ww.sciencedirect.com

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4

Available online at w

journal homepage: www.JournalofSurgicalResearch.com

Mechanical ventilation increases the inflammatory responseinduced by lung contusion

Karlijn J.P. van Wessem, MD,a,* Marije P. Hennus, MD,b Linda van Wagenberg, MD,a

Leo Koenderman, PhD,c and Luke P.H. Leenen, MD, PhDa

aDepartment of Trauma Surgery, University Medical Center Utrecht, Utrecht, The NetherlandsbDepartment of Pediatric Intensive Care, Wilhelmina Children’s Hospital, Utrecht, The NetherlandscDepartment of Respiratory Medicine, University Medical Center Utrecht, Utrecht, The Netherlands

a r t i c l e i n f o

Article history:

Received 24 October 2012

Received in revised form

10 December 2012

Accepted 20 December 2012

Available online 22 January 2013

Keywords:

Blast injury

Mechanical ventilation

Systemic inflammation

Pulmonary inflammation

Neutrophils

* Corresponding author. Department of TrauNetherlands. Tel.: þ31 88 75 598 82; fax: þ31

E-mail address: kwessem@umcutrecht.n0022-4804/$ e see front matter ª 2013 Elsevhttp://dx.doi.org/10.1016/j.jss.2012.12.042

a b s t r a c t

Background: Posttraumatic lung contusion is common after blunt chest trauma, and

patients often need ventilatory support. Lung contusion induces an inflammatory response

signified by primed polymorph neutrophil granulocytes (PMNs) in blood and tissue.

Mechanical ventilation (MV) can also cause an inflammatory response. The aim of this

study was to develop an animal model to investigate the effect of high-volume ventilation

on the inflammatory response in blunt chest trauma.

Materials and methods: We assigned 23 male Sprague-Dawley rats to either MV or bilateral

lung contusion followed by MV. We used three extra rats as controls. Lung contusion was

induced by a blast generator, a device releasing a single pressure blast wave centered on

the chest. We determined tissue and systemic inflammation by absolute PMN numbers in

blood and bronchoalveolar lavage fluid (BALF), myeloperoxidase, interleukin (IL)-6, IL 1b,

growth-related oncogeneeKC, and IL-10 in both plasma and BALF.

Results: Survival after blunt chest trauma was correlated to the distance to the blast

generator. Compared with controls, both MV and blast plus MV rats showed increased

systemic and pulmonary inflammation, expressed by higher PMNs, myeloperoxidase

levels, and cytokine levels in both blood and BALF. Blast plus MV rats showed a higher

systemic and pulmonary inflammatory response than MV rats.

Conclusions: The blast generator generated reproducible blunt chest trauma in rats.

Mechanical ventilation after lung contusion induced a larger overall inflammatory

response than MV alone, which indicates that local damage contributes not only to local

inflammation, but also to systemic inflammation. This emphasizes the importance of lung

protective ventilation strategies after pulmonary contusion.

ª 2013 Elsevier Inc. All rights reserved.

1. Introduction accidental deaths. However, they have resulted in an

Trauma is the leading cause of death in men under 40 y of age

[1]. Improvements in seatbelts, airbags, and vehicle

constructions have led to a decline in the number of

ma Surgery, University M88 75 550 15.

l (K.J.P. van Wessem).ier Inc. All rights reserved

increased number of surviving victims with blunt chest

trauma. In addition, trauma represents the second leading

cause of life-threatening acute lung injury [1]. After initial

nonfatal injury (thoracic), trauma can induce a systemic

edical Center Utrecht, Heidelberglaan 100, 3584 CX Utrecht, The

.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4378

inflammatory response by activating the immune system

[2]. This activation is mainly caused by primed poly-

morphonuclear leukocytes, which are prone to home and

become activated in the tissues when they encounter addi-

tional local inflammatory stimuli. The lung is a preferred site

for homing because of the large and narrow microvascular

bed and long transit time [3,4]. Excessive immune activation

can result in a systemic inflammatory response syndrome and

multiple organ dysfunction syndrome [2e5].

