Protocols for the management of hemorrhagic shock include aggressive IV fluid resuscitation to restore critical tissue perfusion and oxygen delivery, although there is a continuing debate regarding the best type of solution (i.e., crystalloid versus colloid infusion) and the timing of fluid resuscitation (i.e., prehospital versus delayed fluid resuscitation).1 Rapid infusion of small amounts of hypertonic saline solutions has proven to be effective in managing hypotensive patients suffering from severe traumatic brain injury and burn shock.2–4 The recommended infusion of 4 mL/kg 7.5% hypertonic solution has been reported to be as effective as infusion of 2–3 L of crystalloid solution in restoring intravascular volume.4 Hypertonic saline/hydroxyethyl starch (HS-HES) has been shown to be more beneficial with regard to hemodynamic and oxygen transport than is the administration of 6% HES alone.5 However, all IV fluids cause dose-related dilutional coagulopathy and exhibit intrinsic effects on the hemostatic system. Published data show that administration of HS alone alters plasma clotting times and platelet aggregation.6 Nevertheless, because small-volume resuscitation requires a limited volume and colloid content for effective restoration of hemodynamic status, it might impair coagulation less than commonly used amounts of gelatin or HES solutions. Until now, only a few studies have investigated the effect of hypertonic solutions on the hemostatic process, and have mainly used standard coagulation tests.7,8 To test the hypothesis that 4 mL/kg HS-HES exerts fewer effects on the entire coagulation and clot formation process during massive hemorrhage than do commonly used amounts of 4% gelatin solution or HES 130/0.4, we performed an experimental study in pigs using the functional measurement technique of thrombelastometry (ROTEM®) and whole blood impedance aggregometry (Multiplate®), in addition to standard coagulation tests and measurement of coagulation factors FII and FXIII.
The study was approved by the Austrian Federal Animal Investigation Committee, and the animals were managed in accordance with the American Physiological Society institutional guidelines, and the Position of the American Heart Association on Research Animal Use, as adopted on November 11, 1984. Anesthesia was used in all surgical interventions, all unnecessary suffering was avoided, and research was terminated if unnecessary pain or distress resulted. Our animal facilities meet the standards of the American Association for Accreditation of Laboratory Animal Care.
Surgical Preparations and Measurements
This study was performed in 30 healthy, 12–16-wk-old pigs weighing 35–45 kg each. The animals were fasted overnight, but had free access to water. The pigs were premedicated 1 h before surgery with 4 mg/kg azaperone IM and 0.1 mg/kg atropine IM. Anesthesia was induced with 20 mg/kg ketamine IM and IV administration of 1–2 mg/kg propofol. Anesthesia was maintained with IV administration of 6–8 mg · kg−1 · h−1 propofol, 30 mg piritramide IV, and 0.2 mg/kg pancuronium. After endotracheal intubation, the pigs' lungs were ventilated with oxygen 35% using a volume-controlled ventilator (Draeger Evita, Luebeck, Germany) at a rate of 20 bpm with tidal volume adjusted to maintain normocapnia. A 6 F catheter and a 12 F large-bore catheter were inserted for collection of blood samples, continuous arterial blood pressure monitoring and for blood withdrawal and intravascular volume resuscitation. Hemodynamic variables were obtained with a pulmonary arterial catheter. The baseline fluid requirement (4 mg · kg−1 · h−1 IV) was substituted with a crystalloid solution (lactated Ringer's) via peripheral venous access during the entire course of the procedure. Body temperature was maintained between 38.0°C and 39.0°C throughout the study. Finally, midline laparotomy was performed for liver incision.
The study's time course is displayed in Figure 1. After surgical procedures were completed, baseline variables (measurement time point 1; baseline) were recorded. The animals subsequently underwent hemorrhage by withdrawing 60% of the estimated total blood volume (70 mL/kg) (time point 2; hemorrhage). Shed blood was collected and processed in a continuous cell saver system (CATS®, Fresenius, Vienna, Austria) using sodium citrate 3.8% in a ratio of 1:7 as anticoagulant.
Using a computer-generated randomization list, after blood withdrawal each, 10 animals randomly received the tested solutions as follows: 4 mL/kg of a 7.2% HS/6% HES 200/0.62 (Hyperhes®, Fresenius Kabi, Graz, Austria) as recommended for treatment of severe hemorrhage,9 or 6% HES 130/0.4 (Voluven®, Fresenius Kabi), or 4% gelatin solution (Gelofusin®, Braun Melsungen, Melsungen, Germany) in amounts calculated to replace the lost blood volume while taking into consideration the different volume-expanding effects of 4% gelatin (70%) and HES 130/0.4 (100%).10,11 All fluids were administered continuously over 30 min. Concomitantly, all animals received their autologous processed red cell concentrate over 30 min, independently of the actual hemoglobin level, to prevent early death from anemia. After fluid administration, all variables were measured again (time point 3; fluid resuscitation).
