вход по аккаунту


Incomplete versus complete cerebral ischemia Improved outcome with a minimal blood flow.

код для вставкиСкачать
Incomplete versus Complete
Cerebral Ischemia: Improved Outcome
with a Minimal Blood Flow
Petter A. Steen, MD, John D. Michenfelder, MD, and James H. Milde
It has been reported that incomplete cerebral ischemia with cerebral blood flow less than 10% of control may be
more damaging than an equal period of complete ischemia. I n this study, the effects of severe, incomplete cerebral
ischemia on neurological outcome and cerebral metabolism were studied in dogs anesthetized with nitrous oxide.
The results were compared with those of a previous study concerned with the effects of complete ischemia. Dogs
could sustain only 8 to 9 minutes of complete ischemia with return of normal neurological function, whereas
maintenance of a cerebral blood flow rate less than 10% of control extended this limit to 10 to 12 minutes.
Following a 10-minute exposure, only dogs undergoing incomplete ischemia regained a normal cerebral oxygen
consumption within 90 minutes; similarly, animals subjected to incomplete ischemia enjoyed a faster return of EEG
activity than dogs exposed to complete ischemia of the same duration. Cerebral metabolite levels did not prove to be
a good index of return of neurological function. Within periods of cerebral ischemia in which meaningful neurological recovery might be expected, we conclude that some blood flow is better than no flow.
Steen PA, Michenfelder JD, Milde JH: Incomplete versus complete cerebral ischemia: improved outcome
with a minimal blood flow. Ann Neurol 6:389-398, 1979
The brain can endure only short periods of complete
cerebral ischemia with return of normal neurological
function. Recent studies of metabolic [ 2 5 , 301 and
electrophysiological [9, 101 functions after ischemia
have suggested that incomplete ischemia with less
than 10% of normal cerebral blood flow (CBF) is
even more detrimental to the brain than complete
ischemia. One explanation has been that in incomplete ischemia, particles of slowly moving blood
might form aggregates, preventing adequate recirculation after ischemia [7, 251. Other explanations include the possibility of excessive lactate production
during incomplete ischemia because of a continued
supply of glucose 130, 3 11 or membrane damage from
free radicals that are produced during hypoxia, but
not if oxygen is totally absent [ 2 5 , 301.
To our knowledge, the recovery of behavioral
functions following severe degrees of incomplete
versus complete ischemia has not previously been
compared. Elements of metabolic and electrophysiological recovery have been reported, but
these have not been correlated with known neurological outcome [7, 10, 2 5 , 301. In the present study,
both metabolic and neurological effects of severe incomplete ischemia (CBF less than 10% of normal)
were evaluated and compared in the same model to
the previously reported effects of complete cerebral
ischemia [ 3 3 , 341. In those earlier studies [ 3 3 , 341 we
determined that complete ischemia in dogs is tolerated for only 8 to 7 minutes. Ten minutes of complete ischemia resulted in severe postischemic neurological damage when the animals were evaluated 48
hours later and grossly altered cerebral metabolism
for at least 6 hours following the ischemic injury.
From the Department of Anesthesiology, Mayo Medical School,
Rochester, MN.
Accepted for publication Feb 14, 1079.
Address correspondence to Dr Michenfelder, Department of
Anesthesiology, Mayo Clinic, 200 First St, SW, Rochester, MN
5 5901.
Materials and Methods
Twenty-three unmedicated, fasting adult mongrel dogs
weighing 15 to 23 kg were studied in the prone position.
Five dogs (Group A ) were used to study postischemic cerebral blood flow and metabolism, 6 dogs (Group B) to
examine postischemic brain biopsies for changes in brain
metabolites, and 12 dogs (Group C) to evaluate neurological outcome 48 hours after ischemia (Fig 1).
Group A
In the 5 dogs used to study CBF and metabolism, anesthesia was induced and maintained throughout the surgical
preparation with 1c/o halothane in 60 to 70%;nitrous oxide
and oxygen. The surgical procedure required 20 to 30
minutes, after which halothane was discontinued for at least
20 minutes before induction of ischemia. Succinylcholine,
2 mg per kilogram of body weight, was given intravenously
before tracheal intubation with a cuffed endotracheal tube
0364-5134/79/110389-10$01.25 @ 1978 by John D. Michenfelder 389
Arterial and venous cannulation, EEG, temperature monitoring
Sham thoracotomy
5 dogs for CBF and
CMRo, measurements
Group C
Group B
12 dogs for neurological studies
ICP probe in 3 of the dogs
Sagittal sinus ( annulation
ICP probe In 2 of the dogs
Hemorrhagic hypotension for
10 minutes
Hemorrhagic hypotension for 10 minutes
in 6 dogs, 12 minutes in 3 dogs,
and 14 minutes in 3 dogs
, Brain biopsies at 0, 2, 4,6,
! and 8 minutes after ischemia
CBF and CMRo, followed for
90 minutes after ischemia
Neurological evaluation
48 hours after ischemia
Brain biopsies at 90 minutes
after ischemia
and was thereafter continuously infused (at a rate of 7.5
mg/kg/hr) to maintain muscle paralysis. Ventilation was
controlled with a Harvard pump.
Cannulas were inserted into a femoral vein for drug and
fluid administration and into a brachial artery for arterial
pressure measurements. Because the previous studies of
complete ischemia 133, 341 had required a thoractomy (for
temporary ligation of the aorta and venae cavaej, a sham
thoracotomy was performed with local infiltration of 30 ml
of 0.5% procaine. This step, as well as all other anesthetic
management, was identical to that used in the previous
studies [33, 341 concerned with complete ischemia.
