Secondary Logo

Journal Logo

EDITORIAL: Editorial

Treatment of Uncontrolled Hemorrhagic Shock After Liver Trauma: Fatal Effects of Fluid Resuscitation Versus Improved Outcome After Vasopressin

Raedler, Claus, MD*; Voelckel, Wolfgang G., MD*; Wenzel, Volker, MD*; Krismer, Anette C., MD*; Schmittinger, Christian A., DVM*; Herff, Holger, BS*; Mayr, Viktoria D., MD*; Stadlbauer, Karl H., MD*; Lindner, Karl H., MD*; Königsrainer, Alfred, MD

Author Information
doi: 10.1213/01.ANE.0000117150.29361.5A
  • Free

Resuscitation of patients in uncontrolled hemorrhagic shock is one of the most challenging aspects of trauma care. The traditional strategy is to infuse large amounts of IV fluids to maintain circulatory homeostasis. Whereas fluid management is definitely established in controlled hemorrhagic shock, its role is controversial in uncontrolled hemorrhagic shock (1). Although improving and maintaining arterial blood pressure may prevent circulatory shock during uncontrolled hemorrhage, it may worsen bleeding as well because increased arterial blood pressure may impair the formation of new blood clots or dislodge existing ones (2). In fact, in patients with penetrating torso injuries and subsequent uncontrolled hemorrhagic shock, delaying aggressive fluid resuscitation resulted in a survival benefit (3). Thus, it may be beneficial to delay fluid resuscitation in victims of uncontrolled hemorrhagic shock until intervention is performed (4).

In previous experiments in our research laboratory, we found that vasopressin shifts blood during low-flow states, such as cardiopulmonary resuscitation and hemorrhagic shock, from the gut, muscle, and skin towards the heart and brain, thus improving vital organ blood flow, coronary perfusion pressure, resuscitability, and long-term survival (5–9). Subsequently, this finding was questioned on the basis that decreased perfusion of the gut (10) may result in tissue necrosis and subsequent translocation sepsis in the postresuscitation phase. Because we found that vasopressin-mediated transient hypoperfusion of abdominal organs (11) may not impede long-term survival, we speculated that vasopressin's effects could be used in a therapeutic manner. Because severe uncontrolled abdominal bleeding may be a pathophysiologic state that can be managed with vasopressin, knowledge about regional abdominal perfusion in this setting is of utmost importance.

Therefore, we chose to study the effects of vasopressin versus saline placebo versus fluid resuscitation in a porcine model of uncontrolled hemorrhagic shock because of a simulated severe liver injury on regional abdominal perfusion, hemodynamic variables, and survival. Our hypothesis was that there would be no differences between groups with regard to the study end-points (regional abdominal organ blood flow, hemodynamic variables, and survival) throughout the experimental protocol.


This project was approved by the Austrian Federal Animal Investigation Committee. Animal care and use were performed by qualified individuals supervised by veterinarians, and all facilities and transportation complied with current legal requirements and guidelines. Anesthesia was used in all surgical interventions, all unnecessary suffering was avoided, and research was terminated if unnecessary pain resulted.

This study was performed according to the Utstein-style guidelines (12) on 21 healthy, 12- to 16-wk-old swine (Tyrolean domestic pigs) of either sex, weighing 30–40 kg. The pigs were fasted overnight but had free access to water. The pigs were premedicated with azaperone (neuroleptic drug; 4 mg/kg IM) and atropine (0.1 mg/kg IM) 1 h before surgery, and anesthesia was induced with propofol (1–2 mg/kg IV). After tracheal intubation during spontaneous respiration, the pigs were ventilated with a volume-controlled ventilator (Draeger EV-A, Lü beck, Germany) with 35% O2 at 20 breaths/min and with a tidal volume adjusted to maintain normocapnia. Anesthesia was maintained with propofol (6–8 mg · kg−1 · h−1) and a single dose of piritramide (0.5 mg/kg). Lactated Ringer's solution (6 mL · kg−1 · h−1) was administered continuously throughout the preparation period to replace fluid loss during instrumentation. A standard lead II electrocardiogram was used to monitor cardiac rhythm. Depth of anesthesia was judged according to arterial blood pressure, heart rate, and electroencephalographic monitoring (Neurotrac; Engström, Munich, Germany). Body core temperature was maintained with a heating blanket between 38.0°C (100.4°F) and 39.0°C (102.2°F) during surgical preparation. If clinical assessment or physiological measurements indicated a decreasing level of anesthesia, additional propofol and piritramide were given.

