1590
Mild Hypothermia After Cardiac Arrest
in Dogs Does Not Affect Postarrest
Multifocal Cerebral Hypoperfusion
Ken-ichi Oku, MD; Fritz Sterz, MD; Peter Safar, MD; David Johnson, MD; Walter Obrist, PhD;
Yuval Leonov, MD; Kazutoshi Kuboyama, MD; Samuel A. Tisherman, MD; S.W. Stezoski
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
Background and Purpose: Although mild resuscitative hypothermia (34°C) immediately after cardiac
arrest improves neurological outcome in dogs, its effects on cerebral blood flow and metabolism are
unknown.
Methods: We used stable xenon-enhanced computed tomography to study local, regional, and global
cerebral blood flow patterns up to 4 hours after cardiac arrest in dogs. We compared a normothermic
(37.50C) control group (group I, n=5) with a postarrest mild hypothermic group (group II, n=5). After
ventricular fibrillation of 12.5 minutes and reperfusion with brief cardiopulmonary bypass, the ventilation, normotension, normoxia, and mild hypocapnia were controlled to 4 hours after cardiac arrest. Group
II received (minimal) head cooling during cardiac arrest, followed by systemic bypass cooling (to 34°C)
during the first hour of reperfusion after cardiac arrest.
Results: The postarrest homogeneous transient hyperemia was followed by global hypoperfusion from 1
to 4 hours after arrest, with increased "no-flow" and "trickle-flow" voxels (compared with baseline),
without group differences. At 1 to 4 hours, mean global cerebral blood flow in computed tomographic slices
was 55% of baseline in group I and 64% in group II (NS). No flow (local cerebral blood flow <5 mL/100
cm3 per minute) occurred in 5±2% of the voxels in group I versus 9±5% in group II (NS). Trickle flow
(5 to 10 mL/100 cm3 per minute) occurred in 10±3% voxels in group I versus 16±4% in group II (NS).
Cerebral blood flow values in eight brain regions followed the same hyperemia-hypoperfusion sequence as
global cerebral blood flow, with no significant difference in regional values between groups. The global
cerebral metabolic rate of oxygen, which ranged between 2.7 and 4.5 mL/100 cm3 per minute before arrest
in both groups, was at 1 hour after arrest 1.8±0.3 mL in normothermic group I (n=3) and 1.9±0.4 mL
in still-hypothermic group II (n=5); at 2 and 4 hours after arrest, it ranged between 1.2 and 4.2 mL in
group I and between 1.2 and 2.6 mL in group II.
Conclusions: After cardiac arrest, mild resuscitative hypothermia lasting 1 hour does not significantly
affect patterns of cerebral blood flow and oxygen uptake. This suggests that different mechanisms may
explain its mitigating effect on brain damage. (Stroke. 1993;24:1590-1598.)
KEY WoRDs
A
*
cerebral blood flow
*
cerebral ischemia
chieving optimal neurological recovery after cardiac arrest probably requires a multifaceted
treatment' to mitigate the postresuscitation
syndrome,2 which includes (1) reduced cerebral blood
flow (CBF) in relation to the cerebral metabolic rate of
oxygen (CMRO2; ie, 02 uptake) and (2) reoxygenation
injury chemical cascades resulting in lipid peroxidation
of tissues. Hypothermia may mitigate both.3 With normothermia, CBF changes after cardiac arrest include
multifocal no-reflow4 and transient diffuse hyperemia
Received December 30, 1992; final revision received April 14,
1993; accepted May 28, 1993.
From the International Resuscitation Research Center (IRRC)
(K-i.O., F.S., P.S., Y.L., K.K., S.A.T., S.W.S.) and the Departments of Anesthesiology/Critical Care Medicine (K-i.O., F.S., P.S.,
Y.L., K.K., S.W.S.), Radiology (D.J.), Surgery (S.A.T.), and
Neurosurgery (W.O.), Presbyterian-University Hospital, University of Pittsburgh (Pa).
Address for correspondence: Peter Safar, MD, International
Resuscitation Research Center, University of Pittsburgh, 3434
Fifth Avenue, Pittsburgh, PA 15260.
*
hypothermia
*
resuscitation
See Editorial Comment, page 1598
followed by delayed, protracted reduction in CBF,56
which is inhomogeneous.7-11
Moderate hypothermia (with total body, ie, core, or
central venous temperature [Tc] of 30°C), without cardiac arrest, reduces CBF and cardiac output parallel
with oxygen consumption.'1213 It reduces microcirculation by increasing viscosity and hematocrit (Hct),14
which could offset its beneficial effects at the cellular
level. Moderate hypothermia can cause arrhythmias,
and if prolonged, also infection.1516 Mild hypothermia
(Tc 34°C) can be induced more rapidly, has fewer
management problems, and clinically does not seem to
exert deleterious effects on the cardiovascular system.
We have found in a systematic series of four outcome
studies in dogs3'17-19 that mild cerebral hypothermia
after cardiac arrest is more effective than moderate
hypothermia20 in reducing neurological deficit and is
less likely to cause myocardial damage.3"18'20 The effect
of postarrest hypothermia (mild or moderate) on CBF
Oku et al CBF With Mild Resuscitative Hypothermia
1591
gCBF
mil10OgImin
% VOXELS
100
100
_
FLOW AREAS
80 -
mJI10OgImin
E
>120
:1 40-120
60-
F-,
*_
20-40
10-20
40-
a
5-10
*
0-5
20-
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
0
30min
1 h30min
2h
2h30min
TIME
!
3h30min
3h
~
~
~
~
~
4h
MAP
100
120
115
115
100
~
100
~
105
100
PaCO2
31.9
35.2
37.1
33.7
32.9
31.9
37.2
36.5
Temp
37.6
37.4
36.0
34.0
32.0
30.1
28.1
37.8
FIG 1. Graph showing sham experiment without cardiac arrest, with hypothermia. Global cerebral blood flow (gCBF) for the
posterior coronal slice and local CBF values in percent voxels with different flow ranges (left). As central venous temperature was
lowered, gCBF decreased and low-flow voxels increased. Rewarming reversed this process.
and CMRO2 after cardiac arrest and cardiopulmonary
resuscitation (CPR) remains to be studied.
