Journal of Cerebral Blood Flow & Metabolism

23:12391250 2003 The International Society for Cerebral Blood Flow and Metabolism
Published by Lippincott Williams & Wilkins, Inc., Baltimore

Energy Dysfunction as a Predictor of Outcome After Moderate

or Severe Head Injury: Indices of Oxygen, Glucose, and
Lactate Metabolism
*Thomas C. Glenn, Daniel F. Kelly, W. John Boscardin, David L. McArthur, Paul Vespa,
*Matthias Oertel, David A. Hovda, Marvin Bergsneider, Lars Hillered, and *Neil A. Martin
From the *UCLA Cerebral Blood Flow Laboratory, California, U.S.A.; the UCLA Division of Neurosurgery, Harbor-UCLA
Medical Center and Research and Education Institute, UCLA Department of Epidemiology, UCLA Department of Biostatistics,
UCLA Department of Molecular and Medical Pharmacology, UCLA Center for Health Sciences, Los Angeles, U.S.A.; and the
Department of Neuroscience, Neurosurgery, Uppsala University Hospital, Sweden.

Summary: The purpose of this study was to determine if the

relationship between abnormalities in glucose, lactate, and oxygen metabolism were predictive of neurologic outcome after
moderate or severe head injury, relative to other known prognostic factors. Serial assessments of the cerebral metabolic
rates for glucose, lactate, and oxygen were performed using a
modified Kety-Schmidt method. In total, 31 normal control
subjects were studied once, and 49 TBI patients (mean age
3616 years, median GCS 7) were studied five times median
per patient from postinjury days 0 to 9. Univariate and multivariate analyses were performed. Univariate analysis showed
that the 6-month postinjury Glasgow Outcome Scale (GOS)
was most strongly associated with the mean cerebral metabolic
rate of oxygen (CMRO2) (P 0.0001), mean arterial lactate
level (P 0.0001), mean arterial glucose (P 0.0008), mean
cerebral blood flow (CBF), (P 0.002), postresuscitation GCS
(P 0.003), and pupillary status (P 0.004). Brain lactate
uptake was observed in 44% of all metabolic studies, and 76%
of patients had at least one episode of brain lactate uptake. By

dichotomized GOS, patients achieving a favorable outcome

(GOS 4-5) were distinguished from those with an unfavorable
outcome (GOS1-3) by having a higher CMRO2 (P 0.003), a
higher rate of abnormal brain lactate uptake relative to arterial
lactate levels (P 0.04), and lesser degrees of blood-brain
barrier damage based on CT findings (P 0.03). Conclusions:
During the first 6 days after moderate or severe TBI, CMRO2
and arterial lactate levels are the strongest predictors of neurologic outcome. However, the frequent occurrence of abnormal
brain lactate uptake despite only moderate elevations in arterial
lactate levels in the favorable outcome patients suggests the
brains ability to use lactate as a fuel may be another key
outcome predictor. Future studies are needed to determine to
what degree nonglycolytic energy production from alternative
fuels such as lactate occurs after TBI and whether alternative
fuel administration is a viable therapy for TBI patients. Key
Words: Cerebral blood flowGlucoseHyperglycolysis
Intracranial hypertensionLactateOxidative metabolism
Traumatic brain injury.

Although much is known regarding posttraumatic impairment of oxidative metabolism and its impact on outcome (Jaggi et al., 1990; Robertson et al., 1992; Tabaddor et al., 1972), less is understood about how alterations
in glucose and lactate metabolism impact recovery of
neurologic function. Cerebral oxidative metabolism generally remains markedly depressed for the first 2 weeks
after severe head injury, and the degree of depression has
been shown to correlate with poor long-term outcome

(Jaggi et al., 1990; Muizelaar et al., 1989; Obrist et al.,

1984; Robertson et al., 1992; Tabaddor et al., 1972).
In contrast, alterations in glycolysis and lactate metabolism appear to have a more variable time course and
less certain impact upon neurologic recovery. Experimental and clinical studies have shown the existence of
an ultra-early posttraumatic activation of glycolysis, also
termed hyperglycolysis, followed by a subacute depression of glucose metabolism (Bergsneider et al.,
1997,2000,2001; Kuroda et al., 1992; Yoshino et al.,
1991). The relationship between these two different epochs of glucose metabolism and long-term outcome is
unclear. The early hyperglycolytic phase appears to reflect increased energy demands for reversal of ionic imbalances because of excitatory amino acid flux (Kata-

Received May 16, 2003; accepted July 8, 2003.

Supported by National Institutes of Neurological Disorders and
Stroke, grant # NS30308.
Address correspondence and reprint requests to Daniel F. Kelly,
UCLA Division of Neurosurgery, David Geffen School of Medicine at
UCLA, 200 Medical Plaza, Suite 504, Box 718224, Los Angeles, CA
90095-7182, U.S.A.; e-mail: dfkelly@ucla.edu.

1239

DOI: 10.1097/01.WCB.0000089833.23606.7F

1240

T. C. GLENN ET AL.

yama et al., 1990; Kawamata et al., 1992, 1995; Nilsson

et al., 1990, 1993; Yoshino et al., 1991, 1992) and has
also been associated with seizure activity after both experimental (Nilsson et al., 1994) and clinical TBI (Vespa
et al., 1998). The subsequent depression of glycolysis
may simply indicate decreased energy demands. However, given that mitochondrial function after TBI is severely impaired (Fiskum, 2000; Verweij et al., 2000), it
is more likely that energy demands remain high and that
injury-related events directly impair glycolysis. Posttraumatic factors such as NAD+ depletion (Ying et al.,
2002), high NADH (Bouzier-Sore et al., 2001), and zinc
influx (Sheline et al., 2000) may further exacerbate a
dearth of ATP and ultimately result in energy failure.
Because lactate is a key intermediary between glycolysis and oxidative metabolism that can be both produced
and used by the brain under pathologic conditions, it may
play a key role in determining cellular viability and longterm outcome after TBI. In severely head-injured patients with cerebral infarctions, anaerobic glycolysis can
result in marked increases in cerebral lactate production
(Robertson et al., 1987). Elevated CSF lactate (DeSalles
et al., 1986) and extracellular lactate measured by cerebral microdialysis (Goodman et al., 1999) has typically
been attributed to ongoing ischemia (Chieregato et al.,
2002; DeSalles et al., 1986; Goodman et al., 1999; Robertson et al., 1987). Lactate elevation caused by ischemia
leads to lactate output from the brain unless arterial lactate levels are very high. However, there is increasing
evidence indicating that lactate uptake also occurs in the
injured brain and other injured tissues, which may then
be used as an alternative fuel source (Amaral et al., 1986;
Inao et al., 1988; Rivers et al., 1991). Several recent
studies show that lactate produced from astrocytic glycolysis can be transported to adjacent neurons and used
in this manner (Alessandri et al., 1999; Chen et al.,
2000a; Magistretti et al., 1999; Ros et al., 2001). Lactate
administration also improves cognitive outcome after experimental TBI (Rice et al., 2002) and reduces the need
for glucose consumption (Chen et al., 2000a). Finally,
lactate may be converted to pyruvate and, instead of
entering the Krebs cycle, may act as a free radical scavenger (Cicalese, 2001; Ochiai et al., 2001). In sum, these
studies suggest that posttraumatic anaerobic and aerobic
metabolism may be more fully understood when carefully considering lactate as both a metabolic by-product
and as a source for fuel.
Based upon the emerging concepts related to glucose
and lactate metabolism described above, this study was
designed to determine how the interrelationships between changes in glucose, lactate, and oxygen metabolism impact long-term outcome. To accomplish this goal,
acute serial measurements of glucose, lactate, and oxygen metabolism were analyzed in relation to 6-month
Glasgow Outcome Scale, while also considering the
J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

outcome predictors of age, Glasgow Coma Scale

(GCS), pupillary abnormalities, hypotension, hypoxia,
computed tomography findings, and patients course of
ICP and CPP.
METHODS
Patient population
Eligible patients included all mechanically ventilated moderate or severe head-injured patients, aged 14 years and older,
who were admitted to UCLA Medical Center within 24 hours
of injury. Moderate or severe head injury was defined as closed
or penetrating injury, including gunshot and stab wounds, with
a postresuscitation GCS less than or equal to 13, or deterioration to a GCS less than or equal to 13 within 24 hours of
admission, and required mechanical ventilation and ICP monitoring. Exclusion criteria included the following: (1) terminal
illness (e.g., advanced cancer, AIDS); (2) severe neurologic
illness (advanced Parkinsons disease, multiple sclerosis, Alzheimers dementia, disabling cerebrovascular event, mental
retardation); and (3) acute complete spinal cord injury. This
protocol was approved by UCLAs institutional review board
and informed consent was obtained from the patients legal
representatives.

