SHOCK, Vol. 41, No. 2, pp. 159Y165, 2014

MIDTERM EFFECTS OF FLUID RESUSCITATION STRATEGIES IN AN EXPERIMENTAL MODEL OF LUNG CONTUSION AND HEMORRHAGIC SHOCK Bertrand Prunet,*† Nicolas Prat,‡ David Couret,†§ Pierre-Yves Cordier,|| Sophie De Bourmont,†§ Dominique Lambert,§ Yves Asencio,¶ Eric Meaudre,* and Pierre Michelet†§ *Intensive Care Unit, Sainte Anne Military Teaching Hospital, Toulon; and † UMR MD2, P2COE, University of Aix-Marseille, School of Medicine, Marseille, France; ‡ U.S. Army Institute of Surgical Research, Fort Sam Houston, Texas; and §Department of Emergency Medicine and Intensive Care, Timone University Hospital, Marseille; and ||Intensive Care Unit, Laveran Military Teaching Hospital; and ¶ Department of Anesthesiology, Pique Military Teaching Hospital, Bordeaux, France Received 18 Jul 2013; first review completed 7 Aug 2013; accepted in final form 4 Oct 2013 ABSTRACT—Background: This study compared three different fluid resuscitation strategies in terms of respiratory tolerance and hemodynamic efficacy in a pig model of blunt chest trauma with lung contusion and controlled hemorrhagic shock. We hypothesized that the choice of fluid resuscitation strategy (type and amount of fluids) may impact differently contused lungs in terms of extravascular lung water (EVLW) 20 h after trauma. Methods: Anesthetized female pigs (n = 5/group) received five bolt shots to the right thoracic cage and allowed to hemorrhage for 30 min, with 25 to 30 mL/kg of blood loss. Pigs were randomly assigned to resuscitation groups that maintained a minimum mean arterial blood pressure of 70 mmHg with one of three methods: normal saline (NS), unrestricted normal saline; NOREPI, low-volume normal saline with norepinephrine; or HS-HES, hypertonic saline with hydroxyethyl starch. Control pigs were anesthetized, but received no injury or treatment. After 20 h, animals were killed to measure EVLW by gravimetry. Results: Fluid loading was significantly different in each group. All three treatment groups had higher EVLW than controls. Moderate, bilateral pulmonary edema was observed in the NS and HS-HES groups. The three treatment groups showed similar reductions in oxygenation. Static pulmonary compliance was diminished in the NS and HS-HES groups, but compliance was similar in NOREPI and control groups. The NOREPI group had pathological lactate levels. Conclusions: This study demonstrated the impact of fluid resuscitation on contused lungs. Twenty hours after the trauma, all three resuscitation approaches showed modest clinical consequences, with moderate lung edema and reduced compliance in response to the infused volume. KEYWORDS—Lung contusion, blunt chest trauma, hemorrhagic shock, fluid resuscitation, hypertonic saline, pulmonary edema, extravascular lung water

Burbank (4) described Bwet lungs[ in soldiers with thoracic injuries during World War II, fluid management has been an area of controversy (5Y7). During the Vietnam conflict, the aggressive use of crystalloids for resuscitation was associated with the appearance of Bshock lung/DaNang lung,[ which was later termed acute respiratory distress syndrome (8). Currently, the treatment of choice remains controversial in civilian (9Y11) and military (5, 6). It is now well established that both traumatic hemorrhage and pulmonary contusion independently induce acute lung injury (ALI) (12Y15). The pathogenesis of the hemorrhage-induced ALI is still partially understood but includes a two-hit induced lung inflammation and an exaggerated inflammatory response (16). Thus, hemorrhagic shock promotes development of lung injury by priming the innate immune system for an exaggerated inflammatory response to a second, often more trivial inflammatory stimulus (17). Acute lung injury is also associated with pulmonary edema, due to an increase in capillary permeability, and infiltration of inflammatory cells into the interstitium and airspaces (13). Consequently, injured lungs are particularly vulnerable to massive fluid infusions (18Y20). Therefore, administering vasopressor agents can help rapidly achieve the target blood pressure and limit volume fluid requirements (21). This experimental animal study aimed to compare midterm respiratory tolerance and hemodynamic efficacy of three strategies for correcting hemorrhagic shock in the context of lung contusion. We hypothesized that the choice of fluid

INTRODUCTION Chest trauma is one of the most common injuries in polytrauma patients. In this group, the incidence of chest trauma is 45% to 65%, and it accounts for 20% to 25% of adult deaths due to trauma (1). Lung contusion is the most frequently diagnosed intrathoracic injury resulting from blunt trauma, and it is commonly caused by a high-kinetic-energy chest wall impact (2). Patients with pulmonary contusions are at high risk of subsequent adverse events, including pneumonia (3) and acute respiratory distress syndrome (2). Hemorrhagic shock is common in patients with chest traumas, due to thoracic blood loss or other bleeding related to polytrauma. Fluid resuscitation aims to correct blood lossYinduced hypovolemia and to restore circulatory efficiency by providing minimal tissue oxygenation. However, resuscitation from hemorrhage in the setting of lung contusion remains a challenge in trauma care, due to the dilemma of providing hemodynamic efficiency without causing respiratory harm. Indeed, since Burford and Address reprint requests to Bertrand Prunet, MD, MSc, Service de Re´animation Hoˆpital d’Instruction des Arme´es, Sainte Anne Blvd, Sainte Anne 83000, Toulon, France. E-mail: [email protected]. This study was supported by institutional funds from UMR MD2 P2COE (University of Aix-Marseille) and French Army Institute of Biomedical Research. Supplemental digital content is available for this article. Direct URL citation appears in the printed text and is provided in the HTML and PDF versions of this article on the journal_s Web site (www.shockjournal.com). DOI: 10.1097/SHK.0000000000000069 Copyright Ó 2013 by the Shock Society

159

Copyright © 2013 by the Shock Society. Unauthorized reproduction of this article is prohibited.