Blast injuries are a special form of blunt trauma that cause

serious internal injuries, often without evidence of external

lesions. Blast injury occurs after a sudden change in pressure,

originating from an explosion, for example. The organs most

affected by blast injuries are the hollow, gas-filled organs such

as ears, lungs, and gastrointestinal tract, and to a lesser

extent, the cardiovascular and central nervous system. The

lungs are almost always affected by blast injuries.

Treatment of blunt thoracic injury includes resuscitative

fluids and respiratory support bymechanical ventilation (MV).

Mechanical ventilation in itself can also cause an inflamma-

tory response in the lung by activating the immune system

[6e8]. In addition, several studies have demonstrated that

when MV was applied in the presence of a pulmonary infec-

tion, pulmonary inflammation was further enhanced [9,10].

This phenomenon is also known as a second hit model.

To investigate a possible synergistic inflammatory response

between traumatic lung injury andMV,we designed a ratmodel

combining isolated bilateral pulmonary contusionwith MV.We

hypothesized that a combination of pulmonary contusion

and MV would cause a higher inflammatory response than

MV alone. A possible synergy is clinically relevant because

an increased inflammatory response might cause more

inflammation-related complications. This could be influential

in choosing MV strategies.

2. Materials and methods

The Animal Care and Use Committee of the University

Medical Center Utrecht, Utrecht, The Netherlands, approved

this. All animal procedures were carried out in compliance

with national and international standards for use of labora-

tory animals.

2.1. Animal preparation

We performed experiments with healthy adult male Sprague-

Dawley rats (Harlan, Zeist, The Netherlands) weighing

300e480 g. The animals acclimated for at least 7 d in our

animal facility with free access to standard rodent food and

water.

2.2. Experimental design

We subjected 23 rats to either MV alone (six rats) or MV after

blast-induced pulmonary contusion (blast plus MV, 17 rats).

Three extra ratswere directly killed after induction anesthesia

and served as controls to obtain physiological baseline levels.

At the start of the experiment, we anesthetized all rats

using inhalation anesthesia (5% isoflurane [Pharmachemie

BV, Haarlem, The Netherlands], 1 L/min oxygen, and 1 L/min

room air). Next, we maintained anesthesia using a mix of

0.9 mL/kg ketamine (100 mg/mL; AST Farma BV, Oudewater,

The Netherlands), 0.5 mL/kg dexmedetomidine (0.5 mg/mL;

Orion Corporation, Espoo, Finland) and 0.05 mL/kg atropine

(1.0 mg/mL; American Regent, Inc, Shirley, NY) intraperito-

neally. We maintained body temperature between 36�C and

38�C with a heating pad to prevent hypothermia. We inserted

a silicone catheter (0.02 � 0.037 inches; Degania Silicone Ltd,

Hatzor HaGlilit, Israel) into the right carotid artery of all rats to

monitor blood pressure and draw blood. Pain was relieved by

0.3 mL buprenorphine, 0.3 mg/mL/h, 10% intramuscularly

(AST Farma BV).We performed blood gas analysis at t¼ 0 (only

MV rats), 30, 90, 180, 240 and 300 min using a pH/blood gas

analyzer (ABL 505; Radiometer, Copenhagen, Denmark).

Rats subjected to blast plus MV had no blood gas analysis at

t ¼ 0 min because that was the time of the blast. Their first

blood gas analysis was done after 30 min.

2.3. Pulmonary contusion

Animals randomized for pulmonary contusionwere subjected

to a pressure blast. We induced blunt chest trauma by a single

blast wave centered on the chest. For these experiments, we

used a blast wave generator previously described by Jaffin et al.

[11]. The blast generator consisted of two parts. The upper

section served as a pressure reservoir and was separated from

the lower nozzle by a 50-mmMylar polyester film (Du Pont, Bad

Homburg, Germany). The pressure reservoir was connected to

a storage tank of compressed air. Between both components,

an electronically releasable high-speed valve and a pressure-

reducing valve set to 15 bar were interposed. In the experi-

ments, the blast wave generator was directed with the nozzle

toward the animal’s chest. The distance between the nozzle

and animal’s chest varied from 3.5 to 6.0 cm. By opening the

high-speed valve, compressed air was delivered into the upper

section of the generator. When the pressure in this compart-

ment exceeded the resistance of the polyester diaphragm, the

film rapidly ruptured toward the nozzle, releasing a repro-

ducible single blast wave. The reproducibility of each blast

wave was assessed by two pressure transducers (TDS3032B;

Tektronix, Maarssen, The Netherlands) on both sides of the

animal’s chest. We recorded both the maximal pressure and

the duration of the blast wave.

