The Amount of Circumscribed Brain Edema
and the Degree of Post-Ischemic
Neuronal Recovery do not Correlate Well
JOHN M. HALLENBECK, M.D.,
D. R. LEITCH, M.B.,
P H . D . , A. J. DUTKA,
M.D.,
AND L. J. GREENBAUM, J R . , P H . D .
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
SUMMARY Fifty-four dogs were exposed to multi-focal ischemia sufficient to maintain suppression of
the Pj — N, amplitude of the cortical sensory-evoked response (CSER) for 60 minutes. Subsequently, the
P, — N, amplitude recovery of the CSER was followed for an additional 15, 60, or 120 min while the dogs
were treated or left untreated. The combination of PGI2, indomethacin, and heparin promoted a statistically significant augmentation of CSER amplitude return relative to: 1) no treatment; 2) PGI2 alone; 3)
indomethacin alone; 4) PGI2 and heparin; 5) indomethacin and heparin; 6) PGI2 and indomethacin.
Percentage gray matter water by the wet weight/dry weight technique was significantly elevated in all
embolized groups compared to ten non-embolized controls, but the percentage recovery of the CSER did
not correlate with the presence or degree of gray matter edema among embolized animals followed for 1
hour. Separation of embolized animals by the presence or absence of "neuron-disabling"flows(defined as
0-15 ml/100 gm/min for gray matter and 0-6 ml/100 gm/min for white matter) did produce significantly
different mean CSER percentage recoveries but percentage gray matter water in the two groups was
comparable. A proposed explanation of the data is that brain ischemia engenders two parallel processes
which may become uncoupled. Ischemia creates metabolic conditions that lead to increased cellular imbibition of water and produces increased vascular leakiness. These perturbations increase brain water content.
Concomitantly, there is an occurrence of further metabolic derangements and a multifactorial interaction
at the blood-endothelial interface which have a direct influence on neuronal function and recovery.
Stroke, Vol 13, No 6, 1982
EXTENSIVE STUDY in a variety of experimental
models has established that brain ischemia leads to
brain edema, an increase in brain tissue volume due to
an increase in its water content. 1,2 When the edema is
massive, as in large cerebral infarctions, it can cause
severe intracranial hypertension and brain herniation at
several sites.3 Under these circumstances, the contribution of increased tissue fluid to brain damage and
dysfunction would seem straightforward and unequivocal. When clinical deterioration follows a cerebrovascular ischemic event, it is common practice to
invoke the development of brain edema as a probable
explanation. However, when the edema is circumscribed rather than massive, its potential for increasing
neuronal injury and hindering recovery is less clear.
Although this form of fluid accumulation would increase local tissue pressure with a corresponding decrease in local perfusion pressure and could increase
cell-capillary separation with a consequent increase in
oxygen diffusion distances,4 the effect of circumscribed edema on neuronal function and recovery remains a subject for experimental inquiry.1 The results
of the present studies demonstrate that, in a model of
reversible, multifocal ischemia, neither the presence
nor amount of circumscribed edema is a good predictor
of the degree of post-ischemic neuronal recovery as
indicated by the cortical sensory-evoked response
(CSER). Brain waters were measured as one aspect of
a study designed to test the therapeutic efficacy of
From the Naval Medical Research Institute, Bethesda, Maryland,
20814.
Address correspondence to: J. M. Hallenbeck, M.D., Hyperbaric
Medicine Program Center, Naval Medical Research Institute, Bethesda,
Maryland 20814.
PGI 2 , indomethacin, and heparin in reversible cerebral
ischemia.5
Methods
Sixty-four conditioned male mongrel dogs weighing
between 8.1 and 16.4 kg were entered into the experimental protocol (54 embolized; 4 brain water controls
prepared but not embolized; 6 brain water controls not
exposed to surgery). After sedation with xylazine 1.1
mg/kg and atropine 0.05 mg/kg subcutaneously, anesthesia was induced by alpha-chloralose (80 mg/kg)
administered intravenously with incremental doses as
needed. The animals were intubated and mechanically
ventilated. End-tidal PC0 2 and P0 2 were continuously
monitored by Beckman LB-2 and OM-11 analyzers.
Rectal temperature was maintained at 37-38°C with
heating pads and infrared light. Two catheters were
placed in the right femoral artery. One was directed
proximally into the aorta and the other was inserted
into the distal femoral artery. The proximal catheter
permitted sampling of arterial blood. When the two
catheters were later joined through a Y-connector, the
arterial flow to the leg was externalized, permitting
rapid sampling of arterial blood for the l4C-iodoantipyrine autoradiographic blood flow assay.6 A catheter
was threaded into the aorta from the left femoral artery, and two catheters were placed in the left femoral
vein. One venous catheter was advanced into the inferior vena cava for infusion of solutions, sampling of
venous blood, and infusion of isotope during the blood
flow study. The other venous catheter was advanced
into the right ventricle and permitted rapid injection of
a solution of saturated potassium chloride to terminate
the blood flow study. The arterial catheter was connected to a strain gauge for monitoring of arterial blood
798
STROKE
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
pressure. Polyethylene catheters were placed percutaneously into the cephalic vein of each foreleg as routes
for administration of i.v. solutions. The right internal
carotid artery was exposed and catheterized with PE 50
tubing. A lead II electrocardiogram was connected for
monitoring heart rate and rhythm.
