Determination of Regional Cerebral Blood
Flow by Inhalation of 133-Xenon
By Walter D. Obrist, Ph.D., Howard K. Thompson, Jr., M.D.,
C. Herschel King, M.D., and Hsioh Shan Wang, M.D.
ABSTRACT
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
A method for estimating regional cerebral blood flow is described in which
tracer amounts of 138Xe are inhaled for 2 min and monitored extracranially
over the next 45 to 60 min. A three-compartment model used to analyze the
resulting clearance curves provided separate blood flows for "gray" and "white"
matter, plus a decay constant for a slow third component, believed to arise
from extracerebral tissue. The analysis included a correction for recirculation of
all three components, based on radioisotope concentration of the expired air
or arterial blood. Results from the analysis were compared with those obtained
by the more traditional two-compartment approach, using data from 15 healthy
young males. Whereas the three-compartment method yielded measurements
comparable with those obtained by established techniques (average CBF =
54.7 ml/100 g/min), the two-compartment analysis gave consistently lower
values (average CBF = 30.2). Comparison of results based on expired air
and arterial sampling suggested that end-expiratory air is a reasonable substitute for arterial blood. Although the I3SXe inhalation method is technically
simpler and less traumatic than other methods, complex analytic treatment of
the data is necessitated by the presence of appreciable recirculation and extracerebral contamination.
ADDITIONAL KEY WORDS
three-compartment analysis
extracerebral contamination
arterial isotope concentration
recirculation
COo inhalation
• In 1963 Lassen and co-workers (1) introduced a method for measuring regional cerebral blood flow in man that has yielded
results (2, 3) consistent with older established
techniques. This method involves the extracranial monitoring of an inert, diffusible radioisotope (85-krypton or 133-xenon) injected
into the internal carotid artery. The clearance
curves so obtained are subjected to a twocompartment analysis (4, 5) in which separate
From Duke University Medical Center, Durham,
North Carolina.
Supported in part by Grants HD-00368, HE07563 and NB-HE-06233 from the National Institutes of Health, U. S. Public Health Service, and by
the Duke University Research Council and the North
Carolina Heart Association.
Part of this material appeared as an abstract in
Clinical Research, April 1966. Preliminary findings
were reported at the American Academy of Neurology, April 1965.
Accepted for publication November 30, 1966.
124
cerebral clearance curves
two-compartment analysis
expired air sampling
man
blood flows are derived for fast and slow tissue phases attributed to gray and white matter. A major drawback of this technique, however, is the necessity of inserting a needle
or catheter into the carotid artery, thus limiting its application to special studies or to patients undergoing cerebral angiography. The
present paper is based on a technically simpler and less traumatic method in which the
radioactive gas is inhaled rather than injected. Such a technique has a potentially
wider application to patients and experimental
situations in which carotid injection is not
feasible.
Two major difficulties beset the inhalation
method that are negligible in rapid internal
carotid infusion: (1) extracerebral contamination and (2) recirculation (6). During inhalation, some of the isotope enters the extracranial vessels so that the clearance curves
CircuUtio* Ruurcb, Vol. XX, ]**usry 1967
CEREBRAL BLOOD FLOW BY " X e
125
INHALATION
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
are influenced by radioactivity from the scalp
and other noncerebral tissues. Also, a significant quantity of radioactivity is taken up
by tissues throughout the body, with a consequent increase in the isotope content of blood
returning to the lungs and a greater amount
of recirculation. Since both of these factors
tend to reduce the clearance rate of isotope
from the head, they must be considered in any
quantitative estimate of cerebral blood flow
involving inhalation.
Although attempted earlier (7), Mallett
and Veall (8, 9) were the first to undertake a
systematic study of regional cerebral blood
flow by the inhalation method and to suggest
a correction for recirculation, based on isotope concentration in the arterial blood or
expired air (10, 11). Using m Xe, they performed a two-compartment analysis that
yielded lower blood flow values than those
usually reported for carotid injection. The
clearance curves were typically elevated at the
tail and gave longer half-times for both slow
and fast tissue compartments. Recently, Jensen and associates (12) compared the inhalation and injection techniques in the same subjects and concluded that the two-compartment
inhalation method underestimates cerebral
blood flow because of the presence of slower
components arising from noncerebral sources.
Similar slow components attributable to extracerebral contamination have also been noted by other investigators (13-16).
In the present study, an attempt has been
made to subtract out the contribution of the
extracerebral tissue by a three-compartment
analysis, at the same time correcting for recirculation. The results of such an analysis are
compared with those obtained by the more
traditional two-compartment approach. Additional observations are reported on the comparability of expired air and arterial blood in
correcting for recirculation.
liters of air were added. This mixture was blended
with additional air in the anesthesia machine to
give an administered concentration of approximately 300 ficl liter. As the isotope decayed
(5.27 days half-life), the proportion of 13SXe to
air was increased to maintain a constant dosage.
This was found to be a satisfactory system that
requires minimal handling of the isotope and
permits more than 15 studies per tank of 133Xe
over a 2-week period. Although as much as 5 me
may be inhaled during a given experiment, the
total absorbed dose to the lungs is only about
20 mrad (17), which is considered a safe level for
repeated exposures.
