877
Reduction of Central Nervous System
Ischemic Injury in Rabbits Using Leukocyte
Adhesion Antibody Treatment
Wayne M. Clark, MD; Ken P. Madden, MD, PhD;
Robert Rothlein, PhD; and Justin A. Zivin, MD, PhD
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Activated leukocytes appear to be directly involved in ischemic central nervous system injury.
A surface glycoprotein (CD18) on the leukocyte is required for endothelial adherence and
subsequent function and can be blocked with leukocyte adhesion antibody treatment. We used
two animal models to determine the efficacy of anti-CD18 antibody treatment in preserving
neurologic function after central nervous system ischemia. We gave a dose of 1 mg/kg
anti-CD18 to treatment rabbits 30 minutes before inducing irreversible ischemia in the brain
with intraarterial microspheres or in the spinal cord using reversible aortic occlusion.
Treatment with anti-CD18 produced a significant reduction in neurologic deficits in the
reversible spinal cord model, but not in the irreversible microsphere model. This protective
effect supports the active role of leukocytes in central nervous system reperfusion ischemic
injury and offers potential for future therapy. {Stroke 1991;22:877-883)
T
he appearance of leukocytes in injured ischemic tissue traditionally has been believed to
represent a pathophysiologic response to existing injury. Recent evidence indicates they also may
be important in the pathogenesis and extension of
ischemic injury.1"4 Two proposed mechanisms of
leukocyte potentiation of ischemia are 1) microvascular occlusion from direct mechanical obstruction
and cytotoxic effects on the endothelium, and 2)
central nervous system (CNS) tissue infiltration and
neuronal cytotoxic injury.5"7 The leukocyte-mediated
tissue damage may be irreversible even if blood flow
is restored. Leukocyte adhesion is an early step in
both of these mechanisms.
Recent studies have identified a specific leukocyte
membrane glycoprotein complex (CD 18) that is critically important for adherence. Monoclonal antibodies (anti-CD 18) to these glycoproteins have been
From the Department of Neurology (W.M.C.), Oregon Health
Sciences University, Portland, Ore.; the Department of Neurology
(K.P.M.), The Marshfield Clinic, Marshfield, Wis.; Boehringer
Ingelheim Pharmaceuticals, Inc. (R.R.), Ridgefield, Conn.; and
the Department of Neurosciences (J.A.Z.), the University of
California, San Diego, San Diego, Calif.
Supported in part by an American Heart Association, California
Affiliate Research Fellowship Award (W.M.C.); a National Stroke
Association/Allied Signal Inc. Award (K.P.M.); and National
Institute of Neurological Disorders and Stroke grant NS-23814.
Address for correspondence: Dr. Wayne M. Clark, Department
of Neurology L226, Oregon Health Sciences University, 3181 SW
Sam Jackson Pk. Rd., Portland, OR 97201.
Received October 1, 1990; accepted March 27, 1991.
shown to inhibit adherence-dependent leukocyte
functions in vitro.89 In vivo studies using anti-CD18
have shown a reduction of ischemic injury and decreased leukocyte tissue infiltration in heart, lung,
intestinal, and systemic shock models.10-14 These
data indicate that leukocyte endothelial adhesion is
probably a critical early event in the process leading
to neutrophil-mediated tissue reperfusion injury. To
date, there have been no published studies using
anti-CD 18 to reduce CNS ischemic injury. It is not
known how the blood-brain barrier influences leukocyte adhesion and migration or if it would alter
anti-CD 18 effects. Because neutrophil infiltratiori is
present early in CNS ischemic injury,15-17 it is anticipated that leukocytes also will have a critical role in
CNS ischemic injury. To test this hypothesis, we
evaluated the therapeutic efficacy of anti-CD18 in
two models of selective CNS ischemia in rabbits.
These models were chosen to compare the treatment
efficacy of anti-CD 18 in a reperfusion model versus
an irreversible microemboli model.