The sagittal sinus was exposed, isolated, and cannulated
as previously described [22, 33-35], The blood flow was
diverted to a reservoir maintained level with the collapsible
veins in the neck ( 1 2 cm below the sagittal sinus). Direct
measurements, in milliters per minute, were made by
timed collections. These were converted to milliliters per
100 gm of tissue per minute by weighing the whole brain at
autopsy and assuming that 5 4 4 of the brain (anterior,
superior, and lateral portions of both hemispheres) was
drained by the cannula, as was shown in a previous study
[35]. A rigid cerebral encasement was thereafter reestablished using Surgicel and cyanoacrylate glue. Since this enclosure was rigid but not airtight, intracranial pressure
(ICP) thereafter could not decrease below zero (ambient
A four-lead bilateral electroencephalogram was obtained
with electrodes glued on the skull. Body temperature was
measured with :in esophageal thermistor and brain temperature with a parietal epidural thermistor, both maintained
390 Annals of Neurology Vol 6 No 5
Fig I . Protocol for the studj of incomplete cerebral ischemia.
at 37.0" k 0.1"C by the use of heat lamps or ice packs, as
indicated. Intracranial epidural pressure was monitored by
a fiberoptic device (Ladd intracranial pressure monitor) in
2 dogs. Pao,, Paco2, arterial p H , and sagittal sinus blood
PO, (Psso,) were determined with Instrumentation Laboratory electrodes at 37°C. Blood oxygen contents were
calculated from measurements of oxyhemoglobin concentrations (Instrumentation Laboratory Model 282 CoOximeter) and oxygen tensions [36]. Cerebral metabolic
rate for oxygen (CMRo,) was calculated as the product of
CBF and the arterial-sagittal sinus blood oxygen content
difference. Blood glucose was determined by standard enzymatic techniques and blood catecholamines by a modified trihydroxyindole method [26].
To produce incomplete ischemia, large-bore (PE 320gauge) cannulas were inserted in both femoral arteries.
These permitted rapid blood withdrawal and reinfusion
using an external heparinized reservoir, thereby facilitating
an easily controlled state of hemorrhagic hypotension. The
dogs were deliberately positioned with the head 15 to 20
cm above heart level; this permitted severe degrees of
hypotension at the head level while still providing adequate
pressures for survival at the heart level. Blood was drawn
off slowly until mean arterial pressure (MAP) reached 60
mm Hg; the arterial pressure was thereafter rapidly reduced (over 20 to 60 seconds) to a level resulting in a
cerebral perfusion pressure (CPP) below 10 mm Hg. CPP
was defined as MAP minus ICP or cerebral venous pres-
November 1979
sure, whichever was highest. Cerebral venous pressure was
not measured directly, but since the head was elevated, the
venous pressure was assumed always to be negative and
numerically equal to the vertical distance between the
sagittal sinus and the collapsible veins in the neck [81 (approximately - 12 cm H,O o r -9 mm Hg). In pilot studies
it was determined that these maneuvers would decrease
CBF to less than 10% of control. T h e time when CBF and
CPP reached these values was recorded as the starting time
for the period of incomplete ischemia. The dogs were
therefore always subjected to a period of abnormally low
CPP and CBF (lasting up to a minute) before the measured
period of severe incomplete ischemia. After 10 minutes of
ischemia, the blood was rapidly reinfused.
CBF and C M k , were measured before, during, and for
90 minutes after the ischemic period. After 90 minutes the
dura overlying the cerebral hemispheres was excised, and
biopsies of brain were taken by a technique that deposits a
sample of brain (200 to 400 mg) into liquid nitrogen within
one second [ 111. The tissue was stored at - 76°C and prepared for analysis in a refrigerated box (-25°C) as described by Folbergrova et al [ 5 ] . Tissue extracts were
analyzed with enzymatic fluorometric methods for phosphocreatine (PCr); adenosine triphosphate, diphosphate,
and monophosphate (ATP, ADP, AMP); glucose [171; lactate; and pyruvate [16]. The sum of the adenine nucleotides
(Ad) was calculated as ZAd = [ATP] + [ADPI [AMP].
The energy state of the tissues was expressed as the energy
charge potential (ECP) of the adenine nucleotide pool according to Atkinson [ 2 ] : ECP = [ATP] 0.5 [ADPIICAd.
Group B
Another group of 6 dogs was studied only for brain biopsies taken at 0 , 2 , 4 , 6, and 8 minutes following 10 minutes
of incomplete ischemia. These dogs did not undergo sagittal sinus cannulation but were otherwise treated identically
to the previous group. The first biopsy was taken from the
left frontal region, the second from the right parietooccipital region, the third from the left parietooccipital region,
the fourth from the right frontal region, and the fifth from
the left parietal region. At least 2 cm of brain was left
between biopsy sites. Kramer et al [l 11 have previously
shown that biopsies can be taken at 2-minute intervals with
this technique without changes in results due to interference from previous biopsies. This conclusion is supported
by the normalization of AMP values following ischemia in
the present and a previous study from this laboratory [34].
Group C
A third and final group of 12 dogs was used for neurological studies. Six dogs were exposed to 10 minutes, 3 dogs to
12 minutes, and 3 dogs to 14 minutes of severe incomplete
ischemia. In these dogs, anesthesia was induced and maintained with 60 to 70% nitrous oxide and oxygen. T h e inspired gas was changed to 100% oxygen 5 minutes after
ischemia and to room air 20 minutes following ischemia.