Figure 1 illustrates the method of administering
the isotope. A one-way valve was attached to a
mouthpiece or mask worn by the subject. The
ISS
Xe-air mixture was inhaled from the anesthesia
machine through leaded tubing, and the subject's
expired air vented to the outside. Provision was
made for rapid switching from 133Xe to room
air by a valve near the mouthpiece. Although
varying durations of inhalation were tried, ranging from one breath to 11 min, a 2-min period
was adopted as standard procedure. Longer inhalations resulted in a greater contribution from
the slow components, while shorter periods increased the relative effect of recirculation. It
should be noted that it is not necessary to assume
saturation of the various tissue compartments in
the present method.
The air at the mouthpiece was continuously
sampled by a microcatheter pump at 800 ml /min.
The sample was passed through a narrow glass
tube over a 2-inch crystal scintillation counter,
and thence through a Beckman CO2 gas analyzer.
This yielded both a measure of expired air isotope
concentration and COL. content. Four blood samples per study were drawn from an indwelling
Methods
EXPERIMENTAL PROCEDURE
The isotope was administered by an anesthesia
machine in a non-rebreathing system. One curie
of m X e was obtained in an evacuated E-tank
(Oak Ridge National Laboratories) to which 300
CircuUtion RtiMrcb, Vol. XX, U*u*r> 1967
FIGURE 1
Administration of the isotope by an anesthesia machine, illustrating the head probes and microcatheter
sampling system for expired air.
126
OBRIST, THOMPSON, KING, WANG
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
needle in the radial artery for determination of
hematocrit and Pco 2 . In three subjects, repeated
1-ml arterial samples were obtained through a
stopcock manifold during both inhalation and
clearance. Blood isotope concentrations were determined by a Packard gamma spectrometer.
Small wafer-thin crystal detectors ( J x S inch,
Nuclear of Chicago) were placed over homologous regions of the brain for extracranial monitoring. The probes were recessed 3 cm in lateral
shielding, which provides some degree of collimation without reducing the total count below
desired levels. Count rate curves were obtained
from Nuclear of Chicago linear rate meters and
inscribed on Beckman strip-chart recorders. Figure
2 illustrates typical count rate curves recorded
from the head and expired air detectors, using a
short time constant for the latter. All washout
curves were recorded continuously for 45 to 60
min following the end of 18HXe inhalation.
SUBJECTS
Fifteen healthy male volunteers aged 20 to 30
years served as subjects. In 12, cerebral blood
flow was estimated from the left parietal region.
In 3, bilateral determinations were made and
arterial isotope concentration was measured simultaneously. Expired air radioactivity and COo content were measured in all. In 10 subjects arterial
Pco., was determined.
TREATMENT OF DATA
After correction for background, the count rate
curves were measured manually, and the digital
data punched on cards for processing by an IBM
7072 computer. The steeper early parts of the
head curve were sampled every 12 seconds, with
later measurements made at longer intervals. Arterial blood was sampled every 5 to 10 seconds,
and expired air at the end of each expiration (Fig.
2). For convenience of analysis, the washout portion of the arterial and expired air curves was
fitted by the sum of two exponentials. Two Fortran computer programs1 were developed in accordance with the mathematical equations outlined below, one for a three-compartment model
and the other for a two-compartment model. An
unweighted least-squares method of curve fitting
was employed throughout.
MATHEMATICAL ANALYSIS
The basic equations used in the present analysis
are derived from the Fick principle and have been
described in detail, along with their underlying
assumptions, by Kety (18) and others (19). A
three-compartment analysis was carried out by
means of the following equation, which is applicable to all three compartments, and includes
a correction for recirculation:
C,(t)=/,e
(u)ekl"du,
(1)
where
t
FIGURE 2
Upper curve: recorded count rale from a detector
over the parietal area of subject T.C. during and
following 2 min of lssXe inhalation. Maximum
counts = 20,000/min; time constant = 2.0 sec. Lower
curve: recorded count rate from continuous sampling
of the expired air in the same experiment. Maximum
counts = 200,000/min; time constant = 0-2 sec. Measurements of the air data are made at the end of
each expiration (downward peaks during IS3Xe inhalation, and upward peaks during clearance).
= a given time after the beginning of
isotope inhalation,
C,(t) = isotope concentration in the ith tissue
component at time t (i— 1,2,3),
= blood flow per unit volume of the ith
fi
tissue component,
= /|X(, where X, is the tissue-blood partition coefficient for the ith component,
previously determined for both gray
and white matter (20, 21), and
CA(t) = isotope concentration of arterial blood
or expired air at any time t.
Let N(t) represent the counts obtained from
the head probe at time t. If the isotope is assumed
to be homogeneously distributed in each tissue
compartment and if differences in counting geometry are neglected, iV(t) will be proportional
to a weighted sum of the isotope concentrations
within each of the three compartments. Symbolically,
(2)
'Fortran programs can be obtained from the authors upon request.
OrcmUtion Rtsurcb, Vol. XX, Jintury 1967
127
CEREBRAL BLOOD FLOW BY '"Xe INHALATION
where a is a proportionality constant and w,
represents a weighting factor for the »'th tissue
compartment (1wt=l).