The dogs were placed in a Kopf® stereotaxic apparatus. Stainless steel screw electrodes were inserted into
the skull. The recording electrode was positioned over
the right sensorimotor cortex and the reference electrode was embedded in the nasal bones at their distal
extreme. Electrode impedances were less than 3
kohms. Stimulating electrodes were positioned in the
left upper foreleg such that the median nerve was included between them. A square wave stimulus of 300
fi sec duration at a rate of 1.1/sec was led through a
Grass photoelectric stimulus isolation unit with constant current output (Model PISU6C) to the stimulating
electrodes. The strength of the stimulus was adjusted
to cause a maximal cortical sensory-evoked response
(CSER) over the contralateral somatosensory cortex.
Potentials from the recording electrodes were initially
led to a Nicolet HGA-100 preamplifier with a 0.25 Hz
to 10 kHz band width and a 104 gain. The output was
led to a Nicolet 1073 computer of average transients
(CAT) and displayed on a Tektronics 5110 oscilloscope. The CAT output was recorded on a HewlettPackard 7045A XY-plotter. A total of 64 evoked potentials (epochs) were averaged to obtain each
recorded response.
As a final preparatory step, a needle was inserted by
percutaneous puncture into the cisterna magna for
measurement of the cerebrospinal fluid pressure.
After each animal was prepared, a series of not less
than five CSER's were recorded as a baseline. The
latency of the first positive peak, P,, was noted and all
recorded amplitudes from P, to N p the first negative
peak, were averaged to yield a control value.
Following these baseline measurements, 30-50 /xl
of room air was injected as a bolus into the right inter-
VOL
13,
No
6,
NOVEMBER-DECEMBER
nal carotid artery catheter and flushed in with 500 /JL\ of
saline. After 2 min, another CSER was recorded. If the
response was suppressed to between 10 and 20% of the
control amplitude, no more air was infused. If the
CSER was only partially suppressed, another 20-50 /x\
of air was delivered. This sequence was repeated until
the CSER amplitude was 10-20% of the control value.
Suppression of the CSER to values less than 10% of
control was avoided. An animal was not eligible for
inclusion in the series if it took longer than 10 min to
achieve the initial suppression. Subsequently, the periodic regrowth of the CSER to values of 20% of control
or greater was suppressed to a level between 10 and
20% of control by periodic infusion of 20-50 JX\ air.
The volume of these incremental infusions of air was
determined by the apparent sensitivity of the evoked
response to each embolization. This cycle of alternating emergence and ischemic suppression of the evoked
response was continued for 1 hour.
At the conclusion of the 60 min ischemic suppression period, animals received no therapy or one of
several drug regimens depending upon the group to
which they were assigned, as shown in table 1. CSERs
were then followed for an additional 15-120 min and
the percent recovery relative to baseline recorded. Ten
animals in Group 1 were not exposed to ischemia. Four
animals in this group underwent surgical preparation
and the remaining six animals were not exposed to
surgery. The brain waters were all tightly clustered
whether or not the animals underwent surgery and
were accordingly grouped together. All other groups
were subjected to 60 min of ischemia. Eleven animals
in Group 2, one animal in Group 9, and two animals in
Group 11 received no therapy. Other animals were
treated as displayed in table 1. PGI2 (prostacyclin, gift
of Upjohn Co., Kalamazoo, MI) was made up as a 25
Aig/ml solution in 0.1 M Tris-HCL, 0.15 M NaCl pH
8.5 and infused by ice-embedded syringe controlled by
a Sage model 355 continuously variable pump. The
rate of infusion was chosen to depress mean arterial
TABLE 1 Definition of the Various Groups
Treatment
Number of
animals
Duration of
recovery (min)
*1
10
2
11
3
4
Group
PGI2
(ng/kg/min)
Indomethacin
(4 mg/kg)
—
0
0
0
60
0
0
0
11
60
90-260
60
90-285
+
+
+
6
75-340
0
0
5
5
60
1982
Heparin
(300 u/kg)
0
6
5
60
0
+
0
7
5
60
50-200
0
8
5
60
0
+
+
+
9
1
15
0
0
0
+
10
1
15
90-150
+
11
2
120
0
0
0
12
2
120
70-240
+
+
*Group 1 was not subjected to ischemia. Four dogs in this group underwent surgical preparation; the remaining six
animals were not exposed to surgery. All other groups were subjected to 60 min of ischemia.
CIRCUMSCRIBED BRAIN EDEMA/Hallenbeck
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
pressure between 10 and 20 mm Hg. No relationship
was noted between the infusion rate of PGI2 in the
range from 50-340 ng/kg/min and the degree of neuronal recovery. Indomethacin (Indocin®-gift of
Merck, Sharp & Dohme, West Point, PA) was prepared shortly before infusion as a 2 mg/ml solution in
0.1 M Tris-HCL, 0.15 M NaCl pH 8.5.
At the conclusion of the recovery period, a 1-min
l4
C-iodoantipyrine autoradiographic blood flow study
was performed.6 Essentially, this required intravenous
infusion of 50 ju.Ci/kg of 14C-iodoantipyrine at a constant rate over a 1-min period while arterial blood was
sampled every 4-6 sec. The cardiac arrest that terminated this procedure was produced by injecting a bolus
of 50 cc of saturated potassium chloride through the
right femoral vein catheter into the right ventricle. The
brain, spinal cord, and heart were then removed, frozen in liquid freon suspended over liquid nitrogen, and
cut into 20-micron sections. The tissue concentration
of the isotope was determined autoradiographically.