From 1986 through 1990 we explored global, regional,
and local CBF and CMRO2 patterns before and after
cardiac arrest, using established dog models21'22 and
stable xenon-enhanced computed tomography (XeCT).23-25 First, we established the CBF method under
normothermia throughout in a model of external CPR
and then reconfirmed the sequence of hyperemia-hypoperfusion after cardiac arrest.8 Second, we documented
the reproducible heterogeneity of delayed postarrest
hypoperfusion after reperfusion with open-chest CPR
or cardiopulmonary bypass (CPB).9 Third, we found
that hypertensive reperfusion plus hemodilution normalized global, regional, and local CBF after cardiac
arrest10; hemodilution, however, by reducing arterial 02
content, did not improve cerebral 02 delivery. Fourth,
we found the aminosteroid U74006F to have no effect
on CBF and CMRO2 after cardiac arrest.'1
In the present (fifth) study,26 using the same model
and methods, we compared global, regional, and local
CBF patterns in a previously reported normothermic
control group (group I, n=5)9 with those in a mild
hypothermic group (group II, n=5). We also compared
global CMRO2 in group I (n=3) with that in group II
(n=5). In previous studies5,6,9 we found that global
CMRO2 reached or exceeded normal values during
hypoperfusion between 2 hours and 10 hours after
cardiac arrest, ie, was mismatched with oxygen delivery.
We hypothesize that mild hypothermia might alter CBF
and CMRO2 patterns after cardiac arrest in a way that
could explain (at least in part) its mitigating effect on
brain damage.3
Materials and Methods
This project was approved by the Animal Use Committee of the University of Pittsburgh. We used 11
healthy, custom-bred male coon hounds from the same
breeding colony, mean age 10 (range, 8 to 12) months
and mean weight 17 (range, 14 to 18) kg.
One sham dog (CT30) was studied without cardiac
arrest to observe the effect of mild (Tc 34°C) and
moderate (Tc 30°C) hypothermia (Fig 1) on multifocal
CBF over 4 hours; this was carried out under the same
light halothane and fentanyl anesthesia and controlled
intermittent positive pressure ventilation (IPPV) as that
used for baseline measurements in the cardiac arrest
experiments. After two CBF CT measurements were
taken during normothermia (Tc 37.5°C), additional
CBF measurements were taken every 30 minutes, with
Tc decreased to 36°C, 34°C, 32°C, 30°C, and 28°C. At 4
hours the Tc was returned to normothermia. All other
physiologic variables (Table 1) were kept constant.
Ten dogs were studied with ventricular fibrillation
(VF) cardiac arrest lasting 12.5 minutes, reperfusion by
full CPB for <5 minutes after cardiac arrest, and IPPV
to 4 hours. Control group I (n=5) comprised the
normothermic control experiments conducted by the
same team about 1 year earlier.9 Hypothermic group II
dogs (n=5) had the head immersed in ice water starting
at VF 3 minutes and then received mild systemic total
body cooling (to Tc 34°C) by low-flow CPB from reperfusion to 1 hour. The head was placed in a specially
designed, plastic head box that served as a fixation
device and for ice-water immersion of the cranium. This
intraischemic cooling "for preservation" -which lowers
1592
Stroke Vol 24, No 10 October 1993
TABLE 1. Physiological Variables Measured at Baseline and After Cardiac Arrest in Dogs
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
Before
Cardiac
Arrest
10 min
Variable
(Baseline)
Group I (normothermic, n=5)
37.0+0.2
37.8+0.6
Tc, °C
110+8
101±+8
MABP, mm Hg
7.34±0.04
7.36±0.03
Arterial pH
34±2
35+2
Paco2, mm Hg
30±8
Hct, %
42±3
Group II (hypothermic, n=5)
37.6±0.1
33.9±0.3
Tc, C
MABP, mm Hg
110±11
139±19
7.37±0.02
7.43±0.06
Arterial pH
33±4
29±8
Paco2, mm Hg
31±4
48±3
Hct, %
Tc indicates core temperature; MABP, mean arterial
After Cardiac Arrest
30 min
1 h
2h
3h
4h
36.8+0.2
110+12
7.30±0.05
35±5
32±6
37.1 --0.2
126+4
7.35±0.03
34±2
32±6
37.9+0.0
121 ±7
7.38±0.03
35±4
34±4
37.9+0.5
119+7
7.38±0.05
37+4
36±3
38.0+0.2
113+9
7.38±0.03
36±4
39±4
33.4±0.3
124±13
7.34±0.04
28±1
32±3
33.9±0.5
117±9
7.34±0.06
34±5
31±5
34.7±0.4
116±7
7.31±0.03
39±3
36±6
36.5±0.7
125±8
7.35±0.05
34±5
35±5
37.4±0.4
129±5
7.36±0.04
37±8
37±4
blood pressure; and Hct, hematocrit. Values are mean±SD.
brain temperature by only about 0.5°C during VF3,1719-followed by immediate postischemic cooling to 34°C
for "resuscitation" over the first hour had improved
cerebral outcome to 96 hours in a previous study.3
Prearrest and postarrest life support were the same in
all dogs.
In the 10 dogs to be arrested, anesthesia was induced
with ketamine (10 mg/kg IM) and maintained with
50:50% N20/02 plus halothane by endotracheal IPPV
and with controlled normocapnia and normoxia. After
preparations and catheter insertions and before the first
baseline CBF measurements, N20 was replaced with
nitrogen, and halothane was discontinued. Analgesia
was provided with a continuous intravenous infusion of
fentanyl.9
Continuously monitored in all experiments were the
electrocardiogram, heart rate, mean arterial blood pressure (MABP), central venous pressure (CVP), end-tidal
CO2, and Tc. Tympanic and nasopharyngeal thermistor
probes were not used because they would have interfered with the local CBF CT measurements. Intermittently monitored were Pao2, Paco2, pHa, arterial base
excess, Hct, hemoglobin, and serum electrolytes. Controlled were MABP at 100± 10 mm Hg (mean±SD) by
adjusting the halothane concentration before and by
using norepinephrine or trimethaphan after cardiac
arrest, CVP at 5 to 15 mm Hg, To at 37.5 ±0.5°C, Pao2 at
> 100 mm Hg, Paco2 at 30 to 35 mm Hg, base excess at
±7 mmol/L (with intravenous titrated NaHCO3), and
blood glucose concentration at 100 to 175 mg/dL before
cardiac arrest.
