Cerebrovascular Resuscitation after Polytrauma and Fluid ... · Cerebrovascular Resuscitation after...

CaSK

AA

CSACDTS

©P

erebrovascular Resuscitationfter Polytrauma and Fluid Restriction

teven A Earle, MD, Marc A de Moya, MD, Jennifer E Zuccarelli, BA, Michael D Norenberg, MD,enneth G Proctor, MD, PhD

BACKGROUND: There are few reproducible models of blast injury, so it is difficult to evaluate new or existingtherapies. We developed a clinically relevant polytrauma model to test the hypothesis thatcerebrovascular resuscitation is optimized when intravenous fluid is restricted.

STUDY DESIGN: Anesthetized swine (42 � 5 kg, n � 35) received blasts to the head and bilateral chests with captive boltguns, followed by hypoventilation (4 breaths/min; FiO2 � 0.21). After 30 minutes, resuscitation wasdivided intophases to simulate typicalprehospital, emergency room,andICUcare.For30 to45minutes,group 1, the control group (n � 5), received 1L of normal saline (NS). For 45 to 120 minutes, additionalNS was titrated to mean arterial pressure (MAP) � 60 mmHg. After 120 minutes, mannitol (1g/kg) andphenylephrine were administered to manage cerebral perfusion pressure (CPP) � 70 mmHg, plus addi-tional NS was given to maintain central venous pressure (CVP) � 12 mmHg. In group 2 (n � 5), MAPandCPPtargetswere thesame,but theCVPtargetwas � 8mmHg.Group3(n � 5)received1LofNSfollowedonlybyCPPmanagement.Group4(n � 5) receivedHextend(AbbottLaboratories), insteadofNS, to the same MAP and CPP targets as group 2.

RESULTS: Polytrauma caused 13 deaths in the 35 animals. In survivors, at 30 minutes, MAP was 60to 65 mmHg, heart rate was � 100 beats/min, PaO2 was � 50 mmHg, and lactatewas � 5 mmol/L. In two experiments, no fluid or pressor was administered; the tachycardia andhypotension persisted. The first liter of intravenous fluid partially corrected these variables, andalso partially corrected mixed venous O2, gastric and portal venous O2, cardiac output, renalblood flow, and urine output. Additional NS (total of 36 � 1 mL/kg/h and 17 � 6 mL/kg/h,in groups 1 and 2, respectively) correlated with increased intracranial pressure to 38 � 4 mmHg(group 1) and 26 � 4 mmHg (group 2) versus 22 � 4 mmHg in group 3 (who received5 � 1 mL/kg/h). CPP was maintained only after mannitol and phenylephrine. By 5 hours,brain tissue PO2 was � 20 mmHg in groups 1 and 2, but only 6 � 1 mmHg in group 3. Incontrast, minimal Hextend (6 � 3 mL/kg/h) was needed; the corrections in MAP and CPPwere immediate and sustained, intracranial pressure was lower (14 � 2 mmHg), and braintissue PO2 was � 20 mmHg. Neuropathologic changes were consistent with traumatic braininjury, but there were no statistically significant differences between groups.

CONCLUSIONS: After polytrauma and resuscitation to standard MAP and CPP targets with mannitol and pressortherapy, we concluded that intracranial hypertension was attenuated and brain oxygenation wasmaintained with intravenous fluid restriction; cerebrovascular resuscitation was optimized withHextend versus NS; and longer term studies are needed to determine neuropathologic consequences.
(J Am Coll Surg 2007;204:261–275. © 2007 by the American College of Surgeons)
fp

RbFSPCC

lmost every day, for as long as the wars in Iraq andfghanistan continue, our soldiers and marines will suf-

ompeting Interests Declared: None.upported by grant N000140210339, from the Office of Naval Research.bstract presented at the American College of Surgeons 92nd Annual Clinicalongress, Surgical Forum, Chicago, IL, October 2006.r de Moya’s current address is: Massachusetts General Hospital, Division ofrauma, Emergency Surgery, and Surgical Critical Care, 165 Cambridge St,
uite 810, Boston, MA 02114.
S

2612007 by the American College of Surgeons

ublished by Elsevier Inc.

er polytrauma, maiming, or death from improvised ex-losive devices (IED). In fact, the majority of combat

eceived October 13, 2006; Revised November 16, 2006; Accepted Novem-er 16, 2006.rom the Dewitt-Daughtry Family Department of Surgery, Divisions ofTrauma andurgical Critical Care (Earle, de Moya, Zuccarelli, Proctor) and the Department ofathology (Norenberg), University of Miami Miller School of Medicine, Miami, FL.orrespondenceaddress:KennethGProctor,PhD,DivisionsofTraumaandSurgicalritical Care, Daughtry Family Department of Surgery, University of Miami Miller
chool of Medicine, Ryder Trauma Center, 1800 NW 10th Ave, Miami, FL 33136.
ISSN 1072-7515/07/$32.00doi:10.1016/j.jamcollsurg.2006.11.014

ciwc

iatrlvwaAptif

tmooseOm

hmrpihwt

tb

MAiifGpmvna

GF(lxatuPwcaPo

wtMatpnsmtu(nttom

262 Earle et al Cerebrovascular Resuscitation after Polytrauma J Am Coll Surg

asualties are now caused by IEDs.1 The recent upsurgen urban terrorist bombings2-6 suggests that blast injuryill become a disturbing reality at both military and

ivilian trauma centers in the 21st century.7

There are few reproducible animal models of blastnjury.8 Most existing models for resuscitation researchre inadequate because there is minimal (if any) softissue injury; trauma is usually equated with hemor-hagic shock, and resuscitation is based on blood volumeoss in a controlled laboratory setting. In contrast, blastictims can have multiple soft-tissue injuries, with orithout active hemorrhage, and resuscitation is directed

t restoring blood pressure rather than blood volume.9,10

ssociated compartment syndromes can confound com-ensatory responses. With no valid models, it is difficulto evaluate new or existing therapies, and it is virtuallympossible to develop specific evidence-based guidelinesor treatment.

