168
Binding of the Hypoxia Tracer [3H]Misonidazole
in Cerebral Ischemia
John M. Hoffman, Janet S. Rasey, Alexander M. Spence, Dennis W. Shaw,
and Kenneth A. Krohn
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
Radiolabelled misonidazole, a radiosensitizing drug that binds to viable hypoxic tumor cells, may be
useful in identifying hypoxic cells in cerebrovascular disease. Its potential was investigated in the
gerbil stroke model. Biodistribution of [3H]misonidazole was measured in normal gerbils and animals
that had been subjected to right common carotid artery ligation to produce cerebral ischemia. The
uptake of [3H]misonidazole in the right cerebral hemisphere and right/left hemispheral uptake ratios
correlated positively with the severity of the stroke when measured 6-9 hours after carotid ligation.
Histologic studies in symptomatic ligated animals showed ipsilateral widespread but patchy acute
iscbemic changes as well as areas that showed no morphological changes. Microscopic autoradiography in these animals showed diffuse heavy labelling only in the ipsilateral hemisphere. This was
over areas that had histological damage as well as in adjacent areas that appeared intact. Studies
comparing blood flow measured with [ C]iodoantipyrine and [3H]misonidazole retention in gerbils
with carotid artery ligations indicated thatflowis not a major determinant of retention of this hypoxia
tracer. We conclude that a misonidazole congener labelled with a gamma- or positron-emitting isotope
may be useful in nuclear imaging of the degree and regional distribution of hypoxic tissue in cerebrovascular disease. (Stroke 1987;18:168-176)
R
ADIATION biologists have investigated the
potential for imaging hypoxic tumor tissue
with nitroimidazole derivatives, a class of
drugs that sensitizes hypoxic tumor cells to ionizing
radiation.1-2 Misonidazole, the most thoroughly studied nitroimidazole, binds covalently to viable hypoxic
tumor cells in rodent solid tumors and in tissue culture
monolayers or spheroids.3-4 It does not bind in necrotic, irreversibly damaged tumor cells and may therefore serve as a marker of sublethal hypoxic cellular
injury. To date, there have been no investigations of
misonidazole uptake in hypoxia associated with cerebral ischemia. Wereporthere such a study in the gerbil
stroke model, which was initially described by Levine
and Payan.3 Unilateral carotid occlusion produces homolateral ischemia and/or infarction in approximately
30-50% of adult male animals because of anatomic
variations in the circle of Willis.6"9
Our studies of the biodistribution of radiolabelled
misonidazole in gerbils with or without carotid ligation
showed preferential uptake of the compound in the
ischemic hemisphere of those animals with clinical
evidence of cerebral ischemia. These results were confirmed using microscopic autoradiographic techFrom the Departments of Medicine (Neurology) (J.M.H.,
A.M.S.), Radiation Oncology (J.S.R.), Pathology (Neuropathology) (A.M.S.), and Radiology (D.W.S., K.A.K.), University of
Washington, School of Medicine, Seattle, Washington.
Supported in part by Research Grants #R01 CA34570 and #P01
CA 42045 from the National Institutes of Health, Department of
Health and Human Services.
Address for reprints: Janet S. Rasey, PhD, Department of Radiation Oncology, RC-09, University of Washington, School of
Medicine, Seattle, WA 98195.
Received January 15, 1986; accepted September 22, 1986.
niques. Based on these encouraging initial findings,
further work is underway to produce a radiolabelled
misonidazole congener to function as a positron emission tomography imaging agent of hypoxia in cerebrovascular disease.
Materials and Methods
Radiolabelled Misonidazole
[3H]Misonidazole has been synthesized in our laboratory10 according to the method of Bom and Smith"
by reduction of 8 mg of the precursor ketone l-(3methoxy-2-oxopropyl)-2-nitroimidazole with sodium
[3H]borohydride (Amersham, 8.4 mCi/mmol). This
synthesis produces a tritium atom on the side chain at
the second carbon from the imidazole ring. The radiochemical purity of the product was greater than 98%,
determined by high performance liquid chromatography (HPLC). The first lot crystallized from the reaction mixture had a specific activity of 77 mCi/mmol
(383 /xCi/mg); a second lot had a specific activity of 6
mCi/mmol (29.7 /xCi/mg).
Animals and Surgical Procedures
Adult male gerbils (Meriones unguiculatus) weighing 66-122 g were obtained from Tumblebrook Farms
(West Brookfield, Mass.) and housed in controlled
animal facilities for a minimum of 2 weeks prior to any
studies. They were given food and water ad libitum.