Materials and Methods
We used male New Zealand White rabbits weighing 2-3 kg. We selected the rabbit because of the
high CD18 leukocyte density and because the majority of previous anti-CD 18 studies have used this
species. All animal procedures were approved by our
Internal Animal Care Use Committee. The rabbit
spinal cord ischemia model has been described previously in detail.18
878
Stroke Vol 22, No 7 July 1991
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
We anesthetized rabbits with halothane and
placed a snare ligature occluding device around the
abdominal aorta just below the left renal artery,
leaving the end of the occluder accessible through
the skin. The rabbit was allowed to recover for a
minimum of 2 hours. To induce ischemia, we tightened and clamped the occluder, making all rabbits
completely paraplegic within 2 minutes. The animals
showed no evidence of discomfort, and the procedure
appeared to be painless, based on a lack of previously
measured changes in heart rate, blood pressure, or
circulating catecholamine levels. At the end of a
variable predetermined occlusion period, the device
was undamped, removed, and the skin closed with
surgical staples.
Animals were exposed to varying durations of
ischemia. Those with shorter occlusion times tended
to regain function, whereas those with longer occlusion durations remained permanently paraplegic.
Each animal then was evaluated 18 hours later for
degree of neurologic impairment by an observer
blinded to treatment and ischemic duration. Each
animal was scored as paraplegic (no hind limb movement) (0) or functional (normal/paretic) (1). We then
observed the animals for 3 additional days for evidence of changes in neurologic function, performing
manual expression daily and maintaining adequate
hydration as necessary.
To induce irreversible CNS ischemia, we have
devised previously a rabbit multiple cerebral embolism model using microspheres.19 The rabbit was
anesthetized with halothane and, through a lateral
neck incision, a 3-cm long 20-gauge catheter was
implanted into the common carotid artery after the
external carotid was ligated. The end of the catheter
was left externally exposed and plugged with an
injection cap. Microspheres (50 /xm) were mixed
(1:100 iodine-125 labeled to unlabeled spheres),
suspended in acetone, and dried. For each rabbit, a
variable weight of dried spheres was added to 100 fj\
0.05% Tween in normal saline. The amount of radioactivity present was determined and the specific
activity calculated. After the animals fully recovered
from anesthesia, the microsphere suspension was
carefully injected into the catheter. The animals
showed no evidence of pain with the microsphere
injection. Animals that received a small amount of
microspheres remained essentially normal; those that
received a large amount showed persistent behavioral abnormalities or death at 18 hours. An examiner blinded to group and amount of spheres observed the rabbits at 18 hours and scored them as
grossly abnormal (hemiplegia or obtundation)/dead
(0) or essentially normal (mild circling or nystagmus)/
normal (1). The animals then were killed, the brains
quickly removed and sectioned, and the radioactivity
present measured in a gamma counter. From the
specific activity of the injected mixture of microspheres and the amount of radioactivity recovered
from each brain, we calculated the total weight of
micTOspheres delivered to each brain. For each
TABLE 1. Efficacy of Leukocyte Adhesion Antibody Treatment in
Rabbit Spinal Cord Ischemia Model
Ischemic
duration (min)
Control group
Neurologic status
Paraplegic
Normal/paretic
(«)
(«)
0
0
18
20
22
1
1
1
1
1
23
24
0
1
1
0
25
26
28
0
30
34
35
40
0
0
0
8
13
0
1
0
1
0
0
Anti-CD18 group
8
10
18
20
24
26
2S
30
31
32
33
34
36
37
39
40
(}
0
0
0
0
0
1
0
0
1
0
1
0
0
0
0
0
1
0
1
0
1
1
1
For each duration of ischemia, the number of animals exhibiting
each neurologic grade is shown.
group, a range of microsphere amounts was used to
generate an ischemia dose-response curve.
The anti-CD 18 monoclonal antibody used in all of
these experiments was a monoclonal mouse IgGl
immunoglobulin specific for the /3 unit of CD18 (a
spontaneous class switch variant of R 3.3), which was
a gift from Boehringer Ingelheim Pharmaceuticals,
Inc. (Ridgefield, Conn.). This antibody binds to all
leukocyte classes. It has been used previously in
multiple experimental studies in other organs (dose,
1 mg/kg) and inhibits leukocyte adhesion for up to 18
hours.20 Thirty minutes before ischemia was induced,
1 mg/kg anti-CD18 in saline (5.5 mg/ml) was given as
an intravenous bolus. Control animals received a
similar injection of saline. We chose a 30-minute
pretreatment period based on the cardiac studies13
Clark et al
CNS Ischemia Treatment With Leukocyte Adhesion Antibody
879
TABLE 2. Efficacy of Leukocyte Adhesion Antibody Treatment in
Rabbit Multiple Cerebral Emboli Model
Neurologic status
Microsphere
weight (mg)
y
u
Ui
Ui
Q.
o
cc
UJ
a.