160

SHOCK VOL. 41, NO. 2

resuscitation strategy (type and amount of fluids) may impact differently contused lungs in terms of extravascular lung water (EVLW) 20 h after trauma. MATERIALS AND METHODS This protocol was performed in accordance with the Guiding Principles in the Care and Use of Animals of the American Physiologic Society and was approved by the Animal Ethics Committee of the French Army Institute of Biomedical Research.

Animal preparation After a 12-h fasting period with free access to water, 21 white female pigs (4 months old) were premedicated with intramuscular ketamine (30 mg/kg), anesthetized, and paralyzed with an intravenous infusion of midazolam (0.2 mg/kg per hour), sufentanil (2.5 Hg/kg per hour), and cisatracurium (0.8 mg/kg per hour). Basal fluid requirement (5 mL/kg per hour normal saline) was infused continuously throughout the experiment via a venous catheter placed on the ear. The lungs were ventilated via a cuffed tracheostomy tube (inner diameter 7 mm, Tracheosoft; Malindkrodt Medical, Athlone, Ireland) with a Vela ventilator (VIASYS CareFusion, San Diego, Calif). The ventilator was set to deliver a constant fraction of inspired oxygen (FIO2) of 0.4, a positive end-expiratory pressure of 5 cmH2O, an inspiration/expiration ratio of 1:2, a tidal volume of 8 mL/kg, and a flow rate that maintained arterial partial pressure of CO2 (PaCO2) between 35 and 45 mmHg. A catheter (8F, BD Exacta; Becton Dickinson Critical Care Systems, Singapore) was inserted into the left external jugular vein for blood withdrawal, fluid loading, and, when necessary, norepinephrine infusion. A pulmonary artery catheter (TDQ CCO Catheter 8F; Hospira, San Diego, Calif) was inserted via the right external jugular vein into a branch of the pulmonary artery under pressure-waveform guidance for measurements of pulmonary arterial pressures and core temperature. Systemic and pulmonary arterial pressures and pulmonary artery occlusion pressure (PAOP) were measured at the end of expiration. Study data corresponded to the mean of three measurements. Lung recruitment maneuvers were performed every 3 h (30 cmH2O for 30 s) to re-expand collapsed lung tissue. Pigs were monitored continuously with an electrocardioscope (Hewlett Packard, Palo Alto, Calif) and a capnogram (ULTIMA II infrared spectrophotometer; Datex, Helsinki, Finland). Arterial pH, arterial partial pressure of oxygen (PaO2), PaCO2, lactate concentration, and hemoglobin levels were measured with a blood gas analyzer (ABL90 Flex; Radiometer, Copenhagen, Denmark). A thermistor-tipped catheter (Pulsiocath, 5F thermodilution catheter 20 cm; Pulsion Medical System, Munich, Germany) was placed in the aorta via the left femoral artery, and it was connected to the PiCCO monitor (Pulsion Medical System). It was used to monitor systemic pressure, perform arterial sampling, record stroke volume variation (SVV), and measure EVLW indexed to body weight based on transpulmonary thermodilution (EVLWiTT). Cardiac output was measured by injecting a 20-mL bolus of saline (0-CY5-C) into the superior vena cava via the left external jugular catheter. Urine samples were collected with a suprapubic catheter (Cystofix CH10; B-Braun, Melsungen, Germany). Body temperature was maintained at 37-C to 38-C with a heating blanket (Bair Hugger; Arizant Inc, Eden Prairie, Minn).

Experimental protocol Pigs were randomly allocated to four groups (five animals each). The first group was traumatized and resuscitated with normal saline only (NS). The second group was traumatized and resuscitated with hypertonic saline and hydroxyethyl starch (HS-HES). The third group was traumatized and resuscitated with normal saline and norepinephrine (NOREPI). The fourth group remained anesthetized but did not undergo chest trauma, hemorrhage, or resuscitation (control). Pigs were assigned randomly with a 20-ball ballot box, where balls with pig IDs were blindly selected. After induction of anesthesia and instrumentation for monitoring hemodynamic variables, the animals were allowed to rest for 15 min to achieve stabilization (stable heart rate, blood pressure, cardiac output, and end-tidal CO2). Once steady state was achieved, baseline measurements were obtained (T +0 min). The protocol duration was 20 h, divided into four distinct phases (except the control group): blunt chest trauma (30 min), hemorrhagic shock induction (30 min), hemorrhagic shock maintenance (60 min), and resuscitation (18 h), followed by euthanasia. Blunt chest trauma—As in our previous study (22), we delivered a blunt chest trauma with a five-bolt shot (70 J) from a .22-caliber charge. To prevent direct myocardial injury, the right side was chosen. Impact shots were distributed on the seventh right costal arch, along anterior to posterior axillary lines; the endotracheal tube was clamped at full inspiration. Immediately after the trauma, to reduce the risk of tension pneumothorax, a right pleural drainage was inserted with a water-sealed chamber that applied negative suction at j20 cmH2O.