To determine and compare the surrogate markers for hemostasis to those of clinical consequences (bleeding tendency) provoked by blood withdrawal and fluid administration, as previously reported,12,13 a hepatic incision (7-cm long and 1.5-cm deep standardized; using a template; always performed by the same investigator who was unaware of the group to which the animals were assigned) was made in the right liver lobe (central to the Lig. falciforme above the central lobe) to induce uncontrolled hemorrhage. After hepatic incision, the amount of shed blood and the time to death from hemorrhagic shock were determined. During the phase of uncontrolled hemorrhage from liver injury, animals received neither catecholamines nor further intravascular volume supply. If an animal died within the first 120 min, the last blood sample was taken immediately before anticipated death, which was defined as pulseless electrical activity, mean arterial blood pressure below 10 mm Hg, or end-tidal carbon dioxide below 10 mm Hg. Animals surviving more than 120 min were killed with an overdose of piritramide, propofol, and potassium chloride (measurement time point 4; end-point) (120 min after liver incision or immediately before anticipated death).
Blood Sampling and Measurements
All blood samples were drawn from the femoral artery, whereby the first 10 mL of blood was discarded. Blood samples were analyzed by thrombelastometry (ROTEM®, Pentapharm, Munich, Germany); prothrombin time (PT), concentrations of fibrinogen (Clauss method), prothrombin (FII), factor XIII (FXIII), blood cell count, and colloid osmotic pressure (COP) were determined by standard laboratory methods using the appropriate tests from Dade Behring, Marburg, Germany, and the Amelung Coagulometer, Baxter, UK. ROTEM assays were performed at 37°C immediately after drawing arterial blood into 3 mL citrated tubes (Sarstedt, Numbrecht, Germany) using the extrinsically activated ExTEM® test and the FibTEM® assay, which also contains the platelet-blocking substance cytochalasin D, thereby enabling measurement of the fibrinogen/fibrin content of the clot. Reference variables for ROTEM analysis and standard laboratory tests in pigs have been published.14 A typical ROTEM tracing and its interpretation are given in Figure 2.
Finally, whole blood platelet impedance aggregometry was analyzed (Multiplate®, Dynabyte Medical, Munich, Germany). Multiplate technology is based on impedance aggregometry, assessing platelet aggregation in whole blood after activation with adenosine diphosphate (0.2 mmol) or collagen (100 μg/mL) in the presence of 15 μg/mL direct thrombin inhibitor (hirudin in a final concentration of 25 μg/mL). The results of the dual measurements in hirudin-containing tubes [300 μL whole blood, 300 μL saline, and 20 μL activator (collagen) solution] are graphically displayed as the area under the aggregation curve.
The sample size of 10 animals per group was calculated using nQuery Advisor® software (Statistical Solutions, Ireland) with reference to mean (sd) values derived from previous data as usual to achieve 90% power with an anticipated FibTEM-maximum clot firmness difference of 10.4 versus 5.1 mm (common sd 3.4, t Test), thus allowing detection of a difference in ROTEM variables among infusion groups. The power of 90% was used to compensate for the test power lost when applying nonparametric tests. Data showed no normal distribution and were thus analyzed with nonparametric tests. A nonparametric Friedman was applied to analyze a possible time effect in each group. Calculated differences between measurement points 2 and 3 were compared among all three groups using a Kruskal–Wallis test with posthoc pairwise comparisons, using an unpaired Wilcoxon test in the case of a significant overall test. Thrombelastometry parameters and blood loss are presented in box plots (minimum, first quartile, median, third quartile, and maximum). P < 5% was considered statistically significant.
At baseline, no significant differences with regard to hemodynamics and hemostasis were observed among groups. All data are given as medians (25th, 75th percentile).
After withdrawal of about 60% of estimated total blood volume [1484 mL (1369, 1624)], which is 42 mL/kg, pigs in the HS-HES group received 4 mL/kg, which was actually 166 mL.15,16 Those in the HES 130/0.4 group received 41 mL/kg of a 6% HES 130/0.4 solution [1565 mL (1500, 1691)], and animals in the gelatin group received 50 mL/kg of a 4% gelatin solution [1925 mL (1825, 2000)]. In addition, all animals received the total amount [25 mL/kg (23, 26)] of autologous processed red cells [HS-HES group: 1030 mL (990, 1085); gelatin group: 895 mL (786, 1062); HES group: 895 mL (786, 1062)].