To produce cerebral ischemia arightcommon carotid ligation was performed after anesthesia with intraperitoneal (i.p.) ketamine (40 mg/kg) or inhalant
penthrane. With inhalant penthrane, induction and recovery were more rapid and there was less intraoperative mortality. Hence, this became our standard agent.
Hoffman et al
Misonidazole In Cerebral Ischemia
169
Table 1. Modified Stroke Index
Signs
Hair roughed up to tremor
Obtunded or paucity of movement
Hypesthesia of ear
Head cocked
Eye fixed open
Ptosis
Splayed-out hind limb
Circling
Seizures or abrupt explosive movement
Extreme weakness
TOTAL
Stroke index
1
1
1
3
3
1
3
3
3
6
25
Modified stroke index from Reference 12.
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
A midline anterior cervical incision was made. The
right common carotid artery was isolated and ligated
with 6-0 silk suture distally and proximally. The vessel
was then transected to assure no flow. Thereafter, the
incision was closed, the animals allowed to awaken,
and the stroke index described by Ohno et al12 (Table 1)
was calculated at 15, 30,45, 60, 90, 120, 180, and 240
minutes after carotid ligation. When animals were
monitored longer than 4 hours, stroke indices were
determined at 1- to 4-hour intervals. Prior to sacrifice
of any animal, a final stroke index was calculated.
Biodistribution of [3H]misonidazole was studied in
groups of normal gerbils and groups subjected to carotid ligation. The [3H]misonidazole for injection was
prepared by mixing unlabelled misonidazole with
[3H]misonidazole to obtain a solution with a specific
activity of about 10 /iCi/mg. Since misonidazole is
rapidly cleared (see below), the animals were injected
i.p. with 0.1 ml/10 g body wt at 0, 60, and 120 minutes
to mimic a long serum half-life and to minimize the
effect of reduced flow in the ischemic brain. Syringes
were weighed before and after injection to quantify the
amount administered. Blood samples (20-50 /xl) were
obtained from the retro-orbital plexus at 15, 30, 60,
and 120 minutes after the last of the 3 injections. Animals were sacrificed by cervical dislocation 120 minutes after the final injection, 4 hours after initial injection, and 5-6 hours after ligation for those animals
with stroke indices of > 10. Gerbils with lower stroke
indices were entered into injection protocols 2-3 hours
later than markedly symptomatic animals and were
sacrificed 7-9 hours after ligation. Tissue samples of
approximately 100 mg were removed, blotted free of
excess blood, weighed, solubilized in Soluene (Packard, Downers Grove, 111.), and counted in Instagel
(Packard, Downers Grove, 111.) scintillation fluid. The
percent injected dose per gram of tissue (%ID/g) of all
samples was determined. Eleven organs were sampled: brain, submandibular gland, leg muscle, skin,
heart, lung, jejunum, kidney, vagus nerve, thalamus,
and liver.
The brains of ligated animals were processed differ-
ently from those of control animals by subsectioning to
determine the uptake in different regions of the ischemic hemisphere. On sacrifice, the whole brain was
removed, placed on powdered dry ice for 30 seconds,
and then sliced coronally at approximately 2-mm intervals, yielding a total of 7 coronal slices. Each slice was
then cut in half in the midsagittal plane, and the right
and left halves were placed in preweighed, labelled
scintillation vials. Using standard techniques, the
%ID/g was calculated for all brain slices.
Histology
Eight animals were anesthetized with inhalant
penthrane and then subjected to right common carotid
ligation to correlate histologic changes and the stroke
index after varying periods of ischemia. These animals
did not receive [3H]misonidazole. After the 240-minute calculation, the index was obtained at arbitrary
times but always just prior to sacrifice or death. The
minimum sacrifice time post-ligation was 18.5 hours
and the maximum, 46 hours. At death or sacrifice, the
brain was removed and fixed in 10% neutral buffered
formalin, then embedded in paraffin. Coronal sections
were prepared for histology and stained with hematoxylin and eosin or Luxol fast blue with periodic acidSchiff (PAS).
Autoradiography
Brains from another series of carotid-ligated animals were prepared for [3H]misonidazole microscopic
autoradiography to measure the microscopic distribution of the compound in ischemic and nonischemic
brain regions. Carotid ligation was performed, and the
stroke index was calculated at the usual times. Animals
with a stroke index of > 10 were placed in the multiple
injection protocol with [3H] misonidazole (specific activity of 29.7 /u.Ci/mg). Two animals were sacrificed 2
hours after the final injection of drug, which was between 5 and 6 hours after ligation. The brains were cut
into 6 coronal sections, each 3 mm thick. To compare
the left and right hemispheres, alternate sections were
cut sagittally in the midline and processed for liquid
scintillation counting as previously described to allow
calculation of %ID/g. The remaining control sections
were fixed in 10% neutral buffered formalin, embedded in paraffin, and sectioned at 4 /im. Autoradiographs were prepared with Kodak NTB emulsion (Kodak, Rochester, N.Y.), exposed for 6 months, and
developed.