Before cardiac arrest, the fentanyl infusion was discontinued and IPPV was with air for 4 minutes. Then
VF was induced by external transthoracic electric
shock.3.9 In all dogs To was to be maintained at exactly
37.5°C at the start of VF. During cardiac arrest, all
temperatures were observed but not controlled. Immediately before the start of CPB, 0.0125 mg/kg epinephrine was given intra-arterially, and IPPV was restarted
with 100% oxygen. All dogs were kept in VF no flow for
12.5 minutes, after which resuscitation was begun with
CPB high flow > 100 mL/kg per minute (ie, resuscitation
time was 0 minute). The CPB circuit had been primed
earlier with plasma substitute, which resulted in transient moderate hemodilution in both groups. At full
CPB flow of 3 minutes, one or more external defibrillating countershocks of 200 J was used to restore
spontaneous heartbeat. Within a resuscitation time of 5
minutes, partial CPB (25 mL/kg per minute) was substituted for full CPB and continued to 1 hour in both
groups.
To control Tc, CPB and external warming and cooling
were used. In group I, Tc was maintained at 37.50C
throughout the 4 hours after cardiac arrest by using
external means. In group II, ice-water immersion of the
cranium during VF only was followed by full CPB to
restore spontaneous circulation with use of 20°C circuit
temperature to lower Tc to 34°C within 2 to 5 minutes3 '8,19 and then by partial CPB with circuit temperature adjusted to maintain Tc 34°C to resuscitation time
1 hour. All CPB was stopped at 1 hour in both groups.
In group II, external rewarming was performed from 1
hour to 3 hours after cardiac arrest.
Cerebral blood flow was studied sequentially over
time: at 60 and 30 minutes before cardiac arrest; at
resuscitation time '10 minutes" (ie, at 1 to 22 minutes in
group I and 9 to 12 minutes in group II) and 30 minutes;
and 1, 2, 3, and 4 hours after cardiac arrest. Before each
CBF measurement, MABP, Pao2, Paco2, pHa, base
excess, Hct, and Tc were controlled.
The methods for local CBF measurements by XeCT23-25 (two coronal slices) and data analysis were applied to our model as described previously.9-11 We obtained local CBF values for each 1 x 1x 5-mm voxel of
two 5-mm-thick coronal CT slices studied. We empirically defined no flow as 0 to 5 mL/100 cm3 per minute;
trickle flow, 6 to 10 mL/100 cm3 per minute; low flow, 11
to 20 mL/100 cm3 per minute; normal white-matter flow,
21 to 40 mL/100 cm3 per minute; normal gray-matter
flow, 41 to 120 mL/100 cm3 per minute; and hyperemic
flow, > 120 mL/100 cm3 per minute (Table 2). We report
global CBF in percent of the baseline value taken imme-
Oklu et al CBF With Mild Resuscitative Hypothermia
1593
TABLE 2. Global CBF and Local CBF Ranges Before and After Cardiac Arrest in Dogs
Before Cardiac Arrest
(Baseline Values)
After Cardiac Arrest
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
1 h
-30 min
10 min
30 min
CBF
-1 h
Group I (normothermic, n=5) (dogs CT13, CT16, CT20, CT33, and CT34)
Global
48±10
28±2t
47±6
Mean±SD
50±7
106±18*
...
104
60
% of baseline
100
227
Local
2.7±1.4*
0.7±0.5
0.1±0.1
3.1±4.7
No flow
0.5±0.5
7.5±2.9t
1.5±0.8
1.6±1.1
0.1±0.1*
3.8±2.8
Trickle flow
5.8±2.9
7.6±3.2
0.2±0.2* 11.8±4.5 24.9±3.4*
Low flow
Normal for WM 29.6+7.1
30.8±6.4
1.4±0.8* 34.2±12.1 46.0±2.8
Normal for GM 61.5±9.9 59.1±10.2 63.6±20.5 42.4±7.9 18.9±4.8*
1.1±1.7
0.1+0.2 34.6±19.9t 4.8±6.9
0.0±0.0
Hyperemic
Group II (hypothermic, n=5) (dogs CT31, CT32, CT35, CT36, and CT37)
Global
46+6*
26±3t
35±5
Mean±SD
41+10
117±18t
...
% of baseline
100
337
132
74
Local
1.6±1.1
0.1±0.0*
0.7±0.4
7.2±1.4t
No flow
0.9±0.8
4.2+2.6
0.1+0.0*
2.8±1.1 14.2±3.4t
3.4±3.3
Trickle flow
18.4±4.8
12.9±9.1
0.2±0.1* 12.0+4.4 27.6+3.6*
Low flow
2.9±1.7* 35.9±5.1* 33.2+4.7
Normal for WM 37.2±8.2 42.7±3.6
Normal for GM 45.4±18.7 33.2±9.4 48.8±16.2 46.4±8.0 17.9±5.1*
0.0±0.0 47.9±17.8t 2.3±2.7t 0.1±0.1
0.2±0.4
Hyperemic
2h
3h
4h
1 to 4 h
(avg)
26±3t
56
24±4t
51
26+4t
55
26±3
55
5.2±2.1*
7.2+2.8*
13.4±5.4*
5.3±3.0*
5.1±2.3
10.1±3.5
24.9±3.7
43.9±4.2
16.0±5.7
0.0±0.0
19±4*
22+3t
55
63
23±3t
66
9.6±2.6*
23.9±3.9t
43.8+3.7*
17.4±7.Ot
0.0±0.0
9.7±3.1*
26.3±3.4* 24.6±4.Ot
40.6±5.4 45.2+4.9t
12.5±6.2* 15.1±4.9*
0.0±0.0
0.0±0.0
22±4
64
14.3±9.5* 8.9±5.1* 7.6±4.3* 9.5±5.1
19.5±4.4t 16.7+4.7* 14.2±2.9t 16.1±3.7
29.0±1.6* 30.2±2.8* 29.7±5.8* 29.1±3.7
28.5+8.1* 32.5±7.5* 36.4±5.6 32.6±6.0
8.8±4.Ot 11.7±3.2t 12.1 ±5.3t 12.6±4.3
0.0+0.0
0.0±0.0
0.0+0.0
0.0±0.0
CBF indicates cerebral blood flow; WM, white matter; and GM, gray matter. Values are mean±SD for global CBF (mL1100 cm3/min).
