Neurogenic Pulmonary Edema Due to Traumatic Brain Injury: Evidence of

Cardiac Dysfunction
Mabrouk Bahloul, Anis N. Chaari, Hatem Kallel, Abdelmajid Khabir, Adnne Ayadi, Hanne
Charfeddine, Leila Hergafi, Adel D. Chaari, Hedi E. Chelly, Chokri Ben Hamida, Noureddine
Rekik and Mounir Bouaziz
Am J Crit Care 2006;15:462-470
2006 American Association of Critical-Care Nurses
Published online http://www.ajcconline.org

Personal use only. For copyright permission information:

http://ajcc.aacnjournals.org/cgi/external_ref?link_type=PERMISSIONDIRECT

Subscription Information
http://ajcc.aacnjournals.org/subscriptions/

Information for authors

http://ajcc.aacnjournals.org/misc/ifora.xhtml

Submit a manuscript
http://www.editorialmanager.com/ajcc

Email alerts
http://ajcc.aacnjournals.org/subscriptions/etoc.xhtml

AJCC, the American Journal of Critical Care, is the official peer-reviewed research
journal of the American Association of Critical-Care Nurses (AACN), published
bimonthly by The InnoVision Group, 101 Columbia, Aliso Viejo, CA 92656.
Telephone: (800) 899-1712, (949) 362-2050, ext. 532. Fax: (949) 362-2049.
Copyright 2006 by AACN. All rights reserved.

Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013

NEUROGENIC PULMONARY EDEMA DUE TO TRAUMATIC
BRAIN INJURY: EVIDENCE OF CARDIAC DYSFUNCTION
By Mabrouk Bahloul, Anis N. Chaari, Hatem Kallel, Abdelmajid Khabir, Adnne Ayadi, Hanne Charfeddine,
Leila Hergafi, Adel D. Chaari, Hedi E. Chelly, Chokri Ben Hamida, Noureddine Rekik, and Mounir Bouaziz.
From Service de Ranimation Mdicale (MB, ANC, HK, LH, ADC, HEC, CBH, NR, MB), Service dAnatomopathologie
(AK), and Service de Mdecine Lgale (AA), Centre Hospitalier Universitaire Habib Bourguiba, Sfax, Tunisia,
and Service de Cardiologie, Centre Hospitalier Universitaire Hdi Chaker, Sfax, Tunisia (HC).

BACKGROUND Acute neurogenic pulmonary edema, a common and underdiagnosed clinical entity, can
occur after virtually any form of injury of the central nervous system and is a potential early contributor
to pulmonary dysfunction in patients with head injuries.
OBJECTIVE To explore myocardial function in patients with evident neurogenic pulmonary edema after
traumatic head injury.
METHODS During a 1-year period in a university hospital in Sfax, Tunisia, information was collected
prospectively on patients admitted to the 22-bed intensive care unit because of isolated traumatic head
injury who had neurogenic pulmonary edema. Data included demographic information, vital signs, neu-
rological status, physiological status, and laboratory findings. All of the patients had computed tomogra-
phy and plain radiography of the neck and determination of cardiac function.
RESULTS All 7 patients in the sample had cardiac dysfunction. Evidence of myocardial damage was
confirmed by echocardiography in 3 patients, pulmonary artery catheterization in 3 patients, and/or
postmortem myocardial biopsy in 4 patients. Echocardiography studies, repeated 7 days after the initial
study in one patient and 90 days afterward in another, showed complete improvement in wall motion,
with a left ventricular ejection fraction of 0.65.
CONCLUSION All patients who had neurogenic pulmonary edema due to traumatic head injury had
myocardial dysfunction. The mechanisms of the dysfunction were multiple. The great improvement in
wall motion seen in 2 patients indicated the presence of a stunned myocardium. Further studies are
needed to understand the mechanisms of this cardiac dysfunction. (American Journal of Critical Care.
2006;15:462-470)

A cute neurogenic pulmonary edema (NPE) is

a common and underdiagnosed clinical
entity that can occur after virtually any form
of injury of the central nervous system. NPE has been
described in patients with seizures, stroke, several
Corresponding author: Mabrouk Bahloul, Service de Ranimation Mdicale,
Hpital Habib Bourguiba, Route el Ain Km 1, 3029 Sfax, Tunisia (e-mail:
bahloulmab@yahoo.fr)
To purchase electronic or print reprints, contact The InnoVision Group, 101
Columbia, Aliso Viejo, CA 92656. Phone, (800) 809-2273 or (949) 362-2050
(ext 532); fax, (949) 362-2049; e-mail, reprints@aacn.org.
types of intracranial hemorrhage, infection, and drug
use, among other conditions.1
In patients with traumatic acute head injury,
impaired pulmonary function is a common but poorly
understood complication. NPE is a potential early con-
tributor to the pulmonary dysfunction that occurs in
patients with head injuries.1 Although NPE is a fre-
quent complication of traumatic head injury, its occur-
rence in this specific condition is rarely described. In
addition, despite clinical and experimental studies, the
mechanisms leading to NPE are not fully understood.