Local blood flow was calculated from the following
formula:
I
Q (T) = ^k, J Cae
where Cj(T) = the concentration of tracer substance in
the tissue at time T; X = the tissue-blood partition
coefficient for the tracer material (approximated as 1);
k; = the rate of blood flow per unit weight of tissue
multiplied by the reciprocal of the partition coefficient
for that tissue; and Ca = the concentration of tracer
substance in the arterial blood.
Determination of Tissue Wet Weight and Dry Weight
and Calculation of Percentage Water:
From a 2 mm thick coronal tissue section through
sensorimotor cortex in the right hemisphere, two gray
matter samples were excised from the superolateral
cerebral cortex and two samples of underlying white
matter were also obtained. The samples were individually placed in tared, stoppered weighing vials and
immediately weighed on a five-place Mettler HL52
balance. The weighing vials had previously been
baked to remove adsorbed water and had returned to
room temperature in a dessicator oven Drierite® (calcium sulfate). The tissue samples were dried to constant weight (as determined in preliminary experiments) in an oven at 110°C for 48 hrs and subsequently
brought to room temperature in a dessicator oven
Drierite®. The tared vials were then reweighed to yield
the dry weight. The percentage solids for each sample
was calculated:7
dry weight of tissue (g)
% solids =
100
wet weight of tissue (g)
The percentage water was determined as:
% water = 100 - % solids =
wet weight - dry weight
x 100
wet weight
The duplicate determinations for gray matter percentage water were averaged for each animal and the same
procedure was followed for white matter.
et al.
799
Results
Blood gases and body temperature were maintained
in the normal range throughout the experiment in all
groups. In the embolized groups large enough for statistical analysis (groups 2-8), baseline and final mean
arterial blood pressure (MAP) were virtually identical
(118 ± 17 mm Hg and 117 ± 2 mm Hg respectively;
mean ± SEM): The cerebrospinal fluid pressure
(CSFP) underwent a large increase in these same
groups after embolization. The mean ± SEM CSFP
was 63 ± 9 mm H 2 0 at baseline and rose to 194 ± 15
mm H 2 0 by the conclusion of the experiment.
The mean ± SEM percentage water for gray and
white matter, and the corresponding mean ± SEM percentage cortical sensory evoked response (CSER) recovery are shown for each group in table 2. In gray
matter, mean percentage brain water was increased in
all embolized groups so that 1 h of ischemia in this
model produced edema regardless of the type of infusate subsequently administered. By one-way analysis
of variance (ANOVA), there was no difference in percentage gray matter water among the embolized
groups followed for a one hour recovery period,
groups 2-8 (p > 0.1). A three-way comparison of
percentage gray matter water was made between control animals not subjected to embolization (Group 1),
81.1 ± 0.2 (mean ± SEM), animals receiving effective
drug therapy (Group 3), 82.4 ± 0.3, and animals
receiving an infusate without therapeutic benefit
(Groups 2, 4-8), 82.8 ± 0.2, as shown in table 3. The
difference between these groups was significant atp <
0.001 by one-way ANOVA. Both the effective therapy
group and the ineffectual infusate group differed significantly from controls (p < 0.01) but did not differ
from each other (p > 0.1) by Bonferroni testing.
The mean percentage brain water in white matter of
embolized groups was not significantly different than
controls (table 2). In correlations of brain edema with
other variables, therefore, only associations with gray
matter water were analyzed.
TABLE 2 Tabulation of the Corresponding Mean ± SEM Gray
Matter Water, White Matter Water, and Percentage CSER Recovery
for Each Group
Percentage water (imean ± SEM)
Percentage
recovery CSER
Group
Gray matter
White matter
1
81.1+0.2
66.6±0.3
—
2
83.1 ± 0 . 4
67.7±0.4
22±3
3
82.4 + 0.3
67.3 + 0.8
50±4
4
83.6±0.6
67.5 ± 0 . 7
27 ± 8
5
81.8±0.6
65.4±0.7
21 ± 4
6
81.7±0.8
66.6 + 0.4
22±8
7
83.2±0.2
67.3 ± 1 . 0
18±6
8
82.7 + 0.5
67.3 ± 0 . 4
15±4
9
82.0
66.0
6
10
83.8
68.1
37
11
82.4±0.4
67.1 ± 1 . 0
17±2
12
85.6±0.6
67.1 ± 1 . 0
82 ± 3 5
800
STROKE
TABLE 3 Three-way Comparison of Percentage Gray Matter Water (mean ± SEM) and Associated p-Value by One-way ANOVA and
Bonferroni Testing
Effective
therapy
(group 3)
82.4 ± 0.3.
p < 0.01
p
.
«,„
n , n
I , 1U
H
N.
Ineffectual
infusate
(groups 2 & 4-8)
.82.7 ±0.2
/
p < 0.01
^81.1 ±0.2'
Control
(group 1)
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
No correlation was detected between either MAP or
CSFP and percentage brain water in gray matter. When
percentage gray matter water in embolized Groups 2-8
were correlated with percentage CSER recovery, no
association was found; R = - 0.01, p > 0.10 (fig.