For local CBF, values represent the percentage of voxels (mean±SD) at different flow ranges as defined in Fig 1 and the text.
*P<.05, tP<.01, and *P<.001 compared within groups with baseline values.
diately before cardiac arrest. Global CBF was calculated
as the average of all local CBF voxels in the posterior
coronal slice, which included neocortex, hippocampus,
and midbrain.9 We report local CBF of the defined flow
ranges (above) in percentage of voxels in the same
posterior coronal slice. We also studied regional CBF
values for eight selected bilateral regions of interest
(measuring about 5 x5 x5 mm) in both slices, depicting
them in customary graphic form (Figure 1 of Reference
9).
Data were analyzed for differences between groups at
specific time points as well as changes over time, using
repeated-measures analysis of variance, Scheff6's procedure, and the paired t test.
Global CMRO2 was determined in three dogs in
group I and all five dogs in group II (Table 3). To obtain
this measurement, a PE-50 nonocclusive catheter was
inserted through a sterile 2-cm craniotomy into the
sagittal sinus, without occluding it. The catheter tip was
placed 1 cm rostral to the confluence of the sinuses. For
each CMRO2 determination, the catheter dead space
was cleared, and about 0.5 mL arterial and sagittal sinus
blood was drawn at a constant, slow rate into heparinized syringes, cooled, and analyzed within 2 hours for 02
content with a Co-oximeter (Instrumentation Laborato-
ries, Lexington, Mass). Samples for arterial and sagittal
sinus 02 contents (Cao2 and Csso2, respectively) were
taken just before each xenon inhalation. Global CMR02
was calculated as the product of global CBF of the
posterior coronal slice times the 02 gradient (ie, global
CBFx [Cao2-Csso2]). The cerebral 02 utilization coefficient was calculated as global CMRO2 divided by the
cerebral arterial 02 delivery (ie, global CBFxCao2).
Thus, the cerebral 02 utilization coefficient (also called
02 extraction ratio) is [Cao2-Csso2]/Cao2. Although
global CBF was calculated from only one CT slice and
Cao2-Csso2 from the entire cerebrum, relative changes
in CMRO2 were considered valid. The CMRO2 data
were not analyzed statistically because of the small
sample size in group I. Results are reported as
mean+SD.
Results
All 11 dogs followed protocol. The dog in the sham
experiment showed a temperature-dependent decrease
of global (Fig 1) and regional CBF values. Accordingly,
the percentage of voxels of no flow, trickle flow, and low
flow increased. Other physiological variables were constant throughout the experiment. Global CBF decreased from 38 mL/100 cm3 per minute at Tc 37.5°C to
1594
Stroke Vol 24, No 10 October 1993
TABLE 3. Metabolic Variables Before and After Cardiac Arrest in Dogs: Comparison of Control
Group I With Hypothermic Group 11*
After Cardiac Arrest
Before Cardiac Arrest
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
4h
1 h
2h
-1 h
-30 min
10 min
30 min
Variable
Cao2, mL/dL
15.9±+ 0.4
14.3-+-2.2
15.4-+ 1.6
20.3±+-2.3
20.2-+0.9
13.3-+1.8
18.0+ 1.6
Group
17.4±1.9
14.5±1.0
15.0±2.1
16.3±2.5
21.7+0.5
14.4±0.9
21.7±1.0
Group II
CssO2, mL/dL
5.8±1.6
6.5+3.0
12.2±2.7
12.9±1.9
11.1±2.1
7.8±2.7
13.4±3.4
Group
6.9±1.3
13.8±0.6
13.3±1.8
7.4±1.4
6.5±0.8
13.4±2.8
13.4±2.4
Group II
CaO2-CssO2, mL/dL
12.2±2.3
0.4±0.1
4.8±2.4
6.5+1.6
8.9±3.1
8.0±2.7
6.8±3.0
Group
10.5±2.5
0.6±0.6
7.6±1.8
9.8+2.1
8.3+1.8
8.2±2.7
1.2+1.0
Group II
Cerebral 02 utilization coefficient [(CaO2-CssO2)/CaO21
0.67±0.10
0.34±0.14
0.40±0.13
0.03±0.01
0.30±0.14
0.47+0.13
0.58±0.18
Group
0.38±0.12
0.04±0.04
0.09±0.08
0.50±0.09
0.59±0.05
0.60±0.10
0.38±0.09
Group II
Cerebral arterial 02 transport, mL/100 cm3/min
8.7±1.2
9.8+2.0
12.3±4.5
8.1+1.3
3.9±0.8
4.0+0.3
4.5±0.5
Group
7.6±1.1
16.4±3.2
6.7±1.2
8.5±2.4
2.9±0.6
3.8+0.8
3.8±0.5
Group II
CaO2 indicates arterial oxygen content; CssO2, sagittal sinus oxygen content; and CMRO2, cerebral metabolic rate of oxygen. Values
are mean±SD, but statistical analysis not appropriate (numbers per group too small).
*Group 1, control group (n=3; dogs CT20, CT33, and CT34); Group II, hypothermic group (n=5; dogs CT31, CT32, CT35, CT36, and
CT37).
