CLINICAL STUDIES

Mauro Oddo, MD
Department of Neurosurgery,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

Edjah Nduom, MD
Department of Neurosurgery,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

Suzanne Frangos, RN
Department of Neurosurgery,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

Larami MacKenzie, MD
Department of Neurology,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

Isaac Chen, MD
Department of Neurosurgery,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

Eileen Maloney-Wilensky, MSN

Department of Neurosurgery,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

W. Andrew Kofke, MD
Departments of Neurosurgery
and Anesthesiology and Critical Care,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

Joshua M. Levine, MD
Departments of Neurosurgery, Neurology,
and Anesthesiology and Critical Care,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania

Peter D. LeRoux, MD
Department of Neurosurgery,
University of Pennsylvania Medical Center,
Philadelphia, Pennsylvania
Reprint requests:
Peter D. Le Roux, MD,
Department of Neurosurgery,
Clinical Research Division,
University of Pennsylvania Medical Center,
330 S. 9th Street,
Philadelphia, PA 19107.
E-mail: Peter.LeRoux@uphs.upenn.edu
Received, July 30, 2009.
Accepted, February 3, 2010.
Copyright 2010 by the
Congress of Neurological Surgeons

Acute Lung Injury Is an Independent

Risk Factor for Brain Hypoxia After
Severe Traumatic Brain Injury
BACKGROUND: Pulmonary complications are frequently observed after severe traumatic
brain injury (TBI), but little is known about the consequences of lung injury on brain tissue
oxygenation and metabolism.
OBJECTIVE: We examined the association between lung function and brain tissue oxygen tension (PbtO2) in patients with severe TBI.
METHODS: We analyzed data from 78 patients with severe, nonpenetrating TBI who underwent continuous PbtO2 and intracranial pressure monitoring. Acute lung injury was defined
by the presence of pulmonary infiltrates with a PaO2/FiO2 (PF) ratio less than 300 and the
absence of left ventricular failure. A total of 587 simultaneous measurements of PbtO2 and
PF ratio were examined using longitudinal data analysis.
RESULTS: PbtO2 correlated strongly with PaO2 and PF ratio (P < .05) independent of PaCO2,
brain temperature, cerebral perfusion pressure, and hemoglobin. Acute lung injury was
associated with lower PbtO2 (34.6 13.8 mm Hg at PF ratio >300 vs 30.2 10.8 mm Hg [PF
ratio 200-300], 28.9 9.8 mm Hg [PF ratio 100-199], and 21.1 7.4 mm Hg [PF ratio <100],
all P values <.01). After adjusting for intracranial pressure, Marshall computed tomography score, and APACHE II (Acute Physiology and Chronic Health Evaluation) score, acute
lung injury was an independent risk factor for compromised PbtO2 (PbtO2 <20 mm Hg;
adjusted odds ratio: 2.13, 95% confidence interval: 1.21-3.77; P < .01).
CONCLUSION: After severe TBI, PbtO2 correlates with PF ratio. Acute lung injury is associated with an increased risk of compromised PbtO2, independent from intracerebral and
systemic injuries. Our findings support the use of lung-protective strategies to prevent
brain hypoxia in TBI patients.
KEY WORDS: Acute lung injury, Brain hypoxia, Brain tissue oxygen tension, Lung function, Systemic oxygenation, Traumatic brain injury
Neurosurgery 67:338-344, 2010

DOI: 10.1227/01.NEU.0000371979.48809.D9

ulmonary complications occur frequently

in patients with severe traumatic brain injury
(TBI) and are associated with poor outcome.1-4 Management of mechanical ventilation
is therefore central after severe TBI. In critically ill
patients without primary brain injury, strategies
for mechanical ventilation have evolved over the
ABBREVIATIONS: ABG, arterial blood gas; APACHE II,
Acute Physiology and Chronic Health Evaluation; BT,
brain temperature; CPP, cerebral perfusion pressure;
CT, computed tomography; ICP, intracranial pressure;
ICU, intensive care unit; PbtO2, brain tissue oxygen
tension; PEEP, positive end-expiratory pressure; PF,
PaO2/FiO2; TBI, traumatic brain injury

338 | VOLUME 67 | NUMBER 2 | AUGUST 2010

www.neurosurgery- online.com

past decade from standard ventilation with maintenance of normal blood gas values and high tidal
volumes to protective ventilation with the use of
low tidal volumes and positive end-expiratory
pressure (PEEP) (lung-protective ventilation)5 in
an effort to reduce ventilator-induced lung injury.6
In patients with brain injury, lung-protective ventilation traditionally was avoided because of the risk
of increasing intracranial pressure (ICP) and reducing cerebral perfusion pressure (CPP).7,8 Recent
studies, however, have shown that PEEP may have
no negative effect on ICP and CPP,9,10 suggesting that lung-protective ventilation may be warranted after severe TBI.11 Efforts to improve lung
function and optimize systemic oxygenation may

