183
Quantification of Regional Cerebral Blood
Flow With IMP-SPECT
Reproducibility and Clinical Relevance of Flow Values
I. Podreka, MD, C. Baumgartner, MD, E. Suess, MD, C. Miiller, MD, T. Briicke, MD,
W. Lang, MD, F. Holzner, MD, M. Steiner, PhD, and L. Deecke, MD
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Single-photon emission computed tomography with iV-isopropyl[mI]-p-iodoamphetamine (IMPSPECT) was performed in 14 normal volunteers (seven men and seven women aged 25.1±5.3
years) and 29 patients with cerebrovascular disease (18 men and 11 women aged 54.1±13.7
years). The fluid microsphere model was used to estimate cerebral blood flow (CBF). Normal
subjects were scanned twice, 1 week apart, to determine the reproducibility of the CBF
estimates. Hemispheric blood flow (hCBF) was calculated as the mean of regional cerebral blood
flow (rCBF) values in 16 gray matter regions per hemisphere. In normal subjects mean hCBF
was 68 ml/100 g/min. The highest rCBF was found in the occipital cortex, followed by the
frontal, temporal, and parietal cortexes. CBF values were reproducible (p<0.001 except the
right thalamic region, where p<0.01). Intraindividual variation ranged between 0.3% and
15%. Women exhibited significantly higher (16%, p<0.02) CBF than men. Patients were
subdivided into groups with reversible (n=19) and persistent (n=10) symptoms. Significant
hCBF differences between the affected and the contralateral hemispheres were recorded only in
the group with reversible symptoms (/?<0.005), whereas the group with persistent symptoms
showed a significant bilateral decrease of hCBF compared with normal subjects and patients
with reversible symptoms. Focal CBF was significantly lower in patients with completed stroke
than in patients with transient symptoms (p<0.001). Our results indicate that IMP-SPECT can
be used for the routine estimation of CBF in normal and pathologic states. (Stroke 1989;
20:183-191)
I
n the last few years new tracers, such as Nisopropyl[l23I]-/9-iodoamphetamine (IMP)1-2 and
[""Tclhexamethylpropyleneamineoxime (Tc99m-HMPAO),3-4 for visualizing cerebral blood flow
(CBF) using single-photon emission computed
tomography (SPECT) have been synthesized. It is
generally accepted that the distribution of both
tracers in the brain during the first 2 hours after
their administration reflects regional cerebral blood
flow (rCBF). However, quantification of CBF with
Tc-99m-HMPAO seems to be at present rather
difficult due to partially unknown tracer kinetics.
Single attempts to quantify CBF from IMP-SPECT
images have been undertaken, but the reproducibility and clinical reliability of the obtained values
has never been investigated. Our purpose is to
describe the possibility of CBF quantification with
From the Abteilung fflr Neuronuklearmedizin der Neurologischen, UniversitStsklinik Wien, Wien, Austria.
Address for reprints: Doz. Dr. 1. Podreka, Abteilung fflr
Neuronuklearmedizin der Neurologischen Universitatsklinik
Wien, Lazarettgasse 14, A-1090 Wien, Austria.
Received November 9, 1987; accepted July 15, 1988.
IMP-SPECT using a rotating gamma camera. Furthermore, we report the reproducibility of CBF
values obtained in normal volunteers as well as the
relation between calculated CBF values and the
duration of neurologic impairment in patients with
cerebrovascular disease (CVD).
Subjects and Methods
The reproducibility of known concentrations of
iodine-123 in SPECT images was investigated by an
approach similar to that described by Kuhl et al.7
First, a Lucite cylinder 200 mm in diameter (the
phantom) filled with 0.5 /iCi/ml of an iodine-123
solution was scanned as for patient investigations
(see below). The average number of counts per
voxel (count density) in the reconstructed image
was then converted into microcuries per voxel.
Then, 19 cylindrical plastic bottles 29 mm in diameter (approximately twice full-width half-maximum
[FWHM]8) were placed in a concentric arrangement
inside the phantom. Fourteen bottles contained
solutions of different iodine-123 assays, while five
bottles were filled with water (Figure 1) to investi-
184
Stroke
Vol 20, No 2, February 1989
PHANTOM
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
I
3 am
I
FIGURE 1. Schematic representation of phantom and
regression analysis of true and measured iodine-123
assay concentrations. Upper numbers in small circles
and numbers adjacent to each regression point refer to
position of bottle in phantom. Lower numbers inside
small circles give true iodine-123 assay concentration (as
microcuries per milliliter) in each bottle.
gate the efficiency of the applied scatter correction
in depth.
CBF was quantified using the fluid microsphere
model proposed by Kuhl et al,5 in which CBF=
(RxCb)-rO and R=constant withdrawal rate of arterialized venous blood in milliliters per minute,
Cb=local tracer concentration in microcuries per
100 grams trapped by the brain in the first 5 minutes,
and O=octanol extraction of the (true) tracer concentration in the withdrawn blood in microcuries.