For a particular compartment,
N,(t) =mi)|C,(t) =awlfte-t't
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
CA(u)e*»du, (3)
'o
where Nf(t) represents the counts contributed by
the ith component.
Count rates from the head, iV(t), and from
expired air or arterial blood, CA(t), are recorded
in different units due to variations in geometry,
counting efficiency, etc. These differences are
represented by the proportionality constant, a,
in equation 3, which can either be estimated
experimentally (22) or eliminated through the
use of dimensionless ratios. The latter is accomplished by dividing equation 3 by a particular
ordinate value of the curve that is measurable at
some specific time, t,,. This yields
f
CA(u)ek'"du
<"C4(u)ek"'du
(4)
Note that awji drops out on the right-hand side
of equation 3, and that the discrepancy in units
between head and blood counts is no longer a
problem because the units disappear with the
formation of ratios.
Equation 4 is now applied sequentially to
three different segments of the head curve, starting at the tail and moving backward in time. Depending on which compartment is being analyzed,
tn is allowed to equal t7, t^, or tr, as shown in
Figure 3. tj is selected at a point far out on the
curve (30 min after inhalation), where it is safe
to assume that the isotope concentration in the
brain is negligible and that the recorded counts
arise almost entirely from extracerebral tissue.
Letting i — 3 (third component) and tn = t.,
equation 4, slightly rearranged, becomes:
CA(u)ek»>du
•J^((/)e tM du
(5)
/ '
Ns(tz) is a measurable value on the head curve,
assuming that the ordinate at t7 consists entirely
of third component. N3(t) are data points on the
head curve after tr, and CA(t) is obtained from
the arterial curve. The only unknown is its, which
can be solved for by fitting equation 5 to all
measured data points after tr. Having obtained fca,
Circulation Ristrcb, Vol. XX, Itnutry 1967
IINUTES
FIGURE 3
Digitized raw data from the left parietal area of
tubject S.E. following 2 min of >33Xe inhalation.
Ordinates at times tx, ty and tt divide the curve into
three segments, which are employed in the threecompartment analysis (see text). Computed clearance
rates (kj and half-times (T^) for each of the components are shown above the curve, and include a
correction for recirculation. Note the continued elevation of counts above background 60 min after the
end of inhalation, tx.
the third component is extrapolated back to zero
time by using the right-hand side of equation 5
to calculate values of iV3(t) for all times, t > 0.
The third component is then subtracted from the
empirically determined head points to yield a
curve composed of only first and second components. All subsequent computations are performed on this residual curve.
The second component is analyzed in a similar
manner by means of equation 4 (• = 2). A reference point, tv, is selected approximately 5 min
after the end of inhalation (Fig. 3), where the
contribution of the first component is small. Because the first component has not completely
cleared at tv, the value obtained for k2 is only an
initial estimate. After computing k2, a further subtraction is made from the head data to yield an
approximation of the first component curve. An
initial estimate for k1 is then obtained by the same
process, using tm the end of inhalation, as a reference point in equation 4 (i = 1).
It is now possible to arrive at satisfactory
solutions for kx and k.z by repeated iteration of
the first and second component fits. The value
obtained above for kx is used to subtract out the
small first component contribution to the curve
between tv and tx, after which an improved estimate of k2 is computed. This new value of k2
is then used to derive an improved estimate of £,.
Solutions for kt and k2 are obtained alternately
until values for each of the parameters show no
further change, a process that converges rapidly.
128
OBRIST, THOMPSON, KING, WANG
The resulting estimates may be influenced by
the fact that points on the arterial or expired air
curves do not necessarily correspond to points on
the brain curve at the same moment of recorded
time. Better least-squares fits and a significant refinement of the parameter values accrue from
introducing a variable unknown phase shift, such
that the head curve is displaced relative to the
arterial or expired air curve. This shift, which
has been designated A, is positive or negative
Jo
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
r
(u)ek'"du
depending on whether the arterial and head
curves are moved farther apart or closer together.
Because the parameter estimates (klt k2, kH,
and A) are quite sensitive to noisy data, some
initial smoothing of the curve is desirable. Improved solutions may be obtained by substituting
auxiliary unknown variables for the measured
ordinates at tx, tF, and t., so that the fitted curve
is not forced through erratic data points at these
times.
Blood flows for the fast and slow cerebral compartments (/i and f2) were computed from the
formula, / ( = X^k{, using the values for kx and k2
obtained above and the partition coefficients for
gray and white matter (Xi = 0.80, X2 = 1.51) determined by Mallett and Veall (8). Since no
partition coefficient is available for the slow third
component, / s was not computed.
The average blood flow, /, of the two cerebral
compartments was calculated by means of the
following formula (1):
7
W2
(tf,/U/o)
(6)
The ratio, Wi/w2, in the last expression is obtained
from:
CA(u)ek«'du
2V,(t)
7'
(7)
where N t ( t ) and N2(t) are derived from equation 3, and t is any arbitrary time. Given the
relation, 2u>( = 1, and using equation 3 to form
similar ratios, the relative weights for the three
tissue compartments (wit w2, and u)a) can now
be determined.