462 AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5 http://ajcc.aacnjournals.org

Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013

 In this article we report 7 typical cases of NPE
associated with traumatic head injury. The aim of this
study was to explore myocardial function in patients
with evident NPE after traumatic head injury. The
study was approved by an internal review board.
Acute neurogenic pulmonary edema
can occur after any form of central
nervous system injury.
Materials and Methods
Sample
The sample consisted of nonconsecutive patients
admitted to the 22-bed intensive care unit (ICU) of
Centre Hospitalier Universitaire Habib Bourguiba,
Sfax, Tunisia, during a 1-year period because of NPE
due to isolated traumatic head injury. Patients were
admitted directly from the scene of the injury within 1
hour of injury. All were examined, and the score on the
Glasgow Coma Scale (GCS) was determined at arrival.
Patients underwent cerebral computed tomography
(CT) as soon as feasible. Excluded from the study were
all patients with extracranial injury, patients with other
causes of respiratory distress (eg, pulmonary contu-
sion, vomiting, gastric aspiration), and patients with a
history of cardiorespiratory disease.
The diagnosis of pulmonary edema was based on
the presence of clinical and radiological features of
pulmonary edema and on the presence of arterial
hypoxemia.
Methods
The following information was collected on hos-
pital admission, at ICU admission, and during ICU
stay: age, sex, vital signs (heart rate, respiratory rate
before mechanical ventilation, systolic and diastolic
blood pressure), body temperature, GCS score and
neurological manifestations before the use of mechan-
ical ventilation and before patients were sedated, Sim-
plified Acute Physiology Score II (for adult patients)
or Pediatric Risk of Mortality score (for children) cal-
culated within 24 hours after admission, the use of
mechanical ventilation, the use of inotropic drugs, the
occurrence of shock, the occurrence of cardiac arrest,
volume of fluid intake, and urinary output. Biochemi-
cal parameters measured on admission and during the
ICU stay were arterial blood gases and acid-base state
(pH and bicarbonate level); hemoglobin concentra-
http://ajcc.aacnjournals.org AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5
Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013
tion; platelet counts; serum levels of glucose, sodium,
and urea; and plasma protein concentration.
All patients underwent cranial CT and plain radio-
graphic study of the neck. The CT results were simpli-
fied to the presence or absence of hematoma (whether
extradural, subdural, or intracerebral), meningeal hem-
orrhage, cerebral edema, cerebral contusion, pneumo-
cephalus, intracranial mass lesion, and herniation.
Neurological status was assessed by using the
GCS score at the scene of the accident and the GCS
score at hospital arrival after resuscitation but before
the use of sedative. All patients with a GCS score of 8
or less, respiratory distress, or shock were intubated,
treated with mechanical ventilation, and sedated with
midazolam and fentanyl as necessary. Corticosteroids
were not used. When an extracranial abnormality was
suspected, appropriate investigations were done.
The diagnosis of pulmonary edema was established
by a medical committee of 5 ICU physicians at the time
of admission to the ICU or a few hours later. Pulmonary
edema was diagnosed if the patient had clinical and radi-
ological features of pulmonary edema and arterial
hypoxemia (arterial oxygen saturation measured while
the patient was breathing room air if possible). In
patients receiving mechanical ventilation, arterial hypox-
emia was defined as present when the ratio of PaO2 to the
fraction of inspired oxygen was less than 300.
The medical committee took particularly into
account the presence of signs of respiratory distress
(cyanosis, inspiratory retraction of intercostal spaces)
and the presence of lung crackles on auscultation of one
or both lungs. In addition, the committee looked for
signs of interstitial and/or alveolar pulmonary edema on
the chest radiographs. Manifestations of interstitial pul-
monary edema on radiographs included the loss of the
normal sharp definition of pulmonary vascular mark-
ings, haziness, loss of demarcation of hilar shadows,
thickening of interlobular septa, and peribronchial cuff-
ing. Radiographic manifestations of alveolar pulmonary
edema included unilateral or bilateral confluent acinar
shadows creating irregular patchy increases in parenchy-
mal density in the lower two thirds of the lung.2
Cardiac dysfunction was defined by the presence
of cardiogenic pulmonary edema and/or cardiogenic
shock. In all patients included, cardiac function was
explored. In 3 patients, left ventricular ejection frac-
tion (LVEF) was measured by means of echocardiog-
raphy as soon as feasible (within 24 hours after ICU
admission). In addition, in 3 patients, the measure-
ments of pulmonary artery wedge pressure, cardiac
index, stroke volume index, and systemic vascular
resistance were obtained by using a pulmonary artery
catheter inserted a few hours after admission to the
463
Table 1 Demographic and clinical parameters of the study population at admission