For SPECT, we used a dual-head rotating scintillation camera (Siemens Gammasonics, Uithoorn,
The Netherlands, Dual Rota ZLC37) connected to a
computer (Nodecrest Micas 2000). Due to low sen-
sitivity, fast SPECT scans cannot be performed
with this instrumentation. Therefore, the total
amount of tracer trapped by the brain 1-5 minutes
after intravenous bolus injection of 5-6 mCi IMP
(iodine-123 produced by the p,5n reaction and free
of iodine-124) had to be estimated by conventional
scanning (anteroposterior projection) during this
period. To maintain the method nearly noninvasive,
arterialized venous blood (hand placed in a 44° C
water bath) was withdrawn constantly (R=l ml/
min) from a vein in the hand at the same time.
Thirty minutes after IMP injection, when brain
tracer concentration was constant, another static
scan of the head (anteroposterior projection) during
5 minutes was obtained. Immediately thereafter,
SPECT scanning was started. The scanning modalities and image processing have been reported in
detail,9 and here we give just a brief description.
Sixty projections (2x30 angles; 100 sec/angle) with
a sampling distance of 3.125 mm were achieved.
The camera heads were equipped with low-energy
all-purpose (LEAP) collimators (FWHM 1.4 cm in
the reconstructional plane). The subject was placed
on a reclining dentist's chair with a special head
support to reduce the diameter of rotation to 250
mm. Scatter correction was performed using a
multiple energy window technique. Projections were
filtered with a filter of variable shape and size for
reduction of Poisson noise," and data were reorganized in a sinogram. After attenuation correction,
3.125-mm-thick cross sections were reconstructed
by filtered back projection in 128x128 matrices.
Consecutive summation of seven sections gave a
set of 21.9-mm-thick transverse slices. Regions of
interest (ROIs) were drawn on four adjacent cross
sections. The correct position of these cross sections was established on anteroposterior and lateral
projections (Figure 2).
Since the obtained SPECT images represent IMP
distribution 35-85 minutes after tracer administration (during the stable phase), they must be timecorrected to show total tracer uptake during the
inflow phase (i.e., 1-5 minutes after intravenous
injection of IMP). This correction is performed as
follows: ROIs are placed over each hemisphere on
the early (1-5 minutes) and late (30-35 minutes)
anteroposterior scan, and total counting rates are
obtained in these ROIs. The subsequent calculations are made for each hemispheric ROI. The total
counting rate in the late stable-phase anteroposterior scan (LScts) is proportional to the total tracer
amount (LP%, 100%) delivered to the brain between
1 and 30 minutes after IMP injection. Therefore,
LP%=EP%+CP%, where EP% is the count density
in the early anteroposterior scan expressed as percent (EP%= 100%xEScts/LScts), EScts is the counting rate in the early anteroposterior scan, and CP%
is the unknown percentage of tracer delivered to the
brain between 6 and 29 minutes. CP% is therefore
CP%=LP%-EP% (mean±SD CP% in normal volunteers was 34.5±6.2%). Since the late anteropos-
Podreka et al
CBF Quantified by IMP-SPECT
185
FIGURE 2.
ana-
Top:
For
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
tomic orientation, position of
four transverse single-photon
emission computed tomography slices (second row) is displayed on anteroposterior
(A-P) and left lateral (L-LAT)
projections. Third row. Region
of interest scheme used: SF,
mesiofrontal; MF, laterofrontal; IF, inferior laterofrontal (also containing Broca's area); C, central; SP+IP,
superior and inferior parietal;
SO, superior occipital; 10, inferior occipital; ST, superior temporal; IT, inferior temporal;
BG, anterior basal ganglia
(caudate+lentiform nucleus);
TH, thalamus; HI, hippocampal structures.
terior scan and the SPECT study were both performed during the stable phase, the SPECT image
ROI can be time-corrected by a proportionate reduction in counts, corresponding to the hemispheric
CP%. This gives the counting rate that would result
from a 50-minute SPECT recording of a brain that
contains the amount of tracer trapped in the first 5
minutes.
For estimation of normal CBF and the reproducibility of the estimates, 14 controls (normal volunteers, seven men and seven women aged 19-40
[mean±SD 25.1 ±5.3] years) were scanned twice, 1
week apart, after they had given written informed
consent. In addition, 29 patients (18 men and 11
women aged 27-72 [mean±SD 54.1 ±13.7] years)
suffering from CVD were investigated. SPECT was
performed 1-3 weeks after the ischemic attack.
Four patients had had a transient ischemic attack
(TIA), 15 a reversible ischemic neurologic deficit
(RIND), and 10 a completed stroke (CS). No SPECT
study showed a luxury perfusion. End-tidal Pco2
was in the normal range (controls: first measurement mean±SD 38.0±2.2 mm Hg, second measurement 38.2±2.1 mm Hg; patients: 37.8±2.5 mm Hg).
Subjects were studied with their eyes closed and
their ears unplugged. The room was dimly lit, and
background noise originated from cooling fans.