1). Moreover, sorting and recombining embolized animals into an effective therapy group (Group 3) and an
ineffectual infusate group (Groups 2, 4-8) and analyzing each of these two groupings separately failed to
expose any correlation between amount of brain edema
as indicated by percentage gray matter water and the
percentage CSER recovery; R = - 0.09, p > 0.10
and R = 0.15, p > 0.10, respectively.
Gray matter blood flow was calculated as an
unweighted average from the following structures:
auditory cortex, sensorimotor cortex, visual cortex,
anterior association cortex, anterior cerebral-middle
cerebral watershed cortex, posterior cerebral-middle
cerebral watershed cortex, caudate nucleus, thalamus
VOL 13, No 6, NOVEMBER-DECEMBER
and hippocampus. In embolized animals, gray matter
blood flow was 50 ± 33ml/100gm/min(mean ± SD).
White matter blood flow was 14 ± 7 ml/100 gm/min.
White matter blood flow was calculated as an unweighted average from the following structures: anterior centrum ovale, middle centrum ovale, corpus callosum, internal capsule, and optic radiations. There
were no significant differences in average gray or
white matter blood flow among the various groups.
Mean gray matter blood flows measured at the end of
the one hour recovery period in embolized animals
showed no correlation with percentage gray matter
water. However, mean white matter blood flows and,
in particular, middle centrum ovale blood flow did inversely correlate with gray matter water; R = — 0.31,
p < 0.05 and R = - 0.37, p < 0.02 respectively.
This negative correlation was improved when the animals receiving effective therapy (Group 3) were excluded; R = - 0.40, p< 0.02 and R = - 0.49, p<
0.01 respectively.
Embolized animals were sorted into two groups by
the presence (n = 9) or absence (n = 38) of "neurondisabling" flows defined as 15 ml/100 gm/min or less
for gray matter and 6 ml/100 gm/min or less for white
matter.5 As depicted in table 4, the CSER percentage
recoveries in these two groups were significantly different by Student f-test: "neuron-disabling" flow
present — 13.9 ± 2.9% CSER recovery, "neurondisabling" flow absent — 30.9 ± 2.9% CSER recovery (mean ± SEM), p < 0.01. The two groups did not,
however, differ with regard to percentage gray matter
water: "neuron-disabling" flow present — 83.1 ±
s
85 .
G
R
A
Y
84 .
M
A
T
T
E
R
83
A
A
*
A*
A
A
A
82
AA
W
A
T
E
R
80
0
10
20
1982
30
40
50
60
70
% CSER RECOVERY
FIGURE 1. A scatter plot of the relationship between percentage gray matter water and corresponding percentage CSER recovery within individual animals comprising Groups 2-8.
CIRCUMSCRIBED BRAIN EDEMA/Hallenbeck et al.
TABLE 4
Embolized Animals in Groups 2-8 were Sorted by the
Presence or Absence of ' 'Neuron-disabling'' Flows and their Respective Percentage CSER Recoveries (mean ± SEM) Compared
by Student t-Test. A three-way comparison ofpercentage gray matter water (mean ± SEM) between controls and these 2 groups by
one-way ANOVA and Bonferroni test is also displayed.
Percentage CSER
recovery
"neurondisabling"
flow present
13.9 + 2.9
p < 0.01
II
Percentage CSER
recovery
"neurondisabling"
flow absent
30.9 + 2.9
II
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Percentage gray
matter water
controls
(group 1)
p < 0.01
p < 0.05
y
'
81.1±0.2-v^
II
^ ^ 8 2 . 6 + 0.2
83.1+0.5
.
Percentage gray
Percentage gray
matter water
matter water
NS
"neuron"neuronp > 0.10
disabling"
disabling"
flow present
flow absent
0.5, "neuron-disabling" flow absent — 82.6 ± 0.2
(mean ± SEM), p > 0.10; but they did both differ from
controls: p < 0.05 and p < 0.01, respectively, by
one-way ANOVA and Bonferroni testing.