After instrumentation for hemodynamic variables, a midline laparotomy was performed, and the hepatic and left renal arteries and the portal vein were dissected carefully from their supporting tissues and subsequently instrumented with ultrasound flowprobes (Transonic, Ithaca, NY) to measure regional organ perfusion, as previously described, and validated (13–15). Propofol infusion was adjusted to 2 mg · kg−1 · h−1, and infusion of lactated Ringer's solution was stopped before the induction of shock.

After assessing baseline hemodynamic values and abdominal regional organ blood flow, an incision (width, 12 cm; depth, 3 cm) and subsequent finger fraction were performed across the right liver lobe. During the first, or nonintervention, phase to determine the exact amount of blood loss, blood was continuously removed from the abdominal cavity using a suction device with its tip located approximately 10 cm away from the liver injury. Had a blood clot been formed on the wound surface, removal of this blood clot would not have been possible (16,17). When mean arterial blood pressure was <20 mm Hg and heart rate decreased progressively for more than 30% of its peak value, therapeutic intervention was provided for 30 min to simulate a prehospital or transport phase before surgical intervention. At this time, pigs were randomly assigned to receive either an IV bolus dose of 0.4 U/kg of vasopressin (n =7), an equal volume of saline placebo (n = 7), or 1000 mL of lactated Ringer's solution and 1000 mL of hetastarch by a pressurized infusion (n = 7). This (fluid resuscitation) strategy was used to replace the estimated half of the blood volume lost with an equal volume of hetastarch and the other half with the threefold larger volume of a crystalloid solution, as widely used in European emergency medical service systems. Fluid resuscitation was initially set at ~2 mL · kg−1 · min−1 over the first 10 min. If this approach failed to restore arterial blood pressure, fluid resuscitation was enhanced to ~8 mL · kg−1 · min−1. Fluids were warmed to 38.5°C, reflecting the pig's normal body temperature. Vasopressin and saline placebo were flushed with 20 mL of normal saline; subsequently, a continuous infusion of 0.04 U · kg−1 · min−1 of vasopressin or saline placebo at an equal infusion rate was administered in each respective group. Fluid resuscitation pigs received no further drug therapy but only continuous fluid resuscitation. Investigators were blinded to treatment groups. At 30 min after therapeutic intervention, bleeding in all surviving pigs was controlled by manual compression of the liver injury, simulating emergency laparotomy, and 1000 mL of colloid and 1000 mL of crystalloid solutions were administered IV. All surviving pigs were observed for 1 h during this third or “hospital phase” of the experimental protocol. After finishing the experimental protocol, the pigs were killed with an overdose of fentanyl, propofol, and potassium chloride.

All values are expressed as mean ± SEM. The comparability of weight and baseline data were tested with the t-test for continuous variables. One-way analysis of variance was used to determine statistical significance among the three groups, followed by the Student-Newman-Keuls post hoc test. Because blood flow data were distributed unevenly, the Mann-Whitney U-test was used to determine differences among groups. For multiple comparisons, the P value was subsequently adjusted with the Bonferroni method. Paired Student's t-test (two-tailed) was used for comparisons within groups. Survival was compared using Fisher's exact test. We considered a two-tailed value of P < 0.05 statistically significant.