Recent data from Walter Reed Army Hospital suggesthat occult traumatic brain injury (TBI) is common inany soldiers exposed to blast, even when there are no

bvious external signs or loss of consciousness.10-12 Blastverpressure might also produce lung injuries that re-emble pulmonary contusion.9,10 The resultant hypox-mia can potentiate secondary damage after TBI.13,14

ne goal of this study was to develop a clinically relevantodel of blast-induced polytrauma.The other major goal of this study was to test the

ypothesis that resuscitation after polytrauma is opti-ized with IV fluid (IVF) restriction. Cerebrovascular

esuscitation in a combat environment can be com-licated by a lack of basic monitoring capability, lim-ted resources, and prolonged evacuation.15-17 On oneand, the ability to achieve adequate resuscitationith minimal IVF offers considerable logistic advan-

Abbreviations and Acronyms

BIS EEG � bispectral electroencephalogramCPP � cerebral perfusion pressureCVP � central venous pressureICP � intracranial pressureIED � improvised explosive deviceIVF � intravenous fluidMAP � mean arterial pressureNS � normal salineTBI � traumatic brain injury

ages for the military. On the other hand, IVF restric- V

ion has devastating neurologic consequences afterlast-induced neurotrauma.

ETHODSnimals were housed in a facility approved by the Amer-

can Association of Laboratory Animal Care, with veter-narians available at all times. All procedures were per-ormed according to National Institutes of Healthuidelines for Use of Laboratory Animals and were preap-roved by our Institutional Animal Care and Use Com-ittee. Animals were anesthetized in all surgical inter-

entions, and all efforts were made to minimize theumber of animals involved and to alleviate their painnd distress.

eneral instrumentationarm-raised crossbred fasted swine of both genders42 � 2 kg, n � 35) were sedated with an intramuscu-ar injection of 30 mg/kg ketamine and 3.5 mg/kgylazine. The animals underwent orotracheal intubationnd mechanical ventilation (Impact Portable Adult Ven-ilator Model 754, Impact Systems), with tidal vol-mes � 10 mL/kg and 8 to 16 breaths/min to maintainaCO2 � 40 � 5 mmHg. The FiO2 was 0.4 excepthere otherwise noted. Anesthesia was initiated with

ontinuous intravenous infusions of 10 mg/kg/h ket-mine, 0.5 mg/kg/h xylazine, and 50 g/kg/h fentanyl.ulse oximetry (Nellcor Pulse Oximeter) was continu-usly monitored.

With the animal in the supine position, cathetersere placed in the femoral artery for continuous ar-

erial blood pressure monitoring (Zoll Hemodynamiconitor) and in the external jugular vein for IVF

dministration. Additional catheters were placed inhe urinary bladder to measure urine output, in theulmonary artery to measure mixed venous pulmo-ary artery oxygen saturation, pulmonary artery pres-ures, and cardiac output, and in the portal vein toeasure oxygen saturation (Abbott Critical Care Sys-

ems, Abbott Laboratories). Gastric tissue oxygen sat-ration was measured by suturing a surface probeInSpectra Tissue Spectrometer, Hutchinson Tech-ologies Inc) to the outer surface of the anterior gas-ric wall. Laser Doppler blood flow probes were su-ured to the inferior pole of the kidney and to theuter surface of the anterior wall of the stomach toeasure microcirculatory blood flow (BPM 402,
asamedics).

tPog(scbuidtSar

ttr

IAic

aPw

i(al4wts

bsTiPa

cT

ti

RTpmp4e1c23ItttF

(dbramptaN�

flsrrmsit

Gtmw

263Vol. 204, No. 2, February 2007 Earle et al Cerebrovascular Resuscitation after Polytrauma

The animal was then rotated into the prone posi-ion for the remainder of the experiment. Brain tissueO2 and intracranial pressure (ICP) were continu-usly monitored through an intraparenchymal oxy-en electrode and fiber optic pressure transducerLICOX MCB Oxygen Monitor, Integra Neuro-ciences) that was placed through a small left frontalraniotomy. In addition, leads were placed so that aispectral electroencephalogram (BIS EEG) was contin-ously recorded (Model A-1050 monitor, Aspect Med-

cal Systems). The suppression ratio of the BIS EEG isefined as the percentage of isoelectric EEG activity inhe last minute, and is consistent with cerebral ischemia.o, BIS monitoring allows precise evaluation of second-ry injury to the brain, as we18,19 and others haveeported.20,21

During a 60-minute postinstrumentation stabiliza-ion period, 1L of normal saline (NS) was administeredo all animals. All IVF was stopped and the FiO2 waseduced to 0.21 for 15 minutes before injury.

mprovised explosive device polytrauma simulationnesthetized, fully instrumented, and ventilated swine

n the prone position received blunt TBI, bilateral lungontusions, and hypoventilation for 30 minutes.

The blunt TBI was produced with a commerciallyvailable power actuated tool or “nail gun” (Remingtonower Trigger, Model 479, Desa Specialty Products) thatas activated by a 22-caliber #1 cartridge.Within a minute of blunt TBI, bilateral blunt chest

njuries were produced by firing a captive bolt gunmodel ME, Karl Schermer & Co), first against the rightnd then the left chest. The location was the midaxillaryine at the level of the fourth intercostal space with a5-degree cephalad trajectory. The gun was activatedith a #21 10-mm cartridge. Landmarks were the pos-

erior axillary line at the angular scapular groove, as de-cribed in several previous studies.22-24

Kinetic energy from each of the guns was calculatedy determining the mass of the captive bolt and by mea-uring its velocity with high speed digital photography.hese measurements were made at Vision Research Inc

n the Allan Carlisle Photography laboratory with ahantom Digital High Speed Camera, Analysis System,nd CineViewer 606 (Photosonics).

For 30 minutes after this insult, animals were me-hanically ventilated at 4 breaths/min on FiO2 � 0.21.
his superimposed a hypoxic or hypercapnic stimulus 1
hat would mimic the apnea that might accompany TBIn real life.

esuscitation and treatment groupshe remainder of the experiment was divided intohases that were intended to simulate the types ofedical care that would occur in the prehospital

hase or during emergency medical transport (30 to5 minutes postinjury; 0 to 15 minutes of first aid),mergency room (ER, 45 to 120 minutes postinjury;5 to 90 minutes of medical treatment), and intensiveare unit (ICU, 120 to 300 minutes postinjury; 90 to70 minutes of medical treatment). For 300 to30 minutes after polytrauma, anesthesia and otherVFs were turned off, and the ventilator was switchedo the “assist-control” mode so that a minimal inspira-ory effort would trigger a full breath. There were fourreatment groups, and the protocol is outlined inigure 1.In group 1, the standard of care or control group

n � 5), “prehospital” resuscitation consisted of imme-iate ventilatory support (FiO2 � 0.4; rate � 12 to 20reaths/min) and a 1-L bolus of NS. In the “emergencyoom phase,” PEEP was increased to 5 cm H20, anddditional NS was infused to a resuscitation target ofean arterial pressure (MAP) � 60 mmHg. In the “ICU

hase,” prophylactic mannitol (1 g/kg) was adminis-ered for intracranial hypertension (ICP � 20 mmHg),nd phenylephrine was titrated along with additionalS to maintain a cerebral perfusion pressure (CPP)70 mmHg at a CVP �12 mmHg.Mannitol, rather than hypertonic saline, was chosen

or ICP control based on current evidence-based guide-ines for managing severe brain trauma.14 The consensustatement is that “mannitol is effective for the control ofaised ICP after severe head injury . . . Effective dosesange from 0.25 to 1 g/kg body weight.” The use ofannitol for ICP control after severe brain injury is

upported at the level of a guideline because there arensufficient data to support a treatment standard on thisopic.