The formula for the two-compartment analysis
was derived from equations 2 and 3 above, letting
4 = 1,2. The recorded count rate, N ( t ) , which
is the sum of the counts contributed by each
component, is divided by a particular ordinate
value, N(to), also made up of two components.
Thus,
N(t)
h
N,(t)+2V 2 (t)
CA(u)e«'*du
, {ti)ek*"du
(8)
where wx + w2 — 1, appropriate to the assumption
of a two-compartment model. As in the previous
analysis, the head curve is shifted relative to
the arterial or expired air curve by a variable
time interval, designated A. Because of the
relation f{ = J:4X,, and since w2 = 1 — wu equation
8 contains only three unknown parameters, fu jo,
and u)]. Least-squares estimates of these parameters, along with the phase shift, A, were obtained
by fitting equation 8 directly to all head data
points after the end of inhalation (arbitrarily
selected as time t 0 ).
Results
THREE-COMPARTMENT ANALYSIS
Table 1 presents the results of a three-compartment analysis of clearance curves recorded over the parietal area in 15 normal subjects.
Group means and standard deviations are
shown along with the individual findings. Expired air data were used in these computations. In addition to presenting blood flows
for the fast (/i) and slow (j2) brain compartments, Table 1 gives the average cerebral
blood flow (/) for each subject, based on the
relative weights of the two components (u^/
Wi + w-2 and w2/wi + u>2). Constants for the
very slow third component (fc3) are also
shown in Table 1, along with the corresponding weights (wn). Consistently negative values occur for the time displacement (A)
between head and expired air curves, indicating that the best computer fits are obtained
when expired air recordings lead the head
Circulation Rtstsrcb,
Vol. XX, ]d*m*rj 1967
CEREBRAL BLOOD FLOW BY
13J
129
Xe INHALATION
TABLE 1
Cerebral Blood Flow Values Derived From a Three-Compartment Analysis
CBF
Tissue weight
un
(ml/100 j/min)
Subject
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
92
85
Mean
SD
I:
22
A
A.S.
T.H.
J.M.
T.C.
S.E.
L.F.
R.S.
D.S.
J.L.
L.S.
D.B.
J.S.
G.C.
J.H.
J.B.
k,
/
•i (»ec)«
PCOj
.48
.49
.64
.53
.47
.40
.57
.55
-12
—
-
.62
.54
.50
.58
.59
.58
.50
-12
.60
.70
.027
.032
".022
.041
.033
.027
.043
.029
.022
.044
.048
.037
.032
.024
.031
43
—
43
42
45
35
38
—
33
44
34
—
—
40
.605
.062
.395
.062
.033
.008
.537
.064
57
27
23
30
24
27
.30
24
18
26
22
29
26
22
21
.62
.64
.46
74.5
24.8
54.7
9.9
3.5
6.1
84
83
82
79
77
74
73
70
69
67
65
60
UK
.38
.36
.54
.43
.37
.40
.48
.39
.41
.35
.29
.38
.41
.40
.30
66
64
51
60
61
58
55
55
50
55
55
5.3
49
45
46
.57
.63
.60
.52
.61
.59
.65
.71
.62
.59
6
-12
—6
- 9
- 3
-12
-
9
9
3
9
6
3
3
7.6
3.5
*To the nearest 3 seconds.
For abbreviations, see Methods.
* , • »T*
*»•
(f,
•
7»)
1*7
ponents for a brief time after stopping inhalation. Whereas the first component has largely
cleared in 10 min, an appreciable portion of
the second component is still present at 18
min. The latter does not approach background
until 25 to 30 min, after which the curve is
almost entirely made up of third component.
TWO-COMPARTMENT ANALYSIS
FIGURE 4
Analytic curves derived from the fitted right parietal
data of subject T.H,, representing the three tissue
compartments and their total. The clearance rates (k)
and corresjtonding blood floivs (f) are indicated.
curve by an average of 7.6 seconds. Differences in arterial Pco2 between subjects probably contributed to the variation in blood
flow values. In individual subjects, however,
Pco_. levels were quite stable (±1.0 mm),
which argues for little variation in blood flow
attributable to this source during individual
experiments.
Figure 4 illustrates a three-compartment
analysis of one of the head curves. The effect
of recirculation is apparent in the continued
rise of the total curve and its several comCircuUtio* Ru-rcb, Vol. XX, ]murj
1967
Table 2 presents the cerebral blood flow
findings based on a two-compartment analysis
of the same data previously treated. In all
cases, the values obtained for /i and /* are less
than those resulting from the three-compartment analysis. It is the second component,
however, that is most affected by the method,
f-2 being 6035 lower when only two compartments are assumed. The smaller relative
weights of the first component (u>\) in the
two-compartment analysis contribute to the
consistently lower average flows (/) obtained.
Values for A computed by this method vary
from slightly negative to positive.
COMPARISON OF RESULTS BASED ON EXPIRED AIR
AND ARTERIAL SAMPLING
Figure 5 shows the expired air and arterial
isotope concentrations plotted on the same
OBRIST, THOMPSON, KING, WANG
130
TABLE 2
Cerebral Blood Flow Values Derived From a Two-Compartment Analysis
CBF
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Subject
A
A.S.