24 mL/100 cm3 per minute at 34°C (63% baseline value)
and 18 mL/100 cm3 per minute at 30°C (47% baseline
value); after rewarming to normothermia it increased to
59 mL/100 cm3 per minute. At baseline, no-flow and
trickle-flow values were essentially zero. At 30°C, 16%
of the voxels had no flow and 21% had trickle flow.
In the 10 dogs with cardiac arrest in groups I and II,
anesthesia time before the first baseline CBF measurement was 4 to 5 hours. Before cardiac arrest, the Tc,
MABP, Pao2, Paco2, pHa, base excess, blood glucose,
and Hct did not differ between groups (Table 1). During
and after cardiac arrest, Tc followed protocol in both
groups. In group I, Tc was 37.0±0.2°C at 10 minutes
after reperfusion and remained between 37°C and 38°C
to 4 hours. In group II during VF, Tc remained normothermic during head immersion in ice water (which in
previous experiments lowered brain temperature during
VF by about 0.5°C); with reperfusion by CPB, according
to protocol, Tc decreased to 33.9±0.3°C at resuscitation
time 10 minutes, remained at around 34°C to resuscitation time 1 hour, and with external rewarming reached
36°C to 37°C by resuscitation time 3 hours. Resuscitation by CPB, IPPV, and intensive care followed protocol
in both groups. Spontaneous circulation was restored
within 5 minutes after the start of CPB, with both
groups requiring one to three countershocks. The brief
hypertensive bout immediately upon restoration of
heart beat and the requirements for NaHCO3 and
norepinephrine were the same in both groups. Postarrest requirements for trimethaphan were minimal and
similar in both groups. Norepinephrine was needed only
for up to 15 minutes.
In each dog, the two baseline global CBF measurements were numerically lower in group II than in group
I (not significant [NS]; Table 2 and Fig 2). The first
postarrest local CBF determinations during the hyperemic phase did not show any no-flow or trickle-flow
voxels in either group (Table 2). At resuscitation time
30 minutes, global CBF was similar to baseline values on
its way down from hyperemia to hypoperfusion; at 30
minutes, no-flow and trickle-flow voxels were <5% in
both groups, without statistically significant group differences (Table 2). At resuscitation time 1 hour, with
hypothermia still present in group II, global CBF was
60% of baseline values in group I and 74% in group II
GLOBAL CEREBRAL BLOOD FLOW
NORMOTHERMIA vs MILD HYPOTHERMIA
140
NORMOTHERMIA (n=5)
c 120
._
c:
MILD HYPOTHERMIA (n=5)
8a00
I
O
40
E 200L
oL
-60
1-
-30
10
30
60
120
180
240
TIME POSTARREST (min)
FIG 2. Graph showingglobal cerebral bloodflow (gCBF) for
the posterior coronal slice after ventricularfibrillation cardiac
arrest of 12.5 minutes (mean+±SD) of group I (solid line)
compared with group II (broken line), which was cooled to
34°C from resuscitation time 5 to 60 minutes. No group
difference in gCBF.
Oku et al CBF With Mild Resuscitative Hypothermia
1595
4
mIV100g/min
3.5
100
80
__.
E°
12.5 min:-
a)
CL
CARDIAC'
C:
40
20
M0
2.5
L
ARREST
/'
~~~~~~~~~~~......................
2
C,
_........
(31.5
0
0
NORMOTHERMIA (n=3)
80
MILD HYPOTHERMIA (n=5)
E
80
0.
1
40
-60
20
-30
l~
10
30
60
120
180
240
TIME POSTARREST (min)
FIG 4. Global cerebral metabolic rate of oxygen (CMRO2)
before and after ventricularfibrillation cardiac arrest of 12.5
minutes with no blood flow. Group I (solid line, n=3)
compared with group II (broken line, n=5). Values are
mean+±SD. No group difference in global CMRO2 values at
resuscitation times 1 to 4 hours.
80
6
40
20
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
0
80
40
20
80
60
40
20
nU
-0'
-30'
1 min
30min
1h
2t
3h
4h
FIG 3. Graphs showing regional cerebral blood flow in regions
of interest sized 25x5x5 mm as indicated. Group I, solid line;
group II, broken line; no significant group differences.
(NS); no-flow, trickle-flow, and low-flow voxels did not
show any group differences (Table 2). The hypoperfusion phase continued to the end of observation at 4
hours. The mean values from resuscitation time 1 to 4
hours for global CBF were 55% of baseline values in
group I and 64% in group II (NS), again without
statistically significant group differences in the no-flow
and trickle-flow voxels, which reached 5% to 7% and
10% to 13%, respectively. In each dog over time in both
groups, global, regional, and local CBF values during
the postarrest hyperemia and hypoperfusion phases
were significantly different from baseline values as
indicated in Table 2.
Regional CBF values (Fig 3) in both groups, for the
five regions of interest shown, indicate the same hyperemia-hypoperfusion sequence as that seen in global
CBF and as described previously,9-11 without statistical
differences between groups. However, a numerically
higher regional CBF occurred in the neocortex at
resuscitation time 30 minutes in group II (NS).
Cerebral metabolic variables (Table 3 and Fig 4)
differed considerably between animals at the same time
points. Throughout the study, Pao2 remained at .100
mm Hg. The arterial 02 content decreased transiently
immediately after cardiac arrest as a result of hemodilution and then returned to near-baseline values by 4
hours. The Cao2-Csso2 gradient ranged between 5 and
12 mL/dL before arrest, was near zero during hyperemia
at "10 minutes" after reperfusion, varied widely at 30
minutes, and increased to slightly above baseline values
at 2 and 4 hours after cardiac arrest, with no difference
between groups. The cerebral 02 utilization coefficient
(Table 3) was 0.18 to 0.57 at baseline, decreased during
hyperemia, and increased to mean values of 0.67 in
group I and 0.60 in group II by 4 hours, implying an
oxygen delivery/uptake mismatch. Cerebral arterial 02
transport (Table 3) varied before cardiac arrest, increased less than expected during hyperemia (because
of hemodilution), and decreased to about 50% of
baseline values from 1 to 4 hours after cardiac arrest
because CBF was 50% to 60% of baseline at that time.