PRUNET ET

AL.

Hemorrhagic shock induction—Pigs underwent hemorrhage 30 min after chest trauma by withdrawing blood from the left external jugular vein. Based on hemodynamic tolerance, we performed stepwise withdrawals of 2 mL/kg at 2-min intervals for a total of 25 to 30 mL/kg. The procedure lasted approximately 30 min, and it was terminated when the mean arterial pressure (MAP) fell below 50 mmHg. Maintenance phase—The animals remained in hemorrhagic shock for an additional 60 min; no fluid was administered except a baseline NS infusion (5 mL/kg per hour). This 60-min delay before resuscitation allowed the animals to reach a blood pressure nadir to replicate a military or civilian trauma scenario. Resuscitation phase—At the end of the 90-min shock period, we recorded hemodynamic parameters and obtained arterial blood samples. Next, animals were resuscitated according to the assigned groups. The NS group received normal (0.9% NaCl) at 10 mL/kg for 10 min (from T +120 min to T +130 min); then, the flow rate was adjusted to maintain the MAP above 70 mmHg, with no volume restriction. The HS-HES group received 7.2% NaCl/6% HES 200/0.5 (Hyperhes; Fresenius Kabi, Bad Homburg, Germany) at 4 mL/kg for 10 min (from T +120 min to T +130 min); then, HES 130/0.4 (Voluven; Fresenius Kabi) was infused at a flow rate adjusted to maintain the MAP above 70 mmHg, with no volume restriction, but with a maximum flow of 33 mL/kg as recommended by the manufacturer. The NOREPI group received normal saline (0.9% NaCl) at 10 mL/kg for 10 min (from T +120 min to T +130 min), followed by an intravenous infusion of norepinephrine, adjusted to maintain the MAP above 70 mmHg, with no dose restriction. The control group was monitored like the other groups, but did not receive treatment; the controls remained anesthetized for 20 h, with only normal saline infused at 5 mL/kg per hour.

Measurement of indexed EVLW Extravascular lung water was defined as the fluid that accumulated in the interstitial and alveolar spaces of the lung. Increased EVLW is typically taken as a sign of lung edema. Some authors have defined elevated EVLWi as greater than 7 mL/kg (23). In addition to the in vivo EVLWiTT (measured with the PiCCO monitor), we performed a postmortem, gravimetric measurement of EVLW, indexed to body weight (EVLWiG), as described by Pearce et al. (24). Briefly, before euthanasia, 100 mL of peripheral blood was collected for analysis of blood water content and hemoglobin concentration. As soon as possible after killing the animal with an overdose of thiopental, we performed exsanguination and lung hila clamping to prevent acute cardiogenic edema. Then, the lungs were removed, weighed, and blended with an equal amount of distilled water to a homogeneous state. Aliquots of the whole homogenate were centrifuged at 5,000 revolutions/min at 5-C for 30 min, and supernatants were collected. The water contents of blood, homogenate, and supernatant samples were determined by weighing before and after drying at 80-C for 72 h. The accuracy of the in vivo transpulmonary thermodilution method for measuring lung water was established in 2004 by comparing it with the criterion standard gravimetric method in experimental animal studies (23).

Data recording Study parameters were assessed at baseline and after the chest trauma at T +5, 10, 15, 30, 60, 120, and 130 min and at T +3, 4, 5, 6, 7, 11, 14, 17, and 20 h. Hemodynamics data were collected for heart rate, systolic arterial pressure, MAP, diastolic arterial pressure, cardiac index, SVV, EVLWiTT, and pulmonary arterial and central venous pressures. Indexed systemic vascular resistance (iSVR) and indexed pulmonary vascular resistance (iPVR) were calculated with standard formulas, where BSA = K / weight (kg)2/3 and K was 0.112 for pigs (25). Mechanical ventilation was monitored for tidal volume, plateau pressure, and total positive end-expiratory pressure. We also assessed urine output, core temperature, total fluid infusion, norepinephrine infusion dosage, blood volume withdrawn during the hemorrhagic phase, and number of ribs fractured.

End points The primary end point of this study was the EVLWi at T +20 h, measured by gravimetry and transpulmonary thermodilution. This was compared between the three fluid resuscitation strategies. Secondary end points were cardiocirculatory and respiratory parameters, which were used to compare the consequences of resuscitation strategies.

Statistical analysis Statistical analysis was performed with SPSS version 15.0 (SPSS Inc, Chicago, Ill), and data were assumed to be normally distributed. Nominal variables are presented as numbers (%), and continuous variables are presented as the mean T SD. Groups were compared with analysis of variance and Tukey

Copyright © 2013 by the Shock Society. Unauthorized reproduction of this article is prohibited.

SHOCK FEBRUARY 2014

161

FLUID RESUSCITATION STRATEGIES

post hoc tests. A paired Student t test was used to compare data within groups. For all tests, P G 0.05 was considered statistically significant.