2.4. Mechanical ventilation

We tracheotomized rats and inserted a metal cannula. After

the blast, the cannula was connected to a ventilator (Servo

Ventilator 900C; Siemens-Elema, Solna, Sweden) and rats

were ventilated for 5 h in a pressure-controlled, time-cycled

mode (positive end-expiratory pressure 5 cm H2O, pressure

control þ20 cm H2O, FiO2 0.50, frequency 10e20/min). We

made ventilator adjustments only in frequency to aim for

normocapnia. Muscle relaxationwas attainedwith 2mg/kg/90

min pancuronium bromide (Pavulon; Organon Technika,

Boxtel, The Netherlands).

To prevent hemodynamic instability during mechanical

ventilation, we gave all rats 10 ml/kg/h normal saline, as

proposed by Vreugdenhil et al. [12].

Table 2 e Blast wave characteristics related to distance tonozzle.

Distance tonozzle (cm)

Blast pressure (bar) Blast time (ms)

3.5 2.86 � 0.0 0.5 � 0.03

4.0 3.20 � 0.36 0.6 � 0.04

4.5 3.81 � 0.29 0.5 � 0.03

6.0 4.65 � 0.24* 0.5 � 0.03

Values are expressed as means � SEM.

* P < 0.05; the distance of 6.0 cm showed significantly higher blast

pressure compared with all other distances.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4 379

2.5. Blood and tissue collection

We killed rats after 300 min by cardiac puncture. We collected

blood in a heparinized syringe. The lungs were harvested and

weighted to calculate the lung to body weight ratio (LBWR),

which we used as a quantitative measure for lung injury. We

took photographs and assigned affected lungs numerical

values according to the Thoracic Injury Subjective Severity

Score (TISSS), as described by Jaffin et al. [11] (Table 1). Three of

the authors (K.W., M.H., and L.W.) independently measured

TISSS. In case of any discrepancy (n ¼ 2), the mean TISSS was

calculated. The left lung was used for bronchoalveolar lavage

analysis and the right lung was snap-frozen in liquid nitrogen

and stored at �80�C until further analysis.

Weusedneutrophil count; interleukin (IL)-6; (growth-related

oncogene (GRO-KC), the rodent equivalent of human IL-8); IL1b;

and IL-10 in blood to measure systemic inflammation. The

collected blood samples were centrifuged and the plasma was

collected and stored at �80�C. We measured cytokine levels

with the Luminex multiplexing platform (Millipore, Amster-

dam, The Netherlands) as per the manufacturer’s specifica-

tions. The remaining blood was lysed twice in ice-cold isotonic

NH4Cl solution followed by centrifugation at 4�C. After a final

wash with phosphate-buffered saline with added sodium

citrate [0.38% wt/vol] and isotonic pasteurized plasma proteins

[10% vol/vol], we calculated the neutrophil percentage. Differ-

ential cell countswere enumerated on cytospin-prepared slides

stained with May-Grunwald and Giemsa.

2.6. Bronchoalveolar lavage fluid analysis

After termination, we lavaged the left lung three times with 5

mL isotonic saline at 37�C. Bronchoalveolar lavage fluid (BALF)

samples were centrifuged at 1500 rpm for 5 min at 10�C(Allegra 6R Centrifuge; Beckman Coulter, Woerden, The

Netherlands). We resuspended the pellet in 200 mL phosphate

buffered saline, after which we mixed 20 mL BALF with Turks

suspension (lysis buffer). We determined the total cell count

of viable BAL cells using a grid hemocytometer (Fuchs-

Rosenthal, Marienfeld, Germany). Differential cell counts

were enumerated on cytospin-prepared slides stained with

May-Grunwald and Giemsa.