T.H.
J.M.
T.C.
S.E.
L.F.
R.S.
D.S.
J.L.
L.S.
D.B.
J.S.
G.C.
J.H.
J.B.
40
56
51
53
59
60
44
57
45
58
55
46
46
49
49
Mean
51.2
6.2
SD
(ml/100 g/min)
h
7
9
'11
10
9
11
10
10
5
11
10
8
9
7
8
9.1
1.7
TUisue weight
un
/
26
36
28
31
35
37
27
32
24
34
34
27
26
26
30
un + wj
S (iec)
.59
.57
.42
.48
.52
53
.52
.48
.48
.48
.53
.51
.46
.46
.52
.41
.43
.58
.52
.48
.47
.48
.52
.52
.52
.47
.49
.54
.54
.48
-5
-5
0
-3
0
-4
305
4.1
.503
.044
.497
.044
-1.3
+3
+2
-4
+2
0
+2
-1
-5
-1
2.8
For abbreviations, see Methods.
OFF
CPM
Expired Air
Arttnmt B/ood
2OOK
IS
5 I00K
8
I
3
MINUTES
FIGURE 5
S
Comparison of expired air and arterial isotope concentrations during and following 2 min of "3Xe inhalation in stibject T.C. End-expiratory air was sampled at the mouth; blood samples were drawn from
the radial artery. The curves have been normalized
to give equivalent peak values.
time base for one of the three subjects in
whom simultaneous determinations were obtained. There is a close correspondence between the shapes of the two curves, although
the arterial data consistently lag the expired
air data by approximately 10 seconds. Decay
constants fitted to the clearance portions of the
arterial and expired air curves are quite similar (fci = 1.7 and 1.9, fc2 = .32 and .33, with
relative weights for ki of .85 and .86, respectively). Comparable curves were obtained in
the other two subjects.
Table 3 presents a comparison of the threecompartment cerebral blood flow values calculated from the arterial and expired air data.
In one subject (L.F.), blood flows based on
the two methods are almost identical. In the
other two subjects, the /i values derived from
arterial data are slightly higher than those
based on expired air, although f> remains
about the same. The two methods parallel
each other in every case, in that bilateral differences observed with one are accurately reflected by the other. In contrast to the consistently negative A's obtained with expired
air, either zero or positive values occur when
arterial blood is used. Thus, the blood curve
lags not only the expired air data but also the
head curve. Differences in circulation time,
plus a small delay in drawing blood samples,
may account for this time displacement.
Discussion
In the present study, cerebral blood flows
were computed from 18SXe inhalation data by
CircuUlion Rtiurcb,
Vol. XX, J****ry 1967
CEREBRAL BLOOD FLOW BY
m
X e INHALATION
131
TABLE 3
Comparison of Cerebral Blood Flow Analyses Based on Expired Air and Arterial Blood (Three Compartments)
CBF
Subject ind
side
Tissue weight
(ml/100 g /min)
fa
h
/
un -f- wa
85
91
27
27
64
65
.64
.58
.36
.42
.032
.032
.49
.50
78
83
22
23
58
58
.64
.58
.36
.42
.023
.022
.53
-3
.55
+6
83
88
30
31
60
64
.57
.58
.43
.42
.041
.041
.53
.53
+6
Air
85
90
25
26
57
60
.52
.53
.48
.47
.029
.029
.48
.48
-6
Blood
79
79
27
26
58
57
.60
.58
.40
.42
.027
.027
.40
.42
-3
0
99
98
29
28
67
65
.54
.53
.46
.47
.028
.028
.44
.45
-3
0
u-i + im
A (KC)
T.H.
Left
Air
Blood
Right
Air
Blood
-6
+3
T.C.
Left
Air
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Blood
Right
-6
+6
L.F.
Left
Air
Blood
Right
Air
Blood
For abbreviations, see Methods.
TABLE 4
Comparison of Normal Values for Cerebral Blood Flow Obtained by Different Methods
Mean CBF it SD (ml/100 f/min)
Investigator
Ingvar et al. (2)
Present series
Veall and Mallett (11)
Present series
79.7 ± 10.7
20.9 ± 2.6
7
49.8 ± 4.6*
74.5 ± 9.9
24.8 ± 3.5
54.7 ± 6.1
52.9 ± 8.0
9.9 ± 2.0
31.5 ±4.9
51.2 ± 6.2
9.1 ± 1.7
30.2 ±4.1
Method
133
Xe internal
carotid injection
133
Xe inhalation,
3-compartment analysis
]33
Xe inhalation,
2-compartment analysisf
133
Xe inhalation,
2-compartment analysis
/=
*Based on the compartniental blood flows and tissue weights reported for these subjects (N = 7).
fBlood flows were calculated from the published half-times, corrected for recirculation (N = 6), utilizing
X, = 0.80, Xo = 1.51 and mean tissue weights from the present study.