www.neurosurgery-online.com

LUNG FUNCTION AND BRAIN TISSUE OXYGEN TENSION

be a key factor in reducing secondary brain hypoxic insults in

patients with severe TBI. Conversely, reduced systemic oxygenation
may negatively affect brain tissue oxygenation. Despite the high
frequency and the adverse impact of pulmonary complications on
outcome, the exact relationship between brain tissue oxygen tension (PbtO2) and lung function is only beginning to be elucidated
in humans with TBI.12
Continuous monitoring of PbtO2 allows for the quantification
of cerebral hypoxic events that occur after severe brain injury.13 Brain
hypoxia is a major determinant of secondary brain damage14 and
is associated with poor outcome15-18 after severe TBI. The goals
of monitoring brain-injured patients are to detect harmful physiological events before they cause irreversible damage to the brain,
to allow for diagnosis and treatment of secondary injury, and to
provide on-line feedback regarding efficacy of therapy. A PbtO2
monitor used to detect brain hypoxia may help optimize the management of patients with severe TBI.19,20
This study of a large cohort of patients who underwent PbtO2
monitoring after severe nonpenetrating TBI analyzes the relationship between brain tissue and systemic oxygenation and examines whether acute lung injury was an independent risk factor for
compromised brain oxygenation.

PATIENTS AND METHODS

Patients
Approval for the study was obtained from the University of Pennsylvania
Medical Center Institutional Review Board. Patients with severe nonpenetrating TBI admitted to the Hospital of the University of Pennsylvania,
a level I trauma center, and who underwent ICP and PbtO2 monitoring
were studied as part of a prospective observational database (the Brain
Oxygen Monitoring Outcome database). Patients were monitored if their
admission Glasgow Coma Scale score was 8 or lower or they later deteriorated to that level. At admission, the Acute Physiology and Chronic
Health Evaluation (APACHE II) score and Marshall score based on the initial head computed tomography (CT) scan21 were calculated for all patients.
For the purpose of this study, patients who had PbtO2 monitoring for at
least 24 hours in the neurointensive care unit and more than 2 arterial
blood gas (ABG) samples were retrospectively identified. Patients with
left ventricular failure or a history of chronic lung disease were excluded.

Intracranial Monitoring
ICP (Camino, Integra Neurosciences, Plainsboro, New Jersey), brain
temperature (BT), PbtO2 (LICOX, Integra Neuroscience), and intraarterial blood pressure were monitored continuously. CPP was calculated
(CPP = mean arterial pressure ICP). Intraparenchymal probes (ICP,
BT, and PbtO2) were inserted at the bedside in the neurointensive care
unit through a burr hole in the frontal lobe and secured with a triple
lumen bolt. The monitors were placed in white matter that appeared
normal on admission head CT and on the side of maximal pathology.
When there was no asymmetry in brain pathology on CT, the probes
were placed in the right frontal region. If the patient had undergone a
craniotomy, the probes were placed on the same side as the injury if the
craniotomy flap permitted. Follow-up head CT scans were performed
in all patients within 24 hours of admission to confirm correct placement of the various monitors (eg, not in a contusion, hemorrhage, or

NEUROSURGERY

infarct). Probe function and stability were confirmed by an appropriate

PbtO2 increase after an oxygen challenge (FiO2 1.0 for 5 minutes). To
allow for probe equilibration, data from the first 6 hours after PbtO2
monitor insertion were discarded. ICP and PbtO2 monitors were removed
once the ICP was normal without treatment for more than 24 hours or
care was withdrawn because of injury severity.

General Clinical Management

All patients were managed in the neurointensive care unit according to
a local algorithm consistent with the Brain Trauma Foundation TBI
Guidelines.22 This included early evacuation of space-occupying mass
lesions in the operating room. Each patient was fully resuscitated according to Advanced Trauma Life Support guidelines from the American
College of Surgeons, intubated, and mechanically ventilated with the head
of bed initially elevated approximately 30 degrees. Fio2 and minute ventilation were adjusted to maintain SaO2 at more than 93% and to avoid
PaO2 less than 60 mm Hg and PbtO2 less than 20 mm Hg. PaCO2 was
set at approximately 35 to 45 mm Hg unless the ICP was elevated when
PaCO2 was maintained between 30 and 40 mm Hg. Volume resuscitation
was achieved with 0.9% normal saline. Therapeutic targets were adjusted
to avoid ICP more than 20 mm Hg and CPP less than 60 mm Hg. After
adequate fluid resuscitation, phenylephrine (10-100 g/min) was administered when CPP was 60 mm Hg or less and ICP was normal. A standard stair-step approach was used to treat intracranial hypertension (ICP
>20 mm Hg). Initial management consisted of head of bed elevation,
sedation (lorazepam), analgesia (fentanyl), intermittent cerebrospinal fluid
drainage using an external ventricular drain if inserted, and moderate
hyperventilation (to avoid PaCO2 <30 mm Hg and PbtO2 <20 mm Hg).
If ICP remained higher than 20 mm Hg for more than 10 minutes despite
the initial management, osmotherapy was started, provided that serum
osmolarity was less than 320 mOsm/L and serum sodium was less than 155
mmol/L. Osmotherapy with mannitol was more frequently administered
as first-line treatment of intracranial hypertension, whereas hypertonic
saline was used as a second-line therapy of recurrent episodes of sustained
elevated ICP when mannitol did not control ICP. Second-tier therapies
included optimized hyperventilation, decompressive craniectomy, and
additional sedation with propofol or barbiturates.