We measured BALF IL-6, GRO-KC, IL1b, and IL-10 levels

with the Luminex multiplexing platform as per the manu-

facturer’s specifications.

2.7. Lung myeloperoxidase activity

We thawed part of the frozen right lung, homogenized it in

a buffer agent (4-[2-hydroxyethyl]-1-piperazineethanesulfonic

Table 1 e Thoracic Subjective Severity Score according toJaffin et al. [11].

TSSS Macroscopic injury to lungs

0.5 <10% surface of pleural redness

1 <50% surface of pleural redness

2 50% surface of pleural redness

3 >50% surface of pleural redness

acid), and centrifuged it for 30 min at 5000 rpm. The pellet

was dissolved in 10 mmol/L sodium citrate buffer containing

25% hexadecyltrimethyl ammonium bromide (CTAC; Sigma-

Aldrich, Zwijndrecht, The Netherlands). We assayed an

aliquot of the supernatant by measuring H2O2-dependent

oxidation of tetramethylbenzidine (Sigma-Aldrich) in phos-

phoric acid. We measured absorbance at 450 nm of visible light

and calculated myeloperoxidase (MPO) activity in units per

microgram lung tissue.

2.8. Statistical analysis

Data are expressed as mean � standard error (SEM). P < 0.05

was considered statistically significant. We performed calcu-

lations using Graphpad Prism software (version 5; San Diego,

CA). For analysis of differences between groups, we used the

Kruskal-Wallis test with post hoc Dunn’s correction and per-

formed further subanalyses with the Mann-Whitney test.

3. Results

3.1. Reproducibility and injury severity generated byblast generator waves

We tested four different distances (3.5, 4.0, 4.5, and 6.0 cm)

between the nozzle of the blast generator and the animal’s

chest. Analysis of the peak pressures generated showed

a significant increase in blast pressure at a distance of 6.0 cm

compared with all other distances (Table 2). However, the

duration of the blast wave showed no difference between

distances (Table 2), indicating a highly reproducible blast

wave. We observed no abdominal injury. Nine of 17 rats (53%)

that underwent a blast survived. Four rats died instantly after

Table 3 e Thoracic injury severity and survival related todistance to nozzle.

Distance to nozzle (cm) TSSS Survival (%)

3.5 3.0 � 0.0 1/3 (33%)

4.0 2.0 � 1.0 1/2 (50%)

4.5 1.8 � 0.4 2/6 (33%)

6.0 1.4 � 0.5 5/6 (83%)

Values are expressed asmeans� standard error of themean (TSSS)

or percentage (survival).

Fig. 1 e Pulmonary contusion in rat lungs: TSSS compared with normal rat lungs. (Color version of figure is available online.)

non-s

urviv

ors

surviv

ors

0

1

2

3

*

TS

SS

Fig. 2 e Relation between TSSS and survival (*P < 0.05).

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4380

the blast because of great vessel rupture and/or cardiac

rupture. Four other rats died during the course of the experi-

ment (after 40, 90, 120, and 150 min, respectively, owing to

cardiac tamponade and/or massive bilateral hemothorax).

They were all excluded from further analysis. Closer distance

to the nozzle showed a higher Thoracic Subjective Severity

Score (TSSS) and was also associated with higher mortality

rates (Table 3). A distance of 6.0 cm was optimal; the survival

rate was high (83%) even though sufficient injury was created

(TSSS, 1.4 � 0.5) (Table 3). Figure 1 shows examples of con-

tused lungs with different TSSS. The TSSS was significantly

higher in non-survivors compared with survivors (Fig. 2).

Furthermore, LBWR was significantly increased in

blast plusMV rats compared withMV rats and controls (Fig. 3).

Rats subjected to lung contusion had lower pH and higher

PaCO2 after 90 min than MV rats. Their PaO2 was lower than

the PaO2 of MV rats from 30 to 240 min. At death there was no

longer a difference between groups. Therewere no differences

between base deficits for MV and blast plusMV rats during the

whole experiment (Table 4).