Before liver injury and the induction of hypovolemic shock, there were no differences in weight, temperature, hemodynamic variables, regional organ blood flow, and blood gases among groups. After 30 min of uncontrolled hemorrhagic shock, total blood loss per kilogram of body weight in the vasopressin versus saline placebo versus fluid resuscitation pigs was 35 ± 2 versus 35 ± 1 versus 37 ± 2 mL/kg, respectively (not significant). Criteria for therapeutic intervention were reached in the vasopressin versus saline placebo versus fluid resuscitation group after 34 ± 2 versus 36 ± 3 versus 32 ± 2 min, respectively (not significant).

Because arterial blood pressure deteriorated and blood loss increased in the fluid-resuscitated pigs within the first 10 min, the fluid resuscitation protocol of ~2 mL · kg−1 · min−1 was subsequently changed in all pigs to ~8 mL · kg−1 · min−1. However, fluid resuscitation failed to stabilize arterial blood pressure, and median time to death in the fluid resuscitated pigs was 15 min (range, 10–24 min) (Table 1).

Table 1
Table 1:
Fluid Resuscitation Protocol in the Fluid Resuscitation Group

Vasopressin, but not saline placebo or fluid resuscitation, resulted in a significant increase of mean arterial blood pressure (Fig. 1). Total blood loss after therapeutic intervention was significantly more in the fluid resuscitation pigs, whereas no further bleeding was noted in the vasopressin and saline placebo groups (Fig. 1). Blood flow to the liver in the vasopressin-treated pigs was temporarily impaired at 2.5 min after drug administration when compared with saline placebo and fluid resuscitation but started to increase shortly thereafter (Fig. 2). Mean arterial blood pressure and heart rate further deteriorated in all fluid resuscitation and saline placebo pigs. Accordingly, median time to death in 7 of 7 saline placebo pigs was 10 min (range, 3–20 min), whereas 7 of 7 vasopressin pigs survived until bleeding was surgically controlled and 60 min thereafter (P < 0.01). Blood gas values are shown in Table 2.

Figure 1.
Figure 1.:
Mean ± SEM heart rate, mean arterial blood pressure, and total blood loss before, during, and after the administration of a 0.4-U/kg bolus dose and a 0.04-U · kg−1 · min−1 continuous infusion of vasopressin (squares) versus equal volumes of saline placebo (circles) versus fluid resuscitation (triangles). Uncontrolled hemorrhage indicates the no-treatment interval after liver injury; liver tamponade indicates the manual compression of the liver injury to control bleeding; ‡P < 0.05 for vasopressin and fluid resuscitation versus saline placebo; †P < 0.05 for vasopressin versus saline placebo; ΔP < 0.05 for vasopressin and fluid resuscitation versus saline placebo and vasopressin versus fluid resuscitation; *P < 0.05 for fluid resuscitation versus vasopressin and saline placebo. No statistical analysis was performed after Minute 10 during therapeutic intervention because of death of all fluid resuscitation and saline placebo pigs.
Figure 2.
Figure 2.:
Mean ± SEM portal vein, hepatic artery, and left renal artery blood flow before, during, and after administration of a 0.4-U/kg bolus dose and a 0.04-U · kg−1 · min−1 continuous infusion of vasopressin (squares) versus equal volumes of saline placebo (circles) versus fluid resuscitation (triangles). Uncontrolled hemorrhage indicates the no-treatment interval after liver injury; liver tamponade indicates the manual compression of the liver injury to control bleeding; *P < 0.05 for fluid resuscitation versus vasopressin and saline placebo; #P < 0.05 for fluid resuscitation versus saline placebo; §P < 0.05 for vasopressin versus fluid resuscitation and saline placebo; †P < 0.05 for vasopressin versus saline placebo. No statistical analysis was performed after Minute 10 during therapeutic intervention because of death of all fluid resuscitation and saline placebo pigs.
Table 2
Table 2:
Arterial Blood Gas Variables, Hemoglobin, and Lactate Values During Severe Liver Injury in Pigs