ROTEM Analysis and Blood Loss
Throughout the study, coagulation time variables remained unchanged and largely within normal ranges in the HS-HES group, but increased significantly after administration of gelatin or HES 130/0.4, showing median variables above the upper normal. Similar changes were observed for increasing clotting formation time and decreasing α angle. In addition, the decrease in clot firmness and fibrinogen polymerization was significantly less in the HS-HES group when compared with the gelatin and HES 130/0.4 groups. All these ROTEM variables showed no major differences between gelatin and HES 130/0.4 (Fig. 3).
Blood loss after liver incision was comparable in the gelatin and 6% HES 130 groups, but significantly less after HS-HES administration (Fig. 4). Survival time after liver incision was 120 min (40, 120) in the HS-HES group (6 animals survived) when compared to 106 min (75, 120) in the gelatin group (5 animals survived); the shortest survival time was observed in the HES 130/0.4 group, being 38 min (30, 101) (3 animals survived). As expected, and because the study was not designed to evaluate differences in survival, no significant differences among groups were observed (Fig. 4).
After blood withdrawal, all laboratory variables decreased. As expected, after fluid administration, all these variables decreased still further, but least in the HS-HES group. COP increased in the HES 130/0.4 and gelatin groups, but decreased after HS-HES. With blood loss from liver injury at the end of the observation period, all these variables remained grossly unchanged. COP was significantly higher and FXIII concentration significantly lower in the pigs treated with gelatin than in the HES 130/0.4 group. No significant changes in calcium concentration or pH values were observed (data not given) (Table 1).
After blood withdrawal, all animals developed hemorrhagic shock. All hemodynamic variables improved after fluid administration, some even exceeded baseline levels. In the HS-HES group, the increase in cardiac index, central venous pressure (CVP), and pulmonary artery occlusion pressure was significantly less than in the gelatin group; CVP was also significantly less than in the HES 130/0.4 group. After blood loss due to liver incision, all animals again developed hemorrhagic shock (Table 2).
For the first time, functional effects of small-volume resuscitation (4 mL/kg HS-HES) on overall hemostasis and bleeding tendency were investigated and compared with those observed after administration of 4% gelatin or 6% HES 130/0.4 solution to replace the total lost blood volume. These findings most likely result from the greater dilutional effect provoked by the administration of gelatin in a ratio of 1:1.3 or HES in a ratio of 1:1 to blood loss, considering that gelatin solution shows a volume efficacy of about 70%, whereas that of HES 130/0.4 is reported to be 100%.9 When using HS-HES, a smaller volume load is needed because the hypertonic solution exerts its final volume-expanding effect in vivo by drawing water from the intracellular and interstitial space. Reed et al. published their finding that each milliliter of HS-HES administered mobilizes an additional 7 mL of free water from the interstitial and intracellular space into the intravascular compartment,6 mathematically resulting in a net amount of 32 mL/kg fluid administered in the present study. However, the exact amounts of plasma expansion after hypertonic/hyperoncotic solutions are not known and may vary among individuals.
Besides the differences in concentrations of coagulation factors and in numbers of platelets, analysis of functional variables of hemostasis showed HS-HES resuscitation to have markedly fewer effects on the entire coagulation and clot formation process as assessable with the ROTEM technique than does gelatin or HES 130/0.4 (Table 3). In addition, blood loss after liver injury was significantly less in the animals treated with HS-HES when compared with the standard colloid groups. We are aware of the increase in CVP in the HES and the gelatin groups and the concomitant increase in systemic blood flow, which might have at least increased consecutive bleeding tendency. However, invasive monitoring is seldom available at the preclinical trauma scene to which our model refers. Measuring heart rate and mean arterial blood pressure might provide the only guiding variables in these situations, and these were comparable among our study groups. Interestingly, in a previous investigation15 using HES 130/0.4 for volume resuscitation, we also observed that CVP and pulmonary artery occlusion pressure increased three- to fourfold and more than twofold, respectively. Despite these findings of increased filling pressures, the animals treated with fibrinogen concentrate and prothrombin complex concentrate exhibited significantly less blood loss than did those receiving placebo, meaning that not only filling pressures but also hemostatic competence influence bleeding tendency. Thus, we hypothesize that the lesser bleeding tendency after small-volume resuscitation results from the combined effects of better-maintained clot formation and limited filling pressures. Clot formation sustainability most likely resulted from the lesser decrease in fibrinogen, which is a key factor in the resulting clot strength and it also influences thrombin generation; Beguin et al. showed that polymerizing fibrin interacts with von Willebrand factor so as to activate the platelet GPIb receptor.16,17 In addition, in the HS-HES-treated animals, platelet and red cell count decreased less than for either colloid. Because platelet count was also not critically reduced in the gelatin and HES 130/0.4 groups, we suggest the differences in fibrinogen concentrations and fibrinogen polymerization as the prime underlying mechanisms of impaired clot formation. Nevertheless, we cannot exclude the possibility that the smaller decrease in red cells contributed to these effects.