Results
Figure 1 shows a schematic representation of the
phantom used and the plot of the calculated correlation and regression analysis of true and measured
tracer concentrations. Iodine-123 concentrations in
the phantom were similar to those found in normal
or ischemic brain tissue when 5-6 mCi of IMP are
applied. Despite these low isotope concentrations,
the measured values were highly reproducible in
volumes of 14.5 ml (r=0.984, /?<0.001; r=0.045
+0.867A0. The slope of the regression line was
somewhat decreased by the effect of Compton
radiation in bottles containing water and placed in
the center of the phantom (Bottles 2 , 3 , and 4).
Controls
Table 1 shows the calculated hemispheric cerebral blood flow (hCBF) and rCBF values obtained
from controls in the two measurements. The means
were comparable for both measurements. As shown
by the correlation coefficient (r), hCBF and rCBF
values were reproducible (p<0.001 for all regions
except the right laterofrontal and right thalamic
regions, where p<0.01). The intraindividual variation between the measurements ranged from 0.3%
to 15% (mean 6.2%). The regression analysis of
hCBF is shown in Figure 3. By calculating the mean
for each lobe, the highest rCBF was found in the
occipital lobe (left [L] 73.2, right [R] 72.0 ml/100 g/
min), followed by the frontal lobe (L 67.9, R 67.9 ml/
100 g/min), temporal lobe (L 66.7, R 66.5 ml/100 g/
min), and parietal lobe (L 63.3, R 62.8 ml/100 g/min).
Since central cortex regions cover tissue pertaining
to the frontal and parietal lobes, they were treated
separately and showed the lowest rCBF (L 60.6, R
61.0 ml/100 g/min).
When hCBF of male and female controls were
compared using the two-tailed independent t test,
differences were found between the sexes. Women
showed significantly higher (16%; p<0.02) hCBF
186
Stroke
Vol 20, No 2, February 1989
TABLE 1. Correlation and Regression Analysis of Reproduclbility of Two Cerebral Blood Flow Values Measured 1 Week Apart in 14
Normal Volunteers Using IMP-SPECT
Left
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
IM
Hemispheric flow
67.8±IO.2
Regional flow
SF
1
58.8±9.9
2
70.4±ll.2
71.7±9.4
3
MF
67.1 + 10.5
IF
71.5±10.5
c
\^
57.5±9.1
1
2
63.7+9.5
SP
59.2±8.8
IP
67.3+10.1
74.3±12.2
SO
72.1±ll.l
10
68.7±10.7
ST
IT
64.8±10.0
66.4±9.2
HI
BG
75.8± 13.9
76.4±11.4
TH
2M
Right
T
b
m
68.4±9.6
0.869*
12.84
0.82
60.5±9.0
71.0±10.8
73.3+10.3
67.2±9.8
72.3±9.9
0.866*
14.20
0.841*
0.872*
0.850*
0.819*
13.71
0.79
0.81
0.96
0.810*
0.844*
0.868*
0.833*
0.802*
0.844*
0.843*
0.844*
0.794*
58.3+8.2
63.8±9.8
60.1 ±9.2
66.6±10.0
74.3+11.0
72.7+9.8
69.6±10.9
65.2±8.6
66.8±9.6
74.9±9.4
77.4±ll.4
0.880*
0.866*
4.59
13.79
0.80
17.17
0.77
16.20
8.58
6.62
11.09
20.75
18.63
10.51
18.08
12.03
29.74
11.21
0.73
0.87
0.90
0.83
0.72
0.75
0.86
0.73
0.83
0.60
0.88
2M
r
b
m
67.2±9.8
68.0±9.7
0.857*
11.10
0.85
57.7± 10.0
68.6+11.1
71.5±9.3
68.9±10.9
72.8±10.9
59.6±10.0
68.8±10.4
72.1+9.8
68.5±10.6
72.9+10.4
0.890*
0.878*
0.850*
0.7601
0.860*
8.43
12.51
8.26
17.84
5.19
0.89
0.82
0.89
0.73
0.94
56.5±8.7
65.4±9.9
58.7±9.3
66.8±10.3
73.4±11.2
70.5±ll.l
70.1 + 10.9
66.1 ±9.0
63.3±9.2
70.2±9.5
74.8±11.4
57.1+8.4
65.6±9.7
60.2±9.1
67.4+9.1
74.6±10.4
72.7+10.8
70.8±10.4
67.1 ±8.7
64.5±IO.3
0.881*
0.857*
0.842*
0.853*
0.782*
0.835*
0.871*
0.791*
0.798*
0.837*
0.689t
8.65
10.48
11.63
17.01
21.19
15.67
12.74
16.51
7.82
7.65
24.25
0.86
0.84
0.83
0.76
0.73
0.81
0.83
0.77
0.90
0.92
0.67
IM
72.1 ±10.5
74.4±ll.l
Data are mean±SD ml/100 g/min. IMP-SPECT, single-photon emission computed tomography with AMsopropyl[l2JI]p-iodoamphetamine; IM, first measurement; 2M, second measurement; r, correlation coefficient; b, intercept; m, slope. SF,
mesiofrontal; MF, laterofrontal; IF, inferior laterofrontal (also containing Broca's area); C, central; SP, superior parietal; IP, inferior
parietal; SO, superior occipital; IO, inferior occipital; ST, superior temporal; IT, inferior temporal; BG, basal ganglia (caudate+lentiform
nucleus); TH, thalamus; HI, mesiotemporal parts of limbic system. Numerals refer to transverse slices (see Figure 2).