DURATION IN MINUTES
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
FIGURE 1. Quantal dose-response curves for anti-CD18 in
rabbit spinal cord ischemia model; percentage of paraplegic
animals vs. duration of spinal cord ischemia. Midpoint of
curve, ETy,, is ischemic duration that causes 50% of rabbits to
be paraplegic. Error bar shows SEM of the ETX.
and because it allows time for maximal leukocyte
binding.
For both the spinal cord ischemia model and the
multiple cerebral embolism model, we constructed
quantal dose-response curves from the neurologic
function ratings at 18 hours. This analysis involves
iterative fitting of a logistic function. The derivation
of this type of analysis has been described in detail,21
and its utility as a pharmacologic screen has been
demonstrated.22'23 For the spinal cord ischemia
model, at each point on the abscissa, the percentage
of paraplegic animals in the group is calculated and
plotted as the ordinate. The total increases from 0%
at the shortest duration of occlusion to 100% at the
longest duration. The resulting sigmoid curve represents the effect of a range of durations on neurologic
outcome. The point at which 50% of the animals are
paraplegic (termed ETso for effective time) is calculated and is the average length of ischemia that
produces impairment in 50% of the animals. An
analogous calculation is made for the multiple cerebral embolism model. In this model, at each point on
the abscissa, the percentage of abnormal animals is
calculated and plotted as the ordinate. The total
increases from 0% at the lowest weight of microspheres to 100% at the highest. The resulting sigmoid
curve represents the effect of a range of microsphere
weights on neurologic outcome. The point at which
50% of the animals are abnormal (termed ES50 for
effective stroke dose) is calculated and is the average
length of ischemia that produces impairment in 50%
of the animals in the group. An agent effective in
reducing neurologic damage will shift the curve (and
ET50 or ESso) to the right; that is, the animals will, on
average, tolerate a longer period of ischemia or
greater amount of spheres. To determine signifi-
Normal
Abnormal or dead
Control group
0.06
0.08
0.09
0.17
0.20
0.22
0.26
0.38
0.40
0.55
0.56
0.61
0.64
0.86
Anti-CD18 group
0.01
0.06
0.12
0.13
0.14
0.20
0.26
0.28
0.35
0.45
0.62
0.64
0.66
0.71
0.73
0.90
1.11
1.17
For each quantity of microspheres in the brain, number of
animals exhibiting each neurologic grade is shown.
cance, we performed t tests and considered a value of
p<0.05 to be significant.
Results
Table 1 shows the results of anti-CD18 treatment
in the spinal cord ischemia model. The average
length of ischemia that produced impairment (ET^)
in the treatment group («=16) was 32.35±2.88 minutes (mean ± SEM), and the ET^ in the control group
(n = 16) was 22.88 ±1.90 minutes. This difference was
significant at the /?=0.02 level (f=2.45). Thus, pretreatment with anti-CD18 produced a significant
reduction in ischemic injury in this model. Figure 1
shows the quantal dose-response curves for these
groups. We observed the animals in our study for a
880
Stroke
Vol 22, No 7 July 1991
w
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
total of 4 days. During this time, none of the animals
changed from their 18-hour category (normal/paretic
[1] or plegic [0]). However, our binary scoring system
may have missed a partial delayed deterioration.
Table 2 presents the results of anti-CD18 treatment in the multiple cerebral embolism model. The
average weight of microspheres that produced impairment (ES50) in the treatment group (n=19) was
0.52±0.16 mg, and the ES50 in the control group
(n=14) was 0.44±0.12 mg. This difference was not
significant (f=0.40).
We performed histopathologic examination on
randomly selected animals (spinal cord ischemia
model only; rc=6) (see Figure 2). A reduction in
hypercellularity occurred in all of the anti-CD18treated animals studied relative to untreated animals
with similar ischemic durations.