For abbreviations, see Methods.
two different methods. One involved the assumption of three tissue compartments, while
the other employed a two-compartment model. As shown in Table 4, the three-compartment analysis yielded blood flows that are in
good agreement with those reported by Ingvar and co-workers (2) for the carotid injecOrcnUsioH Ru—rch, Vol. XX, l.nuttj
1967
tion technique. Consistently lower values were
found, however, when the same clearance
curves were resolved into only two cornponents. These are compared in Table 4 with
similar low values obtained by Veall and Mallett (11) from two-component fits of inhalation curves.
132
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
The assumption previously made (8-12)
that inhalation data can be treated in the
same manner as injection data, i.e., by a twocompartment analysis, does not appear to be
justified. In contrast to the injection method
where the washout curves approach background in 20 min, inhalation curves remain
well above background for an hour or more
after the start of clearance (Fig. 3). Certainly,
there are slower components in the inhalation
data that must be taken into account. It
seems unlikely that these slow components
arise from the brain, since the clearance rates
for various gray and white structures (19, 23)
are well above the range of the k3's obtained
here. An extracerebral origin of the slow components has been postulated by Veall and
Mallett (11) and extensively discussed by
Jensen and associates (12). Because multiple
tissues are probably involved, one of the major questions raised in the present study is
whether the slow washout phase can be
treated monoexponentially. The excellent computer fits, based on a single clearance rate
from 30 to 60 min, suggest that such treatment of the data is feasible.
Although clearance rates for the slow third
component are in the range of those observed
for muscle, skin and bone (24-26), the
precise source of extracerebral contamination
in the inhalation method is not known. Because of the low energy of 13!lXe gamma
radiation (81 kev) and the proportionately
large contribution of sources close to the detector (9), it is reasonable to assume that the
scalp, cranium, and other superficial tissues
are major contributors to the extracerebral
component. Some support for this hypothesis
comes from Ueda and associates (13), who
recorded clearance curves following injection
of ^'Kr into the common carotid artery. They
observed a slow third component with an exponential decay rate of .03 to .05 that is very
similar to that obtained by inhalation. When
they injected into the external carotid artery,
a curve composed largely of the slowest phase
was obtained, suggesting that tissues supplied by this vessel are an important source of
extracerebral counts. In contrast, Mallett and
OBRIST, THOMPSON, KING, WANG
Veall (9) could demonstrate only a slight influence of the scalp circulation by means of a
pressure cuff. They noted, however, that the
second component was affected more by scalp
compression than the first, which is consistent
with the present finding that the second component is more sensitive to the method of
analysis. Recently, Jensen and co-workers (12)
suggested that one of the factors contributing
to the slow decay in inhalation curves might
be the presence of isotope in the air sinuses.
Using a cuffed endotrachial tube, the present
authors made observations on two waking and
five anesthetized subjects. Neither the shape
of the clearance curves nor the computed
blood flows were affected by this procedure,
making it unlikely that the air sinuses are a
significant source of extracerebral counts.
The additional isotope input to the brain
due to recirculation is another factor accounting for the slower clearance rates of
inhalation curves. In contrast to the injection
technique, in which this factor is minimal, an
adequate correction for recirculation is necessary in the present method (27). Its influence
is apparent from the shape of the head curves
(Figs. 2 and 7), which continue to rise for
some time after cessation of inhalation. Figure
6 demonstrates the effect of recirculation on
the fast component of subject T. H. (ft = 78).
The upper curve, which includes recirculation,
was fitted through data points obtained after
subtraction of the second and third components from the raw head curve. These points
were fitted by means of the following equation:
•'»
Ni(t)
= ;
CA(u)e*»du
using the symbols previously defined. The
right-hand side of the equation can be viewed
as being made up of two factors, one a
monoexponential decay and the other a correction for recirculation, consisting of a Tatio
of two integrals. The lower curve in Figure 6
represents the exponential decay without the
CircnUtion Rticrcb, Vol. XX, Jtnmsn 1967
CEREBRAL BLOOD FLOW BY " X e
INHALATION
133
Itt COMPONENT WITH AND
WtTHXX/T REORCULATtQN
LEFT PAIIETAl
3 * CO,
MJNUTES
FIGURE 6
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Upper curve: the analytic first component clearance
of subject T.H. after 2 min of 133Xe inhalation. This
curve is fitted to data points obtained by subtracting
out the second and third components, and includes
the effect of recirculation. Lower curve: the hypothetical monoexponential decay that would have occurred if no recirculation existed, k, is the same for
both curves. See text for equation.
correction factor, and hence would have resulted if recirculation did not exist. Although
recirculation can safely be neglected in extracting the second and third components
from later portions of the curve, it is nevertheless necessary to assess its effect on these
components at earlier times in order to make
an accurate estimate of the first component.
The present three-compartment findings,
which are limited to data from the parietal
area, resemble some of the results obtained
by other investigators for the whole brain.