Management of PbtO2
All patients received cause-directed therapy to maintain PbtO2 at 20
mm Hg or higher in keeping with current National Institutes of Healthfunded trials that examine PbtO2-based care and consistent with previous
studies that examine normal values in humans23 and that suggest 20 mm
Hg is a break point for an unfavorable outcome.24 In our neurointensive
care unit (neuro-ICU), more than 25 therapies are used to correct PbtO2,25
but the following basic protocol is followed. When PbtO2 was low in the
setting of intracranial hypertension (ICP >20 mm Hg), measures were
taken to lower ICP as described above. If ICP less than 20 mm Hg or lowering ICP failed to increase PbtO2, then CPP was increased (usually with
phenylephrine), but if this did not increase PbtO2, nicardipine to correct
hyperemia was considered. If the cause of low PbtO2 was systemic hypoxia,
then pulmonary function was optimized with the mechanical ventilator
(eg, by increasing FiO2 and/or PEEP). If excess metabolic demand was suspected (eg, caused by pain, agitation, fever, or seizures), then analgesic,
sedative, or antiepileptic medications were administered. If these measures
failed and hemoglobin was less than 10 mg/dL, then a blood transfusion
was administered. A decompressive craniectomy was performed if there
was a progressive PbtO2 decline or PbtO2 was 20 mm Hg for longer than
15 minutes despite maximal medical management for elevated ICP.

VOLUME 67 | NUMBER 2 | AUGUST 2010 | 339

ODDO ET AL

General Ventilation Strategies

All patients were intubated and mechanically ventilated based on standard practice in our neuro-ICU and consistent with published data.22,26
Patients received volume-limited assist-control ventilation using a tidal
volume of 8 mL/kg of predicted body weight. Initial PEEP was set at 5
cm H2O. Thereafter, tidal volume, PEEP, and FiO2 were adjusted in
increments as needed to a maximum PEEP of 15 cm H2O. Plateau pressures were maintained at less than 35 cm H2O.

Lung Function and ABG Analysis

Each patient had an indwelling arterial catheter. ABG analysis was performed at approximately 8:00 AM and 8:00 PM each day while the patient
was ventilated and if there was any significant cardiopulmonary change or
at the discretion of the neurointensivist. The arterial samples were analyzed for PaO2, PaCO2, SaO2, and pH and FiO2, SpO2, respiratory rate,
and ventilator settings (ie, ventilatory mode, tidal volume, minute ventilation, and PEEP) were recorded in the neuro-ICU flowsheet every 15 to
30 minutes. PaO2/FiO2 (PF) ratio was calculated and used to assess lung
function. Acute lung injury was defined by the presence of a PF ratio less
than 300, regardless of the PEEP level, and absence of left ventricular dysfunction and chronic lung disease.26 Results of chest radiographic examinations were retrieved retrospectively from patient electronic charts.

Other Cerebral and Systemic Variables

Mean arterial pressure, ICP, CPP, PbtO2, and BT were recorded continuously using a bedside monitor (Component Monitoring System
M1046-9090C; Hewlett Packard, Andover, Massachusetts) and recorded
in the neuro-ICU flowsheet at least every 30 minutes. Hemoglobin concentration was measured when an ABG was obtained in each patient.

Statistical Analysis
Univariate associations between the PF ratio (categorized into 4 separate ranges, ie, >300, 200-300, 100-199, <100) and PbtO2 were first analyzed with analysis of variance for repeated measures. Because of repeated
measurements of each variable among different subjects at different time
points, a secondary analysis was performed to examine independent associations between PbtO2 and PF ratio, PAO2, and SaO2: a multivariate
model for longitudinal/repeated measures (generalized estimating equations) was used, and within-subject and between-subject dependencies
over time were modeled with an autoregressive correlation structure.27,28
Correlations between PbtO2 and PF ratio, PaO2, and SaO2 all were adjusted
for PaCO2, CPP, BT, and hemoglobin concentration, and all variables were
entered as continuous variables. Compromised PbtO2 was defined as a
PbtO 2 <20 mm Hg. 23-25 Association between acute lung injury
(dichotomized at PF ratio <300 or >300) and compromised PbtO2
(dichotomized at PbtO2 <20 or >20 mm Hg) were analyzed and adjusted
for the severity of intracerebral (ICP, Marshall CT score) and systemic
(APACHE II score) injuries. SPSS 15.0 Software (SPSS Inc, Chicago,
Illinois) was used. A P value <.05 was considered statistically significant.

