Neurologic Critical Care

Moderate hypothermia improves imbalances of thromboxane A2

and prostaglandin I2 production after traumatic brain injury in
humans

Mayuki Aibiki, MD, PhD; Souichi Maekawa, MD; Satoshi Yokono, MD, PhD

Objective: To examine the levels of thromboxane B2 (TXB2) and
6-keto prostaglandin F1␣ (6-keto PGF1␣) production in arterial and
internal jugular bulb sera in patients with traumatic brain injury
(TBI). TBI is associated with arachidonate release and may be
associated with an imbalance of vasoconstricting and vasodilat-
ing cyclooxygenase metabolites.
Design: A prospective, randomized study.
Setting: The intensive care unit of a medical university hospi-
tal.
Interventions: Twenty-six ventilated TBI patents (Glasgow
Coma Scale score on admission, <8 points) were divided ran-
domly into two groups: a hypothermic group (n ⴝ 15), in which
the patients were cooled to 32 to 33°C after being giving vecu-
ronium, midazolam, and buprenorphine; and a normothermic
group (n ⴝ 11), in which the patients’ body temperature was
controlled at 36 to 37°C by surface cooling using the same
treatment as the hypothermic group. Body temperature control
including normothermia was started 3 to 4 hrs after injury. The
duration of hypothermia usually lasted for 3 to 4 days, after which
the patients were rewarmed at a rate of approximately 1°C per
day.
Measurements and Main Results: Blood sampling for TXB2 and
6-keto PGF1␣ was started shortly after admission in both groups.
Arterial TXB2 levels on admission in both groups were elevated
remarkably, but not 6-keto PGF1␣, thereby causing an imbalance
of the prostanoids after injury. In the normothermic group, TXB2
decreased transiently, but this prostanoid increased again 3 days
after the injury. In the hypothermic group, such prostanoid dif-
ferences disappeared shortly after therapy, and the condition was
sustained for 10 days. Hypothermia attenuated differences in
TXB2 levels between arterial and internal jugular bulb sera, which
may reflect reduced cerebral prostanoid production. The Glasgow
Outcome Scale score 6 months after the insult in the hypothermic
group was significantly higher than that in the normothermic
group (p ⴝ .04).
Conclusion: The current results from a limited number of
patients suggest that moderate hypothermia may reduce prosta-
noid production after TBI, thereby attenuating an imbalance of
thromboxane A2 and prostaglandin I2. However, it must be clari-
fied whether the changes in the prostanoid after moderate hypo-
thermia are a secondary effect of other mediator changes or
whether they simply represent an epiphenomenon that is mech-
anistically unrelated to damage in TBI. (Crit Care Med 2000; 28:
3902–3906)
KEY WORDS: arterial–jugular bulb difference; moderate hypo-
thermia; normothermia; prostaglandins; traumatic brain injury

