protocol. Thus, absolute
values of these parameters do not correctly reflect the perfusion
status of the ischemic tissue. To compensate for that, the perfusion
Koenig et al
Assessment of Cerebral Ischemia With Perfusion CT
433
TABLE 1. Relative Perfusion Scores* in Areas of Infarction
and Noninfarction
Tissue Outcome
rCBF
rCBV
rTP
Infarction
0.34⫾0.20
0.43⫾0.22
4.8⫾3.0 s
Noninfarction
0.62⫾0.17
0.78⫾0.18
4.3⫾1.9 s
P⬍0.001
P⬍0.001
P⫽0.30
*The relative perfusion score for CBF and CBV indicates the side-to-side
ratio; for the calculation of rTP, the difference was used.
parameters were assessed semiquantitatively by using the results
from mirrored regions within the contralateral hemisphere as reference. For the CBF and CBV, the ratio of the affected brain tissue to
the normally perfused contralateral hemisphere was calculated; for
the TP, the difference was used, thus yielding a relative perfusion
score for each functional parameter. Areas of local hyperperfusion
(both rCBF and rCBV ⬎1) suggestive of early postischemic hyperemia were excluded from the study.
Figure 2. rCBF versus rCBV in ischemic tissue compartments.
Scatterplot shows strong correlation of rCBF and rCBV for
areas of infarction () and areas of reversible ischemia (‘).
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Statistical Analysis
Standard descriptive statistics such as mean⫾SD values of rCBF,
rCBV, and rTP derived from the infarcted and noninfarcted brain
were calculated. Differences between the 2 outcome groups were
examined by the Student t test or the Wilcoxon test. The functional
data from all regions of interest were also correlated by using linear
regression analysis and the Spearman rank correlation coefficient,
including test of significance. Statistical significance was set at
P⬍0.05. Univariate discriminant analysis was used to obtain a
threshold value of each perfusion parameter to discriminate between
areas of reversible ischemia and brain tissue that underwent infarction. We also examined a combination of 2 or 3 perfusion parameters
by calculating a multivariate discriminant function. To test the
contribution of fibrinolytic therapy on tissue outcome, the treatment
modality used in our study (fibrinolysis versus heparin) was introduced into a stepwise multivariate discriminant analysis that also
included all perfusion parameters. Patients treated with rtPA were
assigned to the fibrinolysis group regardless of the degree of
recanalization that had been achieved. Finally, reclassification of our
data were performed on the basis of the best threshold values and
cutoff functions derived from the discriminant analyses to compare
the sensitivity (true positive prediction of infarction), specificity
(true positive prediction of reversible ischemia), and efficiency to
predict tissue outcome (␹2 test).
Results
respect to the rTP. Data analysis of all 83 ROIs revealed a
close relationship between the rCBF and rCBV. Depending
on the severity of hypoperfusion, a decrease of rCBF was
regularly accompanied by a reduction of rCBV, thus leading
to a strong correlation of both parameters (r⫽0.96, P⬍0.001;
Figure 2). Because of the algorithm used in our study, an
increase of rCBV ⬎1 was not commonly seen in areas of
moderate ischemia. There was a roughly linear increase of
rTP with falling rCBF, but the correlation was weak
(r⫽⫺0.496, P⬍0.001; Figure 3).
From discriminant analysis, the thresholds for the discrimination of infarcted and noninfarcted tissue were estimated to
be approximately 0.48 for rCBF and 0.60 for rCBV if both
parameters were used as a single discriminator (Figures 4 and
5). The lowest rCBF and rCBV values among areas not
developing infarcts were 0.29 and 0.40, respectively. rCBV
was superior to rCBF regarding the sensitivity, specificity,
and efficiency to predict tissue outcome, although the differences did not reach statistical significance. However, the
influence of this parameter was confirmed by multivariate
Five of 13 patients who received LIF had complete recanalization (TIMI grade 3) on posttreatment angiogram, whereas
partial recanalization (TIMI grade 2) was achieved in 4
patients. No reperfusion (TIMI grade 0) was observed in 4
patients, including the 2 patients in whom treatment had to be
stopped at an early stage because of bleeding complications.