Intubation, ventilation, venous and arterial cannulations,
and thoracotomy were performed as in the previous two
groups, but succinylcholine was not continuously infused
and heparin was neutralized with an appropriate dose of
protamine sulfate following reinfusion of blood. In these
dogs a four-lead EEG was recorded from needle electrodes.
Body temperature was measured with an esophageal thermistor and maintained at 38.9" & 0.1"C by use of heat
lamps and electrically heated blankets as needed. ICP was
measured in 3 dogs ( 2 exposed to 12 minutes and 1 to 14
minutes of incomplete ischemia). Because ICP was negative during the period of incomplete ischemia with an intact
skull, it was necessary to lower MAP at the head level to
approximately zero to achieve a CPP equal to o r lower than
that obtained during the metabolic studies (7 5 3 mm Hg).
T h e EEG was continuously monitored to ensure an equal
or greater amount of suppression than that seen during the
metabolic studies in which CBF was actually measured.
Arterial blood gases were measured at regular intervals
after ischemia. Ventilation was controlled until spontaneous breathing was deemed adequate (Pac02 < 45 m m Hg).
The dogs were then extubated and observed. Blood gas
determinations were repeated 30 minutes following extubation and, if adequate ( P w o , < 45 mm Hg, P q , > 701,
the dogs were returned to their cages. O n e dog (10 minutes
of incomplete ischemia) developed severe hypoxia (P-, <
50 mm Hg) during this period and was excluded from the
study by the same criteria as in the previous investigation
[34], leaving a total of 11 dogs for neurological evaluation.
At necropsy a pneumothorax was found in this animal.
Forty-eight hours after ischemia, the dogs were evaluated neurologically and assigned to one of four groups.
Grade 1 dogs (no damage) ate and behaved normally, and
their movements were fully coordinated. Grade 2 dogs
(moderate damage) could stand alone but were ataxic or
exhibited partial or complete blindness. Grade 3 dogs (severe damage) could not stand alone or were comatose.
Grade 4 dogs (dead) died within 48 hours. The surviving
dogs were then killed, and at necropsy the thorax was examined grossly to assess any postthoracotomy complications.
Statistical Evaluation
The Fisher exact test [3] was used for statistical comparison
of neurological outcome between dogs in the previous
study [34] exposed to 10 minutes of complete ischemia and
the dogs exposed to 10 minutes of incomplete ischemia.
For all other comparisons between complete and incomplete ischemia, Student's t test for unpaired data was
utilized. For comparison between preischemic and postischemic values in the same group, Student'st test for paired
data was used. No statistical comparisons were carried out
if fewer than 5 determinations were available. A p value of
less than 0.05 was regarded as significant. All mean values
are reported with the standard error of the mean (SEM).
All tables and figures were constructed to permit immediate comparison between the results with complete and incomplete ischemia.
CBF and Metabolic Studies (Group A)
MAP, Icp, AND BLOOD GASES. There were no
significant differences between the c o m p l e t e a n d incomplete ischemia groups at any t i m e before or after
ischemia (Table 1). ICP increased immediately after
Steen et al: Incomplete Cerebral Ischemia
Table I . Mean Arterial Pressure, lntrucranial Pressure, ArteriaL Blood Gases, and Sagittal Sinus
PO, in Dogs Followed 90 Minutes after Ischemia for Metabolic Studies (Group A)
Mean values during
1 Minute postischemia
90 Minutes postischemia
Type of
No. of
124 f 8
145 2 13
7f 3
126 f 12
109 f 10
98f 6
116f 12
(mm Hg)
34 f 4
32 5
2 f1
10 f 1
144 f 12
167 f 9
146 f 14
38 f 0
392 1
20 f 2
7.30 f 0.06 39 2 1
7.33 2 0 . 0 3 41 2 1
7.43 f 0.01 35 2 1
60 2 3
63 f 4
20f 2
207 f 55
140 f 18
139 f 19
155 2 10
58 f 3
59 f 6
39 f 1
40 6
7.06 5 0 . 0 3
7.08 f 0.04
7.21 f 0.02
7.30-t 0.06
33f 2
33 f 1
35 2 1
38 f 1
75 f 9
77 f 5
32 f 2
buffer base
ischemia, concomitant with the rise in CBF and
PWO,, but decreased rapidly to preischemia levels.
CPP during the period of incomplete ischemia was
equal to MAP, 7 k 3 mm Hg. Postischemia MAP
increased to above 100 mm Hg in both groups within
1 minute.
A combined respiratory and metabolic acidosis was
seen in both groups immediately after ischemia.
Pat-, rapidly returned to control levels, whereas the
metabolic component persisted but gradually improved. During the period of incomplete ischemia,
PSSO,decreased to 20 F 2 mm Hg. In both groups,
Psso, increased transiently above control following
ischemia, significantly so at 15 minutes, then progressively decreased to 32 ? 2 and 40 F 8 mm Hg at 90
minutes for complete and incomplete ischemia, respectively.
CBF AND CMRo,. Preischemia CBF (Fig 2) was
114 ? 7 and 100 t 21 m1/100 gm/min in the complete and incomplete ischemia groups, respectively.