Global (slice) CMRO2 (Fig 4) ranged between 2.7 and
4.5 mL/100 cm3 per minute at baseline in both groups
and was extremely low during the hyperemic phase
(values at 10 and 30 minutes ranged between 0.1 and 1.8
mL, with mean values c1.5 mL in both groups). At 1
hour after arrest, CMRO2 was 1.8±0.3 mL in group I
(n=3) and 1.9+0.4 mL in group II (n=5). Values at 2
and 4 hours ranged between 1.2 and 4.2 mL/100 cm3 per
minute in group I and between 1.2 and 2.6 mL in group
II (Fig 4).
Discussion
Our model of VF cardiac arrest, the method of
determining CBF with Xe-CT, and the CBF patterns
achieved with normothermic standard therapy are well
established and have been discussed.9 Both this CBF
study using Xe-CT and our previous outcome studies
examined mild (34°C) resuscitative hypothermia after
cardiac arrest,317-19 which had not been investigated
previously. This is in contrast to moderate (30°C) protective hypothermia for elective circulatory arrest during open-heart or brain surgery.15"16 Reperfusion with
brief CPB, although not clinically realistic, was essential
not only to control the reperfusion blood pressure, flow,
temperature, and composition,27 but also to obtain
repeated CT images and CBF values without head
motion, which is unavoidable with external CPR.89 In
this CBF study, partial CPB was 1 hour in both groups,
and CBF from 1 to 4 hours was the same in both groups.
1596
Stroke Vol 24, No 10 October 1993
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
The partial CPB flow of 25 mL/kg per minute seems to
have no significant cardiovascular effects. Our previous
studies with the same model to 72 or 96 hours after
arrest all showed improved clinical and histological
cerebral outcome. In two studies,3,19 after immediate
postarrest mild hypothermia to 1 hour, during spontaneous circulation, low-flow CPB for 1 hour was used
merely for cooling but not in the normothermic control
groups. In one study17 no CPB was used. In a fourth
study'8 the same CPB time of 3 hours was used in all
groups.
The spatial resolution of the CT method used allows
reliable values for regions of .5 mm in diameter.9'28
The 1 x 1x 5-mm voxel values, particularly at low local
CBF states, may be in part the result from CT noise and
computational errors.928
In the present study we found that in the sham dog
(Fig 1), global CBF decreased to 63% of the normothermic value with mild hypothermia and to 47% with
moderate hypothermia. Hypothermia seemed to increase the heterogeneity of local CBF. Without arrest,
not only moderate hypothermia in adult dogs13 but also
mild hypothermia in young pigs29 reduces CMRO2 and
CBF in parallel. The depressant effect of mild and
moderate hypothermia on local CBF appears reversible
with return to normothermia (Fig 1).
When induced before circulatory arrest, moderate, 1516 deep (15°C),30 or profound (50C)31 hypothermia
does not add to but reduces ischemic brain damage. In
contrast, in our other recent studies, deep resuscitative
hypothermia induced with reperfusion from cardiac
arrest worsened brain damage,18 whereas mild resuscitative hypothermia mitigated brain damage377-19 more
so than moderate resuscitative hypothermia.20 Impaired
microcirculatory reperfusion with low temperatures is
one of the several explanations that remain to be
studied.
In the present study, after cardiac arrest for 4 hours,
global CBF, regional CBF, and the proportion of no-flow
and low-flow voxels of local CBF were the same in both
groups. Mild hypothermia may not affect blood and
plasma viscosity enough to decrease microcirculatory flow
more than postischemic normothermia.7'9 Moderate hypothermia, when induced in dogs after temporary global
ischemia by intracranial hypertension, also had little effect
on postischemic hyperemia-hypoperfusion.32
In previous studies with normothermia, hypertensive
hemodilution improved global and multifocal CBF after
cardiac arrest,10 and an initial hypertensive bout improved outcome.22 For the present study, to test mild
hypothermia during reperfusion with hypertension and
hemodilution'0 we conducted two pilot experiments
(dogs CT21 and CT22), using the same model with VF
of 12.5 minutes. The same reperfusion and cooling
protocol was used, but with additional (slightly hypervolemic) hemodilution to an Hct of about 20% by
Ringer's dextran solution and norepinephrine-induced
transient hypertension to an MABP of .140 mm Hg
immediately after cardiac arrest, as described previously.'1022 CBF baseline values were as in groups I and IL.
Postarrest MABP values at resuscitation time 10 minutes were 145 mm Hg and 155 mm Hg; Hct was 23%
and Tc 34°C in both dogs. At resuscitation time 10
minutes (during the hyperemic phase), global CBF was
88 mL/100 cm3 per minute and 70 mL/100 cm3 per
minute, without no-flow or trickle-flow voxels. From
resuscitation time 1 hour (still with Tc 34°C) to 3 hours
(normothermia), global CBF remained 81% and 86% of
baseline and regional and local CBF values also remained near baseline values-higher than in group II
without hypertensive hemodilution at the same time.
Thus, flow promotion can normalize CBF after arrest
not only during normothermia10 but also during mild
hypothermia.
Does mild hypothermia for 1 hour early after arrest
improve cerebral 02 delivery in relation to 02 demand?
Without cardiac arrest, hypothermia is known to reduce
CBF in parallel with CMRO2.12,13'15,16 In previous studies9-11 with VF of 12.5 minutes, the 50% reduction in CBF
starting at resuscitation time 1 hour lasted to resuscitation
time 12 hours and was mismatched with global CMRO2,
which reached baseline values at resuscitation time 2 to 3
hours. In this study, during the transient hyperemia at 10
minutes, cerebral arteriovenous 02 differences and calculated CMRO2 values were extremely low and varied widely
between dogs. Although this could be a methodological
artifact because this phase is very dynamic, the low values
could also be the result of vasoparalysis, arteriovenous
shunting, and postischemic metabolic silence. At 30 minutes, when cerebral electric activity returns,3 the in-between values are plausible as hyperemia changes to hypoperfusion. Only at 30 minutes did mild hypothermia result
in a suggestion of better O2 delivery in relation to 02
uptake. By 1 hour after arrest, however, with group II still
mildly hypothermic, the 02 gradients, global CBF values,
and CMRO2 values were near baseline values and were
the same in both groups, and remained so to 4 hours after
arrest (Table 3 and Fig 4). If there is any beneficial effect
of mild postarrest hypothermia during the first hour on
brain reoxygenation, it is early and transient and occurs
before the mismatching of low 02 delivery and "normalized" 02 uptake are established at 2 to 4 hours. More
prolonged mild cooling should be explored, with anesthesia and paralysis for pharmacological poikilothermia.