General management protocol

Patients were admitted to the intensive care unit (ICU) after
initial stabilization or surgical evacuation of an intracranial
hematoma and were treated in accordance with a Level I
Trauma Center protocol. Management goals included maintenance of ICP less than 20 mmHg and CPP above 70 mmHg,
in accordance with the Guidelines for the management of
severe head injury (Bullock et al., 1996). All patients had a
jugular bulb catheter inserted as soon as possible following patient admission to allow determination of arterio-venous
differences for glucose (AVDglc), lactate (AVDlac), and oxygen (AVDO2).

Jugular bulb catheterization

The dominant jugular vein was selected based upon the dominant jugular foramen visualized on admission CT scan. Using
standard techniques a 5 Fr Cordis and a 4 Fr Oxymetric catheter
(Baxter Critical care, Baxter Health Care) was inserted to approximately 15 cm until resistance was encountered. Placement
of the catheter was confirmed by lateral skull x-ray. The catheter was calibrated in vivo, and repeat calibration was performed every 12 hours. Display of light intensity and oxygen
saturation was continuously displayed on the monitor.

Cerebral metabolism and blood flow measurements

Serial measurements of AVDglc, arterial lactate, AVDlac,
AVDO2, and CBF were made simultaneously during the first
10 days after injury (postinjury days 0 to 9). Arterial and venous samples were scheduled every 12 hours during postinjury
days 0 to 5, 7, and 9; 133Xenon CBF was scheduled for every
12 hours for the first 48 hours after injury (postinjury days 0
and 1), then daily on postinjury days 2, 3, 4, 5, 7, and 9. In this
study, to assess the relationship of early brain metabolism to
long-term outcome, only metabolic data obtained from postinjury days 0 to 5 was used given that many patients did not have
complete data on post-injury days 7 and 9. Because of the
patients clinical status, such as severely elevated ICP (ICP >
30 mmHg), hemodynamic or respiratory instability, or removal of the jugular catheter or extubation following clinical

METABOLISM AND OUTCOME AFTER HEAD INJURY

improvement, it was not possible for all scheduled studies to be
completed on all patients.

Blood and cerebral spinal fluid sampling and

specimen analysis for glucose and
lactate determinations
Blood for arterial and venous glucose and lactate levels were
collected in the following manner:
1. In preparation, eight 3-cc syringes were rinsed with heparin,
which was then discarded before drawing the blood
samples. Before the draws, both the arterial and jugular lines
had 5 cc of blood withdrawn as waste.
2. 1 cc was then drawn during the CBF15 study and dispensed
into two chilled and labeled 400-L micro centrifuge tubes
coated with heparin and lithium fluoride to prevent coagulation and glycolysis. The tubes were immediately placed on
ice. Three additional paired blood draws were collected at
1-min intervals.
3. The chilled tubes then underwent centrifugation at 1,200 g
for 3 minutes. Once centrifuged, the supernatant/plasma was
transferred to labeled micro centrifuge tubes for storage at
80C or immediate spectrophotometric analysis. Cerebral
spinal fluid (CSF) samples were drawn from the ventriculostomy catheter. Plasma and CSF levels for glucose and
lactate were determined via spectrophotometric assay kits
(Sigma), which measure the reduction of NAD to NADH at
340 nm (Beckman). The production of NADH is proportional to the concentration of glucose or lactate in the
plasma. Samples were run 5 times with a measured error of
< 4%. Plasma concentrations of glucose and lactate were
corrected for corresponding whole blood values by using the
erythrocyte distribution volumes for glucose and lactate, as
previously described (Madsen et al., 1995; Simonsen
et al., 1994).
For oxygen saturation and AVDO2 measurements, syringes
of venous and arterial blood were immediately analyzed using
a Corning 178 pH/Blood gas analyzer. For blood gas analysis,
1- to 2-mL blood samples were collected in heparinized syringes, capped, and immediately placed on ice.
Bedside CBF measurement was performed using the intravenous 133Xenon clearance technique (Obrist et al. 1979). Patients were injected with 20 to 30 mCi of 133Xenon dissolved in
saline. A Ceretronix Cerebrograph Cortexplorer 16 (Randers,
Denmark) measured gamma radiation from 133Xenon using
eight detectors positioned over each cerebral hemisphere. Endtidal 133Xenon was measured simultaneously to estimate arterial 133Xenon concentration. Study duration was 11 minutes.
Data were collected by a laptop computer, and the mean hemispheric and global CBF15 was calculated. The CBF15 is a CBF
parameter calculated using a two-compartment mathematical
model and has proven to be very stable in patients with reduced
flow (Obrist and Marion, 1996). Pertinent clinical and physiologic data (MABP, ICP, CPP, PaCO2, body temperature, and
vasoactive and metabolic suppressive drugs) were recorded at
the time of study. All information was then entered into the
UCLA CBF Laboratory and Core databases.

Determination of CMRglc, CMRlac, and CMRO2

For each metabolic measurement, the cerebral metabolic
rates for glucose, lactate, and oxygen were calculated as the
product of the simultaneously measured AVDglc, AVDlac or
AVDO2, and the CBF15. Studies were excluded from analysis
if at the time of metabolic assessment the patient was receiving metabolic suppressive therapy with high-dose pentobarbital or high-dose propofol, defined as a propofol infusion of

1241

100 gm/kg/min or greater, given that these two agents are

known to significantly suppress both glucose and oxidative
metabolism (Cormio et al., 1999; Oertel et al., 2002). The
metabolic ratio (MR) was determined by calculating
CMRO2/CMRglc in molar units at 12-hour intervals until the
jugular catheter was removed.

Global outcome measurement

Neurologic outcome was assessed 6 months after injury by
investigators blinded to the patients metabolic and other clinical data, using the 5-point Glasgow Outcome Scale (GOS)
(Jennet and Bond, 1975). In assessing predictors of outcome,
the GOS was used as both a linear scale (1 death to 5
good recovery) and as a dichotomized outcome score of favorable (good recovery or moderate disability) versus unfavorable
(severe disability, vegetative, or death) (Choi et al., 1998).

Predictors of global outcome

The relationships of 6-month GOS to mean levels of
AVDglc, arterial glucose and lactate, AVDlac, SjVO 2 ,
AVDO2, CBF, CMRglc, CMRlac, CMRO2, the metabolic ratio, and CSF glucose and lactate levels were examined. The
impact of increased rates of glucose metabolism was also assessed. Absolute hyperglycolysis (> 6.69 mg/100g/min) was
defined as a CMRglc greater than the 97.5th percentile of
normal for an awake human (normal mean 4.46
mg/100g/min1.16, n 30), and relative hyperglycolysis
(MR < 3.44) was defined as a metabolic ratio (MR
CMRO2/CMRglc in molar units) below the 2.5th percentile of
normal (normal MR mean 5.831.41, n 30).
The associations of outcome to other outcome predictors for
TBI were also examined. These clinical predictors included
patient age, postresuscitative GCS score, postresuscitative pupillary examination (both normal, one abnormal, both abnormal), hypotension (systolic blood pressure < 90 mmHg), hypoxia (PaO2 < 60 mmHg or apnea) within 24 hours of injury, the
mean ICP and CPP, percentage of hours ICP greater than 20
mmHg, and percentage of hours CPP less than 60 mmHg (Marmarou et al., 1991; Marshall et al., 1991a; Vollmer et al., 1991).
Previously described CT predictors of outcome seen on patients first or second CT scans included basilar cistern compression, midline shift greater than 5 mm, acute subdural hematoma requiring evacuation, or the presence of subarachnoid
hemorrhage (Marshall et al., 1991b). All patients first two CT
scans were obtained within 12 hours of admission. The degree
of blood-brain barrier damage was also assessed by CT for each
patient, by determining the total number of the following diagnoses: acute subdural hematomas, intracerebral hematomas,
or multiple contusions. Each of these diagnoses has been associated with blood-brain barrier damage in prior experimental
or clinical studies (Barzo et al., 1996; Beaumont et al., 2002;
Fobben et al., 1989; Okauchi et al., 2002; Xi et al., 2001).