3.2. Mechanical ventilation induces an inflammatoryresponse in control rats

The effect of MV on systemic inflammation measured by

absolute neutrophil count in blood showed a significant

increase in MV rats compared with controls (Fig. 4A). This was

accompanied by an increase in blood IL-1b levels (Fig. 5C). In

contrast, blood IL-6, GRO-KC, and IL-10 levels showed no

differences between MV rats and controls (Fig. 5A, B, and D).

Pulmonary inflammation measured by neutrophil counts

in BALF showed significantly more neutrophils in BALF of MV

rats compared with control rats (Fig. 4B). The MPO of MV rats

was also significantly higher than controls (Fig. 6). The BALF

IL-6 and BALF GRO-KC levels were significantly higher in MV

rats compared with controls (Fig. 7A and B). There was no

difference in BALF IL-1b and BALF IL-10 levels (Fig. 7C and D).

These results show that MV induced both a systemic and

pulmonary inflammatory response.

3.3. Both a systemic and pulmonary inflammatoryresponse was induced by mechanical ventilation after lungcontusion

Mechanical ventilation after pulmonary contusion showed an

increase in systemic inflammation, reflected by the absolute

neutrophil count in blood (Fig. 4A), blood IL-6, GRO-KC, and

IL-1b (Fig. 5AeC). Blood IL-10 levels showed no significant

increase compared with controls, although there was

a tendency toward higher levels in blast plus MV rats

(P ¼ 0.056) (Fig. 5D). In addition, pulmonary inflammation of

blast plus MV rats compared with controls was significantly

increased, reflected by higher neutrophil counts in BALF

(Fig. 4B), MPO (Fig. 6), and BALF IL-6, BALF-GRO-KC, and BALF-

IL-1b levels (Fig. 7AeC). We noted no difference in BALF IL-10

levels (Fig. 7D).

3.4. After lung contusion, MV rats showed a higherinflammatory response than after MV alone

There was no additional effect of lung contusion in MV rats on

systemic inflammation, as measured by absolute neutrophil

count in blood (Fig. 4A) and blood IL-10 levels (Fig. 5D).

However, blood IL-6, GRO-KC, and IL-1b levels were signifi-

cantly increased in blast plus MV rats compared with MV rats

(Fig. 5AeC). Pulmonary inflammation measured by BALF

neutrophil counts (Fig. 4B) and BALF IL-10 levels (Fig. 7D)

showed no additional effect of lung contusion n MV rats.

However, MPO (Fig. 6), and BALF IL-6, BALF-GRO-KC, and

BALF-IL-1b levels (Fig. 7AeC) were significantly higher in

blast plus MV rats compared with MV rats.

control

MV

bla

st+

MV

0

5

10

15

20

* *

x10

3

MV=mechanical ventilation, Blast=pulmonary contusion by blast

Fig. 3 e Lung to body weight ratio (*P < 0.05).

A

control

MV

bla

st+

MV

0

2

4

6 **

ab

so

lu

te n

um

ber n

eu

tro

's 10

6/m

l

2

4

6

**

lu

te n

um

ber n

eu

tro

's 10

6/m

l

B

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4 381

4. Discussion

This study has demonstrated the feasibility of combining

pulmonary contusion and MV in a rat model, a common

combination in contemporary trauma care. The blast gener-

ator generated reproducible, isolated blunt chest trauma in

rats. This model is suitable for studying inflammatory

response after lung contusion, provided that the right

distance to the blast generator is used. Mechanical ventilation

after lung contusion induced a larger systemic and pulmonary

inflammatory response thanMV alone, as expressed by higher

Table 4 e Blood gas analysis during experiment.

pH PaO2

(mm Hg)PaCO2

(mm Hg)BE/BD(mEq/L)