Discussion
By fifteen minutes into the recovery period (75 min
after the start of ischemia), the earliest time point at
which brain water was measured in this study, gray
matter percentage water had reached elevated levels in
embolized animals, as depicted in Table 2. Results
from other studies support the inference that fluid accumulation should have begun during the initial period
of ischemia and become detectable after roughly 3060 min. 8 ' 9 Restoration of blood flow during the one
hour recovery period was probably associated with a
decrease in the rate of edema formation as compared
with the rate during ischemia.9 The relevant levels of
residual blood flow for edema production during the
ischemic period have been defined in several studies
and range from about 7-20 ml/100 gm/min.10' " These
flow rates have been shown to exist during the period
of incremental air embolization in this model.5 Some
fraction of the edema would be of the type termed
' 'cytotoxic" by Klatzo12 and would involve such metabolic alterations as Na + — K + ATPase failure,13 release of poly-unsaturated fatty acids14 and K + dependent, HC0 3 -stimulated, Cl-uptake with a consequent
passive imbibition of water.15 In models of brain ischemia in which tissue perfusion has been interrupted by
vessel ligation, increased permeability of the blood
brain barrier (BBB) is a relatively late phenomenon
occurring after a minimum of 3-4 hr.16- " In contrast,
BBB opening is demonstrable in cerebral air embolism
after seconds.18 The increase in permeability to protein-bound tracers does not appear to involve disruption of interendothelial junctions in ischemia produced
by vessel ligation.I6 The apparent mechanism is a massive increase in the shuttling of the vesicular transen-
801
dothelial transfer system, "pinocytosis." 1619 Some
studies indicate that increased pinocytosis with maintained interendothelial integrity is also the mechanism
of permeability increase in cerebral air embolism20 but
other work suggests that disruption of interendothelial
junctions with endothelial cell desquamation can occur
in this condition.21 Mediators of accelerated pinocytosis include serotonin22 and calcium.23
An increase in tissue water could be expected to
increase local tissue pressure and cause a corresponding decrease in local perfusion pressure. This should
have been reflected in a mild to moderate decrease in
local blood flow provided other factors remained unchanged. Some correlation was noted between white
matter blood flow (particularly middle centrum ovale)
and percentage gray matter water in embolized animals
followed for a one hour recovery period and the association was tightened when animals receiving effective
therapy were eliminated from consideration. The absence of correlation between gray matter water and
gray matter blood flow in this series is unexplained. As
previously described,5 no correlation was noted between average gray or white matter blood flow and
CSER recovery. However, separation of the animals
into two groups based on the presence or absence of
"neuron-disabling" blood flows did produce corresponding CSER recoveries which differed significantly, as depicted in Table 4. Despite the different CSER
recoveries in these two groups, the amount of edema as
indicated by percentage gray matter water did not differ significantly. Instead, both groups had gray matter
water percentages that were increased with respect to
controls. In this study, then, neuronal recovery as indicated by the CSER is related to the presence or absence
of "neuron-disabling'' flows but the percentage CSER
recovery did not correlate with the presence or degree
of gray matter edema and increased gray matter water
occurred in all previously ischemic animals with and
without "neuron-disabling" flows as measured at the
end of the recovery period.
If the local increase in tissue fluid that constitutes
circumscribed edema causes neuronal damage or exacerbates existing damage, one would expect a strong
negative correlation between brain water content and a
quantifiable index of neuronal function such as the
CSER percentage recovery. Further, the presence of
brain edema should contribute to poor recovery regardless of whether therapy was given and effective therapy
should produce a concomitant decrease in cerebral
edema. These consequences of the hypothesis, "circumscribed edema per se causes neuronal damage or
exacerbates existing damage" being true are not observed in the present study. It would appear, then, that
in reversible focal ischemia, a degree of cerebral edema such as that observed in this study is not a good
predictor of the degree of post-ischemic neuronal recovery. This does not, of course, exclude the possibility that edema exceeding some higher threshold of
severity would correlate better with neuronal recovery.
An alternative possibility is that brain ischemia engenders two parallel processes which are not necessar-
802
STROKE
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
ily tightly coupled. Ischemia creates the metabolic
conditions that provoke an increased cellular imbibition of water13-15 and produces a state of increased
vascular leakiness through stimulation of pinocytosis16,19 and also, perhaps, through disruption of interendothelial junction integrity.21 These perturbations
lead to an increase in brain tissue water. Concomitantly, parenchymal metabolic derangements (some of
which cause tissue fluid accumulation) and a multifactorial interaction at the blood-endothelial interface (of
which some components cause tissue fluid accumulation) develop which have a direct influence on neuronal function and recovery. A particularly thorough
and lucid description of the parenchymal metabolic
derangements has appeared in a recent speculative review.24 Events at the blood-endothelial interface are
not fully characterized and have been designated
"blood-damaged tissue interaction. "5,25~26 Inferences
by Stewart et al.27 provide a link between increased
vessel permeability and activation of soluble and cellular systems in blood. The increased permeability of the
endothelium would permit access of contact-activated
proteins such as Factor XII to subendothelial collagen
with consequent activation of the myriad blood responses. Factor XII is an 80,000 mw protein which
binds to negatively charged surfaces, undergoes a
VOL 13, No 6, NOVEMBER-DECEMBER