This study was conducted because some investigators have reported negative effects of hypertonic solutions on the hemostatic process. Most of these studies used 7.5% NaCl/dextran solutions and are summarized in a recently published review article.18 The authors concluded that the available data show no increase in blood loss after the administration of recommended doses. Because dextran exerts somewhat different effects on hemostasis than do the various HES preparations,18,19 these data are not directly comparable with ours. Further problems entailed in comparing our results with the literature arise from the fact that most of the data stem from in vitro assays and thus lack a reference to bleeding tendency,6,20–23 or come from animal studies that focused solely on bleeding volume without measuring the impact on the hemostatic system.24
Using only HS, Reed et al. found in vitro that 10% dilution of plasma with HS significantly deteriorated PT, activated partial thromboplastin time and platelet aggregation, while observing no change for dilution with normal saline.6 In a study of 27 patients7 undergoing abdominal surgery and receiving 4 mL/kg HS-HES or 4 mL/kg 10% HES 200, all coagulation variables changed significantly in the HS-HES group, except thrombelastographic reaction time, while in the HES 200 group only PT variables and platelet count changed significantly. The authors hypothesized that with hypertonic solutions the osmotic stress and membrane pleating may aggravate HES-induced changes in membrane fluidity and microviscosity. It was our intention to investigate the overall hemostatic capacity as monitored in clinical practice during massive hemorrhage and as assessed with thrombelastometry and estimation of blood loss. We did not specifically investigate platelet response to intravascular volume regimens in pigs, which, moreover, shares methodological problems (species-related response, influence of thrombocytopenia during hemorrhage and dilution). However, because platelet-related effects of hypertonic solutions were suggested, we measured platelet aggregation with the available Mulitplate device, which results should not be influenced by platelet count until numbers below 50 g/L are reached. We present these data showing that platelet aggregation decreased uniformly after hemorrhage and fluid administration in all groups, with the only exception being that adenosine diphosphate-activated aggregation in the gelatin group increased after hemorrhage, which we cannot explain. Because functional platelet monitoring using the Multiplate device in pigs is not yet well established, and because variability was huge, our findings should be classified as preliminary and interpreted with caution. Our study was powered to detect differences in the ROTEM tests, for which normal values in pigs have been investigated14 and which are also used in clinical practice to monitor patients during massive blood loss. Further studies are needed to answer the question as to whether platelet function per se is influenced by hypertonic solutions. However, the fact that blood loss was significantly less with hypertonic solution management indicates that, even if platelet function is more impaired with hypertonic solutions than with conventional colloids, this effect seems to be counterbalanced by a lesser effect on fibrinogen/fibrin polymerization and fewer dilutional changes.
Some further limitations of the present study need to be discussed. As with any animal model, data cannot be directly transposed to humans. The principal advantage of an animal model is that it ensures identical and standardized treatment, permits measurements at defined time points and lacks the interference of therapy. Although patients showing severe acute blood loss, to which our study refers, are commonly treated first by transfusing red cell concentrates, our results might be influenced by retransfusion of autologous red cells processed by a cell salvage device. However, this intervention was identical in all groups, and the probable influence on the hemostatic process was similar for all groups. As in humans, the health status of our animals may differ. However, as all animals were especially bred for medical research and regularly examined by veterinarians, their health status was beyond our control. Last, our results were obtained with a single administration of HS-HES, and we cannot estimate the consequences of repetitive dosages.
In conclusion, the impact of administration of 4 mL/kg hypertonic/hyperoncotic solution on clot formation is minor when compared with that of fluid resuscitation using 6% HES 130/0.4 or 4% gelatin and also enables effective resuscitation after hemorrhagic shock without aggressive intravascular volume load, which causes profound dilution and/or specific impairment of clot formation, especially fibrinogen polymerization.