To compare [3H] misonidazole uptake and blood
flow in ischemic gerbil brains, 8 male gerbils were
subjected to right carotid artery ligation as described
above. Two hours after ligation, a single i.p. injection
of [3H]misonidazole (50 /nmol/kg) was given. At 4
hours post-ligation and 10 seconds before sacrifice, a
bolus injection of [14C]iodoantipyrine (Amersham,
specific activity 59 mCi/mmol) was administered
through a femoral vein catheter as a blood flow tracer.
The catheter had been implanted 24 hours prior to
carotid ligation. After sacrifice, multiple samples of
brain from both the right and left hemispheres were
170
Stroke
Vol 18, No 1, January-February 1987
0.9
Z
normal
<r
o
ligoied
FIGURE 1.
0.7
v>
o
o
Q
UJ
I-
Biodistribution of [3H]mison-
idazole in blood and non-CNS tissues of
control gerbils and animals subjected to
right common carotid artery ligation and in
brains of control gerbils. Uptake in brains
of ligated animals is shown separately for
each animal in Figure 2. Animals received
3 i.p. injections of [3H]misonidazole (50
fimoUkg) within 2 hours and were sacrificed 2 hours after the last injection. Each
bar is the mean value of tissue uptake for 5
animals, and the lines are SEM.
as
o
UJ
-i
0.3
o
a:
O.I
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
brain
blood subm muscle
rtin
heart
lung
gut
kidney vogut liver
oxidized to separate carbon-14 and tritium; these samples were counted in a liquid scintillation spectrometer. Average [14C]iodoantipyrine uptake into the
multiple samples of the unaffected left hemisphere was
normalized to 1.0, and the relative blood flow (uptake
of carbon-14) into individual samples from therightor
left brain was expressed as a percent of this value.
[3H]misonidazole uptake was calculated as %ID/g for
each sample.
Results
The organ biodistribution data for both normal (unligated) and ligated animals are presented in Figure 1. In
control animals the %ID/g for whole brain (0.137 ±
0.017) and blood (0.175 ± 0.025) were similar. The
brain/blood ratio was 0.80 ± 0.04. Liver, gut, and
kidney showed two- to threefold greater uptake than
blood. For non-CNS tissues and blood there were no
significant differences between ligated and control animals, and therefore these data were compared directly.
The brain data for ligated vs. control animals were not
directly compared due to the difference in tissue preparation. From control animals a single sample of the
mid-right hemisphere was taken (Figure 1), while
brains from ligated animals were cut into serial coronal
sections (Figure 2) as described in "Materials and
Methods." Thus, no brain uptake data for ligated animals are shown in Figure 1.
The increased uptake of [3H]misonidazole in ischemic hemispheres is shown in Figure 2, which contains
brain data on 5 animals with stroke indices ranging
from 0 to 13. There was fairly uniform uptake and no
FIGURE 2.
zole in coronal brain slices from 5 animals
subjected to right common carotid ligation.
Uptake was measured 2 hours after the last
of 3 [3H]misonidazole injections, which
was 5-9 hours after ligation. The brain of
each gerbil was divided into 7 1-2-mm
coronal sections, designated A-G, from
rostral to caudal ends. The stroke index is
indicated above the histogram for each animal.
I I Right Brain
•
A B C D E F 6
A B C D E F G
A B C D E F G
BRAIN
Left Brain
A B C D E F G
SECTION
The uptake of [3H]misonida-
A B C D E F G
Hoffman et al
Misonidazole in Cerebral Ischemia
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
right-left differences in any coronal sections in the
animal with no evidence of ischemia (stroke index =
0) and minimal right-left differences in the animal
with a stroke index of 4. There was no significant
difference in brain misonidazole uptake in all 7 coronal
brain slices from ligated gerbils with low stroke index
(Figure 2) when compared with control animals (Figure 1). The 2 animals which were markedly symptomatic (stroke index = 13) at the time of sacrifice
showed a two- to threefold higher uptake in the midparietal region of the right hemisphere when compared
with the left. The increased uptake in the anterior regions of the brain in the other symptomatic animal
(stroke index = 10) was positively correlated with
stroke index. The left brain of the most symptomatic
animals also showed a one- to twofold increase in
uptake when compared with similar coronal sections in
animals with very low stroke index.