T ¼ 0 min

MV 7.40 � 0.02 218 � 26 39 � 3 �0.8 � 1.9

Blast þ MV

T ¼ 30 min

MV 7.43 � 0.02 246 � 10* 35 � 4 �0.05 � 2.2

Blast þ MV 7.40 � 0.03 175 � 14 36 � 2 �2.1 � 1.7

T ¼ 90 min

MV 7.43 � 0.02* 227 � 15* 34 � 3* �1.1 � 1.6

Blast þ MV 7.33 � 0.04 188 � 14 43 � 5 �4.0 � 1.6

T ¼ 180 min

MV 7.39 � 0.04 239 � 7* 37 � 5 �2.3 � 2.2

Blast þ MV 7.32 � 0.04 179 � 19 38 � 2 �5.1 � 2.1

T ¼ 240 min

MV 7.33 � 0.04 241 � 5* 37 � 3 �5.8 � 2.5

Blast þ MV 7.30 � 0.04 186 � 18 38 � 4 �8.0 � 2.4

T ¼ 300 min

MV 7.31 � 0.02 191 � 29 42 � 7 �5.3 � 2.8

Blast þ MV 7.28 � 0.03 208 � 17 39 � 3 �7.8 � 2.3

PaO2 ¼ partial pressure of arterial O2; PaCO2 ¼ partial pressure of

arterial CO2; BE/BD ¼ base deficit/base excess.

Values are expressed as means � standard error of the mean.

* P < 0.05, significant difference between MV and blast plus MV

rats.

control

MV

bla

st+

MV

0ab

so

Fig. 4 e Absolute numbers of neutrophils in blood (A) and

BALF (B). *P < 0.05.

MPO levels and higher IL-6, GRO-KC, and IL-1b levels in both

blood and BALF.

In corroboration with our previous work [8] and reported

literature [12e15] these results showed that MV caused not

only pulmonary inflammation, but also systemic inflamma-

tion. Furthermore, pulmonary contusion is known to induce

an inflammatory response [16e18], and an interaction

between lung contusion and sepsis as a second hit has also

been described [15e17,19]. Nevertheless, to our knowledge, no

study has been published investigating the combination of

pulmonary contusion and MV. This study is the first to

analyze the possible synergy in both systemic and pulmonary

inflammatory response in rats subjected to pulmonary

contusion combined with MV.

Most studies published [12,20,21] ventilated rats for up to 4

h. In contrast to those studies, our rats were ventilated for 5 h,

to ensure that an inflammatory response caused by MV would

occur.

IL-6 GRO-KC

control

MV

bla

st+

MV

0

2000

4000

6000

8000

10000*

*

pg

/m

l

control

MV

bla

st+

MV

0

2000

4000

6000

8000

10000*

*

pg

/m

l

IL-1β IL-10

control

MV

bla

st+

MV

0

100

200

300

** *

pg

/m

l

control

MV

bla

st+

MV

0

1000

2000

3000

4000

pg

/m

l

A B

C D

Fig. 5 e Cytokine levels in blood (pg/mL) (*P < 0.05). (A) Interleukin-6 levels in blood. (B) Growth-related oncogene levels in

blood. (C) Interleukin-1b levels in blood. (D) Interleukin-10 levels in blood.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4382

Blast plus MV rats showed a tendency toward lower

neutrophils in blood and BALF compared with MV rats,

although the differences were not statistically different.

Furthermore, blast plus MV rats had higher MPO levels than

MV rats, which suggests that there is a difference in neutro-

phil localization between the blast plus MV and MV rats. Blast

induces direct trauma to the lungs, whereas MV causes bio-

trauma by cyclic stretch of lung parenchyma [6,7]. Possibly,

direct trauma to the lung induces neutrophils to migrate to

the tissue, whereas MV alone induces a more generalized

inflammatory response. Blast pus MV rats showed higher

Control

MV

bla

st+

MV

0

2

4

6

8

10

**

*

un

its/

g lu

ng

tissu

e

MV=mechanical ventilation, Blast=pulmonary contusion by blast

Fig. 6 e Myeloperoxidase (units per microgram lung tissue)

(*P < 0.05).

cytokine levels in blood and BALF compared with MV rats.

This suggests that pulmonary contusion primed both

systemic and pulmonary inflammatory responses, with MV

serving as a second hit. Hoth et al. [15,19] reported similar

results; they investigated the immune response in mice with

lung contusion followed by intratracheal administration of

lipopolysaccharide. The authors observed a synergistic

increase in inflammatory mediator expression in blood and

a more severe lung injury in injured animals challenged with

lipopolysaccharide. However, when blunt chest trauma was

combined with hemorrhagic shock, as was done by Seitz et al.