T hromboxane A2 (TXA2) and
prostaglandin I2 (PGI2) have
potent vasoactive effects in the
cerebral circulation, allowing
the prostanoids to play an important role
in the maintenance of cerebral blood flow
at a constant level (1, 2). Injury to the
central nervous system such as to the
spinal cord induces relatively greater
TXA2 production than PGI2 in the tissue,
which results in an imbalance of the cy-
From the Intensive Care Unit (Drs. Aibiki and
Yokono), Kagawa Medical University Hospital; and the
Department of Emergency Medicine (Dr. Maekawa),
School of Medicine, Ehime University, Ehine, Japan.
Address reprint requests to: Mayuki Aibiki, MD,
PhD, Intensive Care Unit, Kagawa Medical University
Hospital, 1750 –1, Ikenobe, Miki, Kita, Kagawa, 761-
0793, Japan.
Copyright © 2000 by Lippincott Williams & Wilkins
clooxygenase metabolites, TXB2, the sta-
ble metabolite of TXA2, and 6-keto pros-
taglandin F1␣ (6-keto PGF1␣), the stable
metabolite of PGI2 (3). Furthermore, in a
human study, some patients with closed
head trauma showed a predominance of
TXA2 production over PGI2 in cerebrospi-
nal fluid (4). These changes may impair
the microcirculation and may produce
neurotoxic mediators, thereby causing
subsequent neuronal damage (3).
Recently, we reported in patients with
traumatic brain injury (TBI) that moder-
ate hypothermia decreases arterial inter-
leukin-6 (IL-6) levels along with the at-
tenuation of the cytokine in internal
jugular bulb plasma (5), which may re-
flect reduced cerebral cytokine produc-
tion (6). Our report (5) supports a recent
paper demonstrating attenuation of cyto-
kine levels in cerebrospinal fluid by mod-
erate hypothermia in TBI patients (7). In
the rat brain, hypothermia has been
shown to attenuate an increase in IL-1␤
RNA expression after brain injury, which
may ameliorate the neurologic outcome
(8). Moderate hypothermia of 30 to 31°C
has been indicated in reducing postisch-
emic edema development and leukotriene
production in rats (9). We demonstrated
recently that elevated TXB2 production in
arterial sera after brain injury is sup-
pressed by moderate hypothermia (10).
However, in brain-injured patients,
changes in differences between TXB2 and
6-keto PGF1␣ have not been determined.
In the current study, using radioimmu-
noassays, we measured concentrations of
TXB2 and 6-keto PGF1␣ in arterial and
internal jugular bulb sera in patients with

3902 Crit Care Med 2000 Vol. 28, No. 12

TBI who underwent moderate hypother-
mia of 32 to 33°C, and in brain-injured
patients who were maintained in a nor-
mothermic state.
SUBJECTS AND METHODS
This study was approved by the Ethics
Committee, Kagawa Medical University, and
informed consent from families or relatives
was obtained before including the patients in
this study. Fifteen head-injured patients un-
derwent moderate hypothermia of 32 to 33°C,
and 11 patients received normothermic ther-
apy of 36 to 37°C. This unequal number of
patients in each group was the result of exclu-
sion of normothermic patients having abdom-
inal or chest trauma. Patients were assigned
randomly to each group. Body temperatures
were monitored at a tympanic membrane us-
ing a cotton-coated probe and/or at an internal
jugular bulb using a catheter inserted into an
internal jugular vein in the rostral direction.
In all patients, after the insertion of the cath-
eter, we confirmed by craniofacial X-ray exam-
ination that the catheter tip was located near
the internal jugular bulb. Inclusion criteria
included a Glasgow Coma Scale (GCS) score ⱕ
8 points on admission to the emergency room
and evidence of injury on computed tomo-
graphic scanning of the brain. The character-
istics of the patients are shown in Table 1. The
GCS score of the patients ranged from 3 to 8
points on admission. Patients who had abdom-
inal or chest trauma or those who sustained
severe pulmonary infection were excluded
from the study because such derangements
could have modulated prostaglandin produc-
tion (11). Patients who were admitted to the

Table 1. Background of each group

GCS GOS Score CT Pupil Abn.