Blood clearance of [3H]misonidazole was calculated
for control and ligated animals. Both groups received
multiple [3H] misonidazole injections as described
above for the animals in the biodistribution studies
(Figure 3). The clearance curves had the same slope in
both groups of animals. Linear regression analysis
I.U
cr
<S> 0.6\
LJ
CO
GERBIL
H Misonidazole
BLOOD CLEARANCE
3
\T
MNT
o
\ \ ,
3 0.41O
LJ
p
^ V-Ligated
Z
\
1-
animals
\
Normal - \
.
animalt \
\
z
LJ
O
\\T
LJ 0.2-
\
Q.
\i
1
30
60
120
TIME, minutes
FIGURE 3. Blood clearance of [3H]misonidazole in ligated
and nonligated gerbils (n = 5 in each group). Each animal
received 3 injections at 60-minute intervals. Blood samples
were taken at 15, 30, 60, and 120 minutes after the final injection. Both curves are described by a monoexponentialfunction.
At each sample time, there was no significant difference between ligated and control animals. Each point is mean ± SEM.
171
yielded two separate best-of-fit curves, but there was
no significant difference in %ID/g of blood samples
between control and ligated animals at any sample
time.
Several animals were subjected to ligation and sacrificed at 18-46 hours in an attempt to compare histologic changes with time and stroke index. There were no
findings in right hemispheres of gerbils with stroke
indices of 10 or less and no findings in any left hemisphere. Animals with scores of 11-17 showed extensive patchy acute infarction in the right hemisphere
from the anterior limit of the corpus striatum to the
level of the upper mesencephalon, which included the
major structures: cerebral cortex, corona radiata, internal capsule, corpus striatum, globus pallidus, thalamus, and upper mesencephalic tegmentum. In the
frankly damaged tissue areas, the neuropil demonstrated pallor of staining and widespread spongiform
changes. Neurons showed ischemic cell change with
shrunken hyperchromatic nuclei and cell bodies, or
early karyorrhexis (Figure 4). All of the above structures also contained areas that lacked histological
changes, i.e., either had not undergone irreversible
necrosis or had done so but failed to evolve the morphological changes in the short time span of our experiments.
In the 2 animals in which the brain was prepared for
autoradiography, the final stroke indices were 17 and
14 and were never less than 10 at any observation time.
These animals were sacrificed 2 hours after the third
injection and approximately 6 hours after ligation.
There was intense labelling in the ischemic right hemisphere and only sparse labelling in the left. The labelling was relatively evenly spread over all right hemispheral areas, which included regions that did not
show the histological changes of acute ischemia as
well as those that did (Figure 5). There did not appear
to be any tendency for labelling to concentrate over
nuclei, cytoplasm, vessels, neuropil, or any particular
cell type.
Alternate coronal brain slices from these 2 animals
were also obtained and processed for liquid scintillation counting. This confirmed that there was greater
uptake of [3H]misonidazole in the right hemisphere
compared with the left (Figure 6). The accumulation of
[3H]misonidazole in the anterior regions of the brain
in these 2 symptomatic animals and in 4 of the 5 animals shown in Figure 2 verifies that there is a trend
for increased uptake with increasing stroke index
(Figure 7).
Figure 8 compares the [3H]misonidazole uptake and
relative blood flow for different areas of the brains of 3
representative gerbils subjected to right carotid artery
ligation. The presacrifice stroke index for each animal
is listed in the figure. On the left side of the brain, there
was variation in blood flow from 65 to 130% of the
mean value for that hemisphere but essentially no variation in [3H]misonidazole uptake. In symptomatic ligated animals, relative flow on the right side was reduced. There was no consistent indication of variation
in misonidazole retention as a function of varying flow
172
>
Stroke
Vol 18, No 1, January-February 1987
:
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
•
^
FIGURE 4. Section of right cerebral gray matter from a gerbil with stroke index of 10 at the time of sacrifice, 30 hours after ligation.
This animal did not receive [3HJmisonidazole. The section shows pallor and spongiform changes in the right half and acute ischemic
cell changes in the left half (HE stain; X 100).
on the right side. However, there was clear evidence
for increased uptake in areas of reduced flow in animals with stroke indices of > 10, a result inferred from
the other animals with high stroke indices but for
which blood flow information was not obtained (Figures 2, 6, and 7).