*tp<0.001, p<0.01, respectively.
than men in both measurements (first measurement
mean±SD for L hCBF: women 73.9+11.3, men
61.8±3.4 ml/100 g/min 0=2.73]; R hCBF: women
73.0±10.8, men 61.3±3.5 ml/100 g/min [f=2.71]).
No significant intraindividual rCBF changes were
found between the two measurements using the
two-tailed paired / test. When rCBF values for the
first measurement were compared between hemispheres, significant side-to-side differences were
found (mesiofrontal in first slice, L>R, r=2.42,
p<0.05; mesiofrontal in the second slice, L>R,
/=3.31, p<0.01; laterofrontal R>L, t=3.13, p<0.01;
central cortex in second slice, R>L, /=3.36, p<0.01;
anterior basal ganglia, L>R, /=3.62, p<0.01; inferior occipital L>R, f=2.37, p<0.05; hippocampal
L>R, /=3.61,p<0.01).
Patients
hCBF of the 29 CVD patients was significantly
lower by two-tailed independent t test than hCBF of
the controls (L: 55.6±17.7 ml/100 g/min, r=2.39,
p<0.025; R: 58.7+16.1 ml/100 g/min, f=2.08,
p<0.05. Mean±SD hCBF values of the control
group are given in Table 1).
CVD patients were divided into a subgroup of 19
patients with reversible symptoms (TIA and RIND)
and a subgroup of 10 patients with persistent neu-
rologic deficit (CS). hCBF values of both subgroups
were compared with those of the controls by oneway analysis of variance (Table 2). hCBF of the
subgroup with reversible symptoms did not differ
from that of controls despite significantly higher
age, whereas hCBF of the subgroup with persistent
neurologic deficit was significantly lower than that
of the controls and of the subgroup with reversible
symptoms (Newman-Keuls test, p<0.0l).
Table 3 shows the comparison of hCBF values
between the affected and contralateral hemispheres
of the 29 CVD patients. hCBF in the affected
hemisphere was significantly lower than that in the
contralateral hemisphere by two-tailed paired t test
(7=3.059, p<0.005) due to a significant interhemispheric difference in hCBF in the subgroup with
reversible symptoms (/=3.234,p<0.005). No significant difference was obtained in the subgroup with
persistent neurologic deficit (f=2.043) since hCBF
was decreased in both hemispheres.
When rCBF side-to-side percentage differences
of the control group were estimated, the mean±2SD
for each ROI did not exceed 12%. Therefore, ischemic regions were defined as those with rCBF
values that were 15% (probability of correct definition of >95%) lower than those of their contralateral counterpart. If ischemia extended over several
Podreka et al CBF Quantified by IMP-SPECT
187
TABLE 3. Comparison of Hemispheric Cerebral Blood Flow
Measured Using IMP-SPECT in Patients With Cerebrovascular
Disease
LEFT HEMISPHERE
Hemisphere
n
Affected
29 55.4±17.2
Total group
TIA+RIND 19 63.5+14.0
10 39.8+10.9
CS
Contralateral
58.9±I6.6
65.7±I5.7
46.1±9.2
-3.059 <0.005
-3.234 <0.005
-2.043 NS
Data are mean+SD ml/100 g/min. IMP-SPECT, single-photon
emission computed tomography using N-isopropyl[ m I]p-iodoamphetamine. TIA+RIND, reversible symptoms; CS, persistent neurologic deficit, t value obtained from paired t test. NS,
not significant.
RIGHT HEMISPHERE
to
X
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
i
•
i
~i
>•
^^
W-
TO-
O
^
/
r«itor • at<Jt
^
-/"^
^x^
S^°
• « •
^
s^
o
O
•
0
o
0
*
FIGURE 3. Regression analysis of hemispheric blood
flow values obtained in two sessions in 14 normal volunteers. IM, first measurement; 2M, second measurement;
•, women; O, men.
ROIs, then the mean of these rCBF values was used
to compare sides. In 17 of the 29 CVD patients,
ROIs with clearly impaired CBF were detected (one
patient with a TIA, eight with a RIND, and eight
with a CS). The results are shown in Table 4.
Patients with a CS had significantly lower focal
rCBF than patients with a TIA or RIND by twotailed independent t test (f=4.409, p<0.001). A
TABLE 2. Comparison of Hemispheric Cerebral Blood Flow Measured Using IMP-SPECT in Normal Volunteers and Patients With
Cerebrovascular Disease
Patients
Normal _
volunteers TIA+R1ND CS
(n=14)
(n=!9)
(n=10)
Cerebral
Left
Right
Age(yr)
significant difference in rCBF between the two
subgroups was also found in ROIs contralateral to
those demonstrating ischemia (f=3.413, p<0.005).