1. Hoyt DB. A clinical review of bleeding dilemmas in trauma. Semin Hematol 2004;41:40–3
2. Wade CE, Kramer GC, Grady JJ, Fabian TC, Younes RN. Efficacy of hypertonic 7.5% saline and 6% dextran-70 in treating trauma: a meta-analysis of controlled clinical studies. Surgery 1997;122:609–16
3. Younes RN, Aun F, Accioly CQ, Casale LP, Szajnbok I, Birolini D. Hypertonic solutions in the treatment of hypovolemic shock: a prospective, randomized study in patients admitted to the emergency room. Surgery 1992;111:380–5
4. Thompson R, Greaves I. Hypertonic saline–hydroxyethyl starch in trauma resuscitation. J R Army Med Corps 2006;152:6–12
5. Sirieix D, Hongnat JM, Delayance S, D'Attellis N, Vicaut E, Berrebi A, Paris M, Fabiani JN, Carpentier A, Baron JF. Comparison of the acute hemodynamic effects of hypertonic or colloid infusions immediately after mitral valve repair. Crit Care Med 1999;27:2159–65
6. Reed RL, Johnston TD, Chen Y, Fischer RP. Hypertonic saline alters plasma clotting times and platelet aggregation. J Trauma 1991;31:8–14
7. Scherer R, Giebler R, Kampe S, Kox WJ. Effects of hypertonic saline hydroxyethyl starch solution on collagen-induced platelet aggregation and ATP secretion. Infusionsther Transfusionsmed 1994;21:310–4
8. Boldt J, Kling D, Herold C, Dapper F, Hempelmann G. Volume therapy with hypertonic saline hydroxyethyl starch solution in cardiac surgery. Anaesthesia 1990;45:928–34
9. Boldt J. Volume replacement in the surgical patient–does the type of solution make a difference? Br J Anaesth 2000;84:783–93
10. Kramer GC. Hypertonic resuscitation: physiologic mechanisms and recommendations for trauma care. J Trauma 2003;54:S89–S99
11. Boldt J. Fluid choice for resuscitation of the trauma patient: a review of the physiological, pharmacological, and clinical evidence. Can J Anaesth 2004;51:500–13
12. Fries D, Krismer A, Klingler A, Streif W, Klima G, Wenzel V, Haas T, Innerhofer P. Effect of fibrinogen on reversal of dilutional coagulopathy: a porcine model. Br J Anaesth 2005;95:172–7
13. Velik-Salchner C, Haas T, Innerhofer P, Streif W, Nussbaumer W, Klingler A, Klima G, Martinowitz U, Fries D. The effect of fibrinogen concentrate on thrombocytopenia. J Thromb Haemost 2007;5:1019–25
14. Velik-Salchner C, Schnurer C, Fries D, Mussigang PR, Moser PL, Streif W, Kolbitsch C, Lorenz IH. Normal values for thrombelastography (ROTEM®) and selected coagulation parameters in porcine blood. Thromb Res 2005;117:597–602
15. Fries D, Haas T, Klingler A, Streif W, Klima G, Martini J, Wagner-Berger H, Innerhofer P. Efficacy of fibrinogen and prothrombin complex concentrate used to reverse dilutional coagulopathy–a porcine model. Br J Anaesth 2006;97: 460–7
16. Hemker HC, Al Dieri R, De Smedt E, Beguin S. Thrombin generation, a function test of the haemostatic-thrombotic system. Thromb Haemost 2006;96:553–61
17. Beguin S, Kumar R, Keularts I, Seligsohn U, Coller BS, Hemker HC. Fibrin-dependent platelet procoagulant activity requires GPIb receptors and von Willebrand factor. Blood 1999;93:564–70
18. Dubick MA, Bruttig SP, Wade CE. Issues of concern regarding the use of hypertonic/hyperoncotic fluid resuscitation of hemorrhagic hypotension. Shock 2006;25:321–8
19. de Jonge E, Levi M. Effects of different plasma substitutes on blood coagulation: a comparative review. Crit Care Med 2001;29:1261–7
20. Wilder DM, Reid TJ, Bakaltcheva IB. Hypertonic resuscitation and blood coagulation: in vitro comparison of several hypertonic solutions for their action on platelets and plasma coagulation. Thromb Res 2002;107:255–61
21. Tan TS, Tan KH, Ng HP, Loh MW. The effects of hypertonic saline solution (7.5%) on coagulation and fibrinolysis: an in vitro assessment using thromboelastography. Anaesthesia 2002;57: 644–8
22. Brummel-Ziedins K, Whelihan MF, Ziedins EG, Mann KG. The resuscitative fluid you choose may potentiate bleeding. J Trauma 2006;61:1350–8
23. Coats TJ, Heron M. The effect of hypertonic saline dextran on whole blood coagulation. Resuscitation 2004;60:101–4
24. Matsuoka T, Hildreth J, Wisner DH. Liver injury as a model of uncontrolled hemorrhagic shock: resuscitation with different hypertonic regimens. J Trauma 1995;39:674–80