RESULTS
Patient Clinical Characteristics
From January 2002 to August 2006, 78 patients who were monitored within the first 24 hours after initial brain insult and studied for an average of 4 3 days were examined. Patient clinical
and neuroradiological characteristics are listed in Table 1.

340 | VOLUME 67 | NUMBER 2 | AUGUST 2010

TABLE 1. Patient Characteristics

Characteristics

Value

No. of patients

78

Age, y, mean standard deviation

44 20

Female sex, no. (%)

17 (22)

Median (range) admission Glasgow Coma

Scale score

5 (3-11)

Marshall computed tomography score, no. (%)

2

17 (22)

22 (28)

5 (6)

34 (44)

Median (range) APACHE II score

20 (8-31)

Outcome at 30 days, no. (%)

Favorable (Glasgow Outcome Score 4-5)

43 (55)

Unfavorable (Glasgow Outcome Score 1-3)

35 (45)

Physiological Variables
At the time of ABG analysis, the mean PF ratio was 288 143,
the mean PaO2 was 171 89 mm Hg, the mean PaCO2 was 36
7 mm Hg, and the mean PEEP was 7 3 cm H2O. The mean
PbtO2 was 31.1 12.5 mm Hg, the mean ICP was 17 9 mm Hg,
and the mean CPP was 76 14 mm Hg. The mean hemoglobin
concentration was 10.3 1.6 g/dL. An episode of compromised
PbtO2 was observed in 57 patients (73%), and, on average, 22%
of PbtO2 samples per patient were less than 20 mm Hg. An unfavorable outcome was associated with compromised PbtO2; the
proportion of samples with PbtO2 levels less than 20 mm Hg was
23 25% in patients with an unfavorable outcome vs 8 13% in
patients with a favorable outcome (P < .001).
Correlation Between PF Ratio and PbtO2
A total of 587 simultaneous measurements of PF ratio and
PbtO2 were analyzed. Compared with a PF ratio of more than
300 (274 episodes; 31 patients), lower PF ratio ranges analyzed
were associated with significantly reduced PbtO2 values: 34.6 13.8
mm Hg vs 30.2 10.8 mm Hg at a PF ratio of 200 to 300 (n =
128; 47 patients), 28.9 9.8 mm Hg at a PF ratio of 100 to 199
(n = 130; 34 patients), and 21.1 7.4 mm Hg when the PF ratio
was less than 100 (n = 55; 14 patients) (Figure). We used generalized estimating equations to further confirm correlations between
brain and systemic oxygenation. PbtO2, PF ratio, PaO2, and SaO2
were entered in the model as continuous variables, and associations were adjusted for other important physiological variables
that may influence brain and systemic oxygenation, including
PaCO2, CPP, BT, and hemoglobin. This analysis allowed us to
demonstrate a strong and independent linear correlation between
PbtO2 and PF ratio (adjusted P < .01), PaO2 (adjusted P < .01),
and SaO2 (adjusted P = .03) (Table 2).

www.neurosurgery-online.com

LUNG FUNCTION AND BRAIN TISSUE OXYGEN TENSION

TABLE 3. Acute Lung Injury Is an Independent Risk Factor of

Compromised Brain Tissue Oxygen Tensiona
Adjusted OR
(95% Wald CI)

Adjusted
P Value

Acute lung injury

2.13 (1.21-3.77)

<.01

Intracranial pressure

1.36 (1.07-1.66)

.01

APACHE II

1.09 (1.0-1.19)

.03

Variable

OR, odds ratio; CI, confidence interval; APACHE II, Acute Physiology and Chronic
Health Evaluation.
Independent associations were determined with generalized estimating equations
analysis, using longitudinal logistic regression link function to account for withinsubject and between-subject dependencies over time. Acute lung injury was
defined by the presence of a PaO2/FiO2 ratio less than 300, regardless of positive endexpiratory pressure level, and absence of left ventricular dysfunction and chronic
lung disease. Compromised brain tissue oxygen tension was defined as brain tissue
oxygen tension less than 20 mm Hg.
a

FIGURE. Histograms of mean (standard deviation) brain tissue oxygen ten-

sion (PbtO2) values according to the PaO2/FiO2 ratio range. **P < .01 for
comparisons with samples obtained at PaO2/FiO2 ratio >300.

TABLE 2. Correlations Between Brain Tissue Oxygen Tension

and Systemic Oxygenationa
Variable

Adjusted P Value

PaO2/FiO2 ratio

<.01

PaO2

<.01

SaO2

.03

a
Correlations between continuous variables were examined using longitudinal data
analysis (generalized estimating equations), and final P values were adjusted for
important physiological covariates that can influence brain tissue oxygen tension,
including PaCO2, cerebral perfusion pressure, brain temperature, and hemoglobin.