The interhemispheric difference of these rCBF values was less pronounced in the CS subgroup by
two-tailed paired t test (f=4.393, p<0.005) than in
the subgroup with TIA or RIND (r=6.956, p<0.001).
Figure 4 shows SPECT studies and rCBF values
recorded in a control, in a patient suffering from a
RIND, and in a CS patient with an occlusion of the
left middle cerebral artery.
blood flow (ml/100 g/min)
67.8±IO.2 63.8±14.9 40.0±1I.O 15.850 <0.001
67.2±9.8
65.4±14.9 45.9±9.2 10.528 <0.001
25.1+5.3
56.2±13.3 5O.l±14.2 30.542 <0.001
Data are mean±SD. IMP-SPECT, single-photon emission
computed tomography using JV-isopropyl[l23I]-p-iodoamphetamine. T1A+RIND, reversible symptoms; CS, persistent neurologic deficit. F value obtained from one-way analysis of variance.
Discussion
Continuous improvement of the SPECT technique has led to a wide clinical application of this
method for the detection of impaired CBF in various neurologic diseases. Lassen et al13 introduced a
fast rotating multidetector system and a model14
suitable for the determination of CBF with xenon133 SPECT.
Opinions concerning CBF measurement with IMPSPECT are rather divergent. Kuhl et al5 quantified
CBF from IMP-SPECT images using the Mark IV
scanner and obtained an overall mean±SD CBF
value of 47.2±5.4 ml/100 g/min in five normal
subjects applying the microsphere model. Other
authors1516 described a redistribution of IMP in
ischemic lesions that occurs 3-5 hours after tracer
administration. From these observations, it was
postulated that quantification of CBF is not possible
TABLE 4. Comparison of Focal and Contralateral Cerebral Blood
Flow Measured Using IMP-SPECT in Patients With Reversible
Symptoms or Persistent Neurologic Deficit
n
Focal
Total group
17 47.7±16.2
TIA+RIND 9 58.8±9.8
CS
8 35.2±I2.3
»t=4.4098
p<0.001
Contralateral
60.3+12.3
67.8+10.6
51.9+8.0
/t=3.413
p<0.005
/*
P
-6.023 <0.001
-6.956 <0.001
-4.393 <0.005
Data are mean+SD ml/100 g/min. IMP-SPECT, single-photon
emission computed tomography using /V-isopropyl['"I]p-iodoamphetamine. T1A+RJND, reversible symptoms; CS, persistent neurologic deficit.
*t from paired / test for interhemispheric comparison.
17 from independent t test for comparison of TIA+RIND and
CS subgroups.
188
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
S IS
Stroke Vol 20, No 2, February 1989
" • « * • -
FIGURE 4. Color: Singlephoton emission computed
tomograms from normal
female volunteer and two
female patients with leftsided reversible ischemic
neurologic deficit (RIND)
and completed stroke (CS).
Bottom: Cerebral blood flow
(CBF) values. I, left; r, right.
CBF 1-37
CBF r-*9
since IMP distribution in the brain changes with
time and might be related to specific binding of the
tracer to aminergic receptors. To our knowledge,
no conclusive or convincing investigation has yet
been performed that could fully elucidate the cause
of this late tracer redistribution.
In applying the described method for CBF quantification we expected the following: 1) the equip-
Podrekaetal CBF Quantified by IMP-SPECT
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
ment and image processing used should show linear
sensitivity to known isotope concentrations; 2) the
calculated hCBF and rCBF values in normal volunteers should be comparable with those obtained
with other, already established methods; 3) repeated
measurements in the same subject should give comparable results (if significant tracer washout or
redistribution occurs during SPECT scanning [i.e.,
35-85 minutes after IMP administration], reproducibility would be seriously affected); 4) the intraindividual variation should be similar to that of cerebral
metabolic values obtained with positron emission
tomography (PET) in the same subject when
repeated measurements are not performed in the
same session; and 5) in stroke patients CBF values
should reflect the degree of CVD and should be
related to the degree of neurologic impairment.
As shown, the response of the scanning system
used to iodine-123 concentrations that are usually
detected in the human brain after intravenous injection of 5-6 mCi IMP is linear. Only in deep ischemic lesions can somewhat overestimated CBF
values due to incompletely corrected scatter radiation be expected. Looking at the measured concentrations of iodine-123 of Bottles 12, 13, 14, and 15 in
Figure 1, doubts about the proper positioning of the
attenuation matrix may arise since these values are
to some extent overestimated. If this is the case,
then the measured values in Bottles 18 and 19
should be higher. In our opinion, these inconsistencies are caused by adsorption of iodine-123 to
particles in the phantom fluid, which may affect the
uniformity of tracer distribution.
Employing PET to measure CBF, the following
normal mean±SD gray matter flow values have
been reported: 65±17 ml/100 g/min," 59±11 ml/100
g/min,'« 65±7 ml/100 g/min,•» and 74±10 ml/100 g/
min.20 Devous et al21 found a mean±SD gray matter
blood flow of 71 ±12 ml/100 g/min using xenon-133
SPECT. Our calculated mean (67-68 ml/100 g/min)
and standard deviations of gray matter blood flow
closely resemble those obtained with established
three-dimensional techniques.