The present model of uncontrolled hemorrhagic shock may be similar to a prehospital investigation (3) studying patients suffering from penetrating torso injuries that cannot be satisfactorily controlled in the field but are controlled exclusively by surgical management in the hospital. Clinical data revealing a survival benefit of delayed fluid resuscitations are in full agreement with the results of our swine model, in which aggressive immediate fluid resuscitation could not maintain vital organ blood flow but resulted in the rapid death in these animals; moreover, no benefit of fluid resuscitation over either vasopressin or saline placebo could be detected.

Immediately after aggressive fluid resuscitation, portal vein and hepatic artery blood flow, as well as mean arterial blood pressure, increased to ~50% of baseline values, but blood loss caused by continuing hemorrhage was ~100% more than in saline placebo control pigs. This indicates that aggressive fluid resuscitation was associated with increased blood loss and no survival benefit. As indicated by a progressively decreasing heart rate and mean arterial blood pressure, fluid resuscitation reflected only a very transient effect, resulting in rapidly declining cardiocirculatory variables. It is likely that increased portal vein blood flow was simply the cause of increased bleeding in the liver and therefore led to increased blood loss after fluid resuscitation. Although not proven in our study, it is likely that fluids diluted coagulation factors to a level that was unable to terminate bleeding, especially in the presence of high hydrostatic pressure on the wound. Interestingly, the strategy of doing nothing in the saline placebo control group and the aggressive fluid resuscitation used in one of the two intervention groups resulted in death at an almost identical time point after therapeutic intervention. Whereas pigs in the fluid resuscitation group succumbed to uncontrolled bleeding in a way that is often observed in severely injured patients in the emergency room, the saline placebo pigs most likely died because of refractory cardiocirculatory shock, as indicated by a continuing mean arterial blood pressure after therapeutic intervention that is closely related to the hydrostatic pressure in the aorta.

When arterial blood pressure during hemorrhagic shock is barely detectable, and pulseless electrical activity or bradyasystolic rhythms become imminent, pharmacologic support strategies are required (18). Although epinephrine is widely used for treatment of periarrest bradyarrhythmias during hemorrhagic shock, the value of its recommendation has been debated by both the American Heart Association and the European Resuscitation Council (19,20). In contrast, vasopressin is a uniquely effective pressor in the irreversible phase of hemorrhagic shock that is unresponsive to volume replacement and catecholamine vasopressor (8,9,21). Thus, vasopressin has already been successfully used in a few patients suffering from intraabdominal bleeding and subsequent shock that was unresponsive to volume replacement (22).

Based on its physiology, vasopressin may be an interesting option for managing uncontrolled hemorrhage in the extremities and below the diaphragm. Vasopressin leads to peripheral vasoconstriction via V1-receptors in the vasculature and shifts blood primarily from the skeletal muscle, cutaneous, and splanchnic bed to the heart and brain (23,24). This indicates that vasopressin may reflect two advantages in uncontrolled hemorrhagic shock in the abdomen: it may decrease bleeding first by shifting blood away from the injury and by improving vital organ blood flow (8,9). In fact, our model showed that vasopressin decreased portal vein blood flow while maintaining hepatic artery blood flow, which resulted in blood loss similar to that of the saline placebo control animals, but fundamentally improved mean arterial blood pressure. We suggest that this short venous no-flow phase after vasopressin allowed the formation of new blood clots, thus decreasing subsequent bleeding, a phenomenon that did not occur in the fluid resuscitation pigs, leading subsequently to rapid death. Moreover, arterial reperfusion of the liver was immediately restored after vasopressin. As such, the observed increase of serum potassium in the vasopressin pigs shortly after hepatic reperfusion was highly indicative of tissue anoxia during the extreme low-flow state during shock and reperfusion washout after hemodynamic stabilization.