Patient
Parameter 1 2 3 4 5 6 7
Age, y 23 12 6 14 12 49 38
Sex Male Male Male Male Male Male Male
Simplified Acute
Physiology Score II 31 ND ND ND ND 32 31
Time between accident
and admission to
intensive care unit, h 1 1 1 1 1 1 1
Score on Glasgow
Coma Scale 3 4 7 7 5 8 10
Body temperature, C 35 37.6 37 36.5 ND 37.8 38.2
Heart rate,
beats per minute 112 120 130 84 120 35 123
Blood pressure, mm Hg
Systolic 92 100 100 100 50 200 140
Diastolic 43 60 50 50 20 80 70
Respiratory rate before
mechanical ventilation,
breaths per minute 28 24 27 35 8 33 34
Arterial oxygen saturation
while breathing air, % ND ND ND 75 54 79 ND
Glucose level, mmol/L
(mg/dL) 5.5 (99) 6.5 (117) 8.7 (157) 6.7 (121) ND 7.9 (142) 9.4 (169)
Shock No No No No Yes No No
pH 7.43 7.42 7.11 7.32 ND 7.45 7.27
PaCO2, mm Hg 33 27.7 34.8 35.2 ND 21.8 38.5
Bicarbonate, mmol/L 26 18.1 11.2 18.6 ND 15.4 24.6
PaO2 during mechanical
ventilation, mm Hg 177 153 78 175 ND 251 109
Fraction of inspired oxygen 1.00 0.60 0.50 0.60 ND 1.00 0.60
Creatine phosphokinase,
U/L 695 ND ND ND ND ND 270
Alanine aminotransferase,
U/L ND 32 ND 20 ND ND ND
Aspartate
aminotransferase, U/L ND 77 ND 40 ND ND ND
Mechanical ventilation Yes Yes Yes Yes Yes Yes Yes
Urea, mmol/L (mg/dL) 5.5 (15) 10 (28) 6.2 (17) 3 (8) ND 7.7 (22) 5.3 (15)
Hemoglobin, g/L 106 110 72 94 ND 139 95
Hospital stay, d 7 5 3 4 1 15 5
Use of catecholamines Yes Yes Yes Yes Yes Yes Yes
Type of catecholamine Dobut/epin Dobut Dobut Epin Dobut/norepin Dobut Epin
Outcome Died Survived Survived Died Died Survived Died

Abbreviations: Dobut, dobutamine; Epin, epinephrine; ND, not determined; Norepin, norepinephrine.

464 AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5 http://ajcc.aacnjournals.org

Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013

Table 2 Results of cerebral computed tomography scans

Patient
Parameter 1 2 3 4 5 6 7
Extradural hematoma No No No Yes No No Yes
Subdural hematoma No Yes Yes No No Yes No
Intracerebral hematoma No No No No No No No
Meningeal hemorrhage No No No Yes Yes Yes No
Cerebral edema Yes No No No Yes Yes No
Cerebral contusion No No No Yes No Yes No
Pneumocephalus No No No Yes No Yes No
Mass lesion No Yes Yes No No No Yes
Herniation No Yes No No No No Yes

Clinical Information
All 7 patients had hyperpnea and tachycardia at
the time of hospital admission. Cardiogenic shock
developed in 1 patient (patient 5, treated with dobu-
tamine and norepinephrine) on hospital admission.
The mean GCS score was 6 (range 3-10). All patients
had already received mechanical ventilation before the
ICU admission. All 7 patients had clinical manifesta-
tions of respiratory distress at the scene of the acci-
dent. The demographic and clinical parameters of the
sample at admission are shown in Table 1. All patients
had received catecholamines on ICU admission to
treat the pulmonary edema (dobutamine at a mean dose
of 10 g/kg per minute plus or minus epinephrine or
norepinephrine).