During the period of incomplete ischemia, CBF
averaged 5.4 2 0.6% of control (or 5.4 m1/100 gm/
min). Both groups showed an increase in CBF during
the first 10 minutes after ischemia. CBF thereafter
gradually declined and stayed significantly below
control values in both groups after 30 minutes. Although CBF tended to be higher (both in percentage
of control and in absolute values) in the incomplete
ischemia group after ischemia, the difference did not
reach statistical significance at any point.
Preischemia CMRo, (Fig 3) was 4.67
0.29 and
3.96 F 0.44 m1/100 gmlmin in the complete and incomplete ischemia groups, respectively. During the
period o f incomplete ischemia, CMRo, averaged 16
_t 2% of conrrol (or 0.65 ? 0.06 m1/100 gm/min).
One-minute postischemia CMRo, values were at
preischemic levels in both groups, but rapidly decreased over the next 5 minutes. During the next 10to 45-minute interval, CMRo, again increased in
both groups but reached control levels only in the
392 Annals of' Neurology
Vol 6 No 5
Incomplete ischemia
Complete ischemia
0 I0
F i g 2. CBF values before ischemia (control) and during and
after 10 minutes of complete or incomplete ischemia (7 and 5
dogs, respectively).The results are presented as mean percentage
of controlvalues 2 SEM (controlvalues, 114 2 7 and 100 ?
2 1 mll100 gmlmin, respectively).After an initial 10-minute
period of hyperemia after the ischemia, CBF stabilized below
control levels in both groups.
incomplete ischemia group. In the complete ischemia
group, CMRo, remained significantly below the
control level and after 30 minutes was also significantly below that of the incomplete ischemia group.
Catecholamines were significantly elevated 5 minutes after ischemia in both groups
(Table 2). There were no significant differences
between the groups.
Cerebral Metabolite Studies (Groups A and B)
MAP AND BLOOD GASES. There were no significant differences between the complete and incomplete ischemia groups used for serial biopsies
November 1979
emia, high-energy phosphates (PCr and ATP) were
greatly depleted in both groups (Table 3, Fig 4).
Most of the restitution of PCr, ATP, ADP, and AMP
toward preischemic levels occurred during the first 4
minutes after ischemia, with a concomitant normalization of the ECP. Thereafter, recovery followed a
slower course of continued increase in PCr and in the
total adenine nucleotide pool. Lactate levels and
lactatelpyruvate ratios were high initially following
ischemia, with a reduction toward normal values after
90 minutes. Glucose levels, very low initially after
ischemia, rapidly increased to above control values
after 8 minutes and remained elevated at 90 minutes
Incomplete ischemia
Fig 3. CMRo, values before ischemia (control) and during and
after 10 minutes of complete or incomplete ischemia (7 and 5
dogs, respectively). The results are presented as mean percentage
of control values ? SEM (control values, 4.7 f 0.3 and 4.0 f
0.4 mlll00 gmlmin, respectively).After an initial high level
the first minute after ischemia, CMRo, decreased t o well below
controllevels with a secondary increase 15 to 4s minutes after
the ischemia. Only dogs exposed t o incomplete ischemia showed
a return to control levels at 90 minutes.
Neurological Studies (Group C)
100 mm Hg before ischemia, with no significant differences between the complete and incomplete ischemia groups. During the period of incomplete ischemia, MAP was - 1 1 mm Hg and ICP -4 k 1 mm
Hg (ICP was measured in only 3 dogs). This resulted
in a CPP similar to that necessary to reduce CBF to
5% of normal in the metabolic studies. MAP returned to preischemia levels in all dogs within 1 minute after ischemia. There were no significant differences in arterial blood gases between the groups
(Table 4). Arterial acid-base status had returned to
normal levels in all groups at the time of extubation.
(Group B), nor were there any significant differences
between Groups A and B.
CEREBRAL METABOLITES. There were no significant differences between dogs exposed to complete
and incomplete ischemia. After 10 minutes of isch-
Table 2. Arterial Catecholamine Levels Before and After 10 Minutes of Ischemia (Group A)
5 Minutes after
Type of
N o . of
1.78 ? 0.56
1.90 2 0.47
10.81 f 2.77"
5.99 ? 1.14"
0.36 ? 0.11
0.51 f 0.14
1.68 f 0.27"
1.82 ? 0.66
2.14 f 0.64
2.41 f 0.60
12.49 f 2.89"
7.81 ? 1.41"
aSignificantly different from preischemic values.
Table .3. Control and Postischemic Cerebral Metabolic Values after
10 Minutes of Complete or Incomplete Ischemia (Groups A and B )
0 Min
4 Min
90 Min
Type of
No. of
3 . 0 4 f 0.17 2.50 f 0.09 0.92
0.01 2.21 f 0.18 1.04 f 0.14
0.04 1.54 f 0.14 0.38 f 0.05 0.24
1.71 f 0.10 0.43 f 0.04 0.52
f 0.20 1.94 & 0.08 0.92 f 0.01 2.40
? 0.27 1.97 ? 0.05 0.91 f 0.02 3.07
2 0.16 2.19 f 0.19 0.92 2 0.01 4.48
f 0.15 2.11 & 0.07 0.91 2 0.01 4.29
f 0.06
PCr = phosphocreatine; Ad = adenine nucleotides; ECP
LIP Ratio
15.4 2 0.4 132
16.4 f 2.0
1 . 8 0 2 0.30 21
1.74 f 0.21 17
f 0.04 14.7 ? 1.0
f 0.15 14.0 f 0.7
f 0.20
f 0.67
f 0.44
f 0.22
f 20
f 23
energy charge potential; L/P ratio = lactatelpyruvate ratio.