Normal control subjects

A cohort of 31 healthy volunteer subjects was studied to
define the normal range of all metabolic parameters, defined as
between 2.5th percentile and the 97.5th percentile. For data
with an approximately normal distribution, the 2.5th and 97.5th
percentiles would correspond to the mean2 SD. These normal
subjects were studied in an awake state. Through use of a
femoral vein approach, a jugular bulb catheter was passed under fluoroscopic guidance into proper position in the jugular
bulb. A radial arterial line was also placed. Single measurements of AVDglc, AVDlac, and AVDO2 were made by taking
simultaneous samples of arterial and venous blood as described
above. A CBF measurement was performed using the intravenous 133Xenon clearance technique as described above. Normal

J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

1242

T. C. GLENN ET AL.

TABLE 1. Head injury cohort demographics and outcome

Mean age (years)
Gender
Median GCS
Injury Severity Score
Craniotomy for hematoma

36 16
27% female
7 (range 3 to 14)
32 10
59.50%

6 Month GOS

20.40%
34.70%
26.50%
0.00%
18.40%
100%

10
17
13
0
9
49

Good Recovery
Moderate Disability
Severe Disability
Vegatative
Dead
Total
N 49

ranges for cerebral metabolic rates for glucose, lactate, and

oxygen were calculated as the product of the simultaneously
measured AVDglc, AVDlac or AVDO2, and the CBF15.

Statistical analysis
Mixed effects linear regression models were used throughout
to account for the uneven numbers of studies on patients at
different time points (Laird and Ware, 1982). To examine the
association of metabolism and 6-month GOS, the measurements of metabolic parameters were modeled as correlated outcomes within patients with different means for each level of the
GOS. A variety of covariance structures for the repeated measures were considered; we report results for models with random patient-level means and continuous time AR(1) correlation. For these statistical models, the estimated means by GOS
category can be thought of as a data-driven compromise between the means of the studies within each GOS category and
the mean of the patient means for each GOS category. We refer
to these quantities as the adjusted means in the results section;
they are adjusted for the unequal number of studies per patient
at different hours postinjury. Similar interpretation can be given
to the adjusted standard deviations for each GOS category.
Logarithmic transformations of the metabolic measurements
were used to obtain more accurate P values where appropriate;
however, adjusted means and standard deviations are reported
for the original scale data for clarity of interpretation. The
univariate mixed effect models were then augmented by controlling for a fixed set of clinical predictors including patient
age, ISS, pupillary findings, CT findings, mean ICP, mean
CPP, and best GCS. Comparisons of TBI patients to normal
volunteers were performed in a similar manner. In several instances, we consider repeated measures of a dichotomous indicator (e.g., lactate uptake above a certain threshold). Generalized Estimating Equations methods were used to analyze
these data (Liang and Zeger, 1986; Zeger and Liang, 1986).
Comparison of TBI patients with normal volunteers was performed in a similar manner. Statistical significance was defined
at P < 0.05, but in some instances, attention is drawn to suggestive results given the small sample sizes.

RESULTS
Head injury cohort
The 49 TBI patients were enrolled in the study between July 1998 and March 2002. The mean patient age
was 3616 years (range 16 to 81 years) and 13 (27%)
were female. (Table 1) The median postresuscitation
J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

GCS score was 7 with a range of 3 to 14; 86% of patients

had an initial GCS less than or equal to 8. Mechanisms of
injury included 16 motor vehicle accidents, 11 pedestrian
versus automobile accidents, 11 falls, 4 assaults, 3 gunshot wounds, 2 motorcycle accidents, and 2 bicycle accidents. In total, 28 patients (57.1%) required a craniotomy for evacuation of an intracranial hematoma,
including 11 with intracerebral hematoma/contusions, 8
with combination lesions, typically subdural hematomas
in combination with cerebral contusions, 6 with subdural
hematomas, and 3 with epidural hematomas.
At 6 months after injury, global neurologic outcome as
measured by the GOS was good recovery in 10 patients
(20.4%), moderate disability in 17 (34.7%), severe disability in 13 (26.5%), and death in 9 (18.4%); no patients
remained in a vegetative state. Overall, 55% of the patients achieved a favorable outcome, and 45% achieved a
poor outcome (Table 1).
Metabolic rates of TBI subjects versus normal
control subjects
Thirty-one normal volunteers (mean age 33 8 years,
26% female) had a single determination of metabolic
rates in an awake state. The 49 TBI patients had a total
of 185 CMRO2, 185 CMRglc, 179 CMRlac, and 174
Metabolic Ratio determinations (median 5 per patient,
range 1 to 13) over postinjury days 0 to 5. As can be seen
in Table 2, the mean value of each parameter over postinjury days 0 to 5 for the 49 TBI patients was significantly
different from those of the control subjects. The parameters most severely depressed relative to the normal subjects values were CMRO2 (45% of normal), AVDO2
(55% of normal), and CMRglc (76% of normal), whereas
the mean arterial lactate level for TBI patients was most
elevated relative to normal (210% of normal). Notably,
global cerebral ischemia as defined by an AVDO2
TABLE 2. Comparison of metabolic parameters for normal
subjects and trauma patients
Predictor
CBF
CMRO2
CMRglc
CMRlac
AVDO2
AVDglc
AVDlac
Art glucose
Art lactate
Metabolic Ratio
Jugo2Sat
Arto2Sat
Age
Gender

Units

Trauma*
(n 49)

Normal
(n 31)

ml/100g/min
40.17 13.2 46.2 10.5
ml/100g/min
1.4 0.43 3.10 0.56
mg/100g/min
3.43 2.32 4.46 1.16
mg/100g/min 0.0355 0.41 0.18 0.21
ml/dl
3.76 1.37 6.89 1.35
mg/dl
8.94 6.57 9.89 2.92
mg/dl
0.0464 0.94 0.40 0.48
mg/dl
122.7 33.8 82.2 8.3
mg/dl
14.05 8.2
6.7 2.1
4.11 2.11 5.83 1.41
%
72.9 8.6
61.7 5.9
%
98.5 1.5
97.5 1.2
years
35.7
33.3 8.3
% Male
73%
74%

p
value
0.01
0.0001
0.0002
0.0001
0.0001
0.002
0.0002
0.0001
0.0001
0.0001
0.0001
0.0001
0.4
1

* For trauma patients, values are adjusted means as described in

Methods. A total of 181 AVD determinations were made, median 5 per
patient (range 113).

METABOLISM AND OUTCOME AFTER HEAD INJURY

1243

greater than or equal to 9.6 mL/dl was not seen in any

TBI patients.

(Table 3). A statistical trend existed for CBF (P 0.1)

and metabolic ratio (P 0.06).

Predictors of global outcome: univariate analysis

The mixed effect linear regression models yielded the
following associations of metabolic and other prognostic
parameters with 6-month GOS (Table 3). For predicting
6-month GOS category, the strongest factors were mean
CMRO2 (P 0.0001), mean arterial lactate level (P
0.0001), mean arterial glucose (P 0.0008), mean CBF
(P 0.002), postresuscitation GCS (P 0.003), and
pupillary status (P 0.004). For predicting a favorable
outcome (GOS 4 or 5) versus an unfavorable outcome
(GOS 1 or 3), the strongest factors were mean CMRO2,
mean arterial lactate level, mean arterial glucose level,
mean CBF, CSF lactate concentration, postresuscitation
GCS, pupillary status, and number of CT diagnoses associated with blood-brain barrier damage.