1982
change in conformation and becomes susceptible to
serine proteases such as kallikrein.28 As a result of
limited proteolytic cleavage of a single bond, activated
factor XII (HFa) is formed without reduction in molecular weight. HFa can activate the intrinsic coagulation
pathway, generate the fibrinolytic enzyme plasmin and
lead to formation of bradykinin. HFa can be further
attacked by kallikrein to form a 28,000 mw Hageman
factor fragment (HFf). Formation of this molecule can
result in generation of kallikrein, plasmin, bradykinin,
and activation of the classical complement pathway by
hydrolysis of C,.29 A hypothetical schematization of
some of the events in "blood-damaged tissue interaction" is shown in figure 2. Data from allied fields
support this systems diagram of the elements potentially involved.19'28"13 Essentially, the process can be
envisioned as an interwoven fabric of soluble biochemical factors and cellular systems which relate in a
complex, often cybernetic fashion. Regulatory proteins control the spread of activation along potential
pathways and it remains for future research to identify
the major participating systems and to gauge the relative contribution of each. The net effect of these processes over time is regarded as a progressive increase
in resistance in the microvessels of the injury zone with
consequent focal shut-down of nutrient flow.525"26 In
POTENTIALLY ACTIVATING VESSEL CHANGES:
<
A. INTERENDOTHELIAL JUNCTION SEPARATION (EXPOSING COLLAGEN)
1. CALCIUM-INDUCED ENDOTHELIAL CELL CONTRACTION
2. ROUNDING UP OF ENDOTHELIAL CELLS DUE TO INSUFFICIENT ATP
B. RELEASE OF TISSUE FACTOR PR0C0AGULANT ACTIVITY AND TISSUE
KALLIKREIN
C. THROMBIN-INDUCED CHANGE IN ENDOTHELIAL MEMBRANES
D. ALTERED CELL-SURFACE MEMBRANE CONSTITUENTS NO LONGER FAVOR
INACTIVATION OF BOUND PROTEASES AS OCCURS IN THE ALTERNATIVE
PATHWAY OF COMPLEMENT ACTIVATION
E. INCREASED VESICULAR TRANSENDOTHELIAL TRANSFER
F. PROSTAGLANDIN RELEASE (FROM BRAIN)
G. RELEASE OF ENDOTHELIAL ENZYMES THAT CLEAVE HAGEMAN FACTOR
DAMAGED ENDOTHELIAL
MEMBRANE
>
DAMAGED ENDOTHELIAL
MEMBRANE
M^T
PLATELET
AGGREGATION
UNACTTVATED
FACTOR XII
COAGULATION
T
PLASMINOGEN
Q
° J
PLASMINOGEN
PR0ACTIVAT0R
° s
PLASMINOGEN
ACTIVATOR
C1INH
— I—
COMPLEMENT
ACTIVATION
CHEMOTAXIS
ATIIII)
11
CI INH[-M
a I
FIGURE 2. A systems diagram depicting hypothetical vessel changes and the multifactorial blood response that may occur in a zone of
acute CNS injury. Such a process of ' 'blooddamaged tissue interaction'' could cause both
the accumulation of edema and microcirculatory shut-down within injury zone. a2M = a2
macroglobulin; AT III = antithrombin III;
CUNH = CI esterase inhibitor; a2 Ap = a?
antiplasmin.
CIRCUMSCRIBED BRAIN EDEMA/Hallenbeck
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
focal brain ischemia, flow rates in a critical range may
cause abeyance of neurologic function, the "ischemic
penumbra"44 but not proceed rapidly to infarction.45
Progressive staunching of nutrient flow in the injury
zone through a process of "blood-damaged tissue interaction" that is measured in hours rather than minutes could convert functionally impaired but viable
neurons to an infarct. This holds out the prospect of
specific emergency medical intervention designed to
sustain microcirculatory perfusion in the injury zone.
The importance of considering the multifactorial network proposed as "blood-damaged tissue interaction"
is that if such processes are operative, curbing them
could be achieved only by interfering with activity of
agonists or enhancing activity of inhibitors at multiple
sites; measures that block single components of the
network would tend to be circumvented.
A zone of CNS tissue in the "ischemic penumbra"
is probably in a metastable state so that progressive
metabolic derangements24 and ' 'blood-damaged tissue
interaction"5,25-26 acting singly or conjointly could
convert the lesion to a state of irretrievable damage. As
the molecular details of the tissue metabolic perturbations and "blood-damaged tissue interaction" emerge,
they ought to suggest effective new therapeutic measures for the acute phase of CNS ischemia.
Acknowledgments
Naval Medical Research and Development Command, Research
Task No. M0099PN001.il 50. The opinions and assertions contained
herein are the private ones of the writers and are not to be construed as
official or reflecting the views of the Navy Department or the Naval
Service at large. The experiments reported herein were conducted according to the principles set forth in the ' 'Guide for the Care and Use of
Laboratory Animals," Institute of Laboratory Resources, National Research Council, DHEW, Pub. No. (NIH) 7823.
The authors gratefully acknowledge the technical assistance of
Messrs. J. A. Miles, Jr., G. E. Sloan, J. R. De Jesus, J. F. Greenbeck,
Mrs. C. B. Jones, Messrs. P. A. Babb, M. A. Byers, J. M. Bertoncini,
and Ms. R. A. Seifert. Also acknowledged is the editorial assistance of
Ms. R. A. Balenger. We thank the Upjohn Company, Kalamazoo, MI
and Merck, Sharp & Dohme, West Point, PA, for supplying drugs used
in this study. Karen D. Pettigrew, Ph.D., National Institute of Mental
Health, N.I.H., provided consultation regarding statistical analysis.
References
1. Katzman R, Clasen R, Klatzo I, Meyer JS, Pappius HM, Waltz
AG: Report of Joint Committee for Stroke Resources IV: brain
edema in stroke. Stroke 8: 512-540, 1977
2. O'Brien MD: Ischemic cerebral edema: a review. Stroke 10: 623628, 1979
3. Fishman RA: Brain edema. New Eng J Med 293: 706-711, 1975
4. Bourke RS, Kimelberg HK: The role of swollen astrocytes in
promoting focal cerebral tissue hypoxia. In Hallenbeck JM, Greenbaum LJ Jr, (eds): The Workshop on Arterial Air Embolism and
Acute Stroke. Bethesda, Undersea Medical Society, Inc., pp 6263, 1977
5. Hallenbeck JM, Leitch DR, Dutka AJ, Greenbaum LJ Jr, McKee
AE: PGI2, indomethacin and heparin promote post-ischemic neuronal recovery in dogs. Ann Neurol 12: 145-156, 1982
6. Sakurada O, Kennedy C, Jehle J, Brown D, Carbin GL, Sokoloff
L: Measurement of local cerebral blood flow with iodo [,4C] antipyrine. Am J Physiol 234: H59-H66, 1978
7. Elliott KAC, Jasper H: Measurement of experimentally-induced
brain swelling and shrinkage. Am J Physiol 157: 122-129, 1949
8. Nishimoto K, Wolman M, Spatz M, Klatzo I: Pathophysiologic
correlations in the blood-brain barrier damage due to air embolism.
et al.