Discussion
In this study, we found that leukocyte adhesion
antibody treatment reduced CNS ischemic injury. We
saw a significant protective effect in a model of
large-vessel reperfusion (spinal cord ischemia model), but not in a model of irreversible microvascular
occlusion (multiple cerebral embolism model). This
difference in therapeutic efficacy may provide a clue
to the mechanisms of leukocyte-mediated injury.
Clark et al
/
CNS Ischemia Treatment With Leukocyte Adhesion Antibody
881
•
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
FIGURE 2. Light photomicrographs of lower lumbar region of the spinal cord cut in transverse plane. Hemataxylin stain, xl54.
Upper left panel: Normal cord from nonischemic control animal. Lower left panel: Ischemic cord from untreated animal
Ischemic duration: 30 minutes; cord removed at 48 hours. Histopathologic changes include neuronal necrosis, vacuolation of the
neuropU, and increased cellularity. This hypercellularity presumably represents active inflammatory cells. These findings are similar
to those of prior studies.24'21 Upper right panel: Ischemic cord from anti-CD 18-treated animal Ischemic duration: 30 minutes;
cord removed at 48 hours. Although some degree of neuronal necrosis and vacuolation is present, there is less hypercellularity.
Leukocytes are relatively large cells with high viscoelastic properties and require considerable deformation on their passage through capillaries. When
chemotactic substances activate them during ischemia,
their cytoplasmic stiffness increases and they develop
adhesion properties to the capillary endothelium.26
Under conditions of reduced perfusion pressures, they
may be unable to traverse the capillary, leading to
obstruction of the microcirculation. Data from white
blood cell perfusion studies on isolated muscle and
kidney preparations show that leukocytes alone may
obstruct the microcirculation under low-flow states,
increasing the resistance by 30%.27 This leukocyte
capillary plugging also may be the major mechanism of
the "no-reflow phenomenon." This phenomenon, first
described in the CNS by Ames et al,24 is defined as the
incomplete restoration of normal blood flow after a
period of ischemia. Areas of parenchyma that might
be viable when blood flow returns are not adequately
reperfused and ultimately die.
This microvascular obstruction also can be potentiated by damage to the endothelium. Leukocyte
granule contents, reactive oxygen metabolites, and
membrane phospholipases have been found to injure
endothelium.25 By constricting capillary lumen and
increasing leukocyte adhesion, these endothelial effects can potentiate low-flow states.5"7 Because reduced perfusion pressures may be seen both in the
ischemic penumbra (static low flow) and during the
initial phase of reperfusion (early reflow), treatment
with leukocyte adhesion antibody could reduce injury
by improving blood flow in both of these conditions.
We chose the two ischemic models in this study to
evaluate which of these possible leukocyte obstruction mechanisms (reduced flow to the penumbra or
no reflow during reperfusion) is of prime importance.
Because the rabbit has poor spinal cord collaterals,
the spinal cord ischemia model produces relatively
little ischemic penumbra during the occlusion time
(only at rostral edge). When the occlusion is released, there is reperfusion to a large ischemic area
(maximal reflow with relatively little ischemic penumbra). In the multiple cerebral embolism model,
the rabbit has excellent cerebral collaterals; therefore, this model produces multiple small areas of
irreversible ischemia with surrounding areas of low
blood flow (maximal penumbra with minimal reflow).
Our finding that treatment with anti-CD18 reduced
ischemic injury in the spinal cord ischemia model, but
not in the multiple cerebral embolism model, suggests that decreased reperfusion may be the primary
obstructive mechanism of acute leukocyte-mediated
ischemic injury. This study used a pretreatment strategy and did not assess the efficacy of anti-CD18 in
delayed "resuscitative" treatment. Further studies
evaluating the efficacy of treatment initiated during
the reperfusion phase are planned.
Although possible, it is unlikely that the differences in treatment efficacy found in these two models
are due to differences in brain and spinal cord
882
Stroke Vol 22, No 7 July 1991
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
leukocyte properties because both areas of the CNS
have similar "blood-brain barriers," vascular endothelium, meninges, and glia, and both have previously
documented leukocyte infiltration.15-17 For the middle cerebral embolism model results, the possibility
that we did not detect a real treatment effect because
of inadequate sample size must be considered (Type
II error). Based on the assumption that the observed
18% difference in ET^ values between the treatment
and control groups is real, if we had chosen to design
the experiment to have a power of 90% and set the a
error level at the usual 5%, we would have needed
227 animals in each group to find a statistically
significant difference.28 Although embolism models
do have greater inherent variability, we previously
have found the multiple cerebral embolism model to
be a sensitive test of therapeutic efficacy. Using
sample sizes comparable to this study, we have found
significant treatment effects with various neuroprotective agents.19-29 Thus, it is unlikely that increasing
our sample size within reasonable limits would have
altered our conclusions.