Age, yrsa Genderb Diagnosis Score, ptc STd (pt)e Classf on Adm.g

Normothermic Group
76 M CC 6 (⫹) V (2) 5 ⫺
EH
19 M EH 5 ND SD (3) 4 ⫹
SH
70 M EH 5 (⫹) D (1) 5 ⫹
38 F TSAH 5 ND GR (5) 2 ⫺
65 M CC 4 (⫹) D (1) 5 ⫹
ICH
76 M ICH 6 (⫹) V (2) 5 ⫺
27 M CC 7 ND GR (5) 2 ⫺
4 F CC 6 ND SD (3) 4 ⫺
16 F DAI 6 ND MD (4) 2 ⫺
18 M SH 5 (⫹) D (1) 5 ⫹
16 M CC 8 ND GR (5) 2 ⫺
Hypothermic Group
12 M CC 5 (⫹) MD (4) 5 ⫹
EH
67 M EH 5 (⫹) SD (3) 5 ⫹
SH
17 M EH 7 (⫹) GR (5) 5 ⫹
16 M SH 8 (⫹) GR (5) 5 ⫺
9 M CC 5 (⫹) GR (5) 5 ⫺
ICH
52 M ICH 6 (⫹) MD (4) 5 ⫹
57 F ICH 5 (⫹) SD (3) 5 ⫺
48 M CC 7 (⫹) GR (5) 5 ⫺
ICH
18 M DAI 7 ND GR (5) 2 ⫺
TSAH
22 M SH 6 (⫹) MD (4) 5 ⫹
CC
76 M ICH 5 (⫹) MD (4) 5 ⫹
72 M ICH 3 ND D (1) 3 ⫹
27 M CC 8 ND GR (5) 2 ⫺
11 F CC 4 ND GR (5) 2 ⫹
17 F ICH 5 ND GR (5) 2 ⫹

M, male; F, female; GCS, Glasgow Coma Scale; ST, surgical treatment; (⫹), treated; ND, not done; GOS, Glasgow Outcome Scale score 6 months after
injury; GR, good recovery (5); MD, moderate disability (4); SD, severe disability (3); V, vegetative (2); D, brain death (1); CT Class, computed tomographic
classification; Pupil Abn. on adm, pupillary abnormalities on admission. CC, cerebral contusion; EH, epidural hematoma; SH, subdural hematoma; TSAH,
traumatic subarachnoid hemorrhage; ICH, intracerebral hematoma; DAI, diffuse axonal injury; Pupil Abn. on Adm., pupil abnormalities on admission.
a
Normothermic group: mean age, 38 ⫾ 8 years. Hypothermic group, mean age, 34 ⫾ 6 years, NS.
b
F/M: Normothermia, 3/8; hypothermia, 3/12; NS.
c
GCS mean score on admission: normothermic group, 5.7 ⫾ 0.3 points; hypothermic group, 5.7 ⫾ 0.4 points; NS.
d
Normothermia, 5 of 11; hypothermia, 10 of 15; NS.
e
Mean GOS score: normothermia, 2.9 ⫾ 0.5 points; hypothermia, 4.2 ⫾ 0.3 points (p ⬍ 0.05, Mann–Whitney U test.
f
CT classification was done according to the following criteria: 1, no visible evidence of injury; 2, cisterns present with a midline shift less than 5 mm
and no lesion ⬍25 mL; 3, cisterns compressed or absent; 4, midline shift of more than 5 mm; 5, a lesion requiring surgical evacuation.
g
Pupillary abnormalities were defined as abnormalities in size or light response in one or both pupils. ⫹, present; ⫺, absent. Normothermic group, 4
of 11; hypothermic group, 9 of 15; NS.