The remaining three groups were fluid restricted.roup 2 (NS, CVP � 8 mmHg, n � 5) received NS to

he same MAP targets, and NS � phenylephrine �annitol to the same CPP target, but the filling pressureas 8 mmHg.Group 3 (NS 1 L, n � 5) received a total of only
L of NS and then mannitol and phenylephrine to

tM

srCn

TBvtBu7ta

PT6cCw

rsDaelaUa

HTlrwb1bTrbbt

raum


he CPP targets, without regard for filling pressure orAP.Group 4 received Hextend (HEX, 6% hydroxyethyl

tarch in lactated electrolyte injection; Abbot Laborato-ies), instead of NS, to the group 2 targets (HEX,VP � 8 mmHg, n � 5). In two additional experiments,o fluid or pressor was administered after polytrauma.

est of cerebrovascular reactivity and complianceefore injury, and at hourly intervals thereafter, cerebro-ascular reactivity and compliance were evaluated. Thisechnique has been previously described in detail.18,19,25

riefly, inhaled CO2 was maintained at 7.5% for 5 min-tes, which produced an average end-tidal CO2 of 65 to0 mmHg. The degree of the CO2-evoked ICP and brainissue oxygenation changes depends on cerebral compli-nce and vascular reactivity in animals and patients.

hysiologic datahe following were monitored continuously for0 minutes before and 360 minutes after polytrauma:ore temperature, end tidal CO2, heart rate, MAP,VP, pulmonary artery pressure, pulmonary capillary

Figure 1. Experimental design. CPP, cerebral perfuspressure; NS, normal saline; PE, phenylephrine; TBI, t

edge pressure, cardiac output, mixed venous O2 satu- h

ation, portal venous O2 saturation, gastric tissue O2

aturation, gastric laser Doppler blood flow, renal laseroppler blood flow, brain tissue PO2, ICP, BIS EEG,

nd urine output. Blood gases (PaO2, PaCO2, pH, basexcess, and arterial O2 saturation), lactate and electro-ytes (Na�, K�, glucose, and osmolarity) were recordedt 15- to 30-minute intervals on a Nova Stat Profileltra. At 360 minutes, an overdose of anesthesia was

dministered for euthanasia.

istologic specimenshe brain was weighed. The presence and size of surface

esions (measured and contrecoup contusions, hemor-hages including epidural and subdural hemorrhages)ere documented, measured, and photographed. Therain was suspended in 10% buffered formalin forweek. The cerebral hemispheres, brainstem, and cere-ellum were coronally cut into 5-mm thick sections.issue blocks were obtained of all gross lesions and from

epresentative areas of the neocortex, hippocampus,asal ganglia, thalamus, hemispheric white matter, mid-rain, and cerebellar hemisphere. Three-millimeterhick sections were processed routinely for histology (de-

ressure; ER, emergency room; MAP, mean arterialatic brain injury.

ion p

ydration in graded alcohol, clearing in xylene, then

psiesetaaifqfwa

ST(eA

tA

RDBcaktldgiwdcg

ere


araffin embedding). Paraffin sections (8 �m) weretained routinely with hematoxylin and eosin to exam-ne for neuronal shrinkage (early ischemia), neuronalosinophilia (evidence of neuronal necrosis), neuropilpongiosis (edematous changes), microglial nuclearnlargement and pallor (microglial activation), endo-helial nuclear pallor and enlargement (endothelialctivation), and presence of neutrophils (evidence ofcute inflammation). Darkened neurons reflected earlyschemic injury.26 Alzheimer type II astrocytosis re-lected early damage to astrocytes after trauma.27 Semi-uantitative assessment of histologic changes were per-ormed “blindly” and graded on a crude 0-to-3 scale,here 0 � no changes, 1 � slight changes, 2 � moder-

te changes, and 3 � marked changes.

tatisticsreatments were applied in random order in two cohortsgroups 1 versus 3; and groups 2 versus 4). Data arexpressed as mean � SEM. Student t-test and one-way

Figure 2. The gross appearance of a typical brain injuhemorrhage, after durotomy in situ (B) showing subdu(C) and contrecoup injury at the base of the brain (D) w

NOVA for multiple comparisons were calculated with a

he SPSS for Windows release 14.0 program (SPSS Inc).two-tailed p � 0.05 was considered significant.

ESULTSescription of polytrauma modellunt TBI was produced by a power actuated piston oraptive bolt (mass � 103 g) striking a 4-cm (diameter)luminum disk at 30 m/s (kinetic energy � 0.5 [0.103g] [30 m/s]2 � 46.4 joules). The disk was taped tohe skin overlying the occipital bone along the mid-ine. This blast produced no skull fracture or obviousamage to the head (other than mild erythema). Theross appearance of a typical injury on autopsy is shownn Figure 2. After craniectomy, a subdural hematomaas obvious (Figs. 2A [before durotomy] and 2B [afterurotomy]), and the contusion (Fig. 2C) and contre-oup injury at the base of the brain (Fig. 2D) wererossly evident.