The patient who had received intravenous thrombolysis
recovered well immediately after treatment and showed no
persisting neurological deficit on follow-up examinations.
Although the clinical benefit may be indicative for a successful recanalization, there was no information available regarding the arterial patency in this case.
There were 46 ROIs classified as gray matter infarction
and 37 ROIs derived from noninfarcted ischemic brain areas.
The mean values of the relative perfusion scores from areas of
infarction and noninfarction are shown in Table 1. Although
in infarcted regions the mean rCBF and mean rCBV were
both significantly lower than in the ischemic but noninfarcted
regions (all P⬍0.001), no differences were found with
Figure 3. rCBF versus rTP in ischemic tissue compartments.
Values for the rTP are absolute values calculated as the side-toside difference between mirrored ROIs. There is a weak negative
correlation between rCBF and rTP measured in areas of infarction () and reversible ischemia (‘).
434
Stroke
February 2001
Figure 4. rCBF (a) and rCBV (b) in infarcted and
noninfarcted ischemic gray matter areas. From
univariate discriminant analysis, the threshold
levels for the best discrimination of infarcted and
noninfarcted tissue were 0.48 and 0.60 for rCBF
and rCBV, respectively.
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
discriminant analysis of all variables included, yielding a
significance level of P⫽0.002 for the inclusion of rCBV and
P⫽0.21 for rCBF. The predominant value of rCBV was
additionally reflected by the magnitude (absolute values) of
the weighting coefficients of the best linear discriminant
function for both rCBF and rCBV: D⫽⫺2.979⫺4.084⫻
rCBF⫹8.29⫻rCBV.
As Table 2 illustrates, the rTP by itself was not a suitable
measure. Compared with the rCBV, the overall performance
of classifying areas of acute ischemia could not be increased
by any combination of the functional parameters. Moreover,
stepwise variable selection analysis did not demonstrate a
significant impact of the treatment modality on the preservation of brain tissue in our study group. Thus, no separate
analysis was performed for heparin-treated patients and for
the LIF group.
Discussion
In the past, various perfusion imaging techniques have been
reported to be efficient for the clinical investigation of acute
stroke patients. Apart from the delineation of cerebral hypoperfusion, much experimental and clinical work has focused on the
discrimination of areas of infarction and reversible ischemia
because of its considerable clinical impact.5,12,13,19 –23 Although
there is no doubt that impaired blood flow below a certain
critical level is the primary cause for functional and structural
neuronal damage, a variety of perfusion-related parameters have
been proposed to describe cerebral ischemia with respect to the
extent and severity of the perfusion deficit. We have shown
recently24 that perfusion CT is a reliable measure for the
depiction of cerebral hypoperfusion in acute stroke. Among
others, this technique has the advantage of providing functional
information from several parameters derived from a single,
routine examination. Results from a recent CT perfusion study
have demonstrated that within the affected vascular territory, the
ischemic findings as shown on the CBF, CBV, and TP maps
may differ in terms of the extent and severity, thus making the
significance of each parameter still a matter of debate.18 This
was also reported by other groups using MR perfusion maps.22,23
Thus, for a more reliable assessment of different states of
hypoperfusion, quantitative analyses of perfusion data must be
added to the qualitative aspects of functional images.
In addition to current radionuclide imaging modalities,
dynamic contrast-based MR and CT techniques have become
increasingly popular for the assessment of cerebral perfusion–related parameters. However, only PET has been considered a reliable method to define absolute viability thresholds of ischemia.7,8,10,25 It is well known that absolute normal
values of CBF and CBV may vary widely, depending on
technical factors such as the imaging modality, the algorithm
used for calculation, and the evaluation technique. In addition, measurements of both CBF and CBV have shown an
interindividual variability of ⬎20% and a decrease with age
in healthy volunteers.26 Hence, in the setting of acute stroke,
the calculation of accurate absolute values becomes impractical, and the use of absolute thresholds for therapeutic
decision making is questionable.
Intrasubject normalization of the data is a measure frequently used to deal with these problems. To meet the clinical
Figure 5. Maps of CBF (a), CBV (b), and
TP (c) in a patients with left middle cerebral artery occlusion. The extent of the
ischemic territory is clearly depicted on
the CBF and TP images. Infarction of the
frontotemporal and insular cortex (arrow)
was indicated by low values of both rCBF
and rCBV (0.18 and 0.28, respectively).