RESULTS Animal characteristics

Of the 21 pigs included, one of them died of cardiac arrhythmia with ventricular fibrillation during animal preparation, before randomization. Five pigs were included in each group. Two pigs died before T +20 h: one in the NS group at T +15 h and one in the NOREPI group at T +6 h. Data acquired from these two animals before death were included in the analysis. The population characteristics are summarized in Table 1. The four groups were not significantly different in weight and baseline MAP. In the three traumatized groups, postmortem examinations revealed a reproducible, severe pulmonary contusion that included the middle and lower right pulmonary lobes in all cases, with no pneumothorax or significant hemothorax. Macroscopic examinations showed no myocardial or left lung contusion and no intra-abdominal injury. The three groups were not significantly different in the degree of trauma and hemorrhage, the number of ribs fractured, the volume of blood withdrawn, the immediate posttraumatic PaO2/ FIO2 ratio (at T +5 min), the SVV, or the lactate levels before resuscitation (T +120 min). The four groups showed significantly different total fluid loading at T +20 h (NS vs. HS-HES: P = 0.020; NS vs. NOREPI: P = 0.021; NS vs. control: P = 0.011; HS-HES vs. NOREPI: P = 0.045; HS-HES vs. control: P = 0.041; NOREPI vs. control: P = 0.007). See Supplemental Digital Content 1, at http://links.lww.com/SHK/A196, for more information. Indexed EVLW

Indexed EVLW measurements at T +20 h are summarized in Table 2. The EVLWiTT measurements revealed that the NOREPI group was significantly different from the NS group (P = 0.019), and the control group was significantly different from the NS and HS-HES groups (P G 0.001 and P = 0.001, respectively). Both lungs EVLWiG measurements also revealed that the NOREPI group was significantly different from the NS group (P = 0.046), and the control group was significantly

different from the NS and HS-HES groups (P = 0.013 and P = 0.033, respectively). In the control group, both lungs were healthy. The three traumatized groups were not significantly different in the right traumatized lungs. Compared with the control group, the NS and HS-HES groups showed a significantly elevated EVLWiG for NS and HS-HES groups (P = 0.019 and P = 0.034, respectively). In the left, healthy lung, the NOREPI group had a significantly lower EVLWiG than the NS group (P = 0.048). Compared with the control group, the NS and HS-HES groups showed significantly elevated EVLWiG values (P = 0.010 and P = 0.039, respectively). In all four groups, there were no significant differences between the right and left lung EVLWiG values. Furthermore, the four groups showed no significant differences between the gravimetrically determined and transpulmonary thermodilution-estimated EVLWi values. Urinary output, respiratory parameters, and hemodynamic parameters at T +20 h

Table 3 shows the results for the secondary end points. Urinary function—The total urinary output at T +20 h in the NOREPI group was significantly lower than that of the other three groups. Respiratory parameters—Twenty hours after the chest trauma, the PaO2/FIO2 ratio was not different among the three traumatized groups, but they all had lower ratios than did the control group. There was no difference in static pulmonary compliance between the NS and HS-HES groups and between the NOREPI and control groups. However, both the NS and HS-HES groups showed significant differences compared with the NOREPI and control groups. Hemodynamic parameters—Arterial lactate levels were normal and similar among the NS, HS-HES, and control groups. Only the NOREPI group had pathologic lactate levels (5.0 T 3.4 mmol/L), and this was significantly different from the other three groups. Cardiac index was significantly higher in the NS group than in the NOREPI and control groups. The SVV was significantly higher in the NOREPI group than in the other three groups, and PAOP was significantly lower in the NOREPI

TABLE 1. Animal characteristics NS n Weight, kg

HS-HES

NOREPI

Control

5

5

5

39.6 T 4.2

39.8 T 3.0

39.4 T 4.7

39.3 T 3,9

5 90 T 6

Baseline MAP, mmHg

86 T 6

88 T 6

85 T 7

Ribs fractures, n

2.2 T 0.5

2.2 T 1.1

2.4 T 1.1

-

27.4 T 1.8

27.5 T 2.3

27.8 T 1.7

-

Blood withdrawal, mL/kg Death before T +20 h, n

0

1

0

Total fluid loading at T +20 h, mL/kg

221 T 24*†‡

1

124 T 4§†‡

110 T 0§*‡

100 T 0§*†

T +5 min Pao2/FIO2 ratio (posttraumatic)

247 T 43‡

234 T 36‡

251 T 29‡

542 T 48§*†

23 T 4‡

24 T 5‡

22 T 3‡

12 T 3§*†

T +120 min SVV (before resuscitation), % T +120 min lactate level (before resuscitation), mmol/L

4.5 T 1.8

‡

4.4 T 2.1

‡

4.2 T 1.6

Statistical significance was accepted at P G 0.05. *P G 0.05 vs. HS-HES group. † P G 0.05 vs. NOREPI group. ‡ P G 0.05 vs. control group. § P G 0.05 vs. NS group.

Copyright © 2013 by the Shock Society. Unauthorized reproduction of this article is prohibited.

1.6 T 0.4§*†

162

SHOCK VOL. 41, NO. 2

PRUNET ET

AL.