803
Adv in Neurol 20: 237-244, 1978
9. Ito U, Ohno K, Nakamura R, Suganuma F, Inaba Y: Brain edema
during ischemia and after restoration of blood flow: measurement
of water, sodium, potassium content and plasma protein permeability. Stroke 10: 542-547, 1979
10. Symon L, Branston NM, Chikovani O: Ischemic brain edema
following middle cerebral artery occlusion in baboons: relationship
between regional cerebral water content and blood flow at one to
two hours. Stroke 10: 184-191, 1979
11. Crockard A, Iannotti F, Hunstock AT, Smith RD, Harris RJ, Symon L: Cerebral blood flood and edema following carotid occlusion in the gerbil. Stroke 11: 494-498, 1980
12. Klatzo I: Presidential Address. Neuropathological aspects of brain
edema. J Neuropathol Exp Neurol 26: 1-14, 1967
13. Leaf A: Regulation of intracellular fluid volume and disease. AmJ
Med 49: 291-295, 1970
14. Chan PH, Fishman RA: Transient formation of superoxide radicals
in Polyunsaturated fatty acid-induced brain swelling. J Neurochem
35: 1004-1007, 1980
15. Bourke RS, Nelson KM, Naumann RA, Young OM: Studies of the
production and subsequent reduction of swelling in primate cerebral cortex under isosmotic conditions in-vivo. Exp Brain Res 10:
427-446,1970
16. Westergaard E, Go G, Klatzo I, Spatz M: Increased permeability of
cerebral vessels to horseradish peroxidase induced by ischemia in
mongolian gerbils. Acta Neuropath (Berl) 35: 307-325, 1976
17. Hossmann KA, Olsson Y: The effect of transient cerebral ischemia
on the vascular permeability to protein tracers. Acta Neuropath
(Berl) 18: 103-112, 1971
18. Johansson B: Damage to the blood-brain barrier in experimental
gas embolism. In Ariel G, Naquet R (eds): L'embolie gazeuse du
systeme carotidien. Paris, Doin, pp 165-170, 1974
19. Simionescu N, Palade GE: Recent studies on vascular endothelium. Ann NY Acad Sci 275: 64-75, 1976
20. Garcia JH, Klatzo I, Archer T, Lossinsky AS: Arterial air embolism: structural effects on the gerbil brain. Stroke 12: 414-421,
1981
21. Hekmatpanah J: Cerebral microvascular alterations in arterial air
embolism. Adv Neurol 20: 245-253, 1978
22. Westergaard E: The blood-brain barrier to horseradish peroxidase
under normal and experimental conditions. Acta Neuropath (Berl)
39: 181-187, 1977
23. Williams SK, Wagner RC: Regulation of micropinocytosis in capillary endothelium by multivalent cations. Microvascular Res 21:
175-182, 1981
24. Siesjo BK: Cell damage in the brain: a speculative synthesis. J
Cerebral Blood Flow and Metabolism 1: 155-185, 1981
25. Hallenbeck JM: Prevention of postischemic impairment of microvascular perfusion. Neurology 27: 3-10, 1977
26. Hallenbeck JM, Furlow TW Jr: Prostaglandin 12 and indomethacin
prevent impairment of postischemic brain reperfusion in the dog.
Stroke 10: 629-637, 1979
27. Stewart GJ, Lynch PR, Reichle FA, Ritchie WGM, Smith A,
Schaub RG: The adhesion of leukocytes, erythrocytes and noncellular material to the luminal surface of natural and artificial
blood vessels in-vivo. Ann NY Acad Sci 283: 179-207, 1977
28. Griffin JH, Cochrane CG: Recent advances in the understanding of
contact activation reactions. Seminars in Thrombosis and Hemostasis 5: 254-273, 1979
29. Ghebrehiwet B, Silverberg M, Kaplan AP: Activation of the classical pathway of complement by Hageman Factor fragment. J Exp
Med 153: 665-676, 1981
30. Spaet Th H, Gaynor E: Vascular endothelial damage and thrombosis. Adv Cardiol 4: 47-66, 1970
31. De Clerck F, De Brabander M, Neels H, Van de Velde V: Direct
evidence for the contractile capacity of endothelial cells. Thrombosis Res 23: 505-520, 1981
32. Barnhart MI, Baechler CA: Endothelial cell physiology, perturbations and responses. Seminars in Thrombosis and Hemostasis 5:
50-86, 1978
33. Wiggins RC, Loskutoff DJ, Cochrane CG, Griffin JH, Edgington
TS: Activation of rabbit Hageman Factor by homogenates of cultured rabbit endothelial cells. J Clin Invest 65: 197-206, 1980
34. Wolfe LS: Eicosanoids: Prostaglandins, thromboxanes, leuko-
804
35.
36.
37.
38.
39.
40.
STROKE
trienes, and other derivatives of carbon — 20 Unsaturated Fatty
Acids. J Neurochem 38: 1-14, 1982
Gingrich RD, Hoak JC: Platelet-endothelial cell interactions.