Interestingly, despite a continuing vasopressin infusion, blood flow to the liver returned to ~25% of baseline values within 10 minutes after drug administration. This is in good relation to the results of Ericsson (25), who found mesenteric artery blood flow to be improved by vasopressin during hypovolemia. As previously observed in a cardiac arrest model in hypovolemic shock (26), renal blood flow was preserved by vasopressin, which could be another beneficial effect because avoiding acute kidney failure during severe shock may improve chances for long-term survival.

Shortly after fluid replacement in the vasopressin pigs, mean arterial blood pressure and blood flow to the gut and kidney returned to baseline values. By contrast, we found hepatic artery blood flow to be more than twofold larger within the postresuscitation phase, a fact that may be related to the degree of trauma and tissue anoxia. With regard to decreasing lactate levels in the postresuscitation phase, two possible mechanisms are a dilution effect or lactate metabolism in the liver; unfortunately, we were unable to prove either one. In this regard, it is noteworthy that we replaced crystalloid and colloid infusions only, thus resulting in hemoglobin levels as small as 29 g/L after achieving a near-normovolemic condition. As a result of this profound anemia and the prolonged shock phase, we observed a continuing acidotic situation. Nevertheless, there was no need for further vasopressor support in the postresuscitation phase, which may be explained by achieving normovolemia after fluid replacement and a continuing vasopressin level increase (27,28).

Some limitations of the present study should be noted. First, different vasopressin receptors in pigs (lysine vasopressin) and humans (arginine vasopressin) may result in a different hemodynamic response to exogenously-administered vasopressin. Second, this is not a long-term survival study; accordingly, the likelihood of a postshock multiorgan dysfunction syndrome cannot be excluded. Also, the present model reflects fundamental but local trauma that is mainly accompanied by venous bleeding originating from the liver injury. Whether our data can be extrapolated to other major bleeding sources, such as pelvic or open humerus fractures, requires investigation. Finally, we are unable to determine whether fluid resuscitation and saline placebo pigs could have survived had we initiated surgical management earlier; moreover, we are unable to compare laboratory variables and vital organ blood flow values in the postresuscitation phase because of death of all fluid resuscitation and saline placebo pigs. In conclusion, in this model of severe liver trauma with uncontrolled hemorrhagic shock, vasopressin, but not saline placebo or fluid resuscitation, improved short-term survival.

We greatly appreciate the outstanding expertise of Günter Klima, MD, PhD, in animal care.