TABLE 2. Indexed EVLW (in mL/kg) at T +20 h measured with gravimetry and transpulmonary thermodilution EVLWiG right lung (contused)

EVLWiG Left lung (healthy)

EVLWiG RL vs. LL

EVLWiG both lungs

EVLWiTT both lungs

EVLWiG vs. EVLWiTT

NS

4.46 T 0.98

4.14 T 0.73

ns

8.60 T 1.24

9.00 T 1.23

ns

HS-HES

4.19 T 0.88

3.69 T 0.84

ns

7.88 T 1.86

8.25 T 0.96

ns

ns

ns

V

ns

ns

V

3.16 T 0.76

2.85 T 0.53

ns

6.01 T 1.28

6.75 T 0.76

ns

ns

0.048

V

0.046

0.019

V

NS vs. HS-HES NOREPI NOREPI vs. NS NOREPI vs. HS-HES Control

ns

ns

V

ns

ns

V

2.55 T 0.72

2.28 T 0.50

ns

4.83 T 1.36

5.20 T 0.74

ns

Control vs. NS

0.019

0.010

V

0.013

G0.001

-

Control vs. HS-HES

0.034

0.039

V

0.033

0.001

V

Control vs. NOREPI

ns

ns

V

ns

ns

V

LL indicates left lung; NS, nonsignificant; RL, right lung.

group than in the NS and HS-HES groups. The iSVR and iPVR were significantly higher in the NOREPI group than in the NS group (Fig. 1).

Indeed, lung repercussions were evidenced by the modest EVLW increase, which appeared to be proportional to the volume of fluid loading. The only clinically relevant negative effect of the three strategies was an alteration in pulmonary compliance, observed in the NS and HS-HES groups, but it did not significantly change hematosis. The thermodilution method showed that EVLWi was significantly higher in the three traumatized groups than in the control group. Postmortem gravimetry also showed that EVLWi was higher in the NS and HS-HES groups than in the control group. Based on a pathological EVLWi threshold of greater than 7 mL/kg (23), both methods indicated that both the NS and HS-HES groups experienced moderate pulmonary edema. Furthermore, although only the right lung was contused, the edema was bilateral, with EVLWi change also detected in the left lung. These results supported previous studies, which showed that a systemic inflammatory mechanism affected the contralateral lung at 8 h after trauma (27Y29). The only significant difference in EVLWi among the traumatized groups was the elevation in the NS group compared with the NOREPI group, which was confirmed by both methods. This was probably related to the different volumes of fluid loading, as shown in a previous study (19). In 2010, Gryth and colleagues (18)

DISCUSSION To our knowledge, this study was the first to compare strategies for hemorrhage resuscitation after pulmonary contusion in pigs over a midterm period of 20 h with modern, complete instrumentation, including the transpulmonary thermodilution technique. Previous studies with similar aims were shorter in duration, where the treatment lasted only 2 (18) or 4 h (26). Nevertheless, it is well known that lung contusions lead to pulmonary pathophysiologic changes that worsen over 24 to 48 h, but then, they generally resolve by 7 days after the injury (2). Our primary goal was to evaluate the extent of damage the different fluid resuscitation strategies might cause to a contused lung, particularly in terms of EVLW. To achieve this objective, we purposely designed the study to compare three very different strategies. Lung injury and pulmonary tolerance

Our results showed that all the tested infusion strategies after chest trauma impacted pulmonary edema and compliance.

TABLE 3. Urine output, respiratory function, and hemodynamic parameters at T +20 h

Group

Total urine output, mL/kg

Pao2/Fio2 ratio

Static pulmonary Arterial Cardiac compliance, lactate levels, index, L/min PAOP, mL/cmH2O mmol/L per m2 SVV, % mmHg

iSVR, dyn/s per cm5 per m2

iPVR, dyn/s per cm5 per m2

NS

35.5 T 9 353 (278Y427)

25 T 3

0.8 T 0.2

4.6 (4.0Y5.0)

12 T 4 13 T 2

1,337 (1,045Y1,714)

300 T 92

HS-HES

31.9 T 5 398 (302Y422)

24 T 4

0.7 T 0.2

3.8 (3.3Y4.3)

14 T 4 11 T 1

ns

ns

ns

ns

34 T 4

5.0 T 2.4

3.3 (3.2Y3.8)

0.031

G0.001

0.042

NS vs. HS-HES NOREPI NOREPI vs. NS NOREPI vs. HS-HES Control

ns

ns

19.4 T 4 396 (318Y474) 0.008

ns

0.029

ns

38.7 T 4 568 (493Y594)

1,975 (1,375Y2,084)

246 T 96

ns

ns

ns

21 T 2

7T2

2,352 (1,758Y2,462)

437 T 89

0.014

0.009

0.030

0.049

0.021

G0.001

ns

0.048

0.049

ns

ns

34 T 2

0.6 T 0.3

3.3 (2.9Y3.8)

13 T 3

9T2

1,435 (1,371Y1,946)

341 T 108

Control vs. NS

ns

0.004

0.026

ns

0.020

ns

ns

ns

ns

Control vs. HS-HES

ns

0.005

0.016

ns

ns

ns

ns

ns

ns

Control vs. NOREPI

0.001

0.023

ns

G0.001

ns

0.018

ns

ns

ns

ns Indicates nonsignificant.

Copyright © 2013 by the Shock Society. Unauthorized reproduction of this article is prohibited.