Oxygen metabolism and outcome

CMRO2 was significantly and ubiquitously depressed
with 97% (179 of 185 studies) of measurements below
the 2.5th percentile (2.23 mL/100g/min) of the normal
cohort. Overall, the mean CMRO2 was lowest in patients
who died. Patients with a good recovery had a mean
CMRO2, which was 44% greater than those patients
who died.
Lactate metabolism and outcome
In the TBI cohort, 80 of 181 (44%) AVDlac measurements were positive, indicating brain uptake of lactate,
and 32 of 42 (76%) patients had at least one episode of
brain uptake of lactate. Positive AVDlac values occurred
across all injury days although they tended to occur earlier after injury. As seen in Fig. 1, the greatest departure
from normal CMRlac values was in the positive direction
(i.e., brain lactate uptake) and occurred most frequently
in the nine patients who died, in which 20 of 30 (67%)
measurements were above zero. Patients who died and
those who remained severely disabled also had the highest systemic arterial lactate levels, as seen in Table 3, and

Multivariate analysis
After controlling for injury characteristics, age, and
parameters of ICP and CPP, the associations between
GOS category and the following factors remained strong:
CMRO2 (P 0.002), arterial lactate (P 0.007), CSF
lactate (P 0.01), and arterial glucose (P 0.02)

TABLE 3. Association of 6-month outcome to metabolic and clinical parameters

Category
Metabolic

Acute
Injury
Char.

Predictor
JUGO2SAT
ARTO2SAT
ARTPO2
AVDO2
CMRO2
Art glucose
AVDglc
CMRglc
Metabolic Ratio
Art lactate
AVDlac
CMRlac
CBF
CSFglc
CSFlac
Pupils

Units

mm Hg
ml/dl
ml/100g/min
mg/dl
mg/dl
mg/100g/min
mg/dl
mg/dl
mg/100g/min
ml/100g/min
mg/dl
mg/dl
% any
abnormal

GCS
Hypot-Hypoxia % any
episode
CT Score
*CT Score BBB damage
ISS
Patient
Age
Characteristics
Gender
Intracranial
Physiology

years

P-values
Univariate

Multivariate

Univariate

Multivariate

GOS 1
(n 9)

GOS 3
(n 13)

GOS 4
(n 17)

GOS 5
(n 10)

linear
trend

linear
trend

good/
bad

good/
bad

75.8 8.0
98.4 1.6
155.1 85.1
3.4 1.3
1.2 0.4
143.8 46.3
9.0 5.7
3.2 1.8
3.45 1.7
21.9 12.0
0.4 1.3
0.18 0.59
36.5 14.9
91.5 41.7
32 10.4

70.4 7.6
98.1 1.8
132.8 49.3
3.9 1.28
1.3 0.3
123.8 33.9
8.8 6.0
3.1 1.7
4.1 2.1
14.6 7.8
0.04 0.76
0.19 0.27
36.2 11.9
83.0 26.2
25.9 10.9

71.9 9.2
98.7 1.3
144.0 42.4
3.9 1.4
1.4 0.4
113.9 30.4
9.6 8.0
3.7 3.0
4.1 2.3
11.5 7.1
0.02
0.02 0.42
40.3 12.2
76.1 26.4
20.7 8.5

75.4 8.8
98.8 1.1
156.2 48.0
3.5 1.3
1.6 0.5
110.4 23.1
7.9 5.3
3.6 2.2
4.7 2.1
10.7 5.8
0.13
0.04
48.6 11.9
76.6 21.2
19.3 7.1

0.9
0.04
0.8
0.8
0.0001
0.0008
0.4
0.1
0.05
0.0001
0.03
0.05
0.002
0.2
0.01

0.5
0.07
0.9
0.4
0.002
0.02
0.9
0.4
0.06
0.007
0.8
0.8
0.1
0.4
0.01

0.8
0.01
0.6
0.9
0.003
0.0008
0.7
0.1
0.2
0.0007
0.1
0.2
0.01
0.2
0.01

0.9
0.03
0.5
0.8
0.07
0.03
0.9
0.5
0.2
0.02
0.9
0.9
0.2
0.4
0.01

89%

62%

42%

30%

0.004

0.02

7.1 3.0
23%

6.9 1.9
29%

8.2 2.6
0%

0.003
0.5

0.03
0.7

2.4 1.0
1.6
1.3 0.9
1.3 1.3
35.7 14.8 32.7 7.3

1.4 1.0
0.8 0.6
32.9 9.3

1.3 1.3
0.8 0.6
26.0 6.1

0.03
0.1
0.2

0.08
0.03
0.2

0.2

0.08

5 1.9
22%

42.7 24.8

39 17.5 34.1 12.1 27.7 9.4

% Male

56%

85%

71%

80%

0.3

Mean CPP

mm Hg

73.7 9.4

82.1 6.8

85.8 7.4

78.5 4.8

0.5

0.2

% CPP < 60
Mean ICP
% ICP > 20

mm Hg

17.5 20.8 2.5 6.8

2.2 6.7 15.9 2.6
20.2 8.9 14.8 3.6 13.5 3.8
3.7 5.1
33.4 36.0 13.1 21.0 15.3 22.5 13.9 16.7

0.2
0.5
0.5

0.2
0.3
0.8

* CT score BBB (blood-brain-barrier) damage total number of following CT diagnoses - acute subdural hematomas, intracerebral hematomas or multiple contusions.

J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

1244

T. C. GLENN ET AL.

arterial lactate levels overall had a strong inverse correlation with 6-month GOS (P 0.0001). As seen in Fig.
2, abnormal brain lactate production (CMRlac below the
2.5th percentile of normal, i.e., < 0.71 mg/100g/min)
was uncommon, occurring in only 4 (2%) of 179 measurements and in four (8%) patients, whereas abnormal
brain lactate uptake (CMRlaclac above the 97.5th percentile of normal, i.e., > 0.11 mg/100g/min) was observed in 51 (28%) of 179 measurements. Despite significantly lower levels of systemic arterial lactate in the
favorable outcome patients (GOS 4 or 5), abnormal brain
lactate uptake still occurred relatively frequently, being
observed in 26% and 33% of the measurements in these
patient groups, respectively (Fig. 3). Overall, a higher
rate of brain lactate uptake relative to the simultaneous
arterial lactate level was observed in the favorable outcome group (GOS 4 and 5) compared with the unfavorable outcome group (GOS 1 and 3) (P .04).
Glucose, hyperglycolysis, and outcome
High plasma glucose levels were associated with poor
outcome (Fig. 3). The mean plasma glucose was 23%
greater in patients who died compared with patients with
a GOS of 5. Relative hyperglycolysis (metabolic ratio <

3.44) was observed in 44% of studies, and at least one

episode of relative hyperglycolysis was seen in 39 of 49
(80%) patients. Nine episodes of absolute hyperglycolysis (5% of 185 measurements) were documented in nine
patients. In these nine patients with absolute hyperglycolysis, five achieved a favorable outcome, and four
achieved a poor outcome by 6-month GOS. Relative hyperglycolysis appeared to be more frequent in patients
who died with 48% (19 of 40) studies showing a metabolic ratio less than 3.44, whereas patients with a good
outcome had only 29% (10 of 35) studies in the relative
hyperglycolysis range. However, neither episodes of absolute or relative hyperglycolysis were strongly associated with 6-month GOS.
Cerebral blood flow and outcome
Using univariate analysis, CBF was a strong predictor
of 6-month GOS but using multivariate analysis, CBF
was a relatively weaker outcome predictor (P .1 for
linear trend and P .2 for good/bad) (Table 3). CBF in
the ischemic range rarely occurred as it was observed in
only four instances. However, three ischemic CBF values (< 20 mL/100g/min) were observed in patients who

FIG. 1. Boxplots of CMRO2, CMRglc, and CMRlac by 6-month Glasgow Outcome Scale. Summary of values in 49 patients collected from
postinjury days 0 to 5, stratified by 6-month GOS and compared with values of 31 normal control subjects. Lines within the box indicate
median, the box interquartile range, whiskers bracket 95% of the data, and straight bars outliers. Overall CMRO2 is higher in patients with
a good outcome but is depressed in all groups compared with control patients. CMRglc is also depressed after injury but to a lesser degree
than CMRO2. CMRlac was positive in almost half of all studies; note that the most positive values are in the patients who died (GOS 1).