Despite a low rCBF (0.40), reversible ischemia was predicted for the posterior
temporal cortex (arrowhead) because of
an rCBV value of 0.68. This was confirmed by the CT scan at day 2 (d).
Koenig et al
Assessment of Cerebral Ischemia With Perfusion CT
TABLE 2. Sensitivity, Specificity, and Efficiency of
Perfusion-Related Parameters and Therapeutic Regimen to
Predict Tissue Outcome
Discriminant Analysis
(Variables to be Included)
Sensitivity
(%)*
Specificity
(%)†
Efficiency
(%)
rTP
52.2
45.9
49.4
rCBF
76.1
73.0
74.7
rCBV
80.4
86.5
83.1
rCBF, rCBV
76.1
89.2
81.9
rCBF, rCBV, rTP
76.1
89.2
81.9
rCBF, rCBV, rTP, therapy‡
76.1
89.2
81.9
*Sensitivity indicates true positive prediction of infarction.
†Specificity indicates true positive prediction of reversible ischemia.
‡Therapy indicates treatment with thrombolytics versus heparin.
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
need for prompt and detailed informations concerning the
hemodynamic state of an ischemic tissue compartment, we
calculated a relative perfusion score for CBF, CBV, and TP
by using the contralateral hemisphere as reference. This was
done on a routine basis without any need for time-consuming
postprocessing.
With this approach we were able to demonstrate significant
lower values of rCBF for areas of infarction compared with
those suffering from reversible ischemia (0.34⫾0.2 and
0.62⫾0.17, respectively). Similar results (0.39⫾0.12 and
0.69⫾0.15, respectively) were reported by Hatazawa et al13
by using the side-to-side ratio of radioactivity derived from
HMPAO SPECT; in other SPECT studies,27,28 mean values
for both infarcted and noninfarcted tissue were found to be
slightly higher. MR perfusion maps have also been used for
the quantitative assessment of ischemic tissue compartments,
but mean values of relative CBF have shown to be substantially lower than those mentioned above.23 The differences
might mainly be due to the different imaging modalities and
evaluation procedure, because for the perfusion MRI study
the ischemic penumbra was operationally defined to those
ischemic regions that developed an infarct as shown on
follow-up diffusion-weighted MR images. Thus the penumbra area was not compatible with those regions that had been
referred to as “reversible ischemia” in our study.
As demonstrated by PET and perfusion MR imaging
studies, an increase of CBV is a common finding in ischemic
areas affected by moderate hypoperfusion.13,23,29,30 Unlike
others who use the area under the tissue concentration–time
curve for the calculation of CBV, we prefer the ratio of the
maximum enhancement values obtained from the parenchyma and from an intravascular voxel not affected from
partial volume averaging. We may reasonably assume that
our approach leads to an underestimation of CBV for areas
perfused mainly by collateral flow. In the present study, brain
tissue undergoing infarction showed a significant reduction of
mean rCBV compared with the surrounding ischemic area
(0.43⫾0.22 versus 0.62⫾0.17), but, as expected theoretically,
we did not observe an elevation of CBV in regions with
moderate hypoperfusion. Schlaug and coworkers23 have demonstrated that various indexes of relative CBV derived from
a dynamic bolus-contrast study may be valid for the discrimination of the infarct core and the peri-infarct ischemic tissue.
435
However, quantitative comparison did show significant differences between the results of these functional estimates of
CBV. This study and ours support the view that for different
states of hypoperfusion the results of CBV measurements
may strongly depend on the algorithm used making a comparison of the data obtained with different techniques a
challenge.