Acute Lung Injury Is an Independent Risk Factor

for Compromised PbtO2
Acute lung injury was observed in 47 of 78 patients (60%). The
main cause of acute lung injury was pneumonia (62% of patients),
followed by atelectasis (19%), pulmonary contusion (9%), and
pleural effusion (2%). A higher percentage of patients with acute
lung injury had compromised PbtO2 (85% vs 54% in patients
with a PF ratio >300, P < .01). Also, over the entire duration of
intracranial monitoring, the proportion of PbtO2 samples less than
20 mm Hg was greater among patients with acute lung injury than
those without (22% vs 12%, P = .07). After adjusting for ICP,
APACHE II score, and the Marshall CT score, a PF ratio less than
300 was an independent risk factor of compromised PbtO2 in our
cohort of severe TBI patients (adjusted odds ratio: 2.13, 95% confidence interval, 1.21-3.77, P = .009; Table 3).

DISCUSSION
We examined the association between lung function and PbtO2
in 78 patients with severe TBI. The main findings can be summa-

NEUROSURGERY

rized as follows: (1) after severe TBI, acute lung injury is frequent
and may cause significant reduction of systemic oxygenation; (2)
PbtO2 correlates strongly with systemic oxygenation, independent of other physiological parameters, including PaCO2, CPP, BT,
and hemoglobin; (3) acute lung injury is associated with lower
PbtO2 levels and is an independent risk factor for compromised
PbtO2, irrespective of intracerebral and systemic injuries. Our data
suggest that lung-protective strategies aimed at improving systemic
oxygenation may help to reduce secondary brain hypoxic insults and
optimize the management of patients with severe TBI.
The Lung and TBI
Pulmonary complications occur frequently in patients with
severe brain injury and are associated with a poor outcome.1,2,4,29
Acute lung injury after severe TBI is usually attributed to pulmonary contusions, increased sympathetic activity,30 acute systemic inflammatory response, 31 or pneumonia. 3 Ventilator
management of TBI patients was, until recently, based on the use
of high minute volume to maintain mild hypocapnia (PaCO2 3438 mm Hg) to treat intracranial hypertension and with low levels of PEEP to optimize oxygenation while preserving cerebral
venous drainage.
Mechanical ventilation is of enormous value; however, ventilation itself can damage the lungs and high tidal volume ventilation can contribute to or exacerbate lung injury.32,33 On the
other hand, TBI management (eg, fluid-based CPP augmentation) may aggravate lung injury in some patients.2 Concepts of
ventilator-induced lung injury have revolutionized the approach
to the ventilatory management of general neuro-ICU patients
with acute lung injury over the past 10 years.34 Whether TBI
patients benefit from lung-protective ventilation strategies is
unclear because these patients were excluded from clinical trials because of the tight CO2 control required in TBI management.5,35,36 Although the effects of PEEP or hyperventilation
on the brain have been examined,10,37,38 there has been less study
of how lung function influences intracranial physiology, and in

VOLUME 67 | NUMBER 2 | AUGUST 2010 | 341

ODDO ET AL

particular PbtO2, in clinical TBI.12,39,40 These studies emphasize the importance of understanding how pulmonary function
and its management influences PbtO2 and metabolism in TBI.
This is important because retrospective studies suggest that brain
hypoxia is associated with poor outcome after TBI.15-18 Pulmonary
function was not examined in these observational outcome studies. Our data suggest that an important factor in the likelihood
of the development of reduced PbtO2 may be the presence of
pulmonary dysfunction and in particular the development of
acute lung injury.
Correlation Between Lung Function and PbtO2
Correcting compromised PbtO2 requires that the specific cause
of reduced PbtO2 be defined. There are many potential causes of
reduced PbtO2. However, most studies to date have examined the
role of cerebral and hemodynamic factors and the relationship
between lung function, and PbtO2 in TBI patients is only beginning to be elucidated.12,41 We undertook this study to examine
the relationship between lung function and PbtO2 and found that
PbtO2 was strongly correlated with systemic oxygenation. In using
a longitudinal data analysis that accounted for within-subject and
between-subject variations, we found that the relationship between
brain and systemic oxygenation was independent of other important physiological covariates that may influence PbtO2 (ie, PaCO2,
CPP, BT, and hemoglobin). Accordingly, reduced PF ratio levels
were associated with lower PbtO2 in our study cohort, and we
identified acute lung injury as an independent risk factor for compromised PbtO2, irrespective of ICP, Marshall CT score, and
APACHE II score. These findings are consistent with recent observations by Rosenthal et al12 who performed an oxygen challenge
(FiO2 100%) in 37 TBI patients and found that patients with
higher PF ratios achieved a higher PbtO2 during oxygen challenge
than those with a low PF ratio. Others have shown that the rate,
or pattern, of PbtO2 increase in response to an oxygen challenge
(FiO2 increase) was associated with outcome after TBI.42 Together
these clinical observations suggest a relationship between systemic
and brain tissue oxygenation. An important implication of our
study is that that reduced lung function, as assessed by the PF
ratio, is per se a risk factor for compromised PbtO2, independent of other important clinical and physiological determinants.
Although PbtO2 is influenced by factors that regulate cerebral
blood flow, and in particular CO2 and mean arterial pressure,43,44
a PbtO2 monitor is not simply an ischemia monitor or a monitor
of blood flow.45-47 Instead, it likely is a marker of the balance
between regional oxygen supply and cellular oxygen consumption. Furthermore, they imply that strategies to improve PbtO2
include efforts to maximize lung function and gas exchange.
Conceivably one strategy may include efforts to keep a stable PaO2
at approximately 100 mm Hg. This may then remove the variability in lung function from influencing continuous PbtO2 measurements, and then the PF ratio could be used separately to monitor
lung function. It should be recognized, however, that although
prolonged overoxygenation may help some patients, it can, in
turn, also be toxic to the lungs.