J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

METABOLISM AND OUTCOME AFTER HEAD INJURY

1245

FIG. 2. Graph showing the percent of CMR measurements that were either below the 2.5th percentile, above the 97.5th percentile, or
within normal range of control subjects, stratified by 6-month GOS score. Note the relatively uncommon finding (2%) of abnormally
negative CMRlactate (C) values (cerebral lactate production). In contrast, abnormally positive CMRlactate values, indicating cerebral
uptake of lactate, were seen in 28% of studies. (A) CMRO2; (B) CMRglc; (C) CMRlac.

died. Mean CBF was lower in patients with GOS 1 and

3, compared with the GOS 4 and 5 outcome groups.
DISCUSSION
Overview of findings
This study of patients with moderate and severe head
injuries reconfirms previous works on the prognostic importance of acute changes in cerebral oxidative metabolism and blood flow (Jaggi et al., 1990; Kelly et al.,
1997; Obrist et al., 1984; Robertson et al., 1992). This
analysis also brings new insights as to how posttraumatic
interactions between glycolysis, lactate metabolism, and
oxidative metabolism impact long-term outcome. The
major findings of this study are as follows:
1. Moderate or severe TBI resulted in an acute reduction
of CMRO2 and, to a lesser extent, CMRglc on average, to 45% and 76% of normal, respectively.
2. Relative hyperglycolysis was common early after
TBI, being observed at least once in 83% of patients.

3. The best clinical predictors of 6-month GOS were

CMRO2, arterial lactate level, arterial glucose level,
CBF, and postresuscitation GCS and pupillary status.
4. Brain lactate uptake was observed in 44% of all metabolic studies, and 76% of patients had at least one
episode of brain lactate uptake.
5. Abnormal uptake of lactate by the brain (CMRlac >
97.5th percentile of normal) occurred in 28% of studies, whereas abnormal lactate production by the brain
occurred in only 2% of studies.
6. Patients achieving a favorable outcome had a higher
rate of abnormal brain lactate uptake relative to arterial lactate levels in comparison with patients achieving an unfavorable outcome.
These findings are discussed below in relation to previous studies on metabolic changes after TBI and implications for future TBI therapies.
Oxygen metabolism
Consistent with the present study, it is well-established
that CMRO2 is reduced by approximately 50% from norJ Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

1246

T. C. GLENN ET AL.

FIG. 3. Mean arterial lactate and glucose levels. Mean arterial plasma glucose and lactate concentrations were lower in patients with
higher GOS (4 and 5), but both were abnormally elevated compared with control volunteers.

mal in comatose head injury patients (Bouma et al.,

1991; Martin et al., 1997; Obrist et al., 1984; Robertson
et al., 1992). This reduction in CMRO2 occurs both in
areas affected by focal traumatic lesions and in remote
areas, which appear normal on computerized tomographic scans (Tenjin et al., 1990). The depression in
CMRO2 is seen when it is first measured after admission
to the hospital, is proportional to the depth of coma, and
appears to be stable at this level (Bouma et al., 1991;
Jaggi et al., 1990; Obrist et al., 1984, 1993). The low
CMRO2 returns at least partially toward normal in patients who recover, indicating this depression is not simply caused by irreversible destruction of brain cells
(Tenjin et al., 1990).
J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

Glucose metabolism
Unlike the depression of oxidative metabolism, an increase in cerebral glucose use has been a consistent finding across animal brain-injury models, including concussive brain injury (Kawamata, 1992; Sunami et al., 1989;
Yoshino et al., 1991), cortical contusions (Sunami et al.,
1989), and experimental subdural hematoma (Inglis et
al., 1992). These studies have revealed that injured cells
are exposed to a massive ionic flux, in part because of the
release of excitatory amino acids (Katayama et al., 1989,
1990). This ionic imbalance transiently increases the use
of cerebral glucose, which activates Na+/K+ pumps in an
effort to restore ionic homeostasis (Hovda, 1996; Hovda
et al., 1992; Yoshino et al., 1991).

METABOLISM AND OUTCOME AFTER HEAD INJURY

The existence of an increased cerebral metabolic rate
of glucose (CMRglc) in the acute posttraumatic human is
supported by several independent methods. Using [18F]fluorodeoxyglucose positron emission tomography
(FDG-PET), we confirmed experimental autoradiographic studies by demonstrating that regional CMRglc
is elevated in the majority of patients with severe head
injuries studied within 1 week of injury (Bergsneider et
al., 1997). A similar finding has been reported using
FDG-PET combined with an (15O) oxygen PET study
(Yamaki et al., 1996). Given the relative depression of
oxidative metabolism after TBI, one might predict
marked compensatory increases in cerebral glycolysis to
maximize ATP production. However, as seen in this
study, glucose metabolism was also globally depressed
but not to the same degree as oxidative metabolism, resulting in the common finding of relative hyperglycolysis. Several factors are likely responsible for this absolute reduction in glycolysis. Lactate uptake by the brain
likely impedes glycolysis if it is converted to pyruvate
(Amaral et al., 1986). Additional posttraumatic factors
such as NAD+ depletion (Ying et al., 2002), high NADH
(Bouzier-Sore et al., 2001), and zinc influx (Sheline et
al., 2000) may also acutely suppress glycolysis. It is also
likely that episodes of absolute hyperglycolysis were
rarely captured because of the few measurements taken
within hours of injury when absolute hyperglycolysis is
typically observed (Kuroda et al., 1992; Yoshino et al.,
1991). Overall, absolute hyperglycolysis was seen in
only 5% of measurements in this study. Such episodes of
hyperglycolysis are generally expected to result in brain
lactate production, but of interest is the fact that in five of
nine such episodes, abnormal lactate uptake by the brain
was observed. The simultaneous presence of absolute
hyperglycolysis, marked brain lactate uptake, and low
CMRO2 is likely indicative of a cerebral energy crises, as
evidenced by the fact that all three patients in this metabolic state had poor outcomes. Perhaps an important caveat to the interpretation of global measurements of glucose metabolism is that the metabolic fate of glucose,
that is, glycolysis or biosynthesis of other compounds,
cannot currently be ascertained using these techniques.
Absence of ischemia and rare brain
lactate production
This study is in contrast to previous reports that have
stressed cerebral production of lactate as a marker of
brain ischemia (Goodman et al., 1999; Robertson et al.,
1987). In the present study, extreme degrees of cerebral
lactate production were seen in only 2% of studies, and
AVDO2 values in the ischemic range were never observed. Previous microdialysis studies (Goodman et al.,
1999; Meixensberger et al., 2001; Reinert et al., 2000)
and our own ongoing microdialysis studies showing elevated microdialysate levels of lactate may in part be