Analyzing the cortical rCBF distribution in our
controls, the highest value was found in the occipital and frontal lobes, followed by that in the temporal and parietal lobes. Comparison with reported
rCBF values is rather difficult since different investigators use different regional schemes, depending
on the spatial resolution of the system employed.
However, high rCBF values in the visual cortex
have been reported by several groups,21-23 and
"hyperfrontality" is commonly known from the
xenon-133 clearance technique.24 If rCBF and
regional glucose metabolism (CMRglu) is proportionally distributed in normal brain tissue, then our
findings are further supported by metabolic PET
measurements.25-28
It must be considered that our time correction of
SPECT images had to be based on the tracer input
function of the whole hemisphere and not on single
189
regional input functions. Since the hemispheric input
curve represents the mean of all regional input
functions, high rCBF might be somewhat underestimated and low rCBF somewhat overestimated if
large differences in regional tracer arrival time during the first 5 minutes of measurement occur. However, this possible source of error seems not to play
a major role since a similar normal distribution
pattern of rCBF was found with other techniques.
Measurements of hCBF and rCBF in controls
were reproducible. Intraindividual rCBF values
did not differ significantly between the SPECT
studies. The intraindividual variation ranged
between 0.3% and 15%, which compares well with
that found by Phelps et al29 for paired studies
separated by 1-8 days.
An unexpected result was the significant difference in CBF between sexes. We found few articles
in the literature reporting a similar finding. Using
the xenon-133 inhalation technique, Gur et al30 and
Shaw et al31 observed significantly higher values in
women than in men. In the subjects investigated by
Devous et al,21 females of different ages had higher
hCBF and rCBF values than age-matched males;
significant differences were found in the 20-29- and
the 30-39-year-old groups. Yoshii et al32 reported
significantly higher CMRglu in young and elderly
females than males. Besides physiologic explanations such as hematocrit variations or a brainvolume-dependent regional tracer uptake,33 this
finding could be related to a different attenuation of
photons by the skulls of females and males.
In comparing interhemispheric rCBF differences
of controls, we expected a rightward shift that could
be related to increased attention or anxiety of
subjects during scanning. The significance level of
interhemispheric rCBF differences found in our
controls was rather low but consistent in the two
measurements. Since controls were accurately positioned for scanning, side-to-side rCBF differences
were not due to a partial volume effect. When
correction for sequential testing was performed, the
level of significance was >95%. Recently, Perlmutter et al34 reported significantly higher right rCBF in
nine brain regions of 32 normal subjects. Thus, it
seems that rCBF is not symmetrically distributed in
the two hemispheres during scanning in the resting
state.
The recorded CBF values in the CS patients
reflected the degree of their neurologic impairment.
The subgroup of patients with reversible symptoms
showed normal hCBF in both hemispheres; hCBF
in the affected hemisphere was only slightly
decreased compared with that in the contralateral
hemisphere. In contrast, the CS subgroup exhibited
a marked decrease of hCBF in both hemispheres.
The high mean difference in hCBF between the
affected and contralateral hemispheres in this subgroup was caused by hCBF values of two patients
with large infarcts in the territory supplied by the
middle cerebral artery (interhemispheric differences
190
Stroke Vol 20, No 2, February 1989
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
56.43% and 40.09%), while the other eight CS
patients showed relatively small interhemispheric
differences (range -14.6% to 5.5% of hCBF in the
contralateral hemisphere, mean±SD -3.5±5.9%).
Due to the small number of CS patients and the
large intraindividual variability of hCBF differences
in this subgroup, side-to-side comparison did not
yield a statistically reliable result.
A clear-cut difference in focal CBF was found
between the two patient subgroups. The results are
in good agreement with previously reported CBF
findings in CVD patients. Heiss et al35 found a clear
relation between the long-term prognosis of stroke
patients and hCBF. Meyer et al36 recorded decreased
hCBF in the nonaffected hemisphere ("diaschisis")
of stroke patients.
Considering the range of the obtained CBF values, their reproducibility, and the clinical results,
1MP-SPECT seems to be suitable for the routine
determination of CBF under normal and clinical
conditions. The IMP we used was free of iodine-124
(physical half-life 4.2 days) contaminant since iodine123 was produced by the p,5n reaction. Therefore,
dosages of 5-6 mCi are commonly administered.
Scatter radiation is substantially decreased by the
absence of iodine-124 (main energy peak 603 keV)
and thereby the signal-to-noise ratio in SPECT
images can be improved. Kuhl et al3 calculated the
radiation exposure for a 70-kg human (administration of 5 mCi IMP containing 4% iodine-124, assumed
average half-life for washout from all tissues 66
hours) to be 0.7 rad in the brain, 4.9 rad in the lungs,
4.1 rad in the liver, and 0.45 rad for the whole body.