Crit Care Med 2000 Vol. 28, No. 12 3903

hospital more than 8 hrs after the injury were
also excluded. Either moderate hypothermia
or normothermia, depending on group assign-
ment, was induced within 3 to 4 hrs after the
injury and was maintained by surface cooling
after administration of midazolam at a dosage
of 3 to 7 ␮g/kg/min, vecuronium at a dosage of
2 to 4 mg/hr, and buprenorphine at a dosage of
0.4 to 0.6 mg/day. These agents were admin-
istered until the end of the study. All patients
were ventilated artificially to adjust PaCO2 to a
level of 30 to 33 mm Hg, which was not
corrected for body temperature (␣-stat man-
agement) (12). Blood sampling via catheters
placed in a radial artery and an internal jug-
ular bulb was started shortly after admission
in both groups. The internal jugular bulb
catheter was inserted via the right internal
jugular vein toward the rostral side. The side
of the catheter inserted was not always con-
sistent with the side of the injury. During the
study, the serum concentrations of TXB2 and
6-keto PGF1␣ were measured daily using ra-
dioimmunoassay (11). The normal levels of
TXB2 and 6-keto PGF1␣ are documented to
range from 14 to 50 pg/mL and 12 to 33 pg/mL
respectively, by the laboratory company that
measured the samples presented here. After
hypothermic therapy, patients were gradually
rewarmed at a rate of approximately 1°C per
day. The rewarming process was started if
there were no signs of brain swelling on com-
puted tomographic findings and no intracra-
nial pressure (ICP) elevation. The duration of
hypothermia was usually 3 to 4 days. In this
study, there were no patients whose rewarm-
ing process was postponed because of the pre-
vious conditions.
Treatment Protocol. Subdural, epidural, or
intracerebral hematomas compressing the
brain were evacuated as soon as possible by
neurosurgeons. Mean arterial pressure was
maintained between 90 and 100 mm Hg to
keep the cerebral perfusion pressure ⬎ 70 mm
Hg. To ensure stable hemodynamics during
hypothermia, volume replacement therapy in
which a positive daily water balance of 1,000 to
1,500 mL was maintained, and dobutamine at
a dosage of 2 to 4 ␮g/kg/min was administered
so that the cardiac index was kept within 3.5 to
4.5 L/min/m2 (using Swan-Gantz CCOmboTM;
Baxter Healthcare, Valencia, CA). This level of
cardiac output was not different from that of
the normothermic group throughout the
study period. In all patients, 600 to 800 mL of
10% glycerol was administered daily for 7
days. This dose of the agent adheres to our
protocol for TBI patients. ICP monitoring
(Codman ICP sensor™; Johnson and Johnson,
Raynham, MA) was performed in five patients
in the normothermic group and in seven pa-
tients in the hypothermic group. According to
the protocol, a daily dexamethasone dose of 6
to 8 mg was administered to all patients for 4
days. This agent itself could potentially reduce
prostaglandin production after the injury (11),
3904
but doses of the agent did not differ between group was significantly higher than that
the two groups. Nonsteroidal antipyretics in the normothermic group.
such as indomethacin, which could reduce As demonstrated in Figure 1, arterial
prostanoid release (11), were not administered TXB2 levels on admission in both groups
to either group. Thiamylal was not adminis-
increased (normothermia, 173 ⫾ 88 pg/
tered, basically because of difficulties experi-
enced in controlling potassium levels during
mL; hypothermia, 133 ⫾ 42 pg/mL; NS),
but not 6-keto PGF1␣, thereby causing an
hypothermic therapy in a preliminary study
(13). However, in Patient 2 in the hypothermic imbalance of the prostanoids after brain
injury. In the normothermic group, TXB2
group and Patient 8 in the normothermic
group, the agent was administered for possible decreased transiently, but increased
effects relating to reduction of ICP. The elec- again 3 days after the injury. However,
trolyte disturbance after thiamylal administra- hypothermia eliminated such prostanoid
tion could possibly cause arrhythmias during