1. Roberts I, Evans P, Bunn F, et al. Is the normalisation of blood pressure in bleeding trauma patients harmful? Lancet 2001;357:385–7.
2. Dries DJ. Hypotensive resuscitation. Shock 1996;6:311–6.
3. Bickell WH, Wall MJ, Pepe PE, et al. Immediate versus delayed fluid resuscitation for hypotensive patients with penetrating torso injuries. N Engl J Med 1994;331:1105–9.
4. Pepe PE, Mosesso VN Jr, Falk JL. Prehospital fluid resuscitation of the patient with major trauma. Prehosp Emerg Care 2002;6:81–91.
5. Lindner KH, Prengel AW, Pfenninger EG, et al. Vasopressin improves vital organ blood flow during closed-chest CPR in pigs. Circulation 1995;91:215–21.
6. Wenzel V, Lindenr KH, Krismer AC, et al. Repeated administration of vasopressin but not epinephrine maintains coronary perfusion pressure after early and late administration during prolonged cardiopulmonary resuscitation in pigs. Circulation 1999;99:1379–84.
7. Wenzel V, Lindner KH, Krismer AC, et al. Survival with full neurological recovery, and no cerebral pathology after prolonged cardiopulmonary resuscitation with vasopressin in pigs. J Am Coll Cardiol 2000;35:527–33.
8. Voelckel WG, Raedler C, Wenzel V, et al. Arginine vasopressin, but not epinephrine, improves survival in uncontrolled hemorrhagic shock after liver trauma in pigs. Crit Care Med 2003;31:1160–5.
9. Stadlbauer KH, Wagner-Berger HG, Raedler C, et al. Vasopressin, but not fluid resuscitation, enhances survival in a liver trauma model with uncontrolled and otherwise lethal hemorrhagic shock in pigs. Anesthesiology 2003;98:699–704.
10. Prengel AW, Lindner KH, Wenzel V, et al. Splanchnic and renal blood flow after cardiopulmonary resuscitation with epinephrine and vasopressin in pigs. Resuscitation 1998;38:19–24.
11. Voelckel WG, Lindner KH, Wenzel V, et al. Effects of vasopressin and epinephrine on splanchnic blood flow and renal function during and after cardiopulmonary resuscitation in pigs. Crit Care Med 2000;28:1083–8.
12. Idris AH, Becker LB, Ornato JP, et al. Utstein-style guidelines for uniform reporting of laboratory CPR research. Resuscitation 1996;33:69–84.
13. Burton RG, Gorewit RC. Ultrasonic flowmeter uses wide beam transit time technique. Med Electronics 1984;15:68–73.
14. D'Almeida MS, Cailmail S, Lebrec D. Validation of transit-time ultrasound flow probes to directly measure portal blood flow in conscious rats. Am J Physiol 1996;271:H2701–9.
15. D'Almeida MS, Gaudin C, Lebrec D. Validation of 1- and 2- mm transit-time ultrasound flow probes on mesenteric artery and aorta of rats. Am J Physiol 1995;268:1368–72.
16. Jeroukhimov I, Jewelewicz D, Zaias J, et al. Early injection of high-dose recombinant factor VIIa decreases blood loss and prolongs time from injury to death in experimental liver injury. J Trauma 2002;53:1053–7.
17. Katz LM, Manning JE, McCurdy S, et al. HBOC-201 improves survival in a swine model of hemorrhagic shock and liver injury. Resuscitation 2002;54:77–87.
18. Shoemaker WC, Peitzman AB, Bellamy R, et al. Resuscitation from severe hemorrhage. Crit Care Med 1996;24:S12–23.
19. Anonymous. Guidelines 2000 for cardiopulmonary resuscitation and emergency cardiovascular care. Circulation 2000;102:I-1-I-384.
20. Anonymous. Guidelines 2000 for cardiopulmonary resuscitation and emergency cardiovascular care: international consensus on science. Resuscitation 2000;46:1–447.
21. Morales D, Madigan J, Cullinane S, et al. Reversal by vasopressin of intractable hypotension in the late phase of hemorrhagic shock. Circulation 1999;100:226–9.
22. Shelly MP, Greatorex R, Calne RY, Park GR. The physiological effects of vasopressin when used to control intra-abdominal bleeding. Intensive Care Med 1988;14:526–31.
23. Granger DN, Richardson PDI, Kvietys PR. Intestinal blood flow. Gastroenterology 1980;78:837–63.
24. Reilly PM, Bulkley GB. Vasoactive mediators and splanchnic perfusion. Crit Care Med 1993;21:S55–68.
25. Ericsson BF. Hemodynamic effect of vasopressin: an experimental study in normovolemic and hypovolemic anesthetized dogs. Acta Chir Scand Suppl 1971;414:1–29.
26. Voelckel WG, Lurie KG, Lindner KH, et al. Vasopressin improves survival after cardiac arrest in hypovolemic shock. Anesth Analg 2000;91:627–34.
27. Reid IA, Schwartz J. Role of vasopressin in the control of blood pressure. In: Martini L, Ganong WF, eds. Frontiers in neuroendocrinology. New York: Raven Press, 1984:177–97.
28. Wenzel V, Lindner KH, Prengel AW, et al. Endobronchial vasopressin improves survival during cardiopulmonary resuscitation in pigs. Anesthesiology 1997;86:1375–81.
© 2004 International Anesthesia Research Society