1247

measuring lactate that is diffusing into or being transported into the interstitial space from the extracranial
circulation (Chen et al., 2000b; Nemoto and Severinghaus, 1974; Rivers et al., 1991). Admittedly, the global
technique used in this study cannot detect areas of regional ischemia where high lactate production may occur.
Brain lactate uptake: two scenarios
The findings regarding lactate metabolism raise two
key questions. First, given that the majority of patients
are in a state of relative hyperglycolysis acutely after
injury in which CMRO2 is decreased to a greater degree
than CMRglc, why does abnormal uptake of lactate by
the brain occur so much more frequently than brain lactate production? Second, why is brain uptake of lactate
associated with both favorable and unfavorable clinical
outcomes? To answer these questions, we propose that
two very different clinical scenarios exist after TBI. In a
favorable outcome scenario (in patients who achieve a
GOS of 4 or 5), uptake of lactate by the brain occurs
primarily by facilitated transport through a largely intact
blood-brain barrier. Given a less depressed CMRO2 in
these patients, the lactate can be effectively used as a fuel
source. In an unfavorable outcome scenario (in patients
who achieve a GOS of 1 to 3), large amounts of lactate
passively enter the brain as a result of high systemic
arterial lactate levels and a damaged blood-brain barrier.
Given a severely depressed CMRO2 in these patients, the
abundant lactate cannot be effectively used.
Several findings in this study support this twoscenario concept of posttraumatic lactate metabolism.
Despite the fact that the favorable outcome patients had
lower arterial lactate levels than the poor outcome patients, these patients had a higher rate of abnormal brain
lactate uptake relative to arterial lactate levels. This finding of a high rate of brain lactate uptake despite lower
arterial lactate levels suggests the lactate was being actively taken into the brain. The favorable outcome group
was further distinguished from the unfavorable outcome
group by having higher values for CMRO2 and CMRglc
and lesser degrees of blood-brain barrier damage based
upon CT findings. The lesser degree of blood-brain barrier damage in the favorable outcome patients also argues against the possibility that the brain lactate uptake
in these patients was simply a result of diffusion into
the brain.
Previous studies confirm that under certain conditions,
the brain and other tissues, such as heart, kidney, liver,
and muscle, can use lactate through so-called lactate
shuttles and facilitated transport mechanisms (Amaral
et al., 1986; Brandt et al., 1984; Brooks, 2002; Rivers et
al., 1991). Others have shown brain uptake of lactate and
its conversion to pyruvate to enter the Krebs cycle for
ATP production (Chen et al., 2000a; Magistretti et al.,
1999; Ros et al., 2001). Lactate may also be used to
J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

1248

T. C. GLENN ET AL.

scavenge free radicals (Cicalese, 2001; Ochiai et al.,

2001). In contrast, under conditions of high arterial lactate levels and a damaged blood-brain barrier, lactate
may passively enter the brain but not be used because the
neurons and glia have severely depressed metabolic capacity (Clausen et al., 2001; Fiskum, 2000; Nemoto and
Severinghaus, 1974; Verweij et al., 2000).
Ways to improve energy production
The correlation of severely depressed cerebral metabolism with poor outcome (Jaggi et al., 1990; Obrist et
al., 1984; Robertson et al., 1992) raises the question of
whether increased energy production after TBI will enhance outcome. Several possible therapeutic avenues
have been previously proposed, including use of hyperbaric oxygen (Rockswold et al., 2001), increased inspired oxygen concentration (Menzel et al., 1999), and
cyclosporin A (Alessandri et al., 2002). Given the increasing evidence that the injured brain can use alternative substrates such as lactate and that lactate administration itself can improve outcome after experimental
TBI (Rice et al., 2002), future studies aimed at identifying safe and effective fuel sources may prove fruitful.
Pyruvate, which can directly enter the Krebs cycle or be
used as a free radical scavenger, may be one such compound (Cicalese, 2001; Ochiai et al., 2001). Two additional compounds, beta-hydroxybutyrate, a ketone body
known to be used by the brain under starvation conditions, and acetyl-L-carnitine also have potential as alternative fuel sources after TBI (Rosenthal et al., 1992;
Suzuki et al., 2001).
Methodologic issues
The modified Kety-Schmidt method used in this study
is a well-established technique that has been used extensively in TBI patients (Jaggi et al., 1990; Muizelaar et al.,
1989; Obrist et al., 1984). Regarding the timing of studies, data from only the first 6 days (postinjury days 0 to
5) after injury were assessed in part because the first
several days after injury have generally been considered
the most dynamic in terms of metabolic and blood flow
changes (Bouma et al., 1991; Jaggi et al., 1990). In the
present study, a median of five metabolic and CBF determinations were performed during these first 6 days
after injury. For practical reasons and patient safety issues, it was not possible to obtain all planned studies
over this time period. Clinical factors such as high ICP or
hemodynamic instability precluded a minority of studies.
Patients being treated for intractable intracranial hypertension with high-dose pentobarbital or propofol were
also excluded from analysis, given that these agents significantly reduce both oxidative and glucose metabolism
(Cormio et al., 1999; Kelly et al., 1999; Oertel et al.,
2002). After postinjury day 5, neurologic improvement
and extubation and removal of the jugular bulb catheter
J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

were the most common reasons for patients leaving

the study.
The statistical analyses used in this study specifically
accounted for the variable numbers of studies nested
within subjects and the relatively small sample size of
TBI subjects. Although it was not feasible to perform
regression analyses of outcome on metabolic and other
predictors, the longitudinal nature of the study minimized the problem of a small sample size by allowing for
regression of metabolic measurements on outcome and
other predictors.
CONCLUSION
This study of acute posttraumatic cerebral metabolic
changes suggests that the interplay between energy supply and demand ultimately dictates global neurologic
outcome. It again shows the prognostic importance of
acute measurements of cerebral oxidative metabolism
and blood flow. However, it also suggests that brain
lactate metabolism may be another key outcome determinant. We propose that patients with relatively less severe injuries as characterized by a higher CMRO2, lesser
degrees of blood-brain barrier damage, and lower systemic lactate and glucose levels can use lactate as an
additional fuel source and ultimately have a favorable
long-term outcome. In contrast, patients with more severe injuries as characterized by a lower CMRO2, higher
degrees of blood-brain barrier damage, and higher systemic lactate levels are unable to use the lactate and
ultimately have a poor outcome. Thus a distinguishing
feature between patients who make a good recovery and
those who do not may be their ability to meet energy
demands by whatever means possible, including the use
of alternative fuel sources such as lactate. Future studies
are needed to better understand alternative fuel use after
head injury and to develop strategies that augment posttraumatic cerebral metabolic capacity.
Acknowledgment: We thank Maria Etchepare, Jill Hutchinson, and Kathy Langlois for their help in data collection. We
also thank Brenda Rinsky, Oscar Barcenas, and Chris Hanuscin
from the UCLA Cerebral Blood Flow Laboratory for their technical support.

REFERENCES
Alessandri B, Doppenberg E, Zauner A, Woodward J, Choi S, Bullock
R (1999) Evidence for time-dependent glutamate-mediated glycolysis in head-injured patients: a microdialysis study. Acta Neurochir
Suppl 75:2528
Alessandri B, Rice AC, Levasseur J, DeFord M, Hamm RJ, Bullock
MR (2002) Cyclosporin-a improves brain tissue oxygen consumption and learning/memory performance after lateral fluid percussion injury in rats. J Neurotrauma 19:829841
Amaral JF, Shearer JD, Mastrofrancesco B, Gann DS, Caldwell MD
(1986) Can lactate be used as a fuel by wounded tissue? Surgery
100:252260
Barzo P, Marmarou A, Fatouros P, Corwin F, Dunbar J (1996) Magnetic resonance imaging-monitored acute blood-brain barrier