differences after the therapy, and the im-
hypothermia therapy of 32 to 33°C (13), but in provement was sustained for 10 days.
the current study there were no patients who
These changes permitted an improve-
showed such complications. During the study,
no patients in either group received diphenyl-
ment in the imbalance of TXB2 and
hydantoin. 6-keto PGF1␣ in the hypothermic group,
Assessment of Neurologic Outcome. Inde- but not in the normothermic group.
pendent neurosurgeons who were not aware of As demonstrated in Figure 2, normo-
the study presented here determined the neu- thermia did not attenuate the arterial–
rologic outcome 6 months after injury. The jugular bulb differences in TXB2 concen-
neurologic outcome was graded according to trations on the second day of injury.
the Glasgow Outcome Scale (14): 1, death; 2,
vegetative state; 3, severe disability (unable to
live independently but able to follow com-
mands); 4, moderate disability (ability to live
independently but unable to return to work or
school); and 5, good recovery (able to return
to work or school).
Complications in Each Group. The diagno-
sis for pneumonia was made from purulent
sputum suctioned from an intratracheal tube,
positive X-ray findings including lung consol-
idations, and/or positive cultures. Pneumonia
developed in both groups (normothermia, 3 of
11; hypothermia, 4 of 15; with no significant
differences between the groups). Coagulation
abnormalities including thrombocytopenia
occurred in the hypothermic group (10 pa-
tients), but no serious bleeding tendency de-
veloped. Such problems usually improved dur-
ing the rewarming period.
Statistical Analysis. Data are expressed as
mean ⫾ SEM. The baseline characteristics and
outcomes in the two groups were compared
with the use of chi-square tests, Mann–
Whitney U tests, or Student’s t-tests as appro-
priate. Other statistical analyses were per-
formed using analysis of variance followed by
Scheffé’s F-tests for comparisons within a
group and Mann–Whitney U tests for compar-
isons between groups. A p value ⬍ 0.05 was
considered significant.
Figure 1. Arterial thromboxane B2 (TXB2) levels
RESULTS at admission in both groups increased (normo-
thermic group, 173 ⫾ 88 pg/mL; hypothermic
As shown in Table 1, there were no group, 133 ⫾ 42 pg/mL; NS), but not 6-keto
significant differences between the nor- prostaglandin F1␣ (6-keto PGF1␣), thereby caus-
mothermic and hypothermic groups in ing an imbalance of the prostanoids after brain
injury. In the normothermic group, TXB2 de-
terms of age, gender, GCS score on ad-
creased transiently, but increased again 3 days
mission, the classification of computed after injury. However, hypothermia eliminated
tomographic findings, or the occurrence such prostanoid differences after therapy and
of pupillary abnormalities on admission, lasted for 10 days. *Differences between TXB2
but the Glasgow Outcome Scale score 6 and 6-keto PGF1␣ concentrations (p ⬍ 0.05).
months after injury in the hypothermic NORMO, normothermic; HYPO, hypothermic.
Crit Care Med 2000 Vol. 28, No. 12
 Figure 3. The arterial–internal jugular bulb dif-
ferences in thromboxane B2 (TXB2; the percent
values normalizing by each arterial TXB2 concen-
tration on the first day) were suppressed contin-
uously by hypothermia but not by normother-
mia. *Differences if compared with the value on
the first day (p ⬍ 0.05). IJ-A Dif, differences
between internal jugular bulb and arterial TXB2
concentrations; Art Baseline, the arterial TXB2
concentration on the first day.
Figure 2. Normothermia did not attenuate the arterial–jugular bulb differences in thromboxane B2
(TXB2) concentrations on the second day after injury. There was a transient reduction in the
differences on the third day, but this transient change disappeared after day 4. Hypothermia reduced glial cells (20). IL-1␤, a central inflam-
such differences in TXB2 levels from the second day after injury to the end of the study. Hypothermia matory cytokine, enhanced adenosine