Discussion
Our studies were designed to test whether there
was increased binding of [3H]misonidazole in ischemic brain that could be detected by autoradiography
or liquid scintillation counting (LSC) and eventually
by nuclear imaging and positron emission tomography. The LSC data show a two- to eightfold increase
in tracer concentration in ischemic brain compared to
contralateral normal brain. The quantity of labelling
correlates with the stroke index (Figures 2 , 6 , and 7).
The autoradiographs confirm this finding and show
that the uptake is diffuse and evenly distributed
throughout the hemisphere ipsilateral to the ligation
(Figure 5). In gerbils with a low stroke index there was
no greater uptake in the right hemisphere compared
with the left.
The rationale for this work stemmed from experience and evidence gathered on misonidazole binding
under conditions of hypoxia in vitro using cells in
culture and in vivo using experimental rodent tu-
mors. 231013 In monolayer cultures, misonidazole binds
to anoxic but not fully oxygenated cells, and the binding rate increases as O2 concentration in the medium
decreases from 180 to 2 /xM (122 to 1.4 mm Hg). 314
Miller et al13 concluded that misonidazole is covalently
bound to cellular macromolecules in hypoxic EMT-6
mouse sarcoma cells in vitro, and the temperature dependency of binding strongly suggests that this is an
enzymatic process. 3 In vitro 0.7-1 mmdiam. V-79and
EMT-6 mouse sarcoma cell spheroids in fully oxygenated medium (Po2 = 145 mm Hg) bind about 50-fold
more misonidazole in the viable rim of cells around the
necrotic center than in the well-oxygenated periphery
or the dead necrotic center.2 Electrode measurements
have shown that Po 2 in spheroids of this size decreased
from 100-130 mm Hg at the spheroid surface to 0-3
mm Hg at or near the center of V-79 spheroids and
5-25 mm Hg in EMT-6 spheroids.16 Several mouse
tumors in vivo behave similarly, with most radiolabel
bound in cells directly adjacent to necrosis, which occurs 150-200 /xm from blood vessels and is presumed
due to insufficient O2 for cell growth2 (J.S. Rasey,
unpublished observations). There is virtual absence of
label in necrotic areas of tumors as well as in necrotic
centers of larger spheroids, even when these areas are
directly adjacent to viable-appearing cells with heavy
labelling.210 This strongly argues that nonviable hy-
173
* »
*
••-
•
* '-aft
•
•
•
«
•
•
•.
|
X* '
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
FIGURE 5. Top. Microscopic autoradiograph of [3H]misonidazole uptake
from the right thalamus of a gerbil with
stroke index of 14 at death, 6hoursafter
right carotid ligation. Note the density
of the silver granules and the shrunken
dark-staining ischemic cellular nuclei.
Whole field shows ischemic change (HE
stain; X400). Center. Autoradiograph
from right hemisphere of same animal
as in 5 top. This shows heavy labelling
over an area with minimal-to-no his tologic change (HE stain; X400). Bottom. Section similar to 5 top from the
contralateral uninfarcted thalamus of
the same gerbil to show the very sparse
tritium-developed granules and large
pale normal neuronal nuclei (HE stain;
X400).
174
Stroke Vol 18, No 1, January-February 1987
1.51
1.3
CE
Stroke
into-17
a.
o l.l-
CE
s.i.=w
LLJ
(/)
O
a
a
09-
UJ
i - 0.7-
O 0.5
CE
LU
CL
o
Q5.
U
O
CE O 3
3
6
9
12
STROKE
LLJ
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
0.1
A
C
E
BRAIN
A
C
E
IB
INDEX
FIGURE 7. Uptake of [3H]misonidazole by anterior brain sections (A-D)from the right hemisphere plotted vs. stroke index.
There is obvious differential uptake and binding in the right
hemisphere with a stroke index of > 10.
SECTION
3.-
FIGURE 6. The %IDIg of coronal brain slices in 2 animals
subjected to right common carotid ligation and sacrificed 2
hours after the third injection of [3H]misonidazole, 6 hours
after ligation. Brains were divided into 6 coronal slices, A-G.
Alternate slices were processed for autoradiography while
slices A, C, and E were processed for liquid scintillation counting. The stroke index is indicated above the histogram for each
animal.
poxic tissue cannot bind the labelled drug because the
required enzymatic reductions cannot occur in dead
cells.
Ours is thefirstinvestigation that applies this knowledge to a stroke model with the eventual goal to regionally image in human tissue sublethal ischemia and hypoxia with a positive radionuclide marker. That our
present results bring us closer to this goal is argued as
follows.