342 | VOLUME 67 | NUMBER 2 | AUGUST 2010

Study Limitations
Our study has several potential limitations. First, the data were
examined retrospectively, and this may bias our results. However,
the patients were treated according to a standard clinical practice
guideline, monitoring was started after a median of 6 hours after
TBI, and the data were entered prospectively. Second, the study
was performed on patients treated at a single institution, so it may
lack external validity. Third, the sample size of 78 patients is relatively small, although we examined a large number (587 simultaneous PF ratio and PbtO2 samples from different patients over
a mean duration of 4 days after injury. Multiple samples from a single patient can skew our results, and because ABGs were obtained
at various time points after TBI, this may influence cerebral blood
flow and therefore PbtO2. However, univariate associations found
between PbtO2 and PF ratio were further examined with longitudinal data analysis (generalized estimating equations), taking into
account within-subject and between-subject variations over time
and after controlling for other important covariates. We believe
that the statistical analysis used for this study appears to strengthen
our main findings. Fourth, we assessed lung function with PF
ratio only, and we did not use other scores such as the lung injury
score.48 However, the PF ratio has been demonstrated prospectively in trauma patients to accurately predict acute lung injury
severity and mortality and may be superior to other lung injury scores
in this context.49 In addition, acute lung injury was assessed with
chest radiographs alone, and no further radiological investigations
(ie, CT scan) were available to better characterize the particular type
of acute lung injury. Fifth, because we do not have information on
cerebral blood flow, oxygen extraction fraction, or cerebral metabolic rate of oxygen consumption, our data do not allow us to
explain how or why the PF ratio interacts with PbtO2. We cannot
be sure that the relationship between systemic and brain tissue
oxygenation remains valid at any cerebral blood flow or cerebral
metabolic rate of oxygen consumption level; this will need further prospective evaluation. Finally, we did not examine how lung
function responded to therapy, and it remains to be determined
whether a therapeutic strategy aimed at preventing and treating brain
tissue hypoxia may eventually improve functional outcome after
TBI, as suggested by some preliminary studies.20,50 Although
many factors can influence PbtO2, our data emphasize that lungprotective strategies are an essential part of severe TBI management and underlines the need to incorporate lung function and
PF ratio into a PbtO2-guided approach to reduce secondary brain
hypoxic insults in patients with severe TBI.

CONCLUSIONS
We found that PbtO2 was strongly correlated with systemic
oxygenation in patients with severe TBI. Acute lung injury was
associated with significantly lower PbtO2 and was an independent risk factor for compromised PbtO2. These data suggest that lungprotective strategies aimed at improving systemic oxygenation
and lung function are needed to limit secondary brain hypoxic
insults and optimize the management of patients with severe TBI.

www.neurosurgery-online.com

LUNG FUNCTION AND BRAIN TISSUE OXYGEN TENSION

Disclosure
Supported by research grants from the SICPA Foundation, Switzerland (M.O.),
the Integra Foundation (P.D.L.R.), Integra Neurosciences (PDLR), and the Mary
Elisabeth Groff Surgical and Medical Research Trust (P.D.L.R.). Dr. Le Roux is a
member of Integra Neurosciences Speakers Bureau. The authors have no personal
financial or institutional interest in any of the drugs, materials, or devices described
in this article.