Approximately half of the radiation exposure is due
to the iodine-124 contaminant. Considering this
radiation exposure, repeated CBF measurements
cannot be performed in the same session (as is
possible with xenon-133 SPECT) but must be separated by 7 or more days. The higher spatial resolution inherent to stable tracer techniques may
compensate for this disadvantage.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
17.
18.
References
1. Winchell HS, Horst WD, Braun L, Oldendorf WH, Hattner
R, Parker H: yv-isopropyl-(123I)p-iodoamphetamine: Single
pass brain uptake and washout; binding to the brain synaptosomes; and localisation in dog and monkey brain. J Nucl
Med \980;21:947-952
2. Winchell HS, Baldwin RM, Lin TH: Development of 123I
labeled amines for brain studies: Localisation of 123Iiodophenylalkyl amines in rat brain. J Nucl Med 1980;
21:940-946
3. Neirinckx RD, Canning LR, Piper IM, Nowotnik DP, Pickett RD, Holmes RA, Volkert WA, Forster AM, Weisner PS,
Marriott JA, Chaplin SB: Technetium-99m </,/-HM-PAO: A
new radiopharmaceutical for SPECT imaging of regional
cerebral blood perfusion. J Nucl Med 1987;28:191-202
4. Holmes RA, Chaplin SB, Royston KG, Hoffman TJ, Volkert
WA: Cerebral uptake and retention of 99Tcm-hexamethylpropyleneamine oxime (99Tcm-HM-PAO). Nucl Med Commun 1985;6:443-447
5. Kuhl DE, Barrio JR, Huang SC, Selin C, Ackerman RF,
Lear JL, Wu JL, Lin CH, Phelps ME: Quantifying local
19.
20.
21.
22.
23.
cerebral blood flow by N-isopropyl-p(mI)iodoamphetamine
(IMP) tomography. J Nucl Med 1982;23:196-203
Matsuda H, Seki H, Sumiya H, Tsuji S, Hisada K, Fujii H,
Higashi S: Quantitative cerebral blood flow measurements
using N-isopropyl-p-(I-123) iodoamphetamine and single photon emission computed tomography with rotating gamma
camera (abstract). J Cereb Blood Flow Metab 1987;
7(suppl I):S545
Kuhl DE, Edwards RQ, Ricci AR, Yacob RJ, Mich TJ, Alavi
A: The Mark IV system for radionuclide computed tomography of the brain. Radiology 1976;121:405-413
Mazziotta JC, Phelps ME, Plummer D, Kuhl DE: Quantification in positron emission computed tomography: 5. Physical-anatomical effects. J Comput Assist Tomogr 1981;
5:734-743
Podreka I, H611 K, Dai-Bianco P, Goldenberg G: Klinische
und technische Aspekte der SPECT-Hirnszintigraphie mit
m
J-N-Isopropyl-Amphetamin. Nuc Compact 1984;
15:305-314
Todd Pokropek A, Di Paola R: The use of computers for
image processing in nuclear medicine. IEEE Trans Nucl Sci
1982;21:1299-1309
King MA, Schwinger RB, Doherty PW, Penney BC: Twodimensional filtering of SPECT images using Metz and
Wiener filters. J Nucl Med 1984;25:1234-1240
Bellini S, Piacentini M, Cafforio C, Rocca F: Compensation
of tissue absorption in emission tomography. IEEE Trans
Acous Speech Sign Proc ASSP I979;27:213-218
Lassen NA, Henriksen L, Paulson O: Regional cerebral
blood flow in stroke by 133 xenon inhalation and emission
tomography. Stroke 1981; 12:284-288
Kano I, Lassen NA: Two methods for calculating cerebral
blood flow from emission computed tomography of inert gas
concentrations. J Comput Assist Tomogr 1979^3:71-76
Creutzig H, Schober O, Gielow P, Friedrich R, Becker H,
Dietz H, Hundeshagen H: Cerebral dynamics of Nisopropyl-(mI)p-iodoamphetamine. J Nucl Med 1986;
27:178-183
Raynaud C, Rancurel G, Soucy JP, Baron JC, Samson Y,
Lassen NA, Cabanis E, Majdalani A, Bourdoiseau M,
Ricard S, Bourguignon M, Syrota A: Study of chronic
cerebral infarcts by 123-1-iodo amphetamine (IMP) and
133-Xe SPECT and by PET, significance of the peripheral
area (abstract). J Nucl Med 1986;27:888
Lenzi GL, Frackowiak RSJ, Jones T, Heather JD, Lammertsma AA, Rhodes CG, Pozzilli C: CMRO and CBF by
the oxygen-15 inhalation technique. Eur Neurol 1981;
20:285-290
Huang SC, Carson RE, Hoffman EJ, Carson J, MacDonald
N, Barrio JR, Phelps ME: Quantitative measurement of local
cerebral blood flow in humans by positron computed tomography and l5O-water. J Cereb Blood Flow Metab 1983;
3:141-153
Frackowiak RSJ, Lenzi GL, Jones T, Heather JD: Quantitative measurement of regional blood flow and oxygen
metabolism in man using 15-0 and positron emission tomography: Theory, procedure, and normal values. J Comput
Assist Tomogr 1980;4:727-736
Yamamoto YL, Thompson C, Meyer E, Feindel W: Positron
emission tomography for the measurement of regional cerebral blood flow. Adv Neurol 1981 ;30:41-53
Devous MD, Stokely EM, Chehabi HH, Bonte FJ: Normal
distribution of regional blood flow measured by dynamic
single-photon emission tomography. J Cereb Blood Flow
Metab 1986;6:95-104
Baron JC, Steinlling M, TanakaT, Cavalheiro E, Saussaline
F, Collard P: Quantitative measurement of CBF, oxygen
extraction fraction (OEF) and CMRO2 with the 15-O continuous inhalation technique and positron emission tomography
(PET): Experimental evidence and normal values in man. J
Cereb Blood Flow Metab 1981;1:S5-S6
Amamo T, Meyer JS, Okabe T, Shaw T, Mortel KF: Stable
xenon CT cerebral blood flow measurements computed by a
Podreka et al
24.