Figure 1 Computed tomography scan of cerebrum of patient
6 shows a subdural hematoma, meningeal hemorrhage, cere-
bral edema, cerebral contusion, and a pneumocephalus.
ICU. In 2 patients, myocardial echocardiography was
repeated. Finally, a myocardial biopsy and/or pul-
monary biopsy was done for all patients who died.
Brain CT
All patients had brain CT upon admission. The
most common abnormalities were meningeal hemor-
rhage, subdural hematoma, and mass lesion, observed
in 3 patients each (Table 2). Figure 1 shows the cere-
bral CT scan for patient 6.
Chest Radiography
The chest radiographs obtained upon admission
showed signs of alveolar pulmonary edema in all
cases (eg, patient 2, Figure 2). In addition, a chest CT
scan obtained on ICU admission in patient 5 (Figure 3)
clearly shows signs of alveolar pulmonary edema.

Results
During the study period, 202 patients were admit-
ted with traumatic head injury. Only 7 patients were
finally included in this study. The mean time between
traumatic head injury and ICU admission was 1 hour.
The traumatic head injury was caused by a traffic acci-
dent in all cases.
Echocardiography
Echocardiographic studies in 3 patients showed
global hypokinesia with a decrease in LVEF (to 0.40)
in 2 patients (patients 2 and 6). In 1 patient (patient 4),
however, a diastolic dysfunction with a reduced veloc-
ity of early filling (E wave), an increase in the velocity
associated with atrial contraction (A wave), and a ratio

http://ajcc.aacnjournals.org AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5 465

Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013

Figure 2 Chest radiograph of patient 2 shows pulmonary edema.
of E to A of less than 1 was observed. For 2 patients
(patients 2 and 6), echocardiography performed 7 and
90 days after the initial study showed complete recov-
ery, with an LVEF of 0.65.
Electrocardiography
The electrocardiograms showed some abnormali-
ties in all patients. The most common abnormality
was tachycardia, with heart rates exceeding 122/min.
Other abnormalities also were observed, including ST-
segment depression in 1 patient and right bundle branch
block in 2 patients. Table 3 shows the changes observed
for each patient on admission.
Pulmonary Artery Catheterization
Pulmonary artery catheterization was done for 3
patients (patients 1, 3, and 7). The findings indicated a
low stroke volume index in all 3 patients. Pulmonary
artery wedge pressure was higher in all patients except
patient 3. In patient 3, who had pulmonary edema evi-
dent on a chest radiograph, the mixed venous oxygen
saturation was low (<70%), and the arteriovenous dif-
ference in oxygen content was higher than in the other
patients (Table 4).
Histological Findings
Postmortem myocardial biopsies were done in 4
cases (patients 1, 4, 5, and 7), and a pulmonary biopsy
was done in 1 case (patient 5). Myocardial biopsy
showed an interstitial edema in all cases without a
focus of necrotic muscle fiber or myocardial infiltra-
tion by inflammatory cells. Figure 4 shows the histo-
logical findings on myocardial biopsy specimen for
patient 5. Examination of the pulmonary biopsy speci-
men from patient 5 showed marked alveolar edema
without signs of inflammatory lesions (Figure 5).
466 AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5
Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013
Figure 3 Computed tomography scan of the chest of
patient 5 obtained on admission to the intensive care unit
shows alveolar pulmonary edema.
Summary
In all patients included in the study, the diagnosis
of cardiac dysfunction was suspected because of the
signs of acute respiratory distress (cyanosis, inspiratory
retraction of intercostal spaces), the presence of crack-
les on auscultation of one or both lungs, and the evi-
dence of pulmonary edema on the chest radiograph. In
patients who underwent echocardiography, the cardio-
genic part was confirmed by a low LVEF (at 0.40 with
catecholamine) in 2 patients and by a reduced velocity
of early filling (E wave), an increase in the velocity
associated with atrial contraction (A wave), and a ratio
of E to A of less than 1 in 1 patient (patient 4).
In patients who underwent pulmonary artery
catheterization, cardiac dysfunction was indicated by a
pulmonary artery wedge pressure greater than 18 mm
Hg in 2 patients, and in another patient cardiac dys-
function was indicated by the low mixed venous oxy-
gen saturation (<70%) and the low stroke volume index
when treated with catecholamines.
Finally, in patient 5, who was admitted for isolated
traumatic head injury with acute respiratory distress
associated with a refractory shock, the diagnosis of
NPE was made by the medical committee on the basis
of clinical manifestations (respiratory distress), findings
on chest radiographs, and postmortem biopsy findings.
All subjects with neurogenic pulmonary
edema had cardiac dysfunction.
http://ajcc.aacnjournals.org
 Table 3 Electrocardiographic changes observed in all patients