Bilateral blunt chest trauma was produced by a power

autopsy: before durotomy in situ (A) showing epiduralemorrhage. Subarachnoid hemorrhage and contusiongrossly evident. Scale in cm.

ry atral h

ctuated piston with a mushroom head (mass � 352 g)

seTtTmtstcpsmh

whf1mwhw

wsrmiv3

awdcOsTsaojltm

PoFl1acbcamgCgw(

aNtanANsy(cp14p

Fh


triking the right and then the left chest at 26 m/s (kineticnergy � 0.5 [0.352 kg] [26 m/s]2 � 119.0 joules).he most common chest injury pattern was a bruise on

he chest, but there was no other obvious deformity.ube thoracostomy was usually not required. At autopsy,ultiple nondislocated rib fractures were common, but

he parietal pleura was not often violated. Lung contu-ions were obvious on gross examination. Figure 3 showshe typical appearance of the lungs. Gross changes in-lude diffuse contusions with stiffened heavy tissue,leural rents, pneumothorax, or mediastinal extrava-ation of air. Despite these sequelae, PaO2 usually re-ained � 200 mmHg on FiO2 � 0.4 for at least 5

ours with standard supportive care.There were 13 (37%) deaths after polytrauma (Fig. 1),

hich defines the severity of the combined insult. Aboutalf of these deaths occurred within 30 minutes, ie, be-ore resuscitation. The other half occurred within20 minutes; typically, there were two patterns. One wasinimal response to the resuscitation fluid, but thereere still signs of life; this was judged as intractableypotension. The other was ICP � 30 mmHg, whichas judged as a nonsurvivable surgical lesion.Animals who survived 30 minutes after polytrauma

ith ICP � 30 mmHg typically manifested hypoten-ion, hypoxia, hypercapnia, and tachycardia. On initialesuscitation (fluid bolus, ventilatory support, supple-ental O2) followed later by CPP management (includ-

ng mannitol and pressor therapy), most physiologicariables corrected in all groups. With no fluid at

igure 3. The gross appearance of the lungs at autopsy. Scalpelandle is reference.

0 minutes (n � 2), there was persistent hypotension (

nd tachycardia. The hypoxia and hypercapnia correctedith ventilatory support and supplemental O2, but car-iac output, mixed venous O2 saturation, and lactatelearance were all depressed. There was no urine output.ne animal survived only 60 minutes, and the other

urvived 300 minutes in these premorbid conditions.his untreated control arm clearly demonstrated that

ome fluid resuscitation was necessary for survival. Everynimal that survived to resuscitation survived to the endf observation with at least some neurologic function, asudged by spontaneous respiratory efforts during venti-ator weaning. This suggests the adequacy of resuscita-ion with the four fluid combinations in these experi-ental conditions.

hysiologic changes with “standard of care”r three different fluid restriction strategiesigure 4 shows MAP and CPP as a function of time. The

eft panel shows that baseline MAP ranged from 85 to00 mmHg and fell to 60 to 65 mmHg by 30 minutesfter polytrauma. The first liter of IVF transiently in-reased MAP 10 to 19 mmHg in the three NS groups,ut the increase was � 35 mmHg (p � 0.059 versusontrol) and more sustained in the Hextend group. Inll groups, MAP increased at 120 minutes, whenannitol was administered and pressor therapy be-

un. From that point until the end of observation,PP was maintained at the � 70 mmHg target in allroups (right panel). But in the Hextend group, CPPas � 70 mmHg immediately on resuscitation

p � 0.023 versus control).Figure 5 shows the IVF required to maintain the MAP

nd CPP targets. Each animal received a bolus of 1 L ofS (approximately 25 mL/kg) during the instrumenta-

ion period and before injury. After polytrauma, eachnimal received 1 L of either NS or Hextend, plus man-itol (1 g/kg), which partly explains the brisk diuresis.ltogether, group 1 required a total of 161 � 6 mL/kgS for resuscitation, administered as a continuous infu-

ion of 36 � 1 mL/kg/h plus 0.18 � 0.04 mg/kg phen-lephrine. Group 2 required about half as much IVF75 � 29 mL/kg; 17 � 6 mL/kg/h, p � 0.002 versusontrol) and also less phenylephrine (0.13 � 0.03 mg/kg;� 0.34 versus control). Group 3 received a single

-L bolus of NS at resuscitation (22 � 1 mL/kg;.8 � 0.2 mL/kg/h; p � 0.001 versus control), thenhenylephrine only for the remainder of the experiment
0.25 � 0.07 mg/kg; p � 0.20 versus control). Group 4

r0eGIIr

t1vCcectpc4wodb

vk1

ts4iv10ca21rwErrs

enou


eceived 28 � 13 mL/kg Hextend (6 � 3 mL/kg/h; p �.001 versus control) and the least amount of phenyl-phrine (0.07 � 0.02 mg/kg; p � 0.06 versus control).roups 3 and 4 were in relative fluid balance because the

VF infusion almost exactly matched the urine output.n contrast, urine output was less than half of the IVFate in groups 1 and 2.

Figure 6 shows other markers of vascular volume. Inhe control group, CVP was maintained at about2 mmHg and hematocrit remained near the baselinealue of about 25, suggesting euvolemia. With Hextend,VP was maintained at 8 mmHg (p � 0.003 versus

ontrol), but hematocrit was also consistent with euvol-mia. In group 2, CVP was 8 mmHg (p � 0.024 versusontrol), but there was a progressive hemoconcentrationo 30 (p � 0.071 versus control), suggesting relative hy-ovolemia. In group 3, hematocrit was similarly in-reased (p � 0.027 versus control) and CVP wasmmHg (p � 0.001 versus control), also consistentith hypovolemia. In all groups, plasma Na�, K�, andsmolarity changes were variable and not considerablyifferent between groups (data not shown). So, the com-

Figure 4. Mean arterial pressure (MAP) and cerebralstandard of care (group 1) and three fluid-restricted gpartially corrected with 1 L of normal saline (NS), butbenefits. One liter of Hextend (Hex) immediately restorefluid was needed to sustain the effects. CVP, central v

ined data in Figures 6 and 7 suggest that intravascular g

olume was similar with 6 mL/kg/h Hextend or 36 mL/g/h NS, but was slightly reduced with either 5 or7 mL/kg/h NS.Figure 7 suggests that at least some IVF was probably

hird spacing into the brain and lungs. The left panelhows that ICP progressively increased to 35 to0 mmHg in group 1, and ICP was 25% to 40% lowern groups 2 and 3 (p � 0.019 and 0.007, respectively,ersus control). With Hextend, ICP remained near5 mmHg for the entire posttrauma observation (p �.001 versus control), except during the brief tests oferebrovascular reactivity each hour (data not shown). Inll groups, increased FiCO2 for 5 minutes provoked a- to 3-mmHg ICP increase before injury and a 6- to0-mmHg increase after injury. This is consistent witheduced cerebrovascular compliance, but the changesere similar in all groups, so the data are not shown.xcept during the CO2 challenges, PaCO2 was tightly

egulated at 40 � 5 mmHg by varying the respiratoryate; there were no differences between groups (data nothown).

The right panel of Figure 7 shows that in the three NS

sion pressure (CPP), right, as a function of time with. After blast-induced polytrauma, MAP and CPP wereional NS was needed to fully restore and sustain theP and CPP and only minimal supplemental intravenouss pressure; PE, phenylephrine.

perfuroupsadditd MA

roups, peak inspiratory pressure on constant tidal vol-

uH5wr

ascf

aline.


me (10 mL/kg/breath) increased from about 20 cm

20 in baseline conditions to about 25 cm H20 afterhours. In the Hextend group, peak inspiratory pressureas about 5 cm H20 higher throughout the posttrauma

esuscitation period (p � 0.01 versus control).