Despite these limitations, our results add to the notion that
in hyperacute stroke the perfusion impairment is associated
with changes of both CBF and CBV. In our patient group a
severe reduction of rCBF was always followed by a marked
decrease of rCBV indicating the core of infarction, while in
the borderzone with only moderate hypoperfusion the rCBV
was maintained to nearly normal or only slightly decreased
levels. The high correlation of both perfusion parameters is in
agreement with the study of Sorensen et al.22
Temporal changes of the local contrast bolus dynamic have
been well documented for hypoperfused tissue compartments
and among those parameters characterizing the bolus transit
the TP may be easily derived from the concentration-time
curves. Our observations confirm that the prolongation of the
TP obviously reflects flow via collateral pathways or sluggish
flow. Comparing the TP values from the affected and nonaffected hemisphere, we found bolus delays up to a maximum
of 10 seconds. Unlike observations from an animal study,21
we did not see areas with a lack of a bolus peak indicating no
flow, which is probably due to some residual vascular
enhancement that could always be noted within large-sized
ROIs even in infarct lesions.
Although in previous studies TP maps have been shown to
provide excellent information in terms of the extent of the
perfusion impairment,18,31 this parameter may not serve as a
good quantitative estimate of flow because of the weak
correlation to the corresponding rCBF values. Areas of
reversible and irreversible ischemia did not significantly
differ with respect to their bolus delay; in contrast to
Neumann-Haefelin et al,12 we therefore suggest that the TP
cannot reliably be used to discriminate between infarcted and
noninfarcted ischemic tissue.
A value of 0.48 for rCBF was found as a best estimate for
the discrimination of the infarct and peri-infarct regions,
yielding an efficiency of 74.7% for the prediction of tissue
outcome. rCBV tended to be more efficient in predicting the
fate of the ischemic tissue (83.1%) if a cutoff value of 0.6 was
used. It is, however, important to recognize that because of
the overlap of the data from the infarcted and noninfarcted
regions, these best cutoff values do not reflect the lowest level
(ie, the real threshold for a given tissue compartment to
survive an ischemic injury). In addition, the data presented
here were obtained from regions that had been treated with
heparin or thrombolytic therapy, and no distinction was made
between whether an area was successfully reperfused or not.
Only a small number of patients had complete arterial
patency (TIMI 3) after LIF, and in those cases with partial
restoration of flow no definite evaluation of the ROIs could
be performed in relation to the reperfused and nonreperfused
arterial branches. We therefore suggest that the cutoff values
would have been even lower if a sufficient number of patients
with full posttreatment recanalization had been recruited.
436
Stroke
February 2001
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
This would also have allowed dealing with the true viable
tissue, ie, the “potentially salvageable brain,” instead of
“reversible ischemia,” as defined in our study. Taking into
account that in noninfarcted regions the lowest perfusion
values (rCBF 0.29 and rCBV 0.4) were obtained from the
fibrinolysis group, these levels can be used as a rough
estimate to separate viable from nonviable tissue. Assuming
normal values of CBF for cerebral gray matter to be in the
range of 55 to 70 mL/100 mL⫺1 䡠 min⫺1 as derived from PET
and xenon CT studies, a rCBF value of 0.29 corresponds to an
absolute value of about 15 to 20 mL/100 mL⫺1 䡠 min⫺1, which
is in agreement with the widely accepted critical threshold of
ischemia.7,8,26,32,33
In a recent comparative MR perfusion study, Sorensen et
al22 have shown that although not perfect, CBV is a better
predictor of the final infarct size than CBF. No quantitative
analysis of the perfusion parameters was performed in this
study; rather, it was based on the assessment of the lesion size
as shown on the perfusion maps. The authors hypothesized
that the distinction of already-infarcted tissue and tissue at
risk would possibly be improved if a combination of various
perfusion-related parameters and the apparent diffusion coefficient were used. Our results from multivariate discriminant
analysis also confirmed the thesis that rCBV is superior to
rCBF to separate areas of infarction and noninfarction.
Regarding the slightly better sensitivity and specificity of
rCBV, we speculate that within a critically perfused area
structural damage may be prohibited by maintenance of a
certain level of CBV. This vascular volume has also been
referred to as perfused CBV, indicating the fraction of the
vascular compartment that still receives some residual blood
supply.34 It probably points the way for the outcome of the
ischemic tissue. Unlike considerations from Sorensen et al,
the combination of rCBF and rCBV did not offer any benefit
in predicting the fate of the hypoperfused brain. Also,
inclusion of rTP did not improve the overall efficiency level
reached by the use of the blood flow and blood volume
parameters derived from perfusion CT.