Steen et al: Incomplete Cerebral Ischemia
or dead. By contrast, all 8 dogs exposed to either 10
o r 12 minutes of severe incomplete ischemia either
were normal or exhibited only slight ataxia. The difference for dogs exposed to 10 minutes of ischemia is
significant. All 3 dogs exposed to 14 minutes of incomplete ischemia were severely damaged. Most
dogs exposed to 10 minutes of incomplete ischemia
were awake again 60 to 70 minutes after the insult,
comparable to the time of return of a normal
E E G , All Studies (Groups A, B , a n d C/
The EEG became isoelectric at 35 ? 2 seconds in
dogs subjected to complete ischemia and at 30 2 5
seconds in those that had incomplete ischemia. During complete ischemia the EEG stayed isoelectric in
all dogs throughout the ischemic period. During incomplete ischemia, most dogs had two or three short
periods of activity. O n e dog had 1- to 2-second bursts
of activity followed by 5 or 6 seconds of suppression
throughout 14 minutes of ischemia, and 2 dogs (1
subjected to 12 minutes of ischemia, 1 studied with
serial biopsies after ischemia) exhibited single or
multiple spikes separated by total suppression of activity throughout the ischemic period. The mean
F i g 4 ATP, ADP, and AMP d u e s In control (nonisrhemzc)
dogs and after r t~henita z n dogs mposed to 10 minutes of complete or rricomplete zsihemta BJ 4 minuter after ischemia, ATP
uiaJ at 8 0 7 of the iontrol lerel und ADP and AMP were
normal There u ere n o dijfirenct~sbetween the groups.
NEUROLOGICAL STATUS 848 HOURS AFTER ISCHEMIA. Most dogs exposed to 8 minutes of complete
ischemia were normal (Table 5). T h e results at 7
minutes were intermediate, while after 1 0 minutes of
complete ischemia 6 of 7 dogs were severely injured
Table 4 . Artenal Blood Gases bcfore Itrhemza and at the Time of Extubation in Dogs FoLouvd 48 Hours (Group Ci
Duration of Ischemia
Complete Ischemia
9 Min
8 Min
( N = 6)
P q , (mm Hg)
At extubation
At extubation
At extubation
At extubation
3 4 6 + 25
374 + 8
7 7 + 6
35 r 1
38 -t 1
37 f 2
33 -I- 2
7.32 + 0.03 7.33 + 0.02
7.32 + 0.02 7.34 + 0.02
41 + 2
40+ 1
392 2
39-c 1
PaCO., (mm Hg)
BB (mEqlL)
Incomplete Ischemia
10 Min
( N = 7)
10 Min
( N = 5)
1 2 Min
( N = 3)
14 Min
( N = 3)
288 + 71
392 + 13
378 + 25
3 9 4 + 23
9 7 2 5
9 7 + 1
9 5 t 4
39+ 2
38+ 0
39+ 1
37 t 1
95 5
33t 2
37+ 2
7.36 + 0.02 7.32 ? 0.02 7.28 t 0.02 7.31 t 0.04
7.37 + 0.02 7.30 + 0.02 7.32 t 0.02 7.36 t 0.01
42 -t 1
40+ 1
38+ 1
39t 2
38+ 1
3 8 + 1
3 7 + 2
49+ 1
buffer base.
Table 5 , Neurological Status 48 Hour.! after Complete or Incomplete Ischemia (Group C )
No. of Dogs with
Grade of Neurological Damage:
Duration of
Type of
No. of
394 Annals of Neurology Vol 6 No 5 November 1979
total duration of such EEG activity during incomplete
ischemia was 55 -+ 14 seconds, and there were no
differences between the groups studied for metabolic
or neurological results. With equal duration of ischemia (10 minutes), EEG activity returned significantly more rapidly after incomplete (12 ? 2 seconds)
than complete ischemia (18.0 2.9 minutes).
A comparison of complete and severe incomplete cerebral ischemia ideally should be performed in the
same model to be both clinically relevant and reproducible. Unfortunately, this has not proved feasible
[ 14, 2 11. The production of reproducible, reversible,
complete cerebral ischemia has been consistently
demonstrated in only two models: by increasing ICP
above systolic pressure [18, 301 or by clamping the
great vessels 17, 14, 21, 33, 341. In our studies of
complete cerebral ischemia [33, 341 we chose the
latter model because it more closely mimics the most
commonly encountered clinical events such as cardiac arrest. The same approach (i.e., partial clamping of the great vessels) could not be used reliably
for incomplete ischemia. Hemorrhagic hypotension
proved to be a reliable means for maintaining severe
incomplete ischemia while all other variables that
might be of importance could be held at levels similar
to those in the complete ischemia model. There were
no differences in the models with regard to degree of
intracranial manipulations. In both models, the return of cardiac function after ischemia was similar.
Finally, the only major extracranial manipulation,
thoracotomy, was performed in both groups. Thus an
immediate comparison of the results with complete
and incomplete ischemia using these models should
be valid.
Complete ischemia in excess of 8 to 9 minutes
resulted in severe neurological damage in our dogs
[34],in agreement with findings reported by others
17, 21, 391. With the retention of even minimal
CBF (i.e., less than 10% of normal) this limit was
extended such that more than 12 minutes of profound incomplete ischemia was necessary to produce
severe injury when judged by neurological function
48 hours after ischemia. One must conclude that
some flow to the brain, no matter how small, is better
than no flow.