Figure 5. Total intravenous fluid (IVF) mL/kg/h, (inclupressure cerebral perfusion pressure targets. The assomannitol. CUP, central venous pressure; NS, normal s

Figure 6. Other markers of vascular volume status (c
right) in the four groups. Hex, Hextend; NS, normal saline;
There was a clear trend to diminished EEG activitynd also a trend to increases in EEG burst suppres-ion, indicating ischemic EEG changes. But thesehanges were highly variable and not statistically dif-erent between groups, so the data are not shown.

the first liter) required to maintain the mean arteriald urine output reflects the osmotic diuresis caused by

l venous pressure [CVP], left and arterial hematocrit,

dingciate

entra
PE, phenylephrine.

biTbw6ooacawt

amatcb

oo

ta

SFscrsCiptws

ccstacn


Figure 8 shows that brain tissue PO2 decreased from aaseline value of about 12 mmHg to � 2 mmHg afternjury and before resuscitation in all groups (left panel).he first liter of IVF corrected brain tissue PO2 back toaseline in all groups. But in group 3, brain tissue PO2

as not maintained. It slowly decreased to aboutmmHg and stabilized at that value until the end of thebservation period (p � 0.045 versus control). In thether three groups, brain tissue PO2 stabilizedt � 20 mmHg. The right panel shows that PaO2 de-reased from a baseline value of about 225 mmHg tobout 50 mmHg and then corrected to � 200 mmHgith resuscitation. There was no PaO2 difference be-

ween treatments.All other markers of resuscitation and tissue oxygen-

tion were similar between groups. Figure 9 shows thatixed venous O2 saturation fell from a baseline value of

bout 70% to about 25% before resuscitation, and lac-ate increased to � 6 mmol/L. Both of these variablesorrected with resuscitation and there was no differenceetween groups.There were quantitatively similar changes in cardiac

utput, renal blood flow, and portal venous and gastric

Figure 7. Intracranial pressure (mmHg), left, and peakspacing of intravenous fluid in the four groups. CVP, cPE, phenylephrine.

xygenation with injury and with resuscitation, and a

here were no differences between groups, so those datare not shown.

emiquantitative neuropathologic changesigure 10 shows representative histologic specimenstained with hematoxylin and eosin from the cerebralortex from an uninjured animal (panel A), and cor-esponding sections after injury and treatment withtandard of care (panel B) and with Hextend (panel). With standard of care (group 1), there was a strik-

ng degree of neuronal shrinkage (arrows) and neuro-il vacuolization (edema). With Hextend (group 4),here were several dark, slightly shrunken neurons,ith prominent perineuronal spaces (arrows). The

cale bar � 120 �m.In an attempt to quantitate the neuropathologic

hanges within and between groups, sections from theerebral cortex were evaluated in a blinded fashion. Table 1hows that every animal had evidence of severe TBI, buthere was large variability within and between individu-ls. Changes included subarachnoid hemorrhage, pete-hiae, ischemia, neutrophil infiltrates with focal areas ofecrosis, shrunken, darkened neurons, Alzheimer type II

iratory pressure ECM H2O; right, reflect possible thirdl venous pressure; Hex, Hextend; NS, normal saline;

inspentra

strocytosis, or spongiosis (neurophil edema). No apop-

ttef

ifsn


osis or spheroids and, or retraction balls were noted, buthat is not surprising at this early time point. Withinach group, the mean (�standard error) was calculatedrom unweighted averages from each category. Accord-

Figure 8. Brain tissue PO2 (mmHg), left, and PaO2 (mmPO2 was not maintained in group 3, but was � 20 msaline; PE, phenylephrine.

Figure 9. Systemic markers of resuscitation. Mixed ve
corrected in all four groups. Hex, Hextend; NS, normal salin
ng to this system, the overall damage was judged asollows: group 2 � group 3 � group 1 � group 4. Onepecimen was not analyzed in group 3 because of tech-ical reasons.

ight, versus time in four groups. Note that brain tissuein the other three groups. Hex, Hextend; NS, normal

O saturation (%), left, and arterial lactate (mm), right,

Hg), rmHg

nous
2
e; PE, phenylephrine.

DOasapfmfdbttspb

fielabw(gy

“

NmprerR3oawttbgatccawdtcuooc

d of c


ISCUSSIONur soldiers suffer blast injuries almost every day in Iraq

nd Afghanistan.28-32 In the best case, if medical re-ources are available, blood loss can be approximated,nd resuscitation is targeted to end points such as bloodressure, heart rate, or urine output. In the worst case, inield situations, emergency care is delayed, there is noonitoring capability, or resuscitation is limited to the

luids carried by the first responder. In either case, it isifficult to evaluate which therapeutic modality is bestecause there are no reproducible large animal models toest the ideas. To our knowledge, this is the first con-rolled laboratory study to compare fluid restriction re-uscitation strategies after the type of complicated severeolytrauma that might result from a blast injury on theattlefield.In this study, the severity of the polytrauma is re-

lected by the 37% (13 of 35) mortality before random-zation (Fig. 1). Animals that survived to the point ofmergency medical care lived at least 6 hours with ateast some neurologic function, which compares favor-bly to reports of casualties evacuated alive from theattlefield.7,9-12,15-17,28-32 In all survivors, MAP and CPPere stable (Fig. 4), there was adequate urine output

Fig. 5), and lactate was cleared (Fig. 9), which, at firstlance, suggests that all four resuscitation strategiesielded satisfactory results.