Although brief mild hypothermia has little effect on
CBF and metabolism (the present study), it does improve cerebral outcome.3'7-19 Although hemodilution
with plasma substitute to an Hct of <25% normalizes
global, regional, and local CBF after cardiac arrest,10 it
does not (without hypertension) seem to improve outcome.22 An Hct of <20% reduces arterial 02 content,
which offsets the improvement of CBF by hemodilution;
thus, 02 delivery is not improved. To achieve improved
02 delivery, hemodilution might be attempted with an
acellular 02 carrying blood substitute solution. Attempts to resolve these apparent inconsistencies should
include long-term measurements of global, regional,
and local CBF, CMRO2, and outcome.
The mechanisms by which brief postarrest mild
hypothermia317-19 more so than moderate postarrest
hypothermia18'20 might improve cerebral outcome are
unclear. This study suggests that improved intracerebral blood flow distribution is not one of these mechanisms. Mild hypothermia probably exerts multiple
beneficial effects at the neuronal level, including suppression of free radical, enzyme, excitotoxicity, and
inflammatory reactions and direct physical protection
of membranes.3'33
We conclude that after prolonged cardiac arrest,
immediate postarrest mild hypothermia (34°C) for 1
Oku et al CBF With Mild Resuscitative Hypothermia
hour exerts neither a harmful nor a beneficial effect on
global, regional, or local CBF patterns. Further large
animal studies are needed to determine the optimal
duration of mild hypothermia that, when combined with
optimized 02 delivery (eg, with hypertension, vasodilation, hemodilution, and other measures) and suppression of shivering, hypermetabolism, and cerebral excitotoxicity and lipid peroxidation cascades, might
optimize the long-term functional and morphological
outcome of the brain.34
Acknowledgments
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
Supported by National Institutes of Health grant NS24446
and the Asmund S. Laerdal Foundation. The Schroedinger
Foundation of Austria and the Fulbright Foundation supported Dr. Sterz.
Stephen Hecht, MD; Robert Tarr, MD; and Ann Radovsky,
DVM, PhD, helped with local CBF imaging. David Gur, PhD;
Sidney Wolfson, MD; and Howard Yonas, MD, gave valuable
advice on local CBF calculations. Janine E. Janosky, PhD,
helped with statistical analysis. Alan Abraham, Henry Alexander, Scott Nagel, Annette Pastula, and Mark Ritchey helped
with the experiments. Teresa Adams, Pat Colorito, Franca
Defelice, Viekie Mavrakis, William Wigginton, and Donna
Wojciechowski were radiotechnologists. Lisa Cohn helped
with editing. Gale Foster helped prepare the manuscript.
References
1. Safar P. Resuscitation from clinical death: pathophysiologic limits
and therapeutic potentials. Crit Care Med. 1988;16:923-941.
2. Negovsky VA, Gurvitch AM, Zolotokrylina ES. Postresuscitation
Disease. Amsterdam, Elsevier, 1983.
3. Leonov Y, Sterz F, Safar P, Radovsky A, Oku K, Tisherman S,
Stezoski SW. Mild cerebral hypothermia during and after cardiac
arrest improves neurologic outcome in dogs. J Cereb Blood Flow
Metab. 1990;10:57-70.
4. Ames A III, Wright RL, Kowada M, Thurston JM, Majno G.
Cerebral ischemia, II: the no-reflow phenomenon. Am J Pathol.
1968;152:437-453.
5. Lind B, Snyder J, Safar P. Total brain ischemia in dogs: cerebral
6.
7.
8.
9.
10.
11.
12.
physiological and metabolic changes after 15 minutes of circulatory
arrest. Resuscitation. 1975;4:97-113.
Snyder JV, Nemoto EM, Carroll RG, Safar P. Global ischemia in
dogs: intracranial pressures, brain blood flow and metabolism.
Stroke. 1975;6:21-27.
Kagstroem E, Smith ML, Siesjo BK. Local cerebral blood flow in
the recovery period following complete cerebral ischemia in the
rat. J Cereb Blood Flow Metab. 1983;3:170-182.
Wolfson SK, Safar P, Reich H, Clark JM, Gur D, Stezoski W, Cook
EE, Krupper MA. Dynamic heterogeneity of cerebral hypoperfusion after prolonged cardiac arrest in dogs measured by the
stable xenon/CT technique: a preliminary study. Resuscitation.
1992;23:1-20.
Sterz F, Leonov Y, Safar P, Johnson D, Oku K, Tisherman SA,
Latchaw R, Obrist W, Stezoski SW, Hecht S, Tarr R, Janosky JE.
Multifocal cerebral blood flow by Xe-CT and global cerebral
metabolism after prolonged cardiac arrest in dogs: reperfusion
with open-chest CPR or cardiopulmonary bypass. Resuscitation.
1992;24:27-47. See also Safar P, Sterz F, Leonov Y, Johnson D,
Oku K, Tisherman SA, Latchow R, Stezoski SW, Janosky JE.
Multifocal cerebral blood flow by stable xenon-computed
tomography and global cerebral metabolism after cardiac arrest in
dogs. J Cereb Blood Flow Metab. 1991;11:S320. Abstract.
Leonov Y, Sterz F, Safar P, Johnson DW, Tisherman SA, Oku K-I.
Hypertension with hemodilution prevents multifocal cerebral
hypoperfusion after cardiac arrest in dogs. Stroke. 1992;23:45-53.
Sterz F, Safar P, Johnson DW, Oku K, Tisherman SA. Effects of
U74006F on multifocal cerebral blood flow and metabolism after
cardiac arrest in dogs. Stroke. 1991;22:889-895.