METABOLISM AND OUTCOME AFTER HEAD INJURY

changes in experimental traumatic brain injury. J Neurosurg
85:11131121
Beaumont A, Marmarou A, Fatouros P, Corwin F (2002) Secondary
insults worsen blood brain barrier dysfunction assessed by MRI in
cerebral contusion. Acta Neurochir Suppl 81:217219
Bergsneider M, Hovda DA, Lee SM, Kelly DF, McArthur DL, Vespa
PM, Lee JH, Huang SC, Martin NA, Phelps ME, Becker DP (2000)
Dissociation of cerebral glucose metabolism and level of consciousness during the period of metabolic depression following
human traumatic brain injury. J Neurotrauma 17:389401
Bergsneider M, Hovda DA, McArthur DL, Etchepare M, Huang SC,
Sehati N, Satz P, Phelps ME, Becker DP (2001) Metabolic recovery following human traumatic brain injury based on FDG-PET:
time course and relationship to neurological disability. J Head
Trauma Rehabil 16:135148
Bergsneider M, Hovda DA, Shalmon E, Kelly DF, Vespa PM, Martin
NA, Phelps ME, McArthur DL, Caron MJ, Kraus JF, Becker DP
(1997) Cerebral hyperglycolysis following severe traumatic brain
injury in humans: a positron emission tomography study. J Neurosurg 86:241251
Bouma GJ, Muizelaar JP, Choi SC, Newlon PG, Young HF (1991)
Cerebral circulation and metabolism after severe traumatic brain
injury: the elusive role of ischemia. J Neurosurg 75:685693
Bouzier-Sore AK, Canioni P, Merle M (2001) Effect of exogenous
lactate on rat glioma metabolism. J Neurosci Res 65:543548
Brandt RB, Waters MG, Rispler MJ, Kline ES (1984) D- and L-lactate
catabolism in CO2 in rat tissue. Proc Soc Exp Biol Med 175:328
335
Brooks GA (2002) Lactate shuttles in Nature. Biochem Soc Trans
30:258264
Bullock R, Chesnut RM, Clifton G, et al. (1996) Guidelines for the
management of severe head injury. J Neurotrauma 13:639734
Chen T, Qian YZ, Di X, Rice A, Zhu JP, Bullock R (2000a)
Lactate/glucose dynamics after rat fluid percussion brain injury. J
Neurotrauma 17:135142
Chen T, Qian YZ, Rice A, Zhu JP, Di X, Bullock R (2000b) Brain
lactate uptake increases at the site of impact after traumatic brain
injury. Brain Res 861:281287
Chieregato A, Marchi M, Zoppellari R, Fabbri E, Cianchi G, Forini E,
Targa L (2002) Detection of early ischemia in severe head injury
by means of arteriovenous lactate differences and jugular bulb
oxygen saturation. Relationship with CPP, severity indexes and
outcome. Preliminary analysis. Acta Neurochir 81 (Suppl):289
293
Choi SC, Marmarou A, Bullock R, Nichols JS, Wei X, Pitts LH (1998)
Primary end points in phase III clinical trials of severe head
trauma: DRS versus GOS. The American Brain Inj Consortium
Study Group. J Neurotrauma 15:771776
Cicalese L (2001) Pyruvate in organ transplantation. J Parenter Enteral
Nutr 25:216218
Clausen T, Zauner A, Levasseur JE, Rice AC, Bullock R (2001) Induced mitochondrial failure in the feline brain: implications for
understanding acute post-traumatic metabolic events. Brain Res
908:3548
Cormio M, Gopinath SP, Valadka A, Robertson CS (1999) Cerebral
hemodynamic effects of pentobarbital coma in head-injured patients. J Neurotrauma 10:927936
DeSalles AA, Kontos HA, Becker DP, Yang MS, Ward JD, Moulton R,
Gruemer HD, Lutz H, Maset AL, Jenkins L, et al. (1986) Prognostic significance of ventricular CSF lactic acidosis in severe head
injury. J Neurosurg 65:615624
Fiskum G (2000) Mitochondrial participation in ischemic and traumatic
neural cell death. J Neurotrauma 17:843855
Fobben ES, Grossman RI, Atlas SW, Hackney DB, Goldberg HI, Zimmerman RA, Bilaniuk LT (1989) MR characteristics of subdural
hematomas and hygromas at 1.5 T. Am J Roentgenol 153:589
595
Goodman JC, Valadka AB, Gopinath SP, Uzura M, Robertson CS
(1999) Extracellular lactate and glucose alterations in the brain
after head injury measured by microdialysis. Crit Care Med
27:19651973
Hovda DA (1996) Metabolic dysfunction. In: Neurotrama (Narayan

1249

RK, Wilberger JE, Povlishock JT, eds), New York, NY: McGrawHill, pp 14591478
Hovda DA, Becker DP, Katayama Y (1992) Secondary injury and
acidosis. J Neurotrauma 9 (Suppl): S47S60
Inao S, Marmarou A, Clarke GD, Andersen BJ, Fatouros PP, Young
HF (1988) Production and clearance of lactate from brain tissue,
cerebrospinal fluid, and serum following experimental brain injury. J Neurosurg 69:736744
Inglis F, Kuroda Y, Bullock R (1992) Glucose hypermetabolism after
acute subdural hematoma is ameliorated by a competitive NMDA
antagonist. J Neurotrauma 9:75
Jaggi JL, Obrist WD, Gennarelli TA, Langfitt TW (1990) Relationship
of early cerebral blood flow and metabolism to outcome in acute
head injury. J Neurosurg 72:176182
Jennett B, Bond M (1975) Assessment of outcome after severe brain
damage. A practical scale. Lancet 1:480484
Katayama Y, Becker DP, Tamura T, Hovda DA (1990) Massive increases in extracellular potassium and the indiscriminate release of
glutamate following concussive brain injury. J Neurosurg
73:889900
Katayama Y, Cheung MK, Alves A, et al. (1989) Ion fluxes and cell
swelling in experimental traumatic brain injury: The role of excitatory amino acids. In: Intracranial Pressure. (Hoff JT, Betz AL,
eds), Berlin, Springer-Verlag, pp 584588
Kawamata T, Katayama Y, Hovda DA, Yoshino A, Becker DP (1992)
Administration of excitatory amino acid antagonists via microdialysis attenuates the increase in glucose utilization seen following
concussive brain injury. J Cereb Blood Flow Metab 12:1224
Kawamata T, Katayama Y, Hovda DA, Yoshino A, Becker DP (1995)
Lactate accumulation following concussive brain injury: the role of
ionic fluxes induced by excitatory amino acids. Brain Res
674:196204
Kelly DF, Goodale DB, Williams J, Herr DL, Chappell ET, Rosner MJ,
Jacobson J, Levy ML, Croce MA, Maniker AH, Fulda GJ, Lovett
JV, Mohan O, Narayan RK (1999) Propofol in the treatment of
moderate and severe head injury: a randomized, prospective
double blinded pilot trial. J Neurosurg 90:10421052
Kelly DF, Martin NA, Kordestani R, Counelis G, Hovda DA, Bergsneider M, McBride DQ, Shalmon E, Herman D, Becker DP
(1997) Cerebral blood flow as a predictor of outcome following
traumatic brain injury. J Neurosurgery 86:633641
Kuroda Y, Inglis FM, Miller JD, McCulloch J, Graham DI, Bullock R
(1992) Transient glucose hypermetabolism after acute subdural
hematoma in the rat. J Neurosurg 76:471477
Laird NM, Ware JH (1982) Random-effects models for longitudinal
data. Biometrics 38:963974
Liang KY, Zeger SL (1986) Longitudinal data analysis using generalized linear models. Biometrika 73:1322
Madsen PL, Hasselbalch SG, Hagemann LP, Olsen KS, Bulow J, Holm
S, Wildschiodtz G, Paulson OB, Lassen NA (1995) Persistent resetting of the cerebral oxygen/glucose uptake ratio by brain activation: evidence obtained with the Kety-Schmidt technique. J
Cereb Blood Flow Metab 15:485491
Magistretti PJ, Pellerin L, Rothman DL, Shulman RG (1999) Energy on
demand. Science 283:496497
Marmarou A, Anderson RL, Ward JD, et al. (1991) Impact of ICP
instability and hypotension on outcome in patients with severe
head injury. J Neurosurg 75 (Suppl): S59S66
Marshall LF, Gautille T, Klauber MR, et al. (1991a) The outcome of
severe closed head injury. J Neurosurg 75 (Suppl):S2836
Marshall LF, Marshall SB, Klauber MR, et al. (1991b) A new classification of head injury based on computerized tomography. J Neurosurg 75 (Suppl):S14S20
Martin N, Patwardhan R, Zane C, Alexander M, Benalcazar H, Hovda
DA, Becker DP (1997) Characterization of the cerebral hemodynamic phases which follow severe head trauma: hyperperfusion,
hyperemia and vasospasm. J Neurosurg 87:919
Meixensberger J, Kunze E, Barcsay E, Vaeth A, Roosen K (2001)
Clinical cerebral microdialysis: brain metabolism and brain tissue
oxygenation after acute brain injury. Neurol Res 23:801806
Menzel M, Doppenberg EM, Zauner A, Soukup J, Reinert MM, Bullock R (1999) Increased inspired oxygen concentration as a factor