[22], no synergy was observed. This observation is in accor-

dance with our experience in a previous animal model, in

which we failed to demonstrate an increased inflammatory

response when hemorrhagic shock and MV were combined

[8]. All of these findings suggest to us that synergistic

inflammatory response occurs when local (pulmonary)

damage is combined.

Initially,we started the experimentwith a distance of 3.5 cm

from the nozzle of the blast generator to the animal’s chest.We

chose this distance according to results reported earlier by

Jaffin et al. [11] and Eichert [23] using the same blast generator

with adequate isolated bilateral lung contusion without

abdominal injury. During the experiment, we increased the

distance from the nozzle to the chest because of highmortality

rates found at the 3.5-cm distance. Ultimately, 6.0 cm was the

distance with acceptable mortality rates and sufficiently bilat-

eral lung contusion without abdominal injuries.

IL-6 GRO-KC

control

MV

bla

st+

MV

0

5000

10000

15000

20000 **

*

pg

/ml

control

MV

bla

st+

MV

0

5000

10000

15000

20000 **

*

pg

/m

l

IL1β IL-10

control

MV

bla

st+

MV

0

20

40

60

80 *

pg

/m

l

control

MV

bla

st+

MV

0

100

200

300

400

500

pg

/m

l

A B

C D

Fig. 7 e Cytokine levels in BALF (pg/mL) (*P < 0.05). (A) Interleukin-6 levels in BALF. (B) Growth-related oncogene levels in

BALF. (C) Interleukin1b levels in BALF. (D) Interleukin-10 levels in BALF.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4 383

It is unsurprising that the distance from the nozzle to the

chest was inversely proportional to the TSSS. However, it is

remarkable that the 6.0-cm distance from nozzle to the chest

showed higher blast pressures than the 3.5-cm distance. The

blast generator produces an outer circular wave moving

outward. Behind it, a column of turbulent air called the plume

flattens out into a Mach disc [11]. The surface area hit by the

blast becomes larger with larger distance. Possibly, this

conical shapedwave influenced themeasurements calculated

by pressure sensors on both sides of the animal’s chest. It is

also possible that a combination of direct pressure by the blast

itself and the reflection of the wave caused higher pressure in

the 6.0-cm distance. Because the variation in blast pressure

within one distance is small, we still think that lung contusion

induced by the blast generator is reproducible.

All of our rats were ventilated, and one might debate

whether rats should have been included that were subjected

to pulmonary contusion without ventilation. During the

experiments, we learned that rats did not survive the bilateral

lung contusion without pulmonary support by MV. Therefore,

we decided to ventilate all rats. Because all rats were venti-

lated and a group was included that only underwent MV, we

believed that the effect of pulmonary contusion could be

studied properly. Furthermore, adding MV relates more to the

human situation, becausemost trauma patients with bilateral

lung contusion need ventilatory support.

In conclusion, this study investigated the combination of

MV and pulmonary contusion. Mechanical ventilation alone

induced both a systemic and pulmonary inflammatory

response. Mechanical ventilation combined with pulmonary

contusion caused a synergistic inflammatory response, which

suggests that local (pulmonary) damage contributes not only

to local inflammation, but also to systemic inflammation.

When extrapolating this to humans, it indicates that MV in

patients with lung contusion could aggravate the pulmonary

and systemic inflammatory response induced by lung contu-

sion. This emphasizes the importance of lung-protective MV.

r e f e r e n c e s

[1] Harbrecht BG, Forsythe RM, Peitzman AB. Management inshock. In: Feliciano DV, Mattox KL, Moore EE, editors, editors.Trauma. 6th Edition. New York: McGraw-Hill Medical; 2008.p. 216.

[2] Adams JM, Hauser CJ, Livingston DH, Lavery RF, Fekete Z,Deitch EA. Early trauma polymorphonuclear neutrophilresponses to chemokines are associated with developmentof sepsis, pneumonia, and organ failure. J Trauma 2001;50:792.

[3] Pillay J, Hietbrink F, Koenderman L, Leenen LPH. Thesystemic inflammatory response induced by trauma isreflected by multiple phenotypes of blood neutrophils. Injury2007;38:1365.