25.
26.
27.
28.
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
29.
30.
single compartment-double integration model in normal aging
and dementia. J Comput Assist Tomogr 1982;6:923-932
Lassen NA, Ingvar DH, Skinhoj E: Brain function and blood
flow. 5c/ Am 1978^39:62-71
Duara R, Margolin RA, Robertson-Tchabo EA, London ED,
Schwartz M, Renfrew JW, Koziarz BJ, Sundaram M, Grady
C, Moore AM, Ingvar DH, Sokoloff L, Weingartner H,
Kessler RM, Manning RG, Channing MA, Cutler NR, Rapoport SI: Cerebral glucose utilisation, as measured with
positron emission tomography in 21 resting healthy men
between the ages of 21 and 83 years. Brain 1983;IO6:761-775
Metter JE, Riege WH, Kuhl DE, Phelps ME: Cerebral
metabolic relationships for selected brain regions in healthy
adults. J Cereb Blood Flow Melab 1984;4:1-7
Heiss WD, Pawlik G, Herholz K, Wagner R, Goldner H,
Wienhard K: Regional kinetic constants and cerebral metabolic rate for glucose in normal human volunteers determined by dynamic positron emission tomography of ["F]2-fluoro-2-deoxy-</-glucose. J Cereb Blood Flow Metab 1984;
4:212-223
Mazziotta JC, Phelps ME, Miller J, Kuhl DE: Tomographic
mapping of human cerebral metabolism: Normal unstimulated state. Neurology 1981^1:503-516
Phelps ME, Mazziotta JC, Kuhl DE, Nuwer M, Packwood J,
Metter J, Engel J Jr: Tomographic mapping of human
cerebral metabolism: Visual stimulation and deprivation.
Neurology 1981;3l:5l7-529
Gur RC, Gur RE, Obrist WD, Hungerbuhler JP, Younkin D,
Rosen AD, Skolnick BE, Reivich M: Sex and handedness
31.
32.
33.
34.
35.
36.
CBF Quantified by IMP-SPECT
191
differences in cerebral blood flow during rest and cognitive
activity. Science 1982217:659-661
Shaw T, Meyer JS, Mortel K, Cutaia M, Sakai F, Yamaguchi F, Amamoto M: Effects of normal aging, sex and risk
factors for stroke on regional cerebral blood flow (rCBF) in
normal volunteers, in Gotoh F, Nagai H, Tazaki Y (eds):
Cerebral Blood Flow and Metabolism. CopenhagenfTokyo,
Munksgaard, 1979, pp 462-463
Yoshii F, Barker WW, Chang JY, Ginsberg MD, ApiceUa A,
Duara R: Sensitivity of CMR^ to gender, age, and cerebrovascular risk factors (abstract). J Cereb Blood Flow
Metab 1987;7(suppl 1):S49
Hatazawa J, Brooks RA, Di Chiro G, Bacharach SL: Glucose utilisation rate versus brain size in humans. Neurology
1987;37:583-588
Perlmutter JS, Powers WJ, Herscovitch P, Fox PT, Raichle
M: Regional asymmetries of cerebral blood flow, blood
volume, and oxygen utilisation and extraction in normal
subjects. J Cereb Blood Flow Metab 1987;7:64-67
Heiss WD, Zeiler K, Havelec L, Reisner T, Bruck J:
Long-term prognosis in stroke related to cerebral blood flow.
Arch Neurol 1977;34:671-676
Meyer JS, Shinohara Y, Kanda T, Fukuuchi Y, Kok NK,
Ericsson AD: Abnormal hemispheric blood flow and metabolism despite normal angiograms in patients with stroke.
Stroke 1970;l:219-223
KEY WORDS
computed
cerebral blood flow •
tomography, emission
Quantification of regional cerebral blood flow with IMP-SPECT. Reproducibility and
clinical relevance of flow values.
I Podreka, C Baumgartner, E Suess, C Müller, T Brücke, W Lang, F Holzner, M Steiner and L
Deecke
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Stroke. 1989;20:183-191
doi: 10.1161/01.STR.20.2.183
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1989 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/20/2/183
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/