Such a conclusion stands in sharp contrast to those
hawn by Hossmann et a1 [9, 101 from electro3hysiological studies and by Nordstrom et a1 [25,
301 from metabolic studies. Hossmann et a1 [9, 101
ntended to expose cats and monkeys to complete
i erebral ischemia of 60 minutes’ duration by tying off
he main arteries to the head and reducing MAP to
30 mm Hg. Some of the animals, however, had a
remaining CBF estimated to be approximately 5 ml/
100 gm/min. Among the 45 animals that did not regain any electrophysiological activity after the insult,
9 had been exposed to incomplete ischemia, in contrast to only 2 of the 42 animals in which spontaneous
EEG activity returned. Nordstrom et a1 [25, 301 induced incomplete ischemia for 30 minutes in rats
anesthetized with 70% nitrous oxide by clamping
both carotid arteries and reducing MAP to 50 mm
Hg. This diminished cortical blood flow to approximately 5 m1/100 gm/min (5% of control) while retaining higher flows to other brain regions (6092 of
control in the medulla oblongata, 25% in the cerebellum). The postischemic effects of this insult on
cerebral metabolites were then compared to the results after 30 minutes’ exposure to complete cerebral
ischemia, achieved by increasing cerebrospinal fluid
pressure to above systolic blood pressure [24]. By 90
minutes after ischemia all rats exposed to complete
ischemia showed extensive restitution of cerebral
energy metabolites, with normalization of lactate,
PCr, ADP, and AMP, and 99% recovery of the ECP.
Following the same duration of incomplete ischemia,
no rats under nitrous oxide anesthesia showed a corresponding restitution in any of these metabolites. In
neither study was there any attempt to assess the
neurological function of the animals during recovery
from ischemia.
Postischemic neurological function was examined
in rabbits by Marshall et al, who found that functional
recovery [ 181 and brain morphology [ 191 were
significantly better after 15 minutes of incomplete
than of complete ischemia. However, they did not
produce severe degrees of oligemia since CBF was
only reduced to 26 m1/100 gm/min, or 38% of control. Their results therefore do not rule out the possibility of a deleterious effect of more severely reduced CBF. The same criticism cannot apply to the
present study. CBF was reduced to the same degree
as reported by Hossmann and associates [9, 101 and
Nordstrom et a1 [25, 301, namely, 5 ml/100 gm/min,
or less than 10% of control. Both the neurological
and the metabolic studies indicated better recovery
after incomplete ischemia, with better neurological
function, faster return of EEG activity, and reestablishment of a normal CMRo, 1 hour after ischemia.
A likely explanation for the varying results and
conclusions might be found in the different durations
of ischemia. Nordstrom et a1 [25, 301 studied 30
minutes of ischemia; Hossmann et a1 [9, 10],60 minutes. In our previously published study [341, only 10
minutes of complete cerebral ischemia in dogs was
required to produce consistent, severe neurological
deficits. For clinical relevance we therefore chose to
study incomplete ischemia of the same duration. It is
noteworthy that while serious neurological injury occurred after 10 minutes of complete ischemia, this
Steen et al: Incomplete Cerebral Ischemia
duration was not sufficient to block return of electrical activity or normalization of brain metabolites in a
manner similar to that seen following 10 minutes of
incomplete ischemia. It would thus appear that assessments conhned to electrophysiological activity or
metabolic recovery may not provide valid indices of
neurological recovery following periods of ischemia,
whether complete o r incomplete. Granting that if
ischemia lasts long enough, incomplete circulatory
interruption can be more deleterious than complete
ischemia (as judged by effects on the EEG or
metabolites), this appears to occur only long after
irreversible brain damage develops, making meaningful neurological recovery of the organism impossible.
Since recovery of electrophysiological phenomena
or metabolic functions cannot be correlated with
overall neurological recovery following ischemia, the
clinical value of investigations limited to such indices
must be questioned. This concern clearly applies to
the studies by Nordstrom et a1 [25, 301, since Ljunggren and co-workers [ 151 reported that rats failed to
recover functionally after 5 minutes of complete ischemia despite biochemical restitution, even after 15 minutes of complete ischemia; according to Schutz et a1
[28], the latter animals also had partial restitution of
the EEG. The same criticism applies to the studies by
Hossmann et al. All evidence from patients indicates
that a GO-minute period of ischemia is far too long to
allow any meaningful functional recovery. Lack of
correlation between EEG, metabolic, and functional
recovery has also been reported by others. Salford et
a1 [27] found that adenine nucleotides and carbohydrate intermediates returned to normal in the early
stages of hypoxic-oligemic injury in rats even though
histological abnormalities already had developed. In
rats exposed to hypoxia, Nilsson and Busto [23]
found a strikingly poor correlation between recovery
of mitochondrial function and restitution of the EEG
or clinical recovery. Similar observations were reported by Yatsu et al [ 3 7 , 381 and Marshall and associates 1201 after incomplete ischemia in rabbits.
The lack of correlation might be explained in part
by the fact that static levels of cerebral metabolites
need not give a complete picture of mitochondrial
function since these are influenced by the rate of
production and utilization of high-energy phosphates. However, in vitro studies of brain mitochondria previously exposed to ischemia in vivo have indicated high ATP production and tight coupling [38,
391. It is therefore likely that the neuronal cell function first damaged by hypoxia is not mitochondrial
energy production, but rather an energy-dependent
process 133, 391. This inference is supported by the
results of our previous study of complete ischemia
[34],in which CMRo, was the only metabolic factor
396 Annals of Neurology
Vol 6 No 5
significantly better (higher) in dogs that recovered
functionally compared with those that did not recover when studied 48 hours after ischemia. This is
also in agreement with the present study, in which
CMRo, was normalized 60 minutes after incomplete
ischemia but not after complete ischemia.