We were not able to demonstrate an additional benefit of
the therapeutic regimen on the prediction of tissue outcome,
which is apparently due to the small number of patients in
whom full arterial patency was achieved after thrombolytic
therapy. To avoid erroneous conclusions from the slightly
limited predictability of noncritical (ie, reversible) ischemia,
it must be considered that during the days after the perfusion
study there might have been various unknown events that led
to further deterioration of blood flow and subsequent metabolic damage in certain cases. This limitation, however, holds
true for all functional modalities commonly used for the
assessment of acute stroke. Nevertheless, results from
diffusion-weighted MR imaging may add further information
to a certain number of regions (eg, areas of postischemic
hyperemia) that may be already definitely infarcted although
their perfusion values cannot be categorized as critical at the
time of the initial perfusion study.
Regarding the 3D capability of MRI, it might be criticized
that the information obtained with perfusion CT is still
restricted to a single section that has to be carefully set to
cover the level suspected to be ischemic; however, in the near
future the problem may partly be overcome by using modern
multislice CT technology. Furthermore, we suggest that the
expertise gained with perfusion CT may easily be translated
to MR perfusion techniques as far as comparable algorithms
are implemented.
In conclusion, within a certain time window patients
suffering from symptoms of acute stroke may be candidates
for thrombolytic therapy if further criteria of an adequate
selection are fulfilled. In the present study, we quantitatively
described the hemodynamic state of the ischemic brain by
means of perfusion-related parameters derived from perfusion CT.
Although rCBV was shown to be superior to rCBF, both
parameters proved to be efficient for the prediction of
whether an ischemic area survived or to became infarcted.
Quantitative analyses may be restricted to these parameters,
because the rTP values did not allow to discriminate between
infarcts and reversible ischemia. It may be assumed that the
recruitment of a larger number of successfully recanalized
patients will make the results of CT perfusion measurements
even more conclusive in terms of defining thresholds of tissue
viability. With this future outlook, quantitative assessment of
the ischemic brain by means of perfusion CT may play an
increasing role for the selection of a subset of patients who
may be successfully treated with potentially harmful therapeutic regimens such as thrombolysis.
References
1. del Zoppo GJ, Higashida RT, Furlan AJ, Pessin MS, Rowley HA, Gent
M, and the PROACT Investigators. PROACT: a phase II randomized trial
of recombinant pro-urokinase by direct arterial delivery in acute middle
cerebral artery stroke. Stroke. 1998;29:4 –11.
2. Furlan A, Higashida R, Wechsler L, Gent M, Rowley H, Kase C, Pessin
M, Ahuja A, Callahan F, Clark WM, Silver F, Rivera F. Intra-arterial
prourokinase for acute ischemic stroke: the PROACT II Study: a randomized controlled trial. JAMA. 1999;282:2003–2011.
3. Ueda T, Hatakeyama T, Kumon Y, Sakaki 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.
4. 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 acute cerebral embolism. Neurosurg Rev. 1996;19:231–236.
5. Ueda T, Sakaki S, Yuh WT, Nochide I, Ohta S. Outcome in acute stroke
with successful intra-arterial thrombolysis and predictive value of initial
single-photon emission-computed tomography. J Cereb Blood Flow
Metab. 1999;19:99 –108.
6. Heiss WD, Grond M, Thiel A, von Stockhausen HM, Rudolf J, Ghaemi
M, Löttgen J, Stenzel C, Pawlik G. Tissue at risk of infarction rescued by
early reperfusion: a positron emission tomography study in systemic
recombinant tissue plasminogen activator thrombolysis of acute stroke.
J Cereb Blood Flow Metab. 1998;18:1298 –1307.
7. Powers WJ, Grubb RL, Darriet D, Raichle ME. Cerebral blood flow and
cerebral metabolic rate of oxygen requirements for cerebral function and
viability in humans. J Cereb Blood Flow Metab. 1985;5:600 – 608.
8. Baron JC. Positron emission tomography studies in ischemic stroke. In:
Barnett HJM, Mohr JP, Stein BM, Yatsu FM, eds. Stroke: Pathophysiology, Diagnosis, and Management. 2nd ed. New York, NY: Churchill
Livingston; 1992:111–123.