Thus, the average blood flow (/) of 54.7 ml/
100 g/min observed here compares favorably with the value of 54 found by Kety and
Schmidt (28) using the N-O technique. Similarly, the relative weight of the fast tissue
compartment (iojwl + tv2) of .605 in the
present series is almost identical with the
value of .603 obtained by Veall and Mallett (20) from in vitro measurements of the
gray matter fraction. Although by no means
supporting the present observations, agreement with these earlier findings lends credence to the results obtained. Before reaching
conclusions on the validity of the three-compartment method, it will be necessary to make
direct comparisons with established techniques on the same subjects. The reliability
of the method is also open to question. Stable
CircuUlion Rtlttrcb,
Vol. XX, Jtumsrj 1967
7<0O
*00
TOO
FIGURE 7
Comparison of count rate curves obtained while continuously breathing 5% CO2 and normal air. >"Xe
was inhaled during the first 1.5 min.
test-retest measurements have been obtained
from inhalation data (9), along with good
bilateral comparability, using a two-compartment analysis. The stability of the three-compartment method, however, remains to be
evaluated.
The inhalation technique appears to be
sensitive to variations in blood flow produced
by cerebral vasodilation. Figure 7 illustrates
the results from 1 of 3 subjects in whom
blood flows were determined while breathing
5% CO;.. The curve recorded during CO^ inhalation rises more abruptly and declines at
a faster rate than the one obtained with normal air, in confirmation of previous findings (11). The sensitivity of the inhalation
method to pathologic states, particularly focal
disturbances in blood flow, also requires systematic investigation. There is some hope that
a three-compartment analysis might provide
useful information on relative tissue weights,
as in the case of carotid injection (29).
Compared with other techniques, the inhalation method has the advantage of technical simplicity. Because it does not involve
puncture of the neck vessels, it is easier to
apply in a variety of experimental and clinical
situations. Substitution of expired air for arterial sampling is a further advantage, although
its applicability to patients with impaired pulmonary function has been questioned (11).
One of the drawbacks of the three-compartment approach is the necessity of monitoring for extended periods of time (40 min
or more). Although improved methods of
134
OBRIST, THOMPSON, KING, WANG
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
analysis may shorten this time somewhat,
clearance curves of sufficient length are prerequisite for a proper evaluation of the slower
components.
A practical limitation of the present approach is the involved computer analysis of
the clearance curves. A graphic solution is
tedious and inaccurate due to the complication of correcting for recirculation in all three
components. The correction for recirculation
proposed by Veall and Mallett (10, 11), although easy to apply, is only approximately
valid because it assumes a constant arterial
concentration of isotope during the inhalation
phase, and is limited to the fast component.
Nevertheless, it is hoped that similar computational shortcuts can be developed that
will permit estimation of cerebral blood flow
with sufficient accuracy for physiological and
clinical studies.
of radioactive inert gas. Circulation Res. 18:
237, 1966.
6. LASSEN, N. A.
8. MALLETT, B. L., AND VEALL, N.: Investigation
of cerebral blood flow in hypertension, using
radioactive-xenon inhalation and extracranial
recording. Lancet 1: 1081, 1963.
9. MALLETT, B. L., AND VEALL, N.: Measurement
of regional cerebral clearance rates in man
using xenon-133 inhalation and extracranial
recording. Clin. Sci. 29: 179, 1965.
10.
References
1. LASSEN, N. A., HOEDT-RASMUSSEN, K., SORENSEN, S. C , SKINHOJ, E., CRONQUIST, S., BODFORSS, B., AND INCVAR, D. H.: Regional
11. VEALL,
values of regional cerebral blood flow in man,
including flow and weight estimates of gray
and white matter. Acta Neurol. Scand. 41,
Suppl. 14: 72, 1965.
3. HOEDT-RASMUSSEN, K.: Regional cerebral blood
flow in man the intra-arterial injection method: procedure and normal values. In Hyperbaric Oxygenation, edited by I. M. Ledingham. London, E. & S. Livingstone Ltd., 1964,
pp. 166-172.
4. SvEiNsDOTTiR, E.: Clearance curves of Kr815 or
Xe l a s considered as a sum of mono-exponential outwash functions. Acta Neurol. Scand.
41, Suppl. 14: 69, 1965.
5.
HOEDT-RASMUSSEN, K., SVEINSDOTTTR, E., AND
LASSEN, N. A.: Regional cerebral blood flow
in man determined by intra-arterial injection
Two-com-
N.,
AND MALLETT, B.
L.:
Regional
cerebral blood flow determination by ls:l Xe
inhalation and external recording: the effect
of arterial recirculation. Clin. Sci. 30: 353,
1966.
12.
JENSEN, K. B., HOEDT-RASMUSSEN, K., SVEINSDOTTrB, E., STEWART, B. M., AND LASSEN,
N. A.: Cerebral blood flow evaluated by
inhalation of l s s Xe and extracranial recording: a methodological study. Clin. Sci. 30:
485, 1966.
13.
UEDA, H., HATANO, S., MOLDE, T., AND CON-
DOAIRA, T.: Discussion on compartmental analysis of the human brain blood flow. Acta Neurol. Scand. 41, Suppl. 14: 88, 1965.
14.
cerebral blood flow in man determined by
krypton^. Neurology 13: 719, 1963.
2. INCVAR, D. H., CRONQUIST, S., EKBERC, R., RISBERC, J., AND HOEDT-RASMUSSEN, K.: Normal
VEALL, N., AND MALLETT, B. L.:
partment model using Xe 183 inhalation and
external counting. Acta Neurol. Scand. 41,
Suppl. 14: 83, 1965.