Hegnauer AH, D'Amato HE. Oxygen consumption and cardiac
output in the hypothermic dog. Am J Physiol. 1954;178:138-142.
1597
13. Rosomoff HL, Holaday DA. Cerebral blood flow and cerebral
oxygen consumption during hypothermia. Am J PhysioL 1954;179:
85-88.
14. Chen RYZ, Chien S. Hemodynamic functions and blood viscosity
in surface hypothermia. Am J Physiol. 1978;234:H136-H143.
15. Dripps RD, ed: The Physiology of Induced Hypothermia. Washington, DC: National Academy of Science; 1956.
16. Rupp SM, Severinghaus JW. Hypothermia. In: Miller RD, ed:
Anesthesia. New York, NY: Churchill Livingstone Inc; 1986;
1995-2022.
17. Sterz F, Safar P, Tisherman S, Radovsky A, Kuboyama K, Oku
K-I. Mild hypothermic cardiopulmonary resuscitation improves
outcome after prolonged cardiac arrest in dogs. Crit Care Med.
1991;19:379-389.
18. Weinrauch V, Safar P, Tisherman S, Kuboyama K, Radovsky A.
Beneficial effect of mild hypothermia and detrimental effect of
deep hypothermia after cardiac arrest in dogs. Stroke. 1992;23:
1454-1462.
19. Kuboyama K, Safar P, Tisherman SA, Radovsky A, Stezoski SW.
Immediate but not delayed mild cerebral hypothermia improves
outcome after cardiac arrest in dogs. Crit Care Med. In press.
20. Leonov Y, Sterz F, Safar P, Radovsky A, Radovsky A. Moderate
hypothermia after cardiac arrest of 17 minutes in dogs: effect on
cerebral and cardiac outcome. Stroke. 1990;21:1600-1606.
21. Safar P, Sterz F, Leonov Y, Radovsky A, Tisherman S, Oku K.
Systematic development of cerebral resuscitation after cardiac
arrest. Three promising treatments: cardiopulmonary bypass,
hypertensive hemodilution, and mild hypothermia. Acta Neurochir.
1993;57:110-121.
22. Sterz F, Leonov Y, Safar P, Radovsky A, Tisherman S, Oku K.
Hypertension with or without hemodilution after cardiac arrest in
dogs. Stroke. 1990;21:1178-1184.
23. Gur D, Good WF, Wolfson SK, Yonas H, Shabason L. In vivo
mapping of local cerebral blood flow by xenon-enhanced computer
tomography. Science. 1982;215:1267-1268.
24. Yonas H, Good WF, Gur D, Wolfson SK Jr, Latchaw RE, Good
BC, Leanza R, Miller SL. Mapping cerebral blood flow by xenonenhanced computed tomography: clinical experience. Radiology.
1984;152:435-442.
25. Wolfson SK Jr, Clark J, Greenberg JH, Gur D, Yonas H, Brenner
RP, Cook EE, Lordeon PA. Xenon-enhanced computed tomography compared with [i4C]iodoantipyrine for normal and low
cerebral blood flow states in baboons. Stroke. 1990;21:751-757.
26. Oku K, Kuboyama K, Safar P, Johnson D, Sterz F, Obrist W,
Leonov Y, Tisherman S. Multifocal cerebral blood flow (CBF) and
global metabolism (CMR) after prolonged cardiac arrest in dogs:
effect of mild hypothermia (34°C). Anesthesiology. 1990;73:A302.
Abstract.
27. Safar P, Abramson NS, Angelos M, Cantadore R, Leonov Y,
Levine R, Pretto E, Reich H, Sterz F, Stezoski SW, Tisherman S.
Emergency cardiopulmonary bypass for resuscitation from prolonged cardiac arrest. Am J Emerg Med. 1990;8:55- 67.
28. Johnson DW, Stringer WA, Marks MP, Yonas H, Good WF, Gur
D. Stable xenon CT cerebral blood flow imaging: rationale for and
role in clinical decision making.Am JNeuroradiol. 1991;12:201-213.
29. Busija DW, Leffler CW. Hypothermia reduces cerebral metabolic
rate and cerebral blood flow in newborn pigs. Am J Physiol 1987;
253:H869-H973.
30. Tisherman SA, Safar P, Radovsky A, Peitzman A, Sterz F,
Kuboyama K. Therapeutic deep hypothermic circulatory arrest in
dogs: a resuscitation modality for hemorrhagic shock with 'irreparable' injury. J Trauma. 1990;30:836-847.
31. Tisherman SA, Safar P, Radovsky A, Peitzman A, Marrone G,
Kuboyama K, Weinrauch V. Profound hypothermia (<10°C)
compared with deep hypothermia (15°C) improved neurologic
outcome in dogs after two hours' circulatory arrest induced to
enable resuscitative surgery. J Trauma. 1991;31:1051-1062.
32. Baldwin WA, Kirsch JR, Hurn PD, Toung WSP, Traystman RJ.
Hypothermic cerebral reperfusion and recovery from ischemia. Am
J PhysioL 1991;261:H774-H781.
33. Ginsberg MD, Sternau LL, Globus MYT, Dietrich WD, Busto R.
Therapeutic modulation of brain temperature: relevance to
ischemic brain injury. Cardiovasc Brain Metab Rev. 1992;4:189-225.
34. Safar P. Cerebral resuscitation after cardiac arrest: research initiatives and future directions. A review. Ann Emerg Med. 1993;22(pt
2):324-349.
Mild hypothermia after cardiac arrest in dogs does not affect postarrest multifocal
cerebral hypoperfusion.
K Oku, F Sterz, P Safar, D Johnson, W Obrist, Y Leonov, K Kuboyama, S A Tisherman and S
W Stezoski
Downloaded from http://stroke.ahajournals.org/ by guest on June 14, 2017
Stroke. 1993;24:1590-1597
doi: 10.1161/01.STR.24.10.1590
Stroke is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231
Copyright © 1993 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/24/10/1590
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/