9. Hossmann KA. Viability thresholds and the penumbra of focal ischemia.
Ann Neurol. 1994;36:557–565.
10. Heiss WD, Herholz K. Assessment of the pathophysiology of stroke by
positron emission tomography. Eur J Nucl Med. 1994;21:455– 465.
11. Touzani O, Young AR, Derlon JM, Beaudouin V, Marchal G, Rioux P,
Mezenge F, Baron JC, MacKenzie ET. Sequential studies of severely
hypometabolic tissue volumes after permanent middle cerebral artery
occlusion. Stroke. 1995;26:2112–2119.
Koenig et al
Assessment of Cerebral Ischemia With Perfusion CT
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
12. Neumann-Haefelin T, Wittsack HJ, Wenserski F, Siebler M, Seitz RJ,
Mödder U, Freund HJ. Diffusion- and perfusion-weighted MRI. The
DWI/PWI mismatch region in acute stroke. Stroke. 1999;30:1591–1597.
13. Hatazawa J, Shimosegawa E, Toyoshima H, Ardekani BA, Suzuki A,
Okudera T, Miura Y. Cerebral blood volume in acute brain infarction: a
combined study with dynamic susceptibility contrast MRI and 99mTCHMPAO-SPECT. Stroke. 1999;30:800 – 806.
14. The National Institute of Neurological Disorders and Stroke rt-PA
Stroke Study Group. Tissue plasminogen activator for acute ischemic
stroke. New Eng J Med. 1995;333:1581–1587.
15. Chesebro JH, Knatterud G, Roberts R, Borer J, Cohen LS, Dalen J,
Dodge HT, Francis CK, Hillis D, Ludbrook P, Markis JE, Mueller H,
Passamani ER, Powers ER, Rao AK, Robertson T, Ross A, Ryan TJ,
Sobel BE, Willerson J, Williams DO, Zaret BL, Braunwald E. Thrombolysis in Myocardial Infarction (TIMI) Trial, phase I: a comparison
between intravenous tissue plasminogen activator and intravenous streptokinase: clinical findings through hospital discharge. Circulation. 1987;
76:142–154.
16. Klotz E, König M. Perfusion measurements of the brain: using dynamic
CT for the quantitative assessment of cerebral ischemia in acute stroke.
Eur J Radiol. 1999;30:170 –184.
17. König M, Klotz E, Heuser L. Zerebrale Perfusions-CT- Theoretische
Grundlagen, methodische Realisierung und praktische Erfahrungen in
der Diagnostik des ischämischen Hirninfarkts. Fortschr Röntgenstr.
2000;172:210 –218.
18. König M, Banach-Planchamp R, Kraus M, Klotz E, Falk A, Gehlen W,
Heuser L. CT-Perfusionsbildgebung beim akuten ischämischen Hirninfarkt: Vergleich von Parameterbildern der zerebralen Perfusion und
Nativ-CT-Befunden. Fortschr Röntgenstr. 2000;172:219 –226.
19. Jones TH, Morawetz RB, Crowell RM, Marcoux FW, FitzGibbon SJ,
DeGirolami U, Ojemann RG. Thresholds of focal cerebral ischemia in
awake monkeys. J Neurosurg. 1981;54:773–782.
20. Röther J, Gückel F, Neff W, Schwartz A, Hennerici M. Assessment of
regional cerebral blood volume in acute human stroke by use of
single-slice dynamic susceptibility contrast-enhanced magnetic resonance imaging. Stroke. 1996;27:1088 –1093.
21. Lo EH, Rogowska J, Bogorodzki P, Trocha M, Matsumoto K, Saffran B,
Wolf GL. Temporal correlation analysis of penumbral dynamics in focal
cerebral ischemia. J Cereb Blood Flow Metab. 1996;16:60 – 68.
22. Sorensen AG, Copen WA, Ostergaard L, Buonanno FS, Gonzalez RG,
Rordorf G, Rosen BR, Schwamm LH, Weisskoff RM, Koroshetz WJ.
Hyperacute stroke: simultaneous measurement of relative cerebral blood
volume, relative cerebral blood flow, and mean tissue transit time.