Prior work by Onno et al has established convincingly that there is a 5- to 12-fold reduction in ipsilateral
cerebral blood flow (rCBF) on the ligated side in gerbils with stroke indices of > 10 and a lesser and variable reduction in animals with stroke indices ranging
from 3 to 9.12 Kelly and Halsey used the platinized
platinum electrode to measure simultaneous rCBF and
oxygen availability (O2a) at a single locus in the gerbil
brain.17 They generally found immediate simultaneous
reduction in both parameters at several anatomic sites
after temporary carotid ligation, although the correlation was better in some animals than others. These
results suggest a correlation between increasing stroke
index, reduced blood flow, and reduced O2a. Our data
show a positive correlation between misonidazole uptake and stroke index (Figure 7) and also indicate that
misonidazole can enter and be retained in ischemic
brain tissue despite reduced blood flow (Figure 8).
The measurements of relative blood flow and mis-
15
2.
O
LEFT
^cocro
—i
I
i
1 —1—1
1
-i
1
I 2
2 3 4 .5 6
3.
UJ
RIOHT
ANIMAL 1
SI • 2
1. •
0
<>
•
1.5
ANIMAL
SI • 13
0
O
2
O CCDO O
IS
I 2
Q. 3-1
5
ANIMAL
3
SI • 12
O 2
I.-
CDCDCD
O
0.3
6
RELATIVE
.9
BLOOD
I 2
1.5
FLOW
FIGURE 8. Uptake of [3H]misonidazole vs. relative blood
flow measured with [l4C]iodoantipyrine in right (0) and left
(o) hemispheres of gerbils 4 hours after right carotid artery
ligation. The presacrifice stroke index for each animal is
shown. Mean relative flow in the left hemisphere was normalized to 1.0 and flow (['4C]iodoantipyrine uptake) for individual
tissue samples was compared to this value, as described in
"Materials and Methods."
Hoffman et al Misonidazole in Cerebral Ischemia
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
onidazole uptake (Figure 8) indicate that the entry of
this lipophilic radiolabelled drug into ischemic brain
and its retention are not principally determined by
flow. Data obtained on blood flow, measured with
rubidium-103 microspheres vs. [3H]misonidazole retention in a dog myocardial infarct model have shown
a threefold variation in flow but constant misonidazole
uptake in the noninfarcted region and an inverse relation between flow and misonidazole binding in the
infarcted tissue (J. Caldwell and G. Martin, Seattle
VAH, personal communication). While these studies
do not rule out a role for drastically reduced blood flow
in limiting the uptake of misonidazole into ischemic
regions, they suggest that the delivery and retention of
such a lipophilic drug will be principally determined
by other parameters. O2 level is likely to be the most
important factor based on the available evidence from
other systems.
This interpretation is further supported by evidence
reported by Horowitz et al.18 They studied [14C]misonidazole distribution in the RT-9 rat brain tumor
model and correlated this with rCBF by means of
double-label quantitative autoradiography (QAR). In
normal gray matter they found uniform distribution of
low levels of misonidazole, with approximately
10-15% lower values in normal white matter at 0.5, 2,
and 4 hours after injection. In tumor implants there
was no consistent correlation between rCBF and regional carbon-14 concentration. In fact, tumor-to-normal brain ratios of carbon-14 increased over the experimental period despite the finding that tumor blood
flow was lower than normal brain blood flow. From
these findings it was concluded that reduced blood
flow in tumors could not explain the distribution of the
compound.
It is equally unlikely that misonidazole is retained in
ischemic brain solely because reduced flow retards its
washout. Flow in the brains of gerbils with severe
infarcts still is about 8-21% of control levels; it does
not fall to zero. 1219 Also, the accumulated labelled
misonidazole in ischemic gerbil brain remains bound
through all procedures required to prepare autoradiographs. This demonstrates conclusively that the compound is tightly and probably covalently bound.
Because misonidazole is lipophilic, its entry into the
brain is probably not limited by the blood-brain barrier, nor does breakdown of the barrier explain entry of
drug into the brain of gerbils with infarcts. One reason
for this is that the barrier in gerbils remains largely
intact for up to 17 hours after ligation, as determined
by the ability of the brain to exclude trypan blue.7
In addition, Brown and Workman studied the effect
of lipophilicity on distribution of misonidazole and
several other nitroimidazoles in BALB/c mice.20 The
brain/plasma ratio was essentially 1.0 for drugs with an
octanol/water partition coefficient of > 0.25. Misonidazole has a partition coefficient of 0.43. Consistent
with our results in both control and ligated gerbils,
these investigators reported brain/plasma ratios of misonidazole of 1.0 at 20, 45, and 75 minutes after a
single i.p. injection. At 2 hours after the last of 3
175
injections we found blood values that were essentially
the same in ligated as in control animals. The brain/
blood ratios range from 0.71 to 0.91 in control animals
and in those ligated gerbils with a low stroke index.