showed a transient but significant increase in 6-keto prostaglandin F1␣ (6-keto PGF1␣) levels on day triphosphate-evoked release of arachi-
5 after injury. *Differences between arterial and internal jugular bulb prostanoid concentrations (p ⬍ donic acid from murine astrocytes (21).
0.05). INT. JUG. B, internal jugular bulb. Moderate hypothermia decreased the
messenger RNA expression of IL-1␤ in
murine-traumatized brain tissue (8). Re-
There was a transient reduction in the nant of subsequent neuronal damage in cently, we demonstrated that in TBI pa-
differences on the third day, but this TBI (15, 16). In the current study we tients, moderate hypothermia decreased
transient change disappeared after the demonstrated that moderate hypother- IL-6 production both in arterial and
fourth day of injury. Hypothermia re- mia suppressed increases between TXB2 jugular bulb plasma, suggesting anti-
duced such differences in TXB2 levels af- and 6-keto PGF1␣ both in arterial and inflammatory effects of the therapy (5).
ter the second day of the injury to the end internal jugular bulb sera. The prosta- Thus, there is a possibility that suppres-
of the study. Patients in the hypothermic noid levels in internal jugular bulb sera sion of the prostanoid differences by hy-
group showed a transient but significant probably reflect prostanoid production in pothermia is mediated through cytokine
increase in 6-keto PGF1␣ levels on day 5 the brain (6). Thus, the improvement of modulation.
after injury. the prostanoid jugular bulb imbalance TXA2 Production and Blood–Brain
As shown in Figure 3, the arterial– Barrier. Cytokines such as IL-6 disrupt
presented here may be, in part, attribut-
internal jugular bulb differences in TXB2 the blood– brain barrier in vitro, probably
able to a decrease of TXA2 production in
(the percent values normalizing by the mediated through cyclooxygenase activa-
the brain. Therefore, further studies are
arterial concentrations on the first day) tion within the endothelial cells (22).
required to define whether such attenu-
were suppressed continuously by hypo- TXA2 receptors have been detected in hu-
ations by hypothermia improve the cere-
thermia, but not by normothermia. man endothelial cells, and stimulation of
bral circulation after injury, and may these receptors may be involved in the
contribute to neuroprotective effects of blood– brain barrier opening (23). In the
DISCUSSION the therapy (17). current study, hypothermia suppressed
Improvement of Prostanoid Imbal- Although the mechanisms of prosta- augmented TXA2 production in internal
ance by Moderate Hypothermia. In hu- noid production in the brain have not yet jugular bulb sera draining from the
man cerebrospinal fluid, TXA2 is pro- been fully understood, a source of brain- brain, which suggests that moderate hy-
duced after TBI, but PGI2 does not always synthesized TXA2 appears to be glial cells pothermia decreases TXA2 release from
increase more than TXA2 even after the such as astrocytes and microglia (18). the cerebral endothelial cells or attenu-
injury, resulting in differences between Cerebral injury may cause TXA2 produc- ates enhanced blood– brain barrier per-
the two prostanoids (4). Such differences tion from neuronal cells as well as glial meability after injury (24). We reported
may play a role in hypoperfusion in the cells, possibly resulting from the direct previously that an augmented IL-6 level
injured area or in the penumbra, which is activation of membrane phospholipids in in internal jugular bulb plasma, which is
possibly mediated through thrombogen- each cell type (4, 19). Alternatively, inju- presumably involved in blood– brain bar-
esis and/or vasoconstriction by excessive ry-induced increases in cerebral cyto- rier disruption, is decreased by moderate
TX production (3). Such circulatory de- kines, such as IL-1 and/or IL-6, may ac- hypothermia (5). Thus, it appears that
rangements would be a crucial determi- tivate the prostanoid release from the moderate hypothermia has beneficial ef-