Radiology. 1999;210:519 –527.
437
23. Schlaug G, Benfield A, Baird A, Siewert B, Lovblad KO, Parker RA,
Edelman RR, Warach S. The ischemic penumbra: operationally defined
by diffusion and perfusion MRI. Neurology. 1999;53:1528 –1537.
24. König 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;209:85–93.
25. 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;122:2387–2400.
26. Leenders KL, Perani D, Lammertsma AA, Heather JD, Buckingham P,
Healy MJ, Gibbs JM, Wise RJ, Hatazawa J, Herold S, et al. Cerebral
blood flow, blood volume and oxygen utilization: normal values and
effect of age. Brain. 1990;113:27– 47.
27. Sasaki O, Takeuchi S, Koizumi T, Koike T, Tanaka R. Complete
recanalization via fibrinolytic therapy can reduce the number of ischemic territories that progress to infarction. AJNR Am J Neuroradiol.
1996;17:1661–1668.
28. Shimosegawa E, Hatazawa J, Inugami A, Fujita H, Ogawa T, Aizawa Y,
Kanno I, Okudera T, Uemura K. Cerebral infarction within six hours of
onset: prediction of completed infarction with technetium-99m-HMPAO
SPECT. J Nucl Med. 1994;35:1097–1103.
29. Powers WJ, Grubb RL, Raichle ME. Physiological responses to focal
cerebral ischemia in humans. Ann Neurol. 1984;16:546 –552.
30. Gückel F, Brix G, Rempp K, Deimling M, Röther J, Georgi M.
Assessment of cerebral blood volume with dynamic susceptibility
contrast enhanced gradient-echo imaging. J Comput Assist Tomogr.
1994;18:344 –351.
31. Reichenbach JR, Röther J, Jonetz-Mentzel L, Herzau M, Fiala A,
Weiller C, Kaiser WA. Acute stroke evaluated by time-to-peak mapping
during initial and early follow-up perfusion CT studies. AJNR Am J
Neuroradiol. 1999;20:1842–1850.
32. Baron JC. Ischemic stroke studied by 15O-labeled compounds: misery
perfusion and luxury perfusion. In: Heiss WD, Pawlik G, Herholz K,
Wienhard K, eds. Clinical Efficacy of Positron Emission Tomography.
Dordrecht, Germany: Martinus Nijhoff, 1987:15-23.
33. Segawa H. Normative values. In: Tomonaga M, Tanaka A, Yonas H,
eds. Quantitative Cerebral Blood Flow Measurements Using Stable
Xenon/CT: Clinical Applications. Armonk, NY: Futura Publishing
Company; 1995:61– 66.
34. Hunter GJ, Hamberg LM, Ponzo JA, Huang-Hellinger FR, Morris PP,
Rabinow J, Farkas J, Lev MH, Schaefer PW, Ogilvy CS, Schwamm L,
Buonanno F, Koroshetz WJ, Wolf GL, Gonzalez RG. Assessment of
cerebral perfusion and arterial anatomy in hyperacute stroke with threedimensional functional CT: early clinical results. AJNR Am J Neuroradiol. 1998;19:29 –37.
Quantitative Assessment of the Ischemic Brain by Means of Perfusion-Related Parameters
Derived From Perfusion CT
Matthias Koenig, Michael Kraus, Carmen Theek, Ernst Klotz, Walter Gehlen and Lothar Heuser
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Stroke. 2001;32:431-437
doi: 10.1161/01.STR.32.2.431
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 2001 American Heart Association, Inc. All rights reserved.
Print ISSN: 0039-2499. Online ISSN: 1524-4628
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://stroke.ahajournals.org/content/32/2/431
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published
in Stroke can be obtained via RightsLink, a service of the Copyright Clearance Center, not the Editorial Office.
Once the online version of the published article for which permission is being requested is located, click
Request Permissions in the middle column of the Web page under Services. Further information about this
process is available in the Permissions and Rights Question and Answer document.
Reprints: Information about reprints can be found online at:
http://www.lww.com/reprints
Subscriptions: Information about subscribing to Stroke is online at:
http://stroke.ahajournals.org//subscriptions/