An additional proposed mechanism of leukocyte
potentiation of ischemia is direct CNS tissue infiltration and neuronal cytotoxic injury. This mechanism is
felt to be involved in the delayed (days) deterioration
seen after ischemic reperfusion (delayed reperfusion
injury) that has been observed previously in the
spinal cord ischemia model.18-30 Because the dose of
anti-CD18 used in this study is felt to block adhesion
for 12-18 hours, some inhibition of early leukocyte
infiltration is possible, which may have contributed to
the treatment effect. This is supported by our histopathologic study. However, a multiple-dose regime
would be required to inhibit the period of maximal
infiltration, 24-48 hours.30 Inhibition of leukocyte
infiltration and delayed reperfusion injury recently
has been produced in a variation of the spinal cord
ischemia model using the nonspecific neutrophil inhibitor colchicine.16 Repeated treatments with antiCD 18 may be equally efficacious, with less potential
for side effects.
We conclude that treatment with leukocyte adhesion
antibody reduces CNS ischemic injury in a reperfusion
model, but not in an irreversible occlusion model.
These findings support the role of leukocytes as active
participants in CNS injury, presumably through potentiating reperfusion injury. Leukocyte adhesion antibody
may offer potential for future CNS ischemic therapy.
Finding an agent that reduces CNS reperfusion injury
would have significant clinical benefit, especially given
the potential role of stroke thrombolytic therapy, with
resulting reperfusion.
References
1. Mehta J, Nichols W: Neutrophils as potential participants in
acute myocardial ischemia: Relevance to reperfusion. / Am
Coll Cardiol 1988;ll:1309-1316
2. Engler R: Consequences of activation and adenosine mediated inhibition of granulocytes during myocardial ischemia.
Fed Proc 1987;46:2407-2412
3. Engler R, Dahlgren M, Morris D, Schmid-Schonbein G: Role
of leukocytes in response to acute myocardial ischemia and
reflow in dogs. Am J Physiol 1986;251:314-322
4. Romson JL, Hook BG, Kunkel SL, Abrams GD, Schork MA,
Lucchesi BR: Reduction of the extent of ischemic myocardial
injury by neutrophil depletion in the dog. Circulation 1983;67:
1016-1023
5. Schmid-Schonbein G, Engler R: Granulocytes as active participants in acute myocardial ischemia and infarction. Am J
Cardiovasc Pathol 1986;l:15-29
6. Harlan J: Leukocyte-endothelial interactions. / Am Soc Hem
1985;65:513-525
7. Harlan J: Neutrophil mediated vascular injury. Acta Med
Scand 1985;715:123-129
8. Zimmerman G, Mclntyre T: Neutrophil adherence to human
endothelium in vitro occurs by CDwl8 glycoprotein dependent
and independent mechanisms. J Clin Invest 1988;81:531-537
9. Smith C, Rothlein R, Hughes B, Mariscalco M, Rudloff H,
Schmaistieg F, Anderson D: Recognition of an endothelial
determinant for CD 18-dependent human neutrophil adherence and transendothelial migration. / Clin Invest 1988;82:
1746-1756
10. Price T, Beatty P, Gorpuz S: In vivo inhibition of neutrophil
function in the rabbit using monoclonal antibody to CD 18. /
Immunol 1987;139:4174-4177
11. Hernandez L, Grisham M, Twohig B, Arfors K, Harlan J,
Granger N: Role of neutrophils in ischemia reperfusion
induced microvascular injury. Am J Physiol 1987;253:
H699-H703
12. Arfors K, Lundberg C, Lindbom L, Lundberg K, Beatty P,
Harlan J: A monoclonal antibody to the membrane glycoprotein complex CD 18 inhibits PMN accumulation and plasma
leakage in vivo. Blood 1987;69:338-340
13. Simpson P, Todd R, Fantoen J, Mickelson J, Griffin J,