Seminars in Hematology 16: 208-220, 1979
Majerus PW, Miletich JP: Relationships between platelets and
coagulation factors in hemostasis. Ann Rev Med 29: 41-49, 1978
Walsh PN: Platelets and coagulation proteins. Federation Proc 40:
2086-2091, 1981
Moake JL, Tang SS, Olson JD, Troll JH, Cimo PL, Davies PJA:
Platelets, von Willebrand Factor, and prostaglandin I2. Am J Physiol 241: (Heart Circ Physiol) 10: H54-H59, 1981
Austen KF: Hageman — Factor-dependent coagulation, fibrinolysis, and kinin generation. Transplantation Proceedings 6: 39-44,
1974
Colman RW: Formation of human plasma kinin. New Eng J Med
V O L 13, N o
6, NOVEMBER-DECEMBER
1982
291: 509-515, 1974
41. Brown DL: Complement and coagulation. Brit J Haematol 30:
377-382, 1975
42. Cochrane CG, Revak SD, Wuepper KD, Johnston A, Morrison
DC, Ulevitch R: Soluble mediators of injury of the microvasculature: Hageman Factor and the kinin forming, intrinsic clotting and
the fibrinolytic systems. Microvasc Res 8: 112-121, 1974
43. Fearon DT, Austen KF: The alternative pathway of complement: a
system for host resistance to microcrobial infection. New Eng J
Med 303: 259-263, 1980
44. Astrup J, Siesjo BK, Symon L: Thresholds in cerebral ischemia:
the ischemic penumbra. Stroke 12: 723-725, 1981
45. Crowell RM, Marcoux FW, DeGirolami U: Variability and reversibility of focal cerebral ischemia in unanesthetized monkeys.
Neurol (NY) 31: 1295-1302, 1981
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
A Comparison of Trends in Mortality from Stroke
in the United States and Rochester, Minnesota
GREGORY L. ANDERSON, M.D. AND JACK P. WHISNANT, M.D.
SUMMARY More uniform methods of diagnosis and determination of the cause of death result in greater
reliability of trends in mortality from stroke. In all categories of stroke except subarachnoid hemorrhage,
the mortality rates in Rochester, Minnesota, are lower and show a more rapid decline with time than in the
United States as a whole. The differential between the rates for United States white population and the
Rochester population is most evident in the older age groups. This difference may be due to the effects of
revisions in the International Classification of Diseases and/or to the assignment of nonstroke deaths to
stroke causes in the United States as a whole to a greater extent than in Rochester. Whether different
patterns of medical care affect the trends in stroke mortality has not been determined. The trends in
mortality from various categories of stroke in Rochester are reliable because some of the inconsistencies of
diagnosis, assignment of cause of death, and coding changes have been overcome. The decline in mortality
rates for Rochester corresponds well with decreasing incidence rates and lack of change in case fatality rates
previously reported.
Stroke Vol 13, No 6, 1982
A DECLINE IN THE MORTALITY from cerebrovascular disease during the last several decades has been
described in several reports.1'6 However, there are inherent difficulties in the interpretation of mortality
data. Problems that can be identified include revisions
in codes for cause of death, changes in terminology
and patterns of diagnosis, low autopsy rates, and low
accuracy of diagnosis, especially in differentiating various categories of cerebrovascular disease.
This study was undertaken to examine the trends in
mortality from stroke in Minnesota and the United
States white population7 and to compare these with the
trends in mortality from stroke in the community of
Rochester, Minnesota. In Rochester, virtually all residents receive their medical care at the Mayo Clinic and
its associated hospitals or the Olmsted Medical Group
and the Olmsted Community Hospital. The records of
all of these facilities are linked for indexing and retrieval. This assures identification of practically all
residents with serious medical problems, whether their
From the Department of Neurology, Mayo Clinic and Mayo Foundation, Rochester, Minnesota. Supported in part by Research Grant NS6663 from the National Institutes of Health, Public Health Service.
Address correspondence to: Dr. Jack P. Whisnant, Department of
Neurology, Mayo Clinic, Rochester, Minnesota 55905.
Received February 9, 1982; revision accepted July 19, 1982.
conditions were diagnosed in the hospital or in one of
the outpatient facilities or at autopsy. The community
has long had a high degree of neurologic expertise, a
high autopsy rate (> 50%), and a relatively uniform
assignment of cause of death. Virtually all death certificates, including those for deaths in which no autopsy is performed, are completed by Mayo Clinic
pathologists.
Methods
All patients were identified who had a clinical or
pathologic diagnosis of stroke at some time and died
while residents of Rochester during the 30-year period
1945-1974. For the purpose of this study, the criteria
for the diagnosis of stroke included (1) signs of focal
neurologic deficit due to a vascular lesion of the central
nervous system present for at least 24 hours and clinical characteristics to suggest that a stroke was the
cause of the lesion or (2) a recent stroke documented at
autopsy. Patients with transient ischemic attacks —
focal cerebral ischemia of less than 24 hours duration
— were not included.
Death certificates for all patients were reviewed for
the date of death, confirmation of Rochester residency
(regardless of whether the patient was actually in the
city at the time of death), presence of any category of
The amount of circumscribed brain edema and the degree of post-ischemic
neuronal recovery do not correlate well.
J M Hallenbeck, D R Leitch, A J Dutka and L J Greenbaum, Jr
Downloaded from http://stroke.ahajournals.org/ by guest on June 18, 2017
Stroke. 1982;13:797-804
doi: 10.1161/01.STR.13.6.797
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1982 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/13/6/797
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/