The major new findings are that with a reasonable

Figure 10. Representative histologic specimens stained with hemaand corresponding sections after injury and treatment with standar

standard of care” (group 1), which includes an initial f

S bolus followed by judicious IVF administration toaintain MAP � 60 mmHg at a moderate filling

ressure (CVP � 12 mmHg), brain oxygenation wasestored (Fig. 8), but ICP reached dangerous lev-ls � 35 mmHg within the 6-hour observation pe-iod (Fig. 7), even with mannitol and pressor therapy.educing IVF by about half (group 2) reduced ICP by0% (to 25 mmHg); this ICP value remained danger-usly high, according to current evidence-based man-gement guidelines for severe TBI.14 Brain oxygenationas maintained, but there was evidence of hemoconcen-

ration and a volume deficit (Fig. 6). If IVF was addi-ionally restricted to a single 1-L bolus of NS (group 3),rain oxygenation was not maintained. In all three NSroups, CPP was maintained at � 70 mmHg onlyfter mannitol and pressor therapy. In striking con-rast, an initial 1-L bolus of Hextend immediatelyorrected CPP � 70 mmHg, which is important be-ause CPP maintenance is the cornerstone of TBI man-gement.33,34 In addition, in group 4, brain oxygenationas maintained, and overall IVF requirements were re-uced by one-sixth relative to group 1 (Fig. 5). Takenogether, these observations are consistent with the con-lusion that intracranial hypertension was acutely atten-ated, and brain oxygenation was acutely maintainednly when IVF was restricted and that resuscitation wasptimized when Hextend was used instead of NS.This isonsistent with accumulating evidence that Hextend of-

in and eosin from the cerebral cortex from an uninjured animal (A),are, group 1 (B) and with Hextend, group 4 (C).

toxyl

ers unique advantage, relative to crystalloid, for combat

rla

atdpbcchcht

at

cpctwmdr(vdibtbi

T

S

G

G

G

G

EP


esuscitation and other situations in which resources areimited, such as emergency medical transport or duringmass casualty event.16,17

Current military doctrine for treating combat injuriesllows permissive hypotensive resuscitation, based onhe assumption that aggressive normalization of hemo-ynamics might exacerbate bleeding.17,35 There are am-le theoretic data to support this concept.36-38 Becauseleeding extremity wounds have been the most commonombat injury throughout history,15 the importanceannot be overstated. On the other hand, permissiveypotensive resuscitation could be catastrophic afteroncussion from an IED because even brief episodes ofypotension or hypoxia can double morbidity and mor-ality after TBI.13,14,33,34

In the current war in Iraq, the majority of casualtiesre caused by IEDs,1 and TBI is common.11 In theory,

able 1. Distribution of Neuropathologic Changes

pecimenDark

neurons SpongiosiPMN

infiltrationAlzheimer

chang

roup 1#12 ��

#14 ��

#15 ��

#17 � ��

#19 ��

Average 1.0 0.2 1.4 0.0roup 2#3 ��

#5 ��

#6 �

#9 � � �

#11 � � ��

Average 1.6 1.2 1.0 1.4roup 3#13 � ��

#16 ��

#18 ��

#20 ��

Average 0.8 0.2 1.8 1.8roup 4#2 ��

#4#7 ��

#8 �

#10 ��

Average 1.0 0.4 0.8 0.4

ach specimen was graded on a 0–3 scale with 0 as normal and �� as sevMN, polymorphonuclear leukocyte; SAH, subarachnoid hemorrhage.

he injury caused by an IED can be divided into four p

omponents.9,10 Primary blast injury is caused by theressure wave and occurs almost exclusively in gas-ontaining organs, such as the lung, ear, and gastrointes-inal tract. The ear is the most sensitive to this pressureave, but damage to the lung is responsible for the mostorbidity and mortality. Gross lung changes include

iffuse contusions with stiffened heavy tissue, pleuralents, pneumothorax, or mediastinal extravasation of aireg, Fig. 3). Microscopic lung changes include intraal-eolar hemorrhage with perivascular or peribronchialisruption and leukocyte infiltration. Secondary blast

njury results from blunt or penetrating trauma causedy energized objects. Tertiary blast injury occurs whenhe whole body collides with fixed objects. Quaternarylast injury results from burns and exposure to toxicnhalants.7,9

The whole body polytrauma caused by an IED de-

II Focalnecrosis

Focalischemia Petechia SAH

Score(mean�SE)

�

� �

��

�

� ��

0.8 1.6 0.8 0.4 0.9 � 0.2

� ��

� ��

��

�

��

0.8 1.4 1.0 1.2 1.2 � 0.1

� ��

��

��

��

0.8 1.5 1.8 1.2 1.2 � 0.2

��

��

�

�

�

0.8 0.4 0.0 0.6 0.6 � 0.1

typee

ere.

ends on several factors, including blast kinetic energy,

fltciNmb

FcfaelaipalmtormfaoorIgtvtnpAisltifr

d

tthtpbvhbtelaweTfntwf

Ftkpraeiseesiprbldclsmtpt


ragments and projectiles loaded in the bomb oraunched by the explosion, proximity of the patient tohe blast, barriers that partially deflect the energy, and ofourse, biologic variability in the patient. So no twonjuries or compensatory responses are exactly alike.evertheless, if the injury is survivable, there are com-on patterns and general principles that can serve as the

asis for model development.Our IED model was based on three assumptions.

irst, we reasoned that blast overpressure would likelyause some structural lung damage, but have no obviousunctional consequence. This reasoning is based onfter-action reports that confirm that refractory hypox-mia from the blast is relatively uncommon in a venti-ated patient evacuated from the battlefield.10,15,28-31 Inny case, the pathophysiology and treatment of blastnjury to the lung are almost exactly the same as theathophysiology and treatment for blunt chest traumand pulmonary contusion.39 In our model, and in realife, symptomatic treatment includes judicious IVF ad-

inistration, supplemental O2, and appropriate ventila-ory support with positive end expiratory pressure. If anyne of these conditions is not met, then a progressiveespiratory failure develops. Second, we reasoned that inost cases, penetrating TBI after an IED is lethal. It

ollows that the most common pattern of survivable TBIfter IED is blunt trauma caused by either energizedbjects striking the head or the head striking a fixedbject. Of course, survivability depends also on the di-ection of the blast and anatomic location of the injury.n our model, the blast was applied in the occipital re-ion and in an anterocaudal direction perpendicular tohe spinal cord. Third, we reasoned that neurotraumaictims in real life would receive current standards ofrauma care, which would include CPP management, ifecessary. We have a unique perspective on currentroblems related to combat casualty care because the USrmy has selected the University of Miami as the train-

ng site for its forward surgical teams. Whenever ouroldiers are in battle, these teams are near the frontines. Since September 11, 2001, most of the army’seams deployed in Iraq and Afghanistan have trainedn our laboratory.40 Feedback and after-action reportsrom these teams have grounded our research firmly ineality.