Sinus tachycardia T-wave change Depression or elevation Branch block

Patient (beats per minute) (localization) of RST configuration
1 Yes (125) No No Yes (right)
2 Yes (140) Inversion of T No No
wave (V1 and V2)
3 Yes (130) No No Yes (right)
4 Yes (122) No No No
5 Yes (120) No No No
6 Yes (150) No ST-segment depression No
(V4-V6)
7 Yes (150) No No No

Table 4 Hemodynamic parameters recorded

Patient
Parameter 1 3 7
Delay before hemodynamic
study, hours after
admission to intensive
care unit 4 7 8
Heart rate, beats per minute 99 120 133
Mean arterial pressure,
mm Hg 94 73 87
Pulmonary artery wedge
pressure, mm Hg 26 6 20
Cardiac index* 3.30 3.01 5.35
Stroke volume index 30 25 40
Systemic vascular resistance,
dynes seconds cm-5 2089 1833 1121
Arterial hemoglobin Figure 4 Photomicrograph of histological examination of
concentration, g/L 106 137 95 myocardial biopsy specimen from patient 5 shows interstitial
myocardial edema (arrows) with no myocyte injury and
Arterial oxygen saturation, % 99.8 97.1 97.8
inflammatory infiltrate (hematoxylin-eosin, original magnifi-
Mixed venous oxygen cation x200).
saturation, % 67.4 67.8 62.2
Arteriovenous oxygen
content difference 3.78 5.68 2.55 injury is a common but poorly understood phe-
Oxygen extraction ratio, % 32.0 31.4 27.2 nomenon. 1,3 Causes of pulmonary dysfunction in
Arterial carbon dioxide patients with head injury include pneumonia, aspira-
tension, mm Hg 31.1 37.3 40.9 tion, and pulmonary embolus, but seldom NPE.4,5 NPE
Mixed venous carbon is a form of pulmonary edema that develops rapidly
dioxide tension, mm Hg 34.2 39.3 42.9
after a cerebral injury.1,6 It has been described in trauma
Venoarterial carbon dioxide patients as parenchymal edema, hemorrhage, and con-
tension difference, mm Hg 3.1 2.0 2.0
gestion without evidence of chest trauma in patients
*Calculated as cardiac output in liters per minute divided by with isolated head injury.1,7
body surface area in square meters.
Calculated as stroke volume in milliliters per beat divided by The true incidence of NPE after acute head injury is
body surface area in square meters. difficult to estimate because much of the information
comes from small autopsy series or isolated case reports.
Discussion Furthermore, the clinical relevance of NPE in
In this study we confirmed the presence of cardiac patients with nonfatal head injury remains to be eluci-
dysfunction in patients with NPE due to traumatic dated, because NPE seems to be rare in patients who sur-
brain injury. Pulmonary dysfunction after acute brain vive. However, according to Rogers et al,1 the incidence

http://ajcc.aacnjournals.org AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5 467

Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013


Figure 5 Photomicrograph of histological examination of
pulmonary biopsy specimen from patient 5 shows uniform
edematous change in the alveolar septa with dilated lymphatic
channels, accumulation of red blood cells, and intraalveolar
edema (hematoxylin-eosin, original magnification x400).
of NPE in patients with isolated head injury was 32%
in patients who died at the scene and 50% in patients
who died within 96 hours of the injury.
In Tunisia, nearly 13 000 persons are involved in
motor vehicle crashes every year. Approximately 1500
of these patients die.8 However, the incidence of NPE
had not previously been established. In the study
reported here, we included only patients with isolated
traumatic head injury and with the typical signs and
symptoms of NPE. In addition, because of a lack of
tools, it was not possible to examine all patients admit-
ted to our ICU with typical NPE. For these reasons,
patients included were nonconsecutive, and we are not
able to establish the incidence of this abnormality.
NPE is characterized by an increase in extravascu-
lar lung water in patients who have sustained a sudden
change in neurological condition.9-12 The mechanism
by which NPE occurs is not clear, and 2 divergent the-
ories have been proposed to explain its development:
increased lung capillary permeability and increased
pulmonary vascular hydrostatic pressures. Increased
permeability as a mechanism of NPE is supported by
some studies in animals that have shown high intersti-
tial (lung lymphatic) or alveolar protein concentra-
tions9,12 and time-dependent ultrastructural changes in
pneumocyte type II cells after brain injury.13 Increased
permeability may be caused by damage of the capil-
lary endothelium or by direct neural influences on
capillary permeability (the blast theory).14-17
On the other hand, hydrostatically induced pul-
monary edema can occur without endothelial damage.
468 AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5
Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013
One possible sequence leading to NPE is an acute
increase in sympathetic tone that abruptly increases
left ventricular afterload and causes intense venocon-
striction, thereby elevating left ventricular filling pres-
sures and inducing elevated pulmonary artery wedge
pressures, leading to hydrostatic pulmonary edema.
This hypothesis was confirmed by experimental stud-
ies18,19 and studies in humans.9,20,21 However, pulmonary
edema can occur with normal pulmonary artery wedge
pressures,22,23 suggesting a neurally mediated pressure
independent of the influence on capillary permeability.
In addition to these 2 hypotheses, NPE can result
from a cardiac dysfunction. In fact, the early hemody-
namic changes that occur in the setting of NPE may lead
to the conclusion that the pulmonary edema is of cardiac
origin. Smith and Matthay9 reported, as have others, that
early analysis of NPE fluid reveals a low fluid-serum
protein ratio consistent with hydrostatic edema. In addi-
tion to the change in vascular resistance described, the
pathogenesis of hydrostatic NPE may involve direct
negative inotropic effects on the heart.24,25
In a retrospective study26 that included 20 patients
with NPE, all 20 required mechanical ventilation; car-
diac index and left ventricular stroke work index were
markedly depressed in 12 of the 20 patients; mean pul-
monary artery wedge pressure was 17 mm Hg; mean
pulmonary artery pressure was 30.5 mm Hg; and mean
systemic vascular resistance index (calculated as sys-
temic vascular resistance in dynes seconds centime-
ters-5 divided by body surface area in square meters)
was 2852. Patients treated with dobutamine had signifi-
cant increases in cardiac index and left ventricular stroke
work index and significant decreases in systemic vascu-
lar resistance index and pulmonary artery wedge pres-
sure at 2 and 6 hours after institution of therapy and a
significantly increased ratio of PaO2 to fraction of
inspired oxygen at 6 hours after the start of therapy. The
authors26 concluded that NPE was generally associated
with severe depression of myocardial function and ele-
vation of pulmonary vascular pressures. This dysfunc-
tion was readily reversed by dobutamine.
Neurogenic pulmonary edema may
result from increased lung capillary
permeability, increased pulmonary
vascular hydrostatic pressure, or
cardiac dysfunction.
This hypothesis of myocardial dysfunction was
supported by some echocardiographic studies.25,27 In a
http://ajcc.aacnjournals.org
case series of 5 patients with subarachnoid hemor-
rhage and no history of heart disease, all with severe
NPE, 5 had a low LVEF requiring inotropic support.27
Such cases illustrate that NPE and abnormalities in
myocardial wall motion can occur concurrently. The