Acknowledgment
The authors are indebted to Dr. Norman Veall,
Dr. Niels A. Lassen and Dr. David H. Ingvar for
their helpful suggestions. Acknowledgment is also
made of the expert technical assistance of Dr. Felix
J. Pircher, Mr. David E. Burton and Mr. C. Frank
Starmer.
AND INGVAR, D. H.: Regional
cerebral blood flow measurement in man.
Arch. Neurol. 9: 615, 1963.
7. CONN, H. L.: Measurement of organ blood flow
without blood sampling (abstr.). J. Clin. Invest. 34: 916, 1955.
ISBISTER, W . H . , SCHOFIELD, P. F . , AND TOR-
RANCE, H. B.: Cerebral blood flow estimated
by ]88 Xe clearance technique. Arch. Neurol.
14: 512, 1966.
15.
HACGENDAL, E., NILSSON, N. J., AND NORBACK,
B.: On the components of Kr8'" clearance
curves from the brain of the dog. Acta Physiol.
Scand. 66, Suppl. 258: 5, 1965.
16. JOHNSON, A. E., AND GOLLAN,
F.:
Cerebral
blood flow monitoring by external detection
of inhaled radioactive xenon. J. Nucl. Med.
6, Suppl. 9: 679, 1965.
17. MATTHEWS,
C.
M. E.,
FOWLER, J.
F., AND
TURNER, P. C. R.: Absorbed doses from Xe 1 M .
Technical Memorandum No. 84, London, MRC
Cyclotron Unit, Hammersmith Hospital, 1963.
18. KETY, S. S.: Theory and applications of the ex-
change of inert gas at the lungs and tissues.
Pharmacol. Rev. 3: 1, 1951.
19. FREYCANC, W.
H.,
JR., AND SOKOLOFT,
L.:
Quantitative measurement of regional circulation in the central nervous system by the use
GrcuUiUm Ruurcb, Vol. XX, lnatry 1967
CEREBRAL BLOOD FLOW BY
n
of radioactive inert gas. In Advan. Biol. Med.
Phys., Vol. 6, edited by J. H. Lawrence and
C. A. Tobias. New York, Academic Press,
1958, pp. 263-279.
20.
135
X e INHALATION
muscle by intramuscular injection of xenon133. Lancet 1: 686, 1964.
25. COOPER, K. E., EDHOLM, O. C , AND MOTTRAM,
R. F.: Blood flow in skin and muscle of the
human forearm. J. Physiol. 128: 258, 1955.
VEALL, N., AND MALAETT, B. L.: Partition of
trace amounts of xenon between human blood
and brain tissues at 37°C. Phys. Med. Biol.
10: 375, 1965.
21. CONN, H. L.: Equilibrium distribution of radioxenon in tissue: xenon-hemoglobin association curve. J. Appl. Physiol. 16: 1065, 1961.
26. COPP, D. H., AND SHIM, S. S.: Extraction ratio
and bone clearance of Sr85 as a measure of
effective bone blood flow. Circulation Res.
16: 461, 1965.
27.
RANCE, H. B.: Study of the arterial clearance of xenon 133 in man. Brit. J. Anaesthesia
37: 153, 1965.
22. LEWIS, B. M., SOKOLOFF, L., WECHSLER, R. L.,
WENTZ, W. B., AND KETY, S. S.: Method for
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
the continuous measurement of cerebral blood
flow in man by means of radioactive krypton
(Kr'»). J. Clin. Invest. 39: 707, 1960.
23. KETY, S. S.: Observations on the validity of a
two compartmental model of the cerebral circulation. Acta Neurol. Scand. 41, Suppl. 14:
85, 1965.
24.
LASSEN, N. A., LJNDBJERG, J., AND MUNCK, O.:
Measurement of blood-flow through skeletal
Circulation Ruurcb, Vol. XX, Jmusry 1967
ISBISTER, VV. H . , SCHOFIELD, P. F., AND TOR-
28.
KETY,
S.
S., AND SCHMIDT,
C. F.:
Nitrous
oxide method for the quantitative determination of cerebral blood flow in man: theory,
procedure and normal values. J. Clin. Invest.
27: 476, 1948.
29. HOEDT-RASMUSSEN, K.,
AND SKINHOJ, E.:
In
vivo measurements of the relative weights of
gray and white matter in the human brain.
Neurology 16: 515, 1966.
Determination of Regional Cerebral Blood Flow by Inhalation of 133-Xenon
WALTER D. OBRIST, HOWARD K. THOMPSON, Jr., HERSCHEL KING and HSIOH SHAN
WANG
Downloaded from http://circres.ahajournals.org/ by guest on June 16, 2017
Circ Res. 1967;20:124-135
doi: 10.1161/01.RES.20.1.124
Circulation Research is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1967 American Heart Association, Inc. All rights reserved.
Print ISSN: 0009-7330. Online ISSN: 1524-4571
The online version of this article, along with updated information and services, is located on the
World Wide Web at:
http://circres.ahajournals.org/content/20/1/124
Permissions: Requests for permissions to reproduce figures, tables, or portions of articles originally published in
Circulation Research 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 Circulation Research is online at:
http://circres.ahajournals.org//subscriptions/