[4] Reutershan J, Ley K. Bench-to-bedside review: acuterespiratory distress syndrome-how neutrophils migrate intothe lung. Crit Care 2004;8:453.

j o u r n a l o f s u r g i c a l r e s e a r c h 1 8 3 ( 2 0 1 3 ) 3 7 7e3 8 4384

[5] Keel M, Trentz O. Pathophysiology of polytrauma. Injury2005;36:691.

[6] Pl�otz FB, Vreugdenhil HAJ, Slutsky AS, Zijlstra J, Heijen CJ,van Vught H. Mechanical ventilation alters the immuneresponse in children without lung pathology. Intensive CareMed 2002;28:486.

[7] Wolthuis EK, Vlaar APJ, Choi G, Roelofs JJTH, Juffermans NP,Schultz MJ. Mechanical ventilation using non-injuriousventilation settings cause lung injury in the absence of pre-existing lung injury in healthy mice. Crit Care 2009;13:R1.http://dx.doi.org/10.1186/cc7688.

[8] van Wessem KJP, Hennus MP, Heeres M, Koenderman L,Leenen LPH. Mechanical ventilation is the determining factorin inducing an inflammatory response in a hemorrhagicshock model. J Surg Res; 2012. http://dx.doi.org/10.1016/j.jss.2012.10.019 [Epub ahead of print].

[9] Lin CY, Zhang H, Cheng KC, Slutsky AS. Mechanicalventilation may increase susceptibility to the developmentof bacteremia. Crit Care Med 2003;31:1429.

[10] Hennus MP, Janssen R, Pennings JL, et al. Host response tomechanical ventilation for viral respiratory tract infection.Eur Respir J 2012;40:1508e15.

[11] Jaffin JH, McKinney L, Kinney RC, et al. A laboratory modelfor studying blast overpressure injury. J Trauma 1987;27:349.

[12] Vreugdenhil HA, Heijnen CJ, Plotz FB, et al. Mechanicalventilation of healthy rats suppresses peripheral immunefunction. Eur J Resp 2004;23:122.

[13] Dreyfuss D, Saumon G. Ventilator induced lung injury:lessons from experimental studies. Am J Respir Crit CareMed 1998;157:294.

[14] Pinhu L, Whitebread T, Evans T, Griffiths M. Ventilator-associated lung injury. Lancet 2003;361:332.

[15] Hoth JJ, Martin RS, Yoza BK, Wells JD, Meredith JW,McCall CE. Pulmonary contusion primes systemic innateimmunity responses. J Trauma 2009;67:14.

[16] Raghavendran K, Notter RH, Davidson BA, Hellinski JD,Kunkel SL, Knight PR. Lung contusion: inflammatorymechanisms and interaction with other injuries. Shock 2009;32:122.

[17] Perl M, Gebhard F, Bruckner UB, et al. Pulmonarycontusion causes impairment of macrophage andlymphocyte immune functions and increases mortalityassociated with subsequent septic challenge. Crit Care Med2005;33:1351.

[18] Raghavendran K, Davidson BA, Woytash JA, et al. Theevolution of isolated bilateral lung contusion from bluntchest trauma in rats: cellular and cytokine responses. Shock2005;24:132.

[19] Hoth JJ, Wells JD, Brownlee NA, et al. Toll-like receptor4-dependent responses to lung injury in a murine model ofpulmonary contusion. Shock 2009;31:376.

[20] Bouadma L, Dreyfuss D, Ricard JD, Martet G, Saumon G.Mechanical ventilation and hemorrhagic shock-resuscitationinteract to increase inflammatory cytokine release in rats.Crit Care Med 2007;35:2601.

[21] Herff H, Paal P, Goedecke von A, Lindner KH, Severing AC,Wenzel V. Influence of ventilation strategies on survival insevere controlled hemorrhagic shock. Crit Care Med 2008;36:2613.

[22] Seitz DH, Perl M, Linere UC, et al. Inflammatory alterations ina novel combination model of blunt chest trauma andhemorrhagic shock. J Trauma 2011;70:189.

[23] Eichert K. Thorax trauma-induced experimental lung injury.Konstanz, Germany: [Dissertation]; 2003.