cardiac dysfunction and wall-motion abnormalities are
temporary, and cardiac function usually returns to nor-
mal.25,28,29 This hypothesis of cardiac dysfunction was
confirmed by Connor,30 who found further postmortem
evidence of myocytolysis and contraction-band necro-
sis of the heart in neurosurgical patients with NPE.
Our findings confirm the hypothesis of myocar-
dial dysfunction. In fact, in our study, cardiac dys-
function was well documented by echocardiography,
hemodynamic changes, and/or postmortem biopsy. In
addition, the control echocardiography study done in 2
patients showed a complete recovery of cardiac func-
tion. This myocardial dysfunction can be related to the
massive release of catecholamines,15,25,31 the hyper-
glycemia that often occurs after traumatic brain injury,8
and/or the massive liberation of cytokines after severe
traumatic brain injury.32 However, we cannot rule out
other phenomena, especially permeability edema, on
the basis of our findings.
The usually reversible myocardial dysfunction
shown by echocardiography is poorly correlated with
ECG changes.25 Our study confirms this last hypothesis;
in fact, electrocardiographic manifestations of myocar-
dial ischemia were observed in only a single patient.
Finally, our study has several limitations. Because
echocardiography was not available in our ICU or in
our hospital, our patients were not examined in a uni-
form way (some were examined with echocardiogra-
phy, some with pulmonary artery catheterization). For
echocardiography to be done, the patient had to be
transferred to another hospital.
All patients included in our study had respiratory
distress; in particular, patients 1, 3, and 7 had shock asso-
ciated with acute respiratory distress with a ratio of PaO2
to fraction of inspired oxygen less than 200. Therefore,
we preferred to use pulmonary artery catheterization
(performed in our ICU) in these patients. In patient 3,
the pulmonary artery wedge pressure was normal.
Despite that normal finding, the hemodynamic study
of these parameters was performed after administra-
tion of a catecholamine (dobutamine). The patients
mixed venous oxygen saturation was 67.8% and the
stroke volume index (calculated as stroke volume in
milliliters per beat divided by body surface area in
square meters) was 25, suggesting the presence of car-
diac dysfunction.33
The high number of patients who were excluded
from the study could be a methodological limitation.
http://ajcc.aacnjournals.org AMERICAN JOURNAL OF CRITICAL CARE, September 2006, Volume 15, No. 5
Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013
Gastric aspiration and pulmonary contusion, which pro-
duce increased permeability and pulmonary edema, are
potential confounding factors in clinical cases of NPE.
We were careful to exclude any cases of observed aspi-
ration and/or chest trauma in order to study typical cases
of NPE (with a sure diagnosis) and to exclude the maxi-
mal number of patients with other therapeutic interven-
tions (eg, fluid infusion). Our sample included young
patients (age <30 years in 70% of cases) with a low risk
factor of cardiac dysfunction and no history of car-
diorespiratory disease. We therefore believe that the ele-
vated pulmonary artery wedge pressures could not be
related to simple cardiogenic pulmonary edema.
In addition, hemodynamic parameters determined by
using pulmonary artery catheters can be difficult to inter-
pret in patients receiving positive pressure ventilation
with or without high levels of positive end-expiratory
pressure. However, in our study, all parameters (in par-
ticular, cardiac index and pulmonary artery wedge pres-
sure) were measured while the patient was receiving
catecholamines, and in all patients mixed venous oxy-
gen saturation was low, suggesting cardiac dysfunction.33
In patient 5, we inferred cardiac dysfunction on the basis
of a postmortem biopsy; the results of such biopsies can
be unreliable and may reflect complications of the pri-
mary brain injury as much as the primary process. In
this patient, who was admitted for isolated traumatic
head injury with acute respiratory distress associated
with a refractory shock, the diagnosis of NPE was ascer-
tained by the medical committee on the basis of clinical
manifestations (respiratory distress), radiological (chest
radiographic) findings, and results of postmortem biopsy.
Conclusion
Our findings confirmed the presence of myocar-
dial dysfunction in patients with NPE due to traumatic
head injury; however, we cannot rule out other phe-
nomena, especially permeability edema. The mecha-
nisms of myocardial dysfunction were multiple. The
great improvement in wall motion seen in 2 patients
indicated the presence of a stunned but viable myo-
cardium. Further studies are needed to understand the