Crit Care Med 2000 Vol. 28, No. 12 3905

fects on the blood– brain barrier function
after brain injury, mediating through
multifactorial mechanisms.
Mechanisms of TXA2 and PGI2 Pro-
duction in the Brain. In humans, PGI2, a
potent vasodilatory prostanoid, has been
shown to be released mainly from cere-
bral endothelial cells (25). This mediator
is not found abundantly in the gray or
white matter of the brain (25). TXA2 in
the brain can be produced from the neu-
ronal cells as well as glial cells by the
activation of the membrane phospholip-
ids (4, 19). This pathophysiologic back-
ground may contribute to the develop-
ment of the prostanoid imbalance after
TBI. Another major source of PGI2 in the
human central nervous system is thought
to be the choroid plexus (26). This pro-
stanoid release would beneficially inhibit
platelet aggregation and blood clot for-
mation in the intrathecal space (26).
Cerebral ischemia elevates TXA2 as
well as PGI2 in cerebral venous blood in
dogs (27). In humans, also, both prosta-
noids increase in cerebrospinal fluid after
cerebral ischemia (4). Thus, ischemic
brain injury has a lesser prostanoid im-
balance than TBI (27, 28). Therefore,
moderate hypothermia of 32 to 33°C,
which has depressive effects on TXA2 but
not on PGI2 production as presented in
the current study, may exert different ef-
fects on cerebral infarction than on TBI.
This hypothesis should be clarified in fu-
ture studies. Recently, moderate hypo-
thermia has been reported to exert bene-
ficial effects on the postischemic brain
edema in severe ischemic stroke in the
middle cerebral artery territory (29).
Thus, we need to define whether the
mechanisms for reducing brain edema
after stroke are mediated through TXA2
suppression by moderate hypothermia as
presented in the current study.
In conclusion, the results from a lim-
ited number of the patients suggest that
moderate hypothermia may reduce pro-
stanoid production after TBI, thereby at-
tenuating an imbalance of TXA2 and
PGI 2 . However, we need to clarify
whether the changes in the prostanoid
after moderate hypothermia are a second-
ary effect of other mediator changes or
whether they simply represent an epiphe-
nomenon that is mechanistically unre-
lated to damage in TBI.
REFERENCES
1. Boullin DJ, Bunting S, Blaso WP, et al: Re-
sponses of human baboon arteries to prosta-
3906
glandin endoperoxides and biologically gen-
erated and synthetic prostacyclin: Their
relevance to cerebral arterial spasm in man.
Br J Clin Pharmacol 1979; 7:139 –147
2. Leffer CW: Prostanoids: Intrinsic modulation
of cerebral circulation. News Physiol Sci
1997; 12:72–77
3. Hsu CY, Halushka PV, Hogan EL, et al: Al-
teration of thromboxane and prostacyclin
levels in experimental spinal cord injury.
Neurology 1985; 35:1003–1009
4. Westcott JY, Murphy RC, Stenmark K: Eico-
sanoids in human ventricular cerebro-
spinal fluid following severe brain injury.
Prostaglandins 1987; 34:877– 887
5. Aibiki M, Maekawa S, Ogura S, et al: Effect of
moderate hypothermia on systemic and in-
ternal jugular plasma IL-6 levels after
traumatic brain injury in humans.
J Neurotrauma 1999; 16:225–232
6. Kossmann T, Hans VHJ, Imhof H-G, et al:
Intrathecal and serum interleukin-6 and the
acute-phase response in patients with severe
traumatic brain injuries. Shock 1995;
4:311–317
7. Marion DW, Penrod LE, Kelsey SF, et al:
Treatment of traumatic brain injury with
moderate hypothermia. N Engl J Med 1997;
336:540 –546
8. Goss JR, Styren SD, Miller PD, et al: Hypo-
thermia attenuates the normal increase in
interleukin 1␤ RNA and nerve growth factor
following traumatic brain injury in the rat.
J Neurotrauma 1995; 12:159 –167
9. Dempsey RJ, Combs DJ, Maley EM, et al:
Moderate hypothermia reduces postischemic
edema development and leukotriene produc-
tion. Neurosurgery 1987; 21:177–181