SHOCK FEBRUARY 2014

FLUID RESUSCITATION STRATEGIES

163

FIGURE 1. Time course of interventions.

compared the effects of fluid resuscitation with either hypertonic saline dextrane (HSD) or Ringer’s acetate (RA) for 2 h after nonhemorrhagic shock caused by pulmonary contusion. They concluded that HSD treatment did not suppress lung edema compared with the control group. However, the RA group had a significantly higher wet-to-dry weight ratio in the right injured lung than did the HSD group. This suggested that fluid resuscitation with RA might have had deleterious effects on the contused lung. The unharmed left lung showed no increase in the wet-to-dry weight ratio; this indicated that 2 L of RA could be administered safely, when the lung tissue was unharmed. Those findings differed from the results we obtained at 20 h after the trauma. Furthermore, they found no significant difference in the PaO2 between the HSD and RA groups. Those authors concluded that, in a prehospital trauma situation, where the patient demonstrates shock with suspected pulmonary contusion, and there is indication for fluid resuscitation, HSD would probably provide more benefit than RA. Very recently, Silva and colleagues (30) observed that intravascular volume expansion with HES led to less lung injury than RA in experimental ALI after a major hemorrhage. However, it may be primarily an effect of the short duration of the study (31). Respiratory mechanics also showed a deterioration in static pulmonary compliance in the NS and HS-HES groups, but not in the other two groups. Previous studies showed a similar decrease in static pulmonary compliance after lung contusions (26). In 1997, Cohn and colleagues (26) published the first experimental study to evaluate the impact of hypertonic saline resuscitation in the setting of a lung contusion. They used a pig model of right lung contusion and hemorrhage, and they applied hypertonic saline resuscitation (7.5% NaCl, 4 mL/kg in 20 min) or an isotonic solution (0.9% NaCl, 90 mL/kg in 20 min). They found that, at 4 h after the trauma, the two treatments had equivalent resuscitative effects, despite the disparity in administered volumes. Neither approach appeared to benefit either oxygenation or lung edema. Moreover, the compliance worsened to similar extents in both groups after the contusion. The authors concluded that small-volume hypertonic saline resuscitation failed to reduce the magnitude of lung injury. Note that, in that study, lung edema was measured at 4 h after injury, which may be too early to draw categorical conclusions.

Concerning the deterioration in static pulmonary compliance, other studies showed that this effect was maximal at 24 h after contusion (28), and it was correlated to alveolar edema (19, 27). Despite those findings, the choice of fluid resuscitation strategy did not influence hematosis after 20 h of treatment, based on the similar PaO2/FIO2 ratio in all three traumatized groups. On the other hand, hematosis was reduced in the three traumatized groups compared with the control group, which indicated relevant lung dysfunction and demonstrated the validity of the chest trauma model. A recent animal study (14) established that the pathophysiology of hypoxemia after pulmonary contusion involved both a true shunt and transient increases in blood flow to very low and low ventilation-perfusion compartments. Finally, this study was the first to compare the Pulsion transpulmonary thermodilution estimated EVLW with gravimetrically determined EVLW in a contused lung model. Our results demonstrated that the thermodilution results matched those of the criterion standard gravimetric method in this context. Hemodynamic efficiency with different resuscitation strategies

We found that the NS and HS-HES resuscitation treatments were not significantly different in terms of hemodynamic parameters or urine output. Nevertheless, the total fluid loading was very different for the two approaches (221 T 24 vs. 124 T 4 mL/kg, P G 0.0001). Similarly, Roch and colleagues (32) previously observed that normal saline resuscitation and smallvolume resuscitation with hypertonic saline associated with colloids were equally able to restore hemodynamic parameters after controlled hemorrhage. We also found that the NOREPI treatment caused significantly worse preload parameters than the other treatments. The NOREPI treatment was the only one associated with persistent elevated arterial lactate levels at 20 h. Previous studies showed that the arterial lactate level provides an early, objective evaluation of a patient’s response to therapy, and it represents a reliable prognostic index for patients with circulatory shock (33). Our findings indicated that fluid loading was not optimized before the use of the vasopressor. With this very-low-volume resuscitation approach, the lung remained fairly dry (free from edema). However, the vasopressor must be used only when the volume

Copyright © 2013 by the Shock Society. Unauthorized reproduction of this article is prohibited.

164

SHOCK VOL. 41, NO. 2

status is optimized by goal-directed fluid management; otherwise, cellular dysoxia may occur because of microcirculatory infringement.

PRUNET ET

2. 3.

Study limitations

This study had several limitations. First, we did not use a blood product transfusion, which is the conventional resuscitation protocol for hemorrhagic shock, because we did not want a potential transfusion-related ALI to interfere with group comparisons and potential lung effects (34). Second, the hemorrhagic shock we induced was controlled, and definitive hemostasis was achieved at T +60 min; thus, the volume of blood loss was not influenced by the modalities of resuscitation (35). Third, we did not evaluate renal function or coagulation parameters. According to recent studies (36Y38) and the US Food and Drug Administration, HES causes kidney failure and increases the risk of death. Our choice was based on the shortness of the study period (20 h) in regard to a follow-up of several weeks in these studies (36Y38). We considered that the increased risk of renal replacement therapy and mortality were not relevant in the first 20 h, but we realized how these missing relevant end points damaged the quality of this study. All these limitations reduced the impact of our results on clinical practice.

4. 5. 6.

7. 8.

9. 10. 11. 12.

13.