REFERENCES
1. Bratton SL, Davis RL. Acute lung injury in isolated traumatic brain injury.
Neurosurgery. 1997;40(4):707-712.
2. Contant CF, Valadka AB, Gopinath SP, Hannay HJ, Robertson CS. Adult respiratory distress syndrome: a complication of induced hypertension after severe head
injury. J Neurosurg. 2001;95(4):560-568.
3. Bronchard R, Albaladejo P, Brezac G, et al. Early onset pneumonia: risk factors
and consequences in head trauma patients. Anesthesiology. 2004;100(2):234-239.
4. Mascia L, Mastromauro I, Viberti S. High tidal volume as a predictor of acute lung
injury in neurotrauma patients. Minerva Anestesiol. 2008;74(6):325-327.
5. Ventilation with lower tidal volumes as compared with traditional tidal volumes
for acute lung injury and the acute respiratory distress syndrome. The Acute
Respiratory Distress Syndrome Network. N Engl J Med. 2000;342(18):13011308.
6. Pinhu L, Whitehead T, Evans T, Griffiths M. Ventilator-associated lung injury.
Lancet. 2003;361(9354):332-340.
7. Shapiro HM, Marshall LF. Intracranial pressure responses to PEEP in head-injured
patients. J Trauma. 1978;18(4):254-256.
8. Burchiel KJ, Steege TD, Wyler AR. Intracranial pressure changes in brain-injured
patients requiring positive end-expiratory pressure ventilation. Neurosurgery.
1981;8(4):443-449.
9. Huynh T, Messer M, Sing RF, Miles W, Jacobs DG, Thomason MH. Positive endexpiratory pressure alters intracranial and cerebral perfusion pressure in severe traumatic brain injury. J Trauma. 2002;53(3):488-492; discussion: 492-483.
10. Mascia L, Grasso S, Fiore T, Bruno F, Berardino M, Ducati A. Cerebro-pulmonary
interactions during the application of low levels of positive end-expiratory pressure. Intensive Care Med. 2005;31(3):373-379.
11. Lowe GJ, Ferguson ND. Lung-protective ventilation in neurosurgical patients.
Curr Opin Crit Care. 2006;12(1):3-7.
12. Rosenthal G, Hemphill JC, Sorani M, et al. The role of lung function in brain tissue oxygenation following traumatic brain injury. J Neurosurg. 2008;108(1):5965.
13. Nortje J, Gupta AK. The role of tissue oxygen monitoring in patients with acute
brain injury. Br J Anaesth. 2006;97(1):95-106.
14. Coles JP. Regional ischemia after head injury. Curr Opin Crit Care. 2004;10(2):120125.
15. van Santbrink H, Maas AI, Avezaat CJ. Continuous monitoring of partial pressure of brain tissue oxygen in patients with severe head injury. Neurosurgery.
1996;38(1):21-31.
16. Dings J, Jager A, Meixensberger J, Roosen K. Brain tissue pO2 and outcome after
severe head injury. Neurol Res. 1998;20 suppl 1:S71-S75.
17. Valadka AB, Gopinath SP, Contant CF, Uzura M, Robertson CS. Relationship of
brain tissue PO2 to outcome after severe head injury. Crit Care Med. 1998;26(9):15761581.
18. van den Brink WA, van Santbrink H, Steyerberg EW, et al. Brain oxygen tension
in severe head injury. Neurosurgery. 2000;46(4):868-876; discussion: 876-868.
19. Bratton SL, Chestnut RM, Ghajar J, et al. Guidelines for the management of severe
traumatic brain injury. IX. Cerebral perfusion thresholds. J Neurotrauma.
2007;24(suppl 1):S59-S64.
20. Stiefel MF, Spiotta A, Gracias VH, et al. Reduced mortality rate in patients with
severe traumatic brain injury treated with brain tissue oxygen monitoring. J Neurosurg.
2005;103(5):805-811.
21. Marshall LF, Marshall SB, Klauber MR, et al. The diagnosis of head injury requires
a classification based on computed axial tomography. J Neurotrauma. 1992;9(suppl
1):S287-S292.
22. Brain Trauma Foundation AAoNS, Joint Section on Neurotrauma and Critical
Care. Management and prognosis of severe traumatic brain injury. J Neurotrauma.
2000;17(6-7):449-597.