10. Maekawa S, Aibiki M, Ogura S, et al: Mild
hypothermia suppresses thromboxane B2
production in brain-injured patients. In: The
Immune Consequences of Trauma, Shock
and Sepsis. Mechanisms and Therapeutic Ap-
proaches. Faist E (Ed). Bologna, Italy, Mon-
duzzi Editore, 1997, pp 135–138
11. Cook JA, Geisel J, Temple GE: Endotoxin
activation of eicosanoid production by mac-
rophages. In: Pathophysiology of Shock, Sep-
sis and Organ Failure. Schlag G, Redl H
(Eds). Berlin, Springer–Verlag, 1993, pp
518 –530
12. Metz C, Holzschuh M, Bein T, et al: Moderate
hypothermia in patients with severe head
injury: Cerebral and extracerebral effects.
J Neurosurg 1996; 85:533–541
13. Asaga T, Aibiki M, Ogura S, et al: Differences
in serum electrolyte changes depending on
sedatives, thiamylal or midazolam, during
mild hypothermia therapy. ICU and CCU
(Jap J Intensive Care Med) 1997; 21:967–973
[abstract in English]
14. Jannet B, Snoek J, Bond MR, et al: Disability
after severe head injury: Observation on the
use of the Glasgow Outcome Scale. J Neurol
Neurosurg Psychiatry 1981; 44:285–293
15. Teasdale GM, Graham DI: Craniocerebral
trauma: Protection and retrieval of the neu-
ronal population after injury. Neurosurgery
1998; 43:723–738
16. Schroder ML, Muizelaar JP, Bullock MR, et
al: Focal ischemia due to traumatic contu-
sion documented by stable xenon CT and
ultrastructural studies. J Neurosurg 1995;
82:966 –971
17. Karibe H, Zarow GJ, Graham SH, et al: Mild
intraischemic hypothermia reduces postisch-
emic hyperperfusion, delayed postischemic
hypoperfusion, blood– brain barrier disrup-
tion, brain edema, and neuronal damage vol-
ume after temporary focal cerebral ischemia
in rats. J Cereb Blood Flow Metab 1994;
14:620 – 627
18. Murphy S, Jeremy J, Pearce B, et al: Eico-
sanoid synthesis and release from primary
cultures of rat central nervous system astro-
cytes and meningeal cells. Neurosci Lett
1985; 61:61– 65
19. Dewitt DS, Kong DL, Lyeth BG, et al: Exper-
imental traumatic brain injury elevates brain
prostaglandin E2 and thromboxane B2 levels
in rats. J Neurotrauma 1988; 5:303–313
20. de Vries HE, Hoogendoorn KH, Dijk J, et al:
Eicosanoid production by rat cerebral endo-
thelial cells: Stimulation by lipopolysaccha-
ride, interleukin-1 and interleukin-6.
J Neuroimmunol 1995; 59:1– 8
21. Stella N, Estellas A, Siciliano J, et al: Inter-
leukin-1 enhances the ATP-evoked release of
arachidonic acid from mouse astrocytes.
J Neurosci 1997; 17:2939 –2946
22. de Vries HE, Blom–Roosemalen MCM, van
Oosten M, et al: The influence of cytokines
on the integrity of the blood– brain barrier in
vitro. J Neuroimmunol 1996; 64:37– 43
23. Kent KC, Collins LJ, Schwerin FT, et al:
Identification of functional PGH2/TXA2 re-
ceptors on human endothelial cells. Circ Res
1993; 72:958 –965
24. Smith S, Hall ED: Mild pre- and posttraumatic
hypothermia attenuates blood– brain barrier
damage following controlled cortical impact
injury in the rat. J Neurotrauma 1996; 13:1–9
25. Abdel-Halim MS, Holst H, Meyrson B, et al:
Prostaglandin profiles in tissue and blood
vessels from human brain. J Neurochem
1980; 34:1331–1333
26. Pickard JD, Walker V, Brandt L, et al: Effect
of intraventricular haemorrhage and re-
bleeding following subarachnoid haemor-
rhage on CSF eicosanoids. Acta Neurochir
(Wien) 1994; 129:152–157
27. Grice SC, Chappel ET, Prough DS, et al:
Ibuprofen improves cerebral blood flow after
global cerebral ischemia in dogs. Stroke
1987; 18:787–791
28. Dempsey RJ, Roy MW, Meyer K, et al: Devel-
opment of cyclooxygenase and lipoxygenase
metabolites of arachidonic acid after tran-
sient cerebral ischemia. J Neurosurg 1986;
64:118 –124
29. Schwab S, Schwarz S, Spranger M, et al:
Moderate hypothermia in the treatment of
patients with severe middle cerebral artery
infarction. Stroke 1998; 29:2461–2466
Crit Care Med 2000 Vol. 28, No. 12