J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

1250

T. C. GLENN ET AL.

in improved brain tissue oxygenation and tissue lactate levels after

severe human head injury. J Neurosurg 91:110
Muizelaar JP, Marmarou A, DeSalles AA, Ward JD, Zimmerman RS,
Li Z, Choi SC, Young HF (1989) Cerebral blood flow and metabolism in severely head-injured children, part 1: relationship with
GCS score, outcome, ICP, and PVI. J Neurosurg 71:6371
Nemoto EM, Severinghaus JW (1974) Stereospecific permeability of
rat blood brain barrier to lactic acid. Stroke 5:8184
Nilsson P, Hillered L, Ponten U, Ungerstedt U (1990) Changes in
cortical extracellular levels of energy-related metabolites and
amino acids following concussive brain injury in rats. J Cereb
Blood Flow Metab 10:631637
Nilsson P, Hillered L, Olsson Y, Sheardown MJ, Hansen AJ (1993)
Regional changes in interstitial K+ and Ca2+ levels following
cortical compression contusion trauma in rats. J Cereb Blood Flow
Metab 13:183192
Nilsson P, Ronne-Engstrom E, Flink R, Ungerstedt U, Carlson H,
Hillered L (1994) Epileptic seizure activity in the acute phase
following cortical impact trauma in rat. Brain Res 637:227232
Obrist WD, Gennarelli TA, Segawa H, Dolinskas CA, Langfitt TW
(1979) Relation of cerebral blood flow to neurological status and
outcome in head-injured patients. J Neurosurg 51:292300
Obrist WD, Langfitt TW, Jaggi JL, Cruz J, Gennarelli TA (1984)
Cerebral blood flow and metabolism in comatose patients with
acute head injury. Relationship to intracranial hypertension. J Neurosurg 61:241253
Obrist WD, Marion DW (1996) Xenon techniques for CBF measurement in clinical head injury. In: Neurotrauma (Narayan RK, Welberger JE Jr, Povlishock JT, eds) New York, NY: McGraw-Hill, pp
471485
Obrist WD, Marion DW, Aggarwal S, et al. (1993) Time course of
cerebral blood flow and metabolism in comatose patients with
acute head injury. J Cereb Blood Flow Metab 13: S571
Ochiai K, Zhang J, Gong G, Zhang Y, Liu J, Ye Y, Wu X, Liu H,
Murakami Y, Bache RJ, Ugurbil K, From AH (2001) Effects of
augmented delivery of pyruvate on myocardial high-energy phosphate metabolism at high workstate. Am J Physiol Heart Circ
Physiol 281:H1823H1832
Oertel M, Kelly DF, Lee JH, McArthur DL, Glenn TC, Vespa P,
Boscardin WJ, Hovda DA, Martin NA (2002) Efficacy of hyperventilation, blood pressure elevation, and metabolic suppression
therapy in controlling ICP after head injury. J Neurosurg
97:10451053
Okauchi M, Kawai N, Nakamura T, Kawanishi M, Nagao S (2002)
Effects of mild hypothermia and alkalizing agents on brain injuries
in rats with acute subdural hematomas. J Neurotrauma 19:741
751
Reinert M, Hoelper B, Doppenberg E, Zauner A, Bullock R (2000)
Substrate delivery and ionic balance disturbance after severe human head injury. Acta Neurochir 76 (Suppl):439444
Rice AC, Zsoldos R, Chen T, Wilson MS, Alessandri B, Hamm RJ,
Bullock MR (2002) Lactate administration attenuates cognitive
deficits following traumatic brain injury. Brain Res 928:156159
Rivers EP, Paradis NA, Martin GB, Goetting ME, Rosenberg JA,
Smithline HA, Appleton TJ, Nowak RM (1991) Cerebral lactate
uptake during cardiopulmonary resuscitation in humans. J Cereb
Blood Flow Metab 11:479484
Robertson CS, Contant CF, Gokaslan ZL, Naravan RK, Grossman RG
(1992) Cerebral blood flow, arteriovenous oxygen difference, and
outcome in head injured patients. J Neurol Neurosurg Psychiatry
55:594603
Robertson CS, Grossman RG, Goodman JC, Narayan RK (1987) The
predictive value of cerebral anaerobic metabolism with cerebral
infarction after head injury. J Neurosurg 67:361368

J Cereb Blood Flow Metab, Vol. 23, No. 10, 2003

Rockswold SB, Rockswold GL, Vargo JM, Erickson CA, Sutton RL,
Bergman TA, Biros MH (2001) Effects of hyperbaric oxygenation
therapy on cerebral metabolism and intracranial pressure in severely brain injured patients. J Neurosurg 94:40311
Ros J, Pecinska N, Alessandri B, Landolt H, Fillenz M (2001) Lactate
reduces glutamate-induced neurotoxicity in rat cortex. J Neurosci
Res 66:790794
Rosenthal RE, Williams R, Bogaert YE, Getson PR, Fiskum G (1992)
Prevention of postischemic canine neurological injury through potentiation of brain energy metabolism by acetyl-L-carnitine. Stroke
23:13121317
Sheline CT, Behrens MM, Choi DW (2000) Zinc-induced cortical neuronal death: contribution of energy failure attributable to loss of
NAD(+) and inhibition of glycolysis. J Neurosci 20:31393146
Simonsen L, Bulow J, Madsen J (1994) Adipose tissue metabolism in
man determined by vein catheterization and microdialysis techniques. Am J Physiol 266:E357E365
Sunami K, Nakamura T, Ozawa Y, Kubota M, Namba H, Yamaura A
(1989) Hypermetabolic state following experimental head injury.
Neurosurg Rev 12:400-411
Suzuki M, Suzuki M, Sato K, Dohi S, Sato T, Matsuura A, Hiraide A
(2001) Effect of beta-hydroxybutyrate, a cerebral function improving agent, on cerebral hypoxia, anoxia and ischemia in mice and
rats. Jpn J Pharmacol 87:14350
Tabaddor K, Bhushan C, Pevsner PH, Walker AE (1972) Prognostic
value of cerebral blood flow (CBF) and cerebral metabolic rate of
oxygen (CMRO2) in acute head trauma. J Trauma 12:10531055
Tenjin H, Ueda S, Nizukawa N, Inahori Y, Hino A, Yamaki T,
Kuboyama T, Ebisu T, Hirakawa K, Yamashita M, Makahashi H
(1990) Positron emission tomographic studies on cerebral hemodynamics in patients with cerebral contusion. Neurosurgery
26:971979
Verweij BH, Muizelaar JP, Vinas FC, Peterson PL, Xiong Y, Lee CP
(2000) Impaired cerebral mitochondrial function after traumatic
brain injury in humans. J Neurosurg 93:815820
Vespa P, Prins M, Ronne-Engstrom E, Caron M, Shalmon E, Hovda
DA, Martin NA, Becker DP (1998) Increase in extracellular glutamate caused by reduced cerebral perfusion pressure and seizures
after human traumatic brain injury: a microdialysis study. J Neurosurg 89:971982
Vollmer DG, Torner JC, Jane JA (1991) Age and outcome following
traumatic coma: why do older patients fare worse? J Neurosurg
75:S37S49
Xi G, Hua Y, Bhasin RR, Ennis SR, Keep RF, Hoff JT (2001) Mechanisms of edema formation after intracerebral hemorrhage: effects
of extravasated red blood cells on blood flow and blood-brain
barrier integrity. Stroke 32:29322938
Yamaki T, Imahori Y, Ohmori Y, Yoshino E, Hohri T, Ebisu T, Ueda
S (1996) Cerebral hemodynamics and metabolism of severe diffuse
brain injury measured by PET. J Nucl Med 37:11661170
Ying W, Chen Y, Alano CC, Swanson RA (2002) Tricarboxylic Acid
Cycle Substrates Prevent PARP-Mediated Death of Neurons and
Astrocytes. J Cereb Blood Flow Metab 22:774779
Yoshino A, Hovda DA, Katayama Y, Kawamata T, Becker DP (1992)
Hippocampal CA3 lesion prevents postconcussive metabolic dysfunction in CA1. J Cereb Blood Flow Metab 12:9961006
Yoshino A, Hovda DA, Kawamata T, Katayama Y, Becker DP (1991)
Dynamic changes in local cerebral glucose utilization following
cerebral concussion in rats: evidence of a hyper- and subsequent
hypometabolic state. Brain Res 561:106119
Zeger SL, Liang K (1986) Longitudinal data analysis for discrete and
continuous outcomes. Biometrics 42:121130