Hextend is an FDA-approved hetastarch solution in-
icated for hypovolemia during elective surgery,41 but a
here have been no randomized clinical trials evaluatinghe use of Hextend for trauma resuscitation. Relative toetastarch in saline, Hextend has a beneficial coagula-ion profile, less antigenicity, and some antioxidantroperties.42,43 Hextend prevents hyperchloremic meta-olic acidosis and improves gastric mucosal perfusionersus saline solutions.44 The potential value of Hextendas been recognized in a review of special issues faced byattlefield medics.17 But it should be emphasized thathe use of Hextend in trauma patients is off-label, itsffects are transient, and there is a potential for coagu-opathy.45 It is regarded as a temporizing measure onlynd is not recommended for those with TBI. Before ourork, there was no evidence that Hextend was safe and

ffective for resuscitation after trauma with or withoutBI. In three different trauma models, we showed that

luid requirements were reduced by at least half, pulmo-ary and cerebrovascular function were improved, andhere was no adverse effect on the coagulation profileith Hextend relative to crystalloid.19,23,46 The results

rom this study affirm those results.There are at least four major limitations to this study.

irst, for ethical reasons, all animals were anesthetized athe time of injury and throughout the experiment. Theetamine or fentanyl anesthesia could impart either arotective or a deleterious effect that may differ from theeal-life nonanesthetized trauma patient. Although annesthesia artifact may have existed, it was imposedqually on all groups. Second, the magnitude of the lungnjury in this study was somewhat less than that in oureveral previous studies,22-24 even though the kinetic en-rgy delivered to the chest wall was similar. The exactxplanation for this difference was not rigorously pur-ued, but the animals were about 15% to 20% larger, thenjury was delivered in the prone rather than supineosition, there was no superimposed hemorrhage, andesuscitation IVF was guided by MAP targets rather thanlood volume lost. Third, there was no longterm fol-

owup, and it is possible that the short-term physiologicifferences between groups have no neurologic or otherlinical consequence. There were also no obvious histo-ogic differences after 6 hours (Table 1), so longtermtudies are clearly indicated. Fourth, the model did notimic the shrapnel, heat, or electromagnetic radiation

hat can be caused by an IED. Also, the captive bolt gunsroduced injuries to the brain and lungs that varied be-ween individuals. On the other hand, relative to an
ctual explosion, recreating the polytrauma with captive

bmti

pTwstfrhmctwsmtstamtsqp

ASAA

DC

Azsf(t(etL

R

1

1

1

1

1

1

1

1

1

1


olt guns was relatively inexpensive in terms of equip-ent and manpower, and the contribution and direc-

ionality of the various injury components could be eas-ly manipulated.

In summary, we described a new model of blast injuryolytrauma using captive bolt guns in anesthetized pigs.he anatomic injury pattern and functional changesere relatively reproducible and appeared to mimic

ome key aspects of (survivable) IED-related neuro-rauma. The 40% mortality before resuscitation re-lected injury severity, but the 0% mortality thereaftereflected the efficacy of current standards of field andospital care with almost immediate definitive treat-ent and CPP-directed therapy. In these nearly ideal

ircumstances, the physiologic sequelae of severe neuro-rauma were attenuated for at least 6 hours when IVFas restricted. Additional work is needed to define re-

ponses with evacuation delays or less definitive treat-ent. Also, the data showed that, by many criteria, Hex-

end was superior to crystalloid, even at the sametandard resuscitation end points. More work is neededo determine if these apparent short-term benefits haveny longterm neurologic consequences. Optimal fluidanagement is essential in the treatment of severe TBI

o reduce the formation of cerebral edema and avoidecondary injury. Until now, there have been no ade-uate models to study these effects after blast-inducedolytrauma.

uthor Contributionstudy conception and design: Earle, de Moya, Proctorcquisition of data: Earle, de Moya, Zuccarellinalysis and interpretation of data: Earle, Norenberg,Proctorrafting of manuscript: Earle, Norenberg, Proctorritical revision: Earle, Norenberg, Proctor

cknowledgment: We appreciate the efforts of Connie Jant-en of Vision Research Inc (Stuart, FL), who made the highpeed digital videos; Victor Castro of Zoll Medical (Chelms-ord, MA), for the hemodynamic monitors; Dave GloverHutchinson Technology, Hutchinson, MN) for providinghe NIR system; George Beck of Impact InstrumentationWest Caldwell, NJ) for providing the ventilators; and Conc-tta Gorski, RN, BS, CCRA, Integra LifeSciences Corpora-ion, (Plainsboro, NJ) for providing the camino monitors and
iCOX probes.
EFERENCES

1. Coalition casualties in Iraq and Afghanistan. Available at: http://icasualties.org/oif/default.aspx. Accessed December 6, 2006.

2. Mallonee S, Shariat S, Stennies G, et al. Physical injuries andfatalities resulting from the Oklahoma City bombing. JAMA1996;276:382–387.

3. Gutierrez de Ceballos JP, Turegano Fuentes F, Perez Diaz D,et al. Casualties treated at the closest hospital in the Madrid,March 11, terrorist bombings. Crit Care Med 2005;33(1 Suppl):S107–112.

4. Rodoplu U, Arnold JL, Tokyay R, et al. Mass-casualty terroristbombings in Istanbul, Turkey, November 2003: report of theevents and the prehospital emergency response. Prehospital Di-saster Med 2004;19:133–145.

5. Kluger Y, Peleg K, Daniel-Aharonson L, et al. The special injurypattern in terrorist bombings. J Am Coll Surg 2004;199:875–879.

6. Ryan J, Montgomery H. The London attacks–preparedness:Terrorism and the medical response. N Engl J Med 2005;353:543–545.

7. Nelson TJ, Wall DB, Stedje-Larsen ET, et al. Predictors of mor-tality in close proximity blast injuries during Operation IraqiFreedom. J Am Coll Surg 2006;202:418–422.

8. Majde JA. Animal models for hemorrhage and resuscitation re-search. J Trauma 2003;54(5 Suppl):S100–10-5.

9. Bowen TE, Bellamy RF, eds. Emergency war surgery: secondUnited States revision of the emergency war surgery NATOhandbook. Chapter 5: Blast injuries. Washington, DC: UnitedStates Department of Defense, United States Government;1988.

0. DePalma RG, Burris DG, Champion HR, Hodgson MJ. Blastinjuries. N Engl J Med 2005;352:1335–1342.

1. Okie S. Traumatic brain injury in the war zone. N Engl J Med2005;352:2043–2047.

2. Montgomery SP, Swiecki CW, Shriver CD. The evaluation ofcasualties from operation Iraqi freedom on return to the conti-nental United States from March to June 2003. J Am Coll Surg2005;201:7–12.

3. Robertson CS, Valadka AB, Hannay HJ, et al. Prevention ofsecondary ischemic insults after severe head injury. Crit CareMed 1999;27:2086–2095.