NEUROSURGERY

23. Pennings FA, Schuurman PR, van den Munckhof P, Bouma GJ. Brain tissue oxygen pressure monitoring in awake patients during functional neurosurgery: the
assessment of normal values. J Neurotrauma. 2008;25(10):1173-1177.
24. Chang JJ, Youn TS, Benson D, et al. Physiological and functional outcome correlates of brain tissue hypoxia in traumatic brain injury. Crit Care Med. 2009;37(1):283
to 290.
25. Spiotta A, Stiefel M, Gracias V, et al. Brain tissue oxygen directed management
and outcome in patients with severe traumatic brain injury. J Neurosurg. 2010 Apr
23. [Epub ahead of print].
26. Bernard GR, Artigas A, Brigham KL, et al. The American-European Consensus
Conference on ARDS. Definitions, mechanisms, relevant outcomes, and clinical
trial coordination. Am J Respir Crit Care Med. 1994;149(3 pt 1):818-824.
27. Zeger SL, Liang KY. An overview of methods for the analysis of longitudinal data.
Stat Med. 1992;11(14-15):1825-1839.
28. Zeger SL, Liang KY, Albert PS. Models for longitudinal data: a generalized estimating equation approach. Biometrics. 1988;44(4):1049-1060.
29. Holland MC, Mackersie RC, Morabito D, et al. The development of acute lung injury
is associated with worse neurological outcome in patients with severe traumatic
brain injury. J Trauma. 2003;55(1):106-111.
30. Rogers FB, Shackford SR, Trevisani GT, Davis JW, Mackersie RC, Hoyt DB.
Neurogenic pulmonary edema in fatal and nonfatal head injuries. J Trauma. 1995;
39(5):860-866; discussion: 866-868.
31. Yildirim E, Kaptanoglu E, Ozisik K, et al. Ultrastructural changes in pneumocyte
type II cells following traumatic brain injury in rats. Eur J Cardiothorac Surg.
2004;25(4):523-529.
32. Gajic O, Dara SI, Mendez JL, et al. Ventilator-associated lung injury in patients without acute lung injury at the onset of mechanical ventilation. Crit Care Med.
2004;32(9):1817-1824.
33. Mascia L, Zavala E, Bosma K, et al. High tidal volume is associated with the development of acute lung injury after severe brain injury: an international observational study. Crit Care Med. 2007;35(8):1815-1820.
34. Tobin MJ. Culmination of an era in research on the acute respiratory distress syndrome. N Engl J Med. 2000;342(18):1360-1361.
35. Brochard L, Roudot-Thoraval F, Roupie E, et al. Tidal volume reduction for prevention of ventilator-induced lung injury in acute respiratory distress syndrome. The
Multicenter Trail Group on Tidal Volume reduction in ARDS. Am J Respir Crit
Care Med. 1998;158(6):1831-1838.
36. Ranieri VM, Suter PM, Tortorella C, et al. Effect of mechanical ventilation on
inflammatory mediators in patients with acute respiratory distress syndrome: a
randomized controlled trial. JAMA. 1999;282(1):54-61.
37. Crossley DJ, McGuire GP, Barrow PM, Houston PL. Influence of inspired oxygen concentration on deadspace, respiratory drive, and PaCO2 in intubated
patients with chronic obstructive pulmonary disease. Crit Care Med. 1997;25(9):15221526.
38. Coles JP, Minhas PS, Fryer TD, et al. Effect of hyperventilation on cerebral blood
flow in traumatic head injury: clinical relevance and monitoring correlates. Crit
Care Med. 2002;30(9):1950-1959.
39. Reinert M, Barth A, Rothen HU, Schaller B, Takala J, Seiler RW. Effects of cerebral perfusion pressure and increased fraction of inspired oxygen on brain tissue
oxygen, lactate and glucose in patients with severe head injury. Acta Neurochir
(Wien). 2003;145(5):341-349; discussion: 349-350.
40. Kalsotra A, Zhao J, Anakk S, Dash PK, Strobel HW. Brain trauma leads to enhanced
lung inflammation and injury: evidence for role of P4504Fs in resolution. J Cereb
Blood Flow Metab. 2007;27(5):963-974.
41. Hlatky R, Valadka AB, Gopinath SP, Robertson CS. Brain tissue oxygen tension
response to induced hyperoxia reduced in hypoperfused brain. J Neurosurg. 2008;
108(1):53-58.
42. van Santbrink H, vd Brink WA, Steyerberg EW, Carmona Suazo JA, Avezaat CJ,
Maas AI. Brain tissue oxygen response in severe traumatic brain injury. Acta Neurochir
(Wien). 2003;145(6):429-438; discussion: 438.
43. Hemphill JC 3rd, Knudson MM, Derugin N, Morabito D, Manley GT. Carbon
dioxide reactivity and pressure autoregulation of brain tissue oxygen. Neurosurgery.
2001;48(2):377-383; discussion: 383-374.
44. Hemphill JC 3rd, Smith WS, Sonne DC, Morabito D, Manley GT. Relationship
between brain tissue oxygen tension and CT perfusion: feasibility and initial results.
AJNR Am J Neuroradiol. 2005;26(5):1095-1100.

VOLUME 67 | NUMBER 2 | AUGUST 2010 | 343

ODDO ET AL

45. Oddo M, Le Roux P. Brain tissue oxygen monitors: more than an ischemia monitor. Crit Care Med. 2008;36(6):1984-1985.
46. Rosenthal G, Hemphill JC 3rd, Sorani M, et al. Brain tissue oxygen tension is
more indicative of oxygen diffusion than oxygen delivery and metabolism in patients
with traumatic brain injury. Crit Care Med. 2008;36(6):1917-1924.
47. Scheufler KM, Rohrborn HJ, Zentner J. Does tissue oxygen-tension reliably reflect
cerebral oxygen delivery and consumption? Anesth Analg. 2002;95(4):1042-1048,
table of contents.

344 | VOLUME 67 | NUMBER 2 | AUGUST 2010

48. Murray JF, Matthay MA, Luce JM, Flick MR. An expanded definition of the adult
respiratory distress syndrome. Am Rev Respir Dis. 1988;138(3):720-723.
49. Offner PJ, Moore EE. Lung injury severity scoring in the era of lung protective mechanical ventilation: the PaO2/FIO2 ratio. J Trauma. 2003;55(2):285289.
50. Meixensberger J, Jaeger M, Vath A, Dings J, Kunze E, Roosen K. Brain tissue oxygen guided treatment supplementing ICP/CPP therapy after traumatic brain injury.
J Neurol Neurosurg Psychiatry. 2003;74(6):760-764.

www.neurosurgery-online.com