mechanisms of this cardiac dysfunction.
REFERENCES
1. Rogers FB, Shackford SR, Trevisani GT, Davis JW, Mackersie RC, Hoyt
DB. Neurogenic pulmonary edema in fatal and nonfatal head injuries. J
Trauma. 1995;39:860-868.
2. Fraser RG, Par JAP. Solitary pulmonary nodules less than 6 cm. Diagnosis
of Diseases of the Chest. Philadelphia, Pa: WB Saunders Co; 1970:854-860.
3. Demling R, Riessen R. Pulmonary dysfunction after cerebral injury. Crit
Care Med. 1990;18:768-774.
4. Helling TS, Evans LL, Fowler DL, Hays LV, Kennedy FR. Infectious com-
plications in patients with severe head injury. J Trauma. 1988;28:1575-1577.
5. NIH Consensus Report. Prevention of venous thromboses and pulmonary
embolism. JAMA. 1986;256:744-749.
6. Demling R, Riessen R. Pulmonary dysfunction after cerebral injury. Crit
469
 Care Med. 1990;18:768-771.
7. Simmons RL, Martin AM Jr, Heisterkamp CA III, Ducker TB. Respiratory
insufficiency in combat casualties (pulmonary edema following head
injury). Ann Surg. 1969;170:39-44.
8. Bahloul M, Chelly H, Ben Hmida M, et al. Prognosis of traumatic head
injury in South Tunisia: a multivariate analysis of 437 cases. J Trauma.
2004;57:255-261.
9. Smith WS, Matthay MA. Evidence for a hydrostatic mechanism in human
neurogenic pulmonary edema. Chest. 1997;111:1326-1333.
10. Simon RP. Neurogenic pulmonary edema. Neurol Clin. 1993;11:309-323.
11. Colice GL, Matthay MA, Bass E, Matthay RA. Neurogenic pulmonary
edema. Am Rev Respir Dis. 1984;130:941-948.
12. Kowalski ML, Didier A, Kaliner MA. Neurogenic inflammation in the air-
ways. I. Neurogenic stimulation induces plasma protein extravasation into
the rat airway lumen. Am Rev Respir Dis. 1989;140:101-109.
13. Yildirim E, Kaptanoglu E, Ozisik K, et al. Ultrastructural changes in pneu-
mocyte type II cells following traumatic brain injury in rats. Eur J Cardio-
thorac Surg. 2004;25:523-529.
14. Theodore J, Robin ED. Speculations on neurogenic pulmonary edema
(NPE). Am Rev Respir Dis. 1976;113:405-411.
15. Wasowska-Krolikowska K, Krogulska A, Modzelewska-Holynska M.
Neurogenic pulmonary oedema in a 13-year-old boy in the course of symp-
tomatic epilepsy: case report. Med Sci Monit. 2000;6:1003-1007.
16. West JB. Stress failure of pulmonary capillaries: role in lung and heart dis-
ease. Lancet. 1992;340:762-767.
17. Bachofen H, Schurch S, Weibel ER. Experimental hydrostatic pulmonary
edema in rabbit lungs: barrier lesions. Am Rev Respir Dis. 1993;147:997-1004.
18. Maron MB. Analysis of airway fluid protein concentration in neurogenic
pulmonary edema. J Appl Physiol. 1987;62:470-476.
19. Johnston SC, Horn JK, Valente J, Simon RP. The role of hypoventilation
in a sheep model of epileptic sudden death. Ann Neurol. 1995;37:531-537.
20. Carlson RW, Schaeffer RJ, Michaels SG, Weil MH. Pulmonary edema fol-
lowing intracranial hemorrhage. Chest. 1979;75:731-734.
Downloaded from ajcc.aacnjournals.org at Mobile User on May 20, 2013
21. Wray NP, Nicotra MB. Pathogenesis of neurogenic pulmonary edema. Am
Rev Respir Dis. 1978;118:783-786.
22. Mayer SA, Fink ME, Homma S, et al. Cardiac injury associated with neu-
rogenic pulmonary edema following subarachnoid hemorrhage. Neurology.
1994;44:815-820.
23. Yabumoto M, Kuriyama T, Iwamoto M, Kinoshita T. Neurogenic pul-
monary edema associated with ruptured intracranial aneurysm: case report.
Neurosurgery. 1986;19:300-304.
24. Samuels MA. Neurogenic heart disease: a unifying hypothesis. Am J Car-
diol. 1987;60:15J-19J.
25. Macmillan CS, Grant IS, Andrews PJ. Pulmonary and cardiac sequelae of
subarachnoid haemorrhage: time for active management? Intensive Care
Med. 2002;28:1012-1023.
26. Deehan SC, Grant IS. Haemodynamic changes in neurogenic pulmonary
oedema: effect of dobutamine. Intensive Care Med. 1996;22:672-676.
27. Mayer SA, Fink ME, Homma S, et al. Cardiac injury associated with neu-
rogenic pulmonary edema following subarachnoid hemorrhage. Neurology.
1994;44:815-820.
28. Parr MJ, Finfer SR, Morgan MK. Reversible cardiogenic shock complicat-
ing subarachnoid haemorrhage. BMJ. 1996;313:681-683.
29. de Chazal I, Parham WM III, Liopyris P, Wijdicks EF. Delayed cardio-
genic shock and acute lung injury after aneurysmal subarachnoid hemor-
rhage. Anesth Analg. 2005;100:1147-1149.
30. Connor RC. Myocardial damage secondary to brain lesions. Am Heart J.
1969;78:145-148.
31. Wittstein IS, Thiemann DR, Lima JA, et al. Neurohumoral features of
myocardial stunning due to sudden emotional stress. N Engl J Med.
2005;352:539-548.
32. Ahn MJ, Sherwood ER, Prough DS, Lin CY, DeWitt DS. The effects of
traumatic brain injury on cerebral blood flow and brain tissue nitric oxide
levels and cytokine expression. J Neurotrauma. 2004; 21:1431-1442.
33. Cariou A, Monchi M, Dhainaut JF. Continuous cardiac output and mixed
venous oxygen saturation monitoring. J Crit Care. 1998;13:198-213.