CONCLUSIONS The results of this study contributed to an understanding of the consequences of fluid resuscitation on contused lungs. The three different resuscitation modalities studied had only very modest consequences at 20 h after a blunt chest trauma. The PaO2/FIO2 ratio did not show significant variations among the three groups. The high-volume crystalloid strategy restored hemodynamic status, but led to pulmonary edema and altered lung compliance. The very-low-volume crystalloid plus vasopressor strategy reduced lung edema and preserved lung function, but altered the circulatory parameters with consecutive hyperlactatemia. The HS plus HES strategy improved hemodynamic status and provided low-volume resuscitation, but produced no more favorable clinical repercussions to the lung compared with the NS strategy. The clinical relevance of these findings was that the choice of infusion strategy in the setting of lung contusion modestly impacted pulmonary edema and compliance but did not significantly changed hematosis 20 h after trauma. Management of patients with hemorrhages and pulmonary contusions requires judicious fluid administration, and more research is needed to find an optimal strategy. Future studies may investigate whether a combination of normal saline with volume optimized by goal-directed fluid management and relevant use of vasopressor might offer more beneficial effects than previous resuscitation approaches. ACKNOWLEDGMENTS The authors thank the Surgical and Physiological Research Unit of the French Army Institute of Biomedical Research for technical support, in particular Mr. Christian Aglioni, Mr. William Menini, and Mr. Joel Mosnier.

14.

15.

16.

17.

18.

19. 20. 21.

22.

23.

24. 25. 26. 27.

REFERENCES 1. Hildebrand F, Giannoudis PV, Griensven M, Zelle B, Ulmer B, Krettek C, Bellamy MC, Pape HC: Management of polytraumatized patients with associated blunt

28.

AL.

chest trauma: a comparison of two European countries. Injury 36(2):293Y302, 2005. Cohn SM, DuBose JJ: Pulmonary contusion: an update on recent advances in clinical management. World J Surg 34(8):1959Y1970, 2010. Michelet P, Couret D, Bre´geon F, Perrin G, D’Journo XB, Pequignot V, Vig V, Auffray JP: Early onset pneumonia in severe chest trauma: a risk factor analysis. J Trauma 68(2):395Y400, 2010. Burford TH, Burbank B: Traumatic wet lung. J Thorac Surg 14:415Y424, 1945. Butler F: Fluid resuscitation in tactical combat casualty care: brief history and current status. J Trauma 70(Suppl 5):S11YS12, 2011. McSwain NE, Champion HR, Fabian TC, Hoyt DB, Wade CE, Eastridge BJ, Rasmussen TE, Roussel RR, Butler FK, Holcomb JB, et al.: State of the art of fluid resuscitation 2010: prehospital and immediate transition to the hospital. J Trauma 70(Suppl 5):S2YS10, 2011. Santry HP, Alam HB: Fluid resuscitation: past, present and the future. Shock 33(3):229Y241, 2010. Rhee P, Koustova E, Alam HB: Searching for the optimal resuscitation method: recommendations for the initial fluid resuscitation of combat casualties. J Trauma 54(Suppl 5):S52YS62, 2003. Roberts L, Alderson P, Bunn F: Colloids versus crystalloids for fluid resuscitation in critically ill patients. Cochrane Database Syst Rev 4:CD000567, 2004. Rizoli SB: Crystalloids and colloids in trauma resuscitation: a brief overview of the current debate. J Trauma 54(Suppl 5):S82YS88, 2003. Kwan I, Bunn F, Roberts I: Timing and volume of fluid administration for patients with bleeding. Cochrane Database Syst Rev 3:CD002245, 2003. Perl M, Kieninger M, Huber-Lang MS, Gross HJ, Bachem MG, Braumu¨ller S, Gebhard F, Ayala A, Kno¨ferl MW: Divergent effects of activated neutrophils on inflammation, Kupffer cell/splenocyte activation, and lung injury following blunt chest trauma. Shock 37(2):210Y218, 2012. Balkamou X, Xanthos T, Stroumpoulis K, Moutzouris DA, Rokas G, Agrogiannis G, Demestiha T, Patsouris E, Papadimitriou L: Hydroxyethyl starch 6% (130/0.4) ameliorates acute lung injury in swine hemorrhagic shock. Anesthesiology 113(5): 1092Y1098, 2010. Batchinsky AI, Jordan BS, Necsoiu C, Dubick MA, Cancio LC: Dynamic changes in shunt and ventilation-perfusion mismatch following experimental pulmonary contusion. Shock 33(4):419Y425, 2010. Nirmalan M, Willard T, Edwards DJ, Dark P, Little RA: Effects of sustained post-traumatic shock and initial fluid resuscitation on extravascular lung water content and pulmonary vascular pressures in a porcine model of shock. Br J Anaesth 91(2):224Y232, 2003. Lomas-Neira J, Perl M, Venet F, Chung CS, Ayala A: The role and source of tumor necrosis factor-! in hemorrhage-induced priming for septic lung injury. Shock 37(6):611Y620, 2012. Chen C, Zhang F, Zhang Z, Peng M, Wang Y, Chen Y: TLR4 signalinginduced heme oxygenase upregulation in the acute lung injury: role in hemorrhagic shock and two-hit induced lung inflammation. Mol Biol Rep 40(2): 1167Y1172, 2013. Gryth D, Rocksen D, Drobin D, Druid H, Weitzberg E, Bursell J, Olsson LG, Arborelius UP: Effects of fluid resuscitation with hypertonic saline dextrane or Ringer’s acetate after nonhemorrhagic shock caused by pulmonary contusion. J Trauma 69(4):741Y748, 2010. Cohn SM, Zieg PM: Experimental pulmonary contusion: review of the literature and description of a new porcine model. J Trauma 41(3):565Y571, 1996. Cohn SM: Pulmonary contusion: review of the clinical entity. J Trauma 42(5): 973Y979, 1997. Poloujadoff MP, Borron SW, Amathieu R, Favret F, Camara MS, Lapostolle F, Vicaut E, Adnet F: Improved survival after resuscitation with norepinephrine in a murine model of uncontrolled hemorrhagic shock. Anesthesiology 107(4): 591Y596, 2007. Couret D, de Bourmont S, Prat N, Cordier PY, Soureau JB, Lambert D, Prunet B, Michelet P: A pig model for blunt chest trauma: no pulmonary edema in the early phase. Am J Emerg Med 31(8):1220Y1225, 2013. Brown LM, Liu KD, Matthay MA: Measurement of extravascular lung water using the single indicator method in patients: research and potential clinical value. Am J Physiol Lung Cell Mol Physiol 297(4):547Y558, 2009. Pearce ML, Yamashita J, Beazell J: Measurement of pulmonary edema. Circ Res 16:482Y488, 1965. Kelley KW, Curtis SE, Marzan GT, Karara HM, Anderson CR: Body surface area of female swine. J Anim Sci 36(5):927Y930, 1973. Cohn SM, Fisher BT, Rosenfield AT, Cross JH: Resuscitation of pulmonary contusion: hypertonic saline is not beneficial. Shock 8(4):292Y299, 1997. Obertacke U, Neudeck F, Majetschak M, Hellinger A, Kleinschmidt C, Schade FU, Hogasen K, Jochum M, Strohmeier W, Thurnher M, et al.: Local and systemic reactions after lung contusion: an experimental study in the pig. Shock 10(1):7Y12, 1998. Raghavendran K, Davidson BA, Woytash JA, Helinski JD, Marschke CJ, Manderscheid PA, Notter RH, Knight PR: The evolution of isolated bilateral