Evidence for the postulated complications of incomplete ischemia, including no-reflow, lactate accumulation, or free radical damage, was not observed
in this study. Had no-reflow [l] occurred, at least a
few of our brain biopsies should have revealed elevated AMP levels at some time after 4 minutes. Since
this sensitive indicator of metabolic recovery [ 2 9 , 331
was always normal, lack of tissue reperfusion seems
unlikely. It would again appear that no-reflow is a
phenomenon observed only after prolonged ischemia. Thus, Levy et a1 [12, 131 observed in rats and
gerbils that ischemic brain damage precedes any evidence of vascular obstruction. Ginsberg et a1 [6]
could not demonstrate the no-reflow phenomenon in
cats after 15 minutes of incomplete ischemia (CBF, 4
m1/100 gm/min), but only after 30 minutes of ischemia, a time when histological changes are pronounced and neurological recovery seems unlikely.
O u r results d o not support the suggestion that lactate accumulation would be greater during incomplete than complete ischemia because of continued
delivery of substrate (glucose). It seems probable that
lactate production would be increased during partial
ischemia, but since static levels of lactate were the
same following complete or incomplete ischemia, it
must be concluded that lactate removal occurs at the
same rate as substrate delivery.
The suggestion that cellular damage by free radicals (molecules with an unpaired electron) is a main
factor in cerebral hypoxia and ischemia [4, 25, 301
was not supported by the present results. Barbiturates protected against the deleterious effect of incomplete ischemia in the study of Nordstrom et a1
[ 2 5 ] ,and quenching of free radical reactions has been
suggested as the mechanism for barbiturate protection [4, 25, 301. W e have previously failed to find any
correlation between cerebral protection with barbiturates and free radical scavenging [32], and in the
present study we failed to confirm the importance of
free radical damage during ischemia since a continued minimal oxygen supply, which should increase
the amount of free radicals [ 4 ] ,had no deleterious
The clinical implications of a beneficial rather than
a detrimental effect of minimal CBF during a period
of ischemia are important. Thus, in the management
of cardiac arrest, immediate efforts should be directed toward restoring at least some circulation to
the brain, even when thought inadequate, rather than
delaying resuscitation attempts until all resources are
November 1979
available for a fully effective effort. Similarly, the
maintenance of at least some cerebral perfusion during procedures such as carotid endarterectomy
should always be sought as a means of minimizing
cerebral complications. Finally, the occasional need
for profound degrees of hypotension in certain surgical procedures should not be replaced by techniques
which temporarily arrest the circulation.
From the present results we conclude: (1) that
metabolic and electrophysiological studies of cerebral function after ischemia are of limited value if not
correlated with evaluation of total neurological recovery; (2) that the retention of a minimal CBF, i.e.,
less than 10% of normal, is beneficial to the ischemic
brain; and (3) that in the clinical management of an
episode of cerebral ischemia, a major goal should be
the maintenance of at least some cerebral perfusion,
no matter how small.
Supported in part by Research Grant NS-7507 from the National
Institutes of Health, US Public Health Service.
1. Ames A 111, Wright RL, Kowada M: Cerebral ischemia. 11.
The no-reflow phenomenon. Am J Pathol 52:437-453, 1968
2. Atkinson DE: The energy charge of the adenylate pool as a
regulatory parameter. Interaction with feedback modifiers.
Biochemistry 7:4030-4034, 1968
3. Croxton FE: Elementary statistics with applications in medicine and the biological sciences. New York, Dover, 1959, pp
4 . Demopoulos HB, Flamm ES, Seligman ML, et al: Antioxidant
effects of barbiturates in model membranes undergoing free
radical damage. Acta Neurol Scand [Suppll 56:152-153,
5 . Folbergrova J, MacMillan V, Siesjo B K The effect of moderate and marked hypercapnia upon the energy state and upon
the cytoplasm NAD/NADH+ ratio of the rat brain. J
Neurochem 19:2497-2505, 1972
6. Ginsberg MD, Budd WW, Welsh FA: Diffuse cerebral ischemia in the cat: I. Local blood flow during severe ischemia
and recirculation. Ann Neurol 3:482-492, 1978
7. Goldstein A Jr, Wells BA, Keats AS: Increased tolerance to
cerebral anoxia by pentobarbital. Arch Int Pharmacodyn Ther
161:138-143, 1966
8. Guyton AC, Jones CE, Coleman TG: Circulatory Physiology:
Cardiac Output and Its Regulation. Second edition. Philadelphia, London, Toronto, Saunders, 1973
9 . Hossmann K-A, Kleihues P: Reversibility of ischemic brain
damage. Arch Neurol 29:375-384, 1973
10. Hossmann K-A, Zimmermann V: Resuscitation of the monkey brain after 1 hour complete ischemia. I. Physiological and
morphological observations. Brain Res 81:59-74, 1974
11. Kramer RS, Sanders AP, Lesage AM, et al: The effect of
profound hypothermia on preservation of cerebral ATP content during circulatory arrest. J Thorac Cardiovasc Surg
56:699-709, 1968
:2. Levy DE, Brierley JB, Plum F: Ischemic brain damage in the
gerbil in the absence of “no reflow.” J Neurol Neurosurg
Psychiatry 38:1197-1205, 1975
13. Levy DE, Brierley JB, Silverman DG, et al: Brief hypoxiaischemia initially damages cerebral neurons. Arch Neurol
32:450-456, 1975
14. Lind B, Snyder J, Kampschulte S, et al: A review of total brain
ischemia models in dogs and original experiments on clamping
the aorta. Resuscitation 4:19-31, 1975
15. Ljunggren B, Granholm L, Schutz H , et al: Energy state of the
brain during and after compression ischemia, in Brock M,
Diett H (eds): Intracranial Pressure. Experimental and Clinical Aspects. Heidelberg, New York, Springer, 1972, pp
16. Lowry OH, Passonneau JV: A Flexible System of Enzymatic
Analysis. New York, London, Academic, 1972
17. Lowry O H , Passonneau JV, Hasselberger FX, et al: Effect of
ischemia on known substrates and cofactors of the glycolytic
pathway in brain. J Biol Chem 239:18-30, 1964
18. Marshall LF, Durity F, Lounsbury R, et al: Experimental cerebral oligemia and ischemia produced by intracranial hypertension. Part I: Pathophysiology, electroencephalography, cerebral blood flow, blood-brain barrier, and neurological function. J Neurosurg 43:308-317, 1975
19. Marshall LF, Graham DI, Durity F, et al: Experimental cerebral oligemia and ischemia produced by intracranial hypertension. Part 2: Brain morphology. J Neurosurg 43:318-322,
20. Marshall LF, Welsh F, Durity F, et al: Experimental cerebral
oligemia and ischemia produced by intracranial hypertension.