Our data show a rapid clearance of misonidazole from
the blood of both ligated and nonligated gerbils after
the multiple injection protocol with essentially no difference in clearance rates between the two groups. The
blood clearance can be described by a monoexponential function over the 2-hour experimental period.
Thus, misonidazole demonstrates the expected pattern
of washout from normal (nonischemic) brain of a lipophilic, freely diffusible compound.
The current data do not clarify whether misonidazole binds above or below the level of hypoxia, at
which normal brain tissue cell viability is destroyed.
A.J. Franko et al (unpublished observations) determined that for several tumor types in short-term organ
culture, [14C]misonidazole binding was inhibited approximately 50% by O2 levels of 2,000-9,000 ppm
(1.5-6.8 mm Hg) relative to binding levels occurring
under anoxic conditions. With our present data it also
is not possible to distinguish whether misonidazole
accumulated 1) in cells that were hypoxic but not yet
irreversibly damaged at a time preceding necrosis, or
2) in cells mat were already irreversibly damaged at the
time the misonidazole reached them. The regions in
the ipsilateral hemisphere showing heavy label over
histologically preserved cells and neuropil (Figure 5
Center) suggest that misonidazole does bind in sublethally hypoxic cells. Although we cannot rule out that
these regions represent irreversibly injured cells that
had not evolved the histologic changes of ischemia, the
absence of binding in necrotic cells of tumors or the in
vitro tumor model, multicell spheroids, argues against
frankly dead cells being able to bind the drug. In addition, labelled histologically damaged cells seen 6-9
hours after carotid ligation may have been viable when
labelled drug was first available but died subsequently.
Misonidazole binds relatively evenly throughout the
underperfused tissue, not just at margins where perfusion is preserved. This distribution of label in autoradiographs suggests that there may be a distinct advantage to using such a compound to produce a positive
image of the regional distribution of sublethally damaged hypoxic tissue rather than a negative image of the
preserved tissue that surrounds a hypoxic area. However, more investigations are required to determine
whether misonidazole is detecting reversible hypoxia
in the brain.
The coronal brain sections (Figures 2 and 6) consistently showed a decreasing rostral-to-caudal binding of
label. Anatomical features of gerbil cerebral vessels
may explain this greater uptake in the forward onethird of the right hemisphere and also why markedly
symptomatic animals (stroke index of > 13) had increased uptake in the anterior regions of the left hemisphere. The gerbil has an incomplete circle of Willis
and, depending on collateral flow, may develop ischemia after carotid ligation. The propensity to develop
ischemia depends on the variability of small connect-
1/0
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
ing vessels between the vertebro-basilar and carotid
circulations. The gerbil also has a unique anterior circulation since the 2 anterior cerebral arteries fuse in the
interhemispheric fissure to form a single pericallosal
artery.6"8 Ischemia may well have been present in the
anterior brain regions on the nonligated side secondary
to the loss of the contributing anterior cerebral flow
from the opposite internal carotid circulation. The
cerebellum, which is supplied by the posterior circulation, was included in the 2 most posterior sections
where there was essentially no difference between left
and right.
Our biodistribution data in non-CNS organs were
consistent with findings from other laboratories.2122
Liver and kidney showed higher uptake than most other normal tissues. The reason for high retention in liver
may relate to misonidazole being metabolized there by
nitroreductase enzymes or by conjugation prior to excretion. Alternatively, liver may contain hypoxic
foci.23 The gerbil has a unique concentrating kidney
that may be responsible for increased uptake in addition to major excretion through the urinary tract. In
preparation of the bowel for analysis the contents were
not removed, and anaerobic bacteria may be responsible for the increased uptake seen in this organ, especially after i.p. injection.
The lipophilicity and hypoxia-mediated binding
properties of misonidazole make it a potentially useful
agent for imaging cerebral ischemia. Based on these
unique properties and our preliminary results, we are
actively investigating a positron-emitting congener
that can be applied to the study of ischemia and hypoxia in man.
Acknowledgments
The technical assistance of Lay Chin, Norma Nelson, Sara Magee, and Zdenka Grunbaum is gratefully
acknowledged.