4. Management and Prognosis of Severe Traumatic Brain Injury. Ajoint project of the Brain Trauma Foundation and the AmericanAssociation of Neurological Surgeons, section on Neurotraumaand Critical Care. Available at http://braintrauma.org. AccessedDecember 6, 2006.

5. Champion HR, Bellamy RF, Roberts CP, Leppaniemi A. A pro-file of combat injury. J Trauma 2003;54(5 Suppl):S13–19.

6. Krausz MM. Fluid resuscitation strategies in the Israeli army.J Trauma 2003;54(5 Suppl):S39–42.

7. Holcomb JB. Fluid resuscitation in modern combat casualtycare: lessons learned from Somalia. J Trauma 2003;54(5 Suppl):S46–51.

8. Sanui M, King DR, Feinstein AJ, et al. Effects of argininevasopressin during resuscitation from hemorrhagic hypoten-sion after traumatic brain injury. Crit Care Med 2006;34:433–438.

9. King DR, Cohn SM, Proctor KG. Changes in intracranial pres-sure, coagulation and neurologic outcome after resuscitationfrom experimental traumatic brain injury with Hetastarch. Sur-
gery 2004;136:355–363.
http://icasualties.org/oif/default.aspx

http://icasualties.org/oif/default.aspx

http://braintrauma.org

2

2

2

2

2

2

2

2

2

2

3

3

3

3

3

3

3

3

3

3

4

4

4

4

4

4

4


0. Hayashida M, Chinzei M, Komatsu K, et al. Detection of cere-bral hypoperfusion with bispectral index during paediatric car-diac surgery. Br J Anaesth 2003;90:694–698.

1. Myles PS, Cairo S. Artifact in the bispectral index in a patient withsevere ischemic brain injury. Anesth Analg 2004;98:706–707.

2. Feinstein AJ, Cohn SM, Sanui M, et al. Early vasopressin im-proves short term survival after pulmonary contusion. J Trauma2005;59:876–883.

3. Kelly ME, Miller PR, Greenhaw JJ, et al. Novel resuscitationstrategy for pulmonary contusion after severe chest trauma.J Trauma 2003;55:94–105.

4. Desselle WJ, Greenhaw JJ, Trenthem LL, et al. Macrophage cyclo-oxygenase expression, immunosuppression, and cardiopulmonarydysfunction after chest trauma. J Trauma 2001;51:239–252.

5. Feinstein AJ, Patel MB, Sanui M, et al. Resuscitation with pres-sors after traumatic brain injury. J Am Coll Surg 2005;201:536–545.

6. Panickar KS, Norenberg MD. Astrocytes in cerebral ischemicinjury: morphological and general considerations. Glia 2005;50:287–298.

7. Norenberg MD. Astrocyte responses to CNS injury. J Neuro-pathol Exp Neurol 1994;53:213–220.

8. Patel TH, Wenner KA, Price SA, et al. A U.S. Army ForwardSurgical Team’s experience in Operation Iraqi Freedom.J Trauma 2004;57:201–207.

9. Chambers LW, Rhee P, Baker BC, et al. Initial experience of USMarine Corps forward resuscitative surgical system during Op-eration Iraqi Freedom. Arch Surg 2005;140:26–32.

0. Gawande A. Casualties of war–-military care for the woundedfrom Iraq and Afghanistan. N Engl J Med 2004;351:2471–2475.

1. Bilski TR, Baker BC, Grove JR, et al. Battlefield casualtiestreated at Camp Rhino, Afghanistan: lessons learned. J Trauma2003;54:814–822.

2. Butler FK Jr, Hagmann JH, Richards DT. Tactical managementof urban warfare casualties in special operations. Mil Med 2000;165(4 Suppl):1–48.

3. Rosner MJ, Rosner SD, Johnson AH. Cerebral perfusion pres-sure: management protocol and clinical results. J Neurosurg
1995;83:949–962.
4. Ling GS, Neal CJ. Maintaining cerebral perfusion pressure is aworthy clinical goal. Neurocrit Care 2005;2:75–81.

5. Bickell WH, Wall MJ Jr, Pepe PE, et al. Immediate versusdelayed fluid resuscitation for hypotensive patients with pen-etrating torso injuries. N Engl J Med 1994;331:1105–1109.

6. Dubick MA, Atkins JL. Small-volume fluid resuscitation for thefar-forward combat environment: current concepts. J Trauma2003;54(5 Suppl):S43–5.

7. Sondeen JL, Coppes VG, Holcomb JB. Blood pressure at whichrebleeding occurs after resuscitation in swine with aortic injury.J Trauma 2003;54(5 Suppl):S110–117.

8. Rafie AD, Rath PA, Michell MW, et al. Hypotensive resuscita-tion of multiple hemorrhages using crystalloid and colloids.Shock 2004;22:262–269.

9. Klein Y, Cohn SM, Proctor KG. Lung contusion-pathophysiologyand management. Curr Opin Anesthesiol 2002;15:65–68.

0. King DR, Patel MB, Feinstein AJ, et al. Simulation training fora mass casualty event: A two year experience at the Army TraumaTraining Center. J Trauma 2006;61:943–948.

1. Roche AM, Mythen MG, James MF. Effects of a new modifiedbalanced hydroxyethyl starch preparation (Hextend) on mea-sures of coagulation. Br J Anaesth 2004;92:154–155.

2. Gan TJ, Bennett-Guerrero E, Phillips-Bute B, et al. Hextend,a physiologically balanced plasma expander for large volumeuse in major surgery: a randomized phase III clinicaltrial. Hextend Study Group. Anesth Analg 1999;88:992–998.

3. Nielsen VG, Tan S, Brix AE, et al. Hextend (hetastarch solution)decreases multiple organ injury and xanthine oxidase releaseafter hepatoenteric ischemia-reperfusion in rabbits. Crit CareMed 1997;25:1565–1574.

4. Wilkes NJ, Woolf RL, Powanda MC, et al. Hydroxyethyl starchin balanced electrolyte solution (Hextend)—pharmacokineticand pharmacodynamic profiles in healthy volunteers. AnesthAnalg 2002;94:538–544.

5. Wiedermann CJ. Hydroxyethyl starch—can the safety problemsbe ignored? Wien Klin Wochenschr 2004;116:583–594.

6. Crookes BA, Cohn SM, Bonet H, et al. Building a better fluidfor emergency resuscitation of traumatic brain injury. J Trauma
2004;57:547–554.

Cerebrovascular Resuscitation after Polytrauma and Fluid ... · Cerebrovascular Resuscitation after...

Documents

Transcript of Cerebrovascular Resuscitation after Polytrauma and Fluid ... · Cerebrovascular Resuscitation after...