Lucchesi B: Reduction of experimental canine myocardial
reperfusion injury by a monoclonal antibody that inhibits
leukocyte adhesion. / Clin Invest 1988;81:624-629
14. Vedder N, Winn C, Rice C, Chi E, Arfors K, Harlan J: A
monoclonal antibody to the adherence promoting leukocyte
glycoprotein, CD 18, reduces organ injury and improves
survival from hemorrhagic shock and resuscitation in rabbits. /
Clin Invest 1988;81:939-944
15. Hallenbeck J, Dutka A, Tanishima T, Kochanek P, Kumaroo
K, Thompson C, Obreovitch T, Contreras T: Polymorphonudear leukocyte accumulation in brain regions with low blood
flow during the early postischemic period. Stroke 1986;17:
246-253
16. Giulian D, Robertson C: Inhibition of mononuclear phagocytes reduces ischemic injury in the spinal cord. Ann Neurol
1990;27:33-42
17. Pozzilli C, Lenzi G, Argentino C, Carolei A, Rasura M,
Signore A, Bozzao L, Pozzilli P: Imaging of leukocyte infiltration in human cerebral infarcts. Stroke 1985;16:251-256
18. Zivin J, DeGirolami U, Hurwitz E: Spectrum of neurological
deficits in experimental CNS ischemia. Arch Neurol 1982;39:
408-412
19. Zivin J, DeGirolami U, Lyden P, Kochhar A, Mazzarella V,
Hemenway C, Henry M: A model for quantitative evaluation
of embolic stroke therapy. Brain Res 1987;435:305-308
20. Barton R, Rothlein R, Ksiazek J, Kennedy C: The effect of
antiintercellular adhesion molecule-1 on phorbol ester
induced rabbit lung inflammation. / Immunol 1989;142:
1278-1282
21. Waud D: On biological assays involving quantal responses. /
Pharmacol Exp Ther 1972;183:577-607
22. Lyden P, Zivin J, Kochhar A, Mazzarella V: Effects of calcium
channel blockers on neurologic outcome after focal ischemia
in rabbits. Stroke 1988;19:1020-1025
23. Zivin J, Venditto J: Experimental CNS ischemia: Serotonin
antagonists reduce or prevent damage. Neurology 1984;34:
469-474
24. Ames A, Wright L, Masayoshi K, Thurston J, Majno G:
Cerebral ischemia: The no-reflow phenomenon. Am J Pathol
1968^2:437-443
Clark et al
CNS Ischemia Treatment With Leukocyte Adhesion Antibody
25. Janoff A, Zweifach B: Production of inflammatory changes in
the microcirculation by cationic proteins extracted from lysosomes. / Exp Med 1964;120:747-764
26. Schmid-Schonbein G: Capillary plugging by granulocytes and
the no-refiow phenomenon in the microcirculation. FASEB J
1987;46:2397-2401
27. Braide M, Amundson B, Chien S: Quantitative studies on the
influence of leukocytes on vascular resistance in a skeletal
muscle preparation. Microvasc Res 1984;27:331-352
28. Snedecore G, Cochran W: Statistical Methods. Ames, Iowa,
Iowa State University Press, 1980, pp 103-105
883
29. Kochhar A, Zivin J, Lyden P, Mazzarella V: Glutamate
antagonist therapy reduces neurologic deficits produced by
focal central nervous system ischemia. Arch Neurol 1988;45:
148-153
30. Jacobs T, Shohami E, Baze W, Burgard E, Gunderson C:
Deteriorating stroke model: Histopathology, edema, and
eicosanoid changes following spinal cord ischemia in rabbits.
Stroke 1987;18:741-750.
KEY WORDS
• rabbits
cell adhesion • cerebral ischemia • leukocytes
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Reduction of central nervous system ischemic injury in rabbits using leukocyte adhesion
antibody treatment.
W M Clark, K P Madden, R Rothlein and J A Zivin
Stroke. 1991;22:877-883
doi: 10.1161/01.STR.22.7.877
Downloaded from http://stroke.ahajournals.org/ by guest on June 17, 2017
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1991 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/22/7/877
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/