References
1. Chapman JD: Hypoxic sensitizers — Implications for radiation
therapy. N Engl J Med 1979;301:1429-1432
2. Chapman JD, Franko AJ, Sharplin J: A marker for hypoxic
cells in tumours with potential clinical applicability. BrJ Cancer 1981;43:546-550
3. Chapman JD, Baer K, Lee J: Characteristics of the metabolism-induced binding of misonidazole to hypoxic mammalian
cells. Cancer Res 1983;43:1523-1528
4. Rasey JS, Krohn KA, Grunbaum Z, Conroy PJ, Bauer K,
Sutherland RM: Further characterization of 4-bromomisonidazole as a potential detector of hypoxic cells. Radiat Res
1985; 102:76-85
Stroke
Vol IS, No 1, January-teoruary 198/
5. Levine J, Payan H: Effects of ischemia and other procedures on
the brain and retina of the gerbil (Meriones unguiculatus). Exp
Neurol 1966;16:255-262
6. Levine S, John D: Cerebral ischemia in infant and adult gerbils. Arch Pathol 1969;87:315-317
7. Kahn K: The natural course of experimental cerebral infarction
in the gerbil. Neurology 1972;22:510-515
8. Levy DE, Brierley JB: Communications between the vertebrobasilar and carotid arterial circulations in the gerbil. Exp
Neurol 1974;45:503-508
9. Levy DE, Brierley JB, Plum F: Ischemic brain damage in the
gerbil in the absence of 'no-reflow.' J Neurol Neurosurg Psychiatry 1975;38:1197-1205
10. Rasey JS, Grunbaum Z, Krohn K, Nelson N, Chin L: Comparison of binding of 3H-misonidazole and 14C-misonidazole in
multicell spheroids. Radiat Res 1985;101:473-479
11. Born JL, Smith BR: The synthesis of tritium labeled misonidazole. J Label Comp Radiopharm 1983;20:429-432
12. Ohno K, Ito U, Inaba Y: Regional cerebral blood flow and
stroke index after left carotid artery ligation in the conscious
gerbil. Brain Res 1984;297:151-157
13. Franko AJ, Chapman JD, Koch CJ: Binding of misonidazole to
EMT-6 and V-79 spheroids. Int J Radiat Oncol Biol Phys
1982;8:737-739
14. Koch CJ, Stobbe CC, Baer KA: Metabolism induced binding
of C-14 misonidazole to hypoxic cells: Kinetic dependence on
oxygen concentration and misonidazole concentration. Int J
Radiat Oncol Biol Phys 1984;10:1327-1332
15. Miller GG, Ngan-Lee J, Chapman JD: Intracellular localization of radioactivity labeled misonidazole in EMT-6 tumor
cells in vitro. Int J Radiat Oncol Biol Phys 1982;8:741-744
16. Mueller-Klieser WF, Sutherland RM: Oxygen tensions in multicell spheroids of two cell lines. BrJ Cancer 1982;45:256-264
17. Kelly RA, Halsey JH: Comparison of local blood flow and
oxygen availability at the same locus in the ischemic gerbil
brain. Stroke 1976;7:274-278
18. Horowitz M, Blasberg R, Molnar P, Strong J, Kornblith P,
Pleasants R, Fenstermacher J: Regional l4C-misonidazole distribution in experimental RT-9 brain tumors. Cancer Res
1983;43:38OO-38O7
19. Nakai K, Welch KM A, Meyer JS: Critical cerebral blood flow
for production of hemiparesis after unilateral carotid occlusion
in the gerbil. J Neurol Neurosurg Psychiatry 1977;40:595-599
20. Brown JM, Workman P: Partition coefficient as a guide to the
development of radiosensitizers which are less toxic than misonidazole. Radiat Res 1980;82:171-190
21. Garrecht BM, Chapman JD: The labeling of EMT-6 tumors in
BALB/c mice with 14C-misonidazole. Br J Radiol 1983;
56:745-753
22. Flockhart IR, Malcolm SL, Marten TR, Parkins CS, Ruane RJ,
Troup D: Some aspects of the metabolism of misonidazole. Br
J Cancer 1978;37(suppl 3):264-267
23. Van Os-Corby DJ, Chapman JD, Koch CJ: In vitro binding of
14
C-misonidazole to hepatocytes and hepatoma cells. Int J Radiat Oncol Biol Phys 1986;12:1251-1254
KEY WORDS • radiolabelled misonidazole • nitroimidazole
• cerebrovascular disease • carotid ligation • gerbil stroke
model
Binding of the hypoxia tracer [3H]misonidazole in cerebral ischemia.
J M Hoffman, J S Rasey, A M Spence, D W Shaw and K A Krohn
Stroke. 1987;18:168-176
doi: 10.1161/01.STR.18.1.168
Downloaded from http://stroke.ahajournals.org/ by guest on June 16, 2017
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1987 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/18/1/168
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/