Part 3: Brain energy metabolism. J Neurosurg 43:323-328,
21. Marshall SB, Owens YC, Swan H : Temporary circulatory occlusion to the brain of the hypothermic dog. Arch Surg
72:98-106, 1956
22. Michenfelder JD, Messick JM, Theye RA: Simultaneous cerebral blood flow measured by direct and indirect methods. J
Surg Res 8:475-481, 1968
23. Nilsson L, Busto R: Brain energy metabolism during the process of dying and after cardiopulmonary resuscitation. Acta
Anaesthesiol Scand 20:57-64, 1976
24. Nordstrom C-H, Rehncrona S, Siesjo BK: Restitution of cerebral energy state, as well as of glycolytic metabolites, citric
acid cycle intermediates and associated amino acids after 30
minutes of complete ischemia in rats anaesthetized with nitrous oxide or phenobarbital. J Neurochem 30:479-486,
25. Nordstrom C-H, Rehncrona S, Siesjo B K Effects of
phenobarbital in cerebral ischemia. Part 11: Restitution of cerebral energy state, as well as of glycolytic metabolites, citric
acid cycle intermediates and associated amino acids after pronounced incomplete ischemia. Stroke 9:335-343, 1978
26. Perry LB, Van Dyke RA, Theye RA: Sympathoadrenal and
hemodynamic effects of isoflurane, halothane, and cyclopropane in dogs. Anesthesiology 40:465-470, 1974
27. Salford LB, Plum F, Siesjo BK: Graded hypoxia-oligemia in
rat brain. I. Biochemical alterations and their implications.
Arch Neurol 29:227-233, 1973
28. Schutz H , Silverstein PR, Vapalahti M, et al: Brain mitochondrial function after ischemia and hypoxia. I. Ischemia induced
by increased intracranial pressure. Arch Neurol 29:408-4 16,
29. Siesjo BK, Ljunggren B: Cerebral energy reserves after prolonged hypoxia and ischemia. Arch Neurol29:400-407, 1973
30. Siesjo BK, Nordstrom C-H, Rehncrona S: Metabolic aspects
of cerebral hypoxia-ischemia. Adv Exp Med Biol78:261-269,
31. Silver IA: Changes in PO, and ion fluxes in cerebral hypoxiaischemia. Adv Exp Med Biol 78:299-312, 1977
32. Steen PA, Michenfelder JD: Cerebral protection with bar-
Steen et al: Incomplete Cerebral Ischemia
biturates. Relation to anesthetic effect. Stroke 9: 140-142,
33. Steen PA, Milde J H , Michenfelder JD: Cerebral metabolic
and vascular effects of barbiturate therapy following complete
global ischemia. J Neurochem (in press)
34. Steen PA, Milde JH, Michenfelder JD: No barbiturate protection in a dog model of complete cerebral ischemia. Ann
Neurol 5:343-349, 1979
35. Takeshita H, Michenfelder JD, Theye RA: The effects
of morphine and N-allyl-morphine on canine cerebral
metabolism and circulation. Anesthesiology 37:605-612,
36. Theye RA: Calculation of blood 0, content from optically
determined H b and HbO,. Anesthesiology 33:653-657,
37. Yatsu FM, Lee L-W, Liao C-L Energy metabolism during
brain ischemia. Stability during reversible and irreversible
damage. Stroke 6:678-683, 1975
38. Yatsu FM, Lindquist P, Graziano C: An experimental model
of brain ischemia combining hypotension and hypoxia. Stroke
5132-39, 1974
39. Young WP, Javid M: Study of the use of intravenous urea
after simulated cardiac arrest in dogs. Surg Forum 10:522524, 1960
398 Annals of Neurology Vol 6 No 5 November 1979
Без категории
Размер файла
919 Кб
flow, outcomes, ischemia, complete, incomplete, versus, improve, minimax, cerebral, blood
Пожаловаться на содержимое документа