Copyright © 2013 by the Shock Society. Unauthorized reproduction of this article is prohibited.

SHOCK FEBRUARY 2014

29.

30.

31. 32.

33.

lung contusion from blunt chest trauma in rats: cellular and cytokine responses. Shock 24(2):132Y138, 2005. Seitz DH, Perl M, Liener UC, Tauchmann B, Braumu¨ller ST, Bru¨ckner UB, Gebhard F, Kno¨ferl MW: Inflammatory alterations in a novel combination model of blunt chest trauma and hemorrhagic shock. J Trauma 70(1):189Y196, 2011. Silva PL, Gu¨ldner A, Uhlig C, Carvalho N, Beda A, Rentzsch I, Kasper M, Wiedemann B, Spieth PM, Koch T, et al.: Effects of intravascular volume replacement on lung and kidney function and damage in nonseptic experimental lung injury. Anesthesiology 118(2):395Y408, 2013. Bagchi A, Eikermann M: Mashed potatoes and maize: are the starches safe? Anesthesiology 118(2):244Y247, 2013. Roch A, Blayac D, Ramiara P, Chetaille B, Marin V, Michelet P, Lambert D, Papazian L, Auffray JP, Carpentier JP: Comparison of lung injury after normal or small volume optimized resuscitation in a model of hemorrhagic shock. Intensive Care Med 33(9):1645Y1654, 2007. Spahn DR, Bouillon B, Cerny V, Coats TJ, Duranteau J, Ferna´ndez-Monde´jar E, Filipescu D, Hunt BJ, Komadina R, Nardi G, et al.: Management of bleeding and coagulopathy following major trauma: an updated European guideline. Crit Care 17(2):R76, 2013.

FLUID RESUSCITATION STRATEGIES

165

34. Starkey K, Keene D, Morrison JJ, Doughty H, Midwinter MJ, Woolley T, Jansen JO: Impact of high ratios of plasma-to-red cell concentrate on the incidence of acute respiratory distress syndrome in UK transfused combat casualties. Shock 40(1):15Y20, 2013. 35. Hildebrand F, Weuster M, Mommsen P, Mohr J, Fro¨hlich M, Witte I, Keibl C, Ruchholtz S, Seekamp A, Pape HC, et al.: A combined trauma model of chest and abdominal trauma with hemorrhagic shock: description of a new porcine model. Shock 38(6):664Y670, 2012. 36. Brunkhorst FM, Engel C, Bloos F, Meier-Hellmann A, Ragaller M, Weiler N, Moerer O, Gruendling M, Oppert M, Grond S, et al.: Intensive insulin therapy and pentastarch resuscitation in severe sepsis. N Engl J Med 358(2): 125Y139, 2008. ˚ neman A, 37. Perner A, Haase N, Guttormsen AB, Tenhunen J, Klemenzson G, A Madsen KR, Møller MH, Elkj&r JM, Poulsen LM, et al.: 6S Trial Group: Hydroxyethyl starch 130/0.42 versus Ringer’s acetate in severe sepsis. N Engl J Med 367(2):124Y134, 2012. 38. Myburgh JA, Finfer S, Bellomo R, Billot L, Cass A, Gattas D, Glass P, Lipman J, Liu B, McArthur C, et al.: Hydroxyethyl starch or saline for fluid resuscitation in intensive care. N Engl J Med 367(20):1901Y1911, 2012.

Copyright © 2013 by the Shock Society. Unauthorized reproduction of this article is prohibited.