Effects of synthetic versus natural colloid resuscitation on inducing dilutional coagulopathy and increasing hemorrhage in rabbits.

BACKGROUND: : On the basis of logistic benefits of colloids over crystalloids, the U.S. military selected Hextend for resuscitation of combat casualties in the field. We investigated the effects of resuscitation with this fluid, as well as other colloids, on coagulation and uncontrolled bleeding in rabbits subjected to a splenic injury. METHODS: : Anesthetized male New Zealand white rabbits (3.3 kg +/- 0.2 kg) were divided into three groups and subjected to hypothermia (35 degrees C +/- 0.5 degrees C) and approximately 40% isovolemic blood exchange (hemodilution) with Hextend (H); Dextran70 (D); or 5% human albumin (A) solution (n = 8/group). Complete blood count, arterial blood gas, and coagulation values were measured before and after hemodilution. Laparotomy was performed and a standard splenic injury causing uncontrolled hemorrhage was made. Rabbits were resuscitated (25 mL/kg) with the same colloid used for hemodilution to restore baseline blood pressure. Animals were monitored for 2 hours or until death. Blood loss and survival times were measured. RESULTS: : There were no differences among groups in pH, Hct, fibrinogen, or platelets before or after hemodilution. Hct, fibrinogen, and platelets were reduced by 45% to 60% in all groups. Prothrombin time (PT) and activated partial thromboplastin time were prolonged in all the rabbits with the greatest increase in A group. Thrombelastograph (TEG) analysis showed longer initial reaction (R) and clotting (K) times, slower clotting rate and lower clot strength in H and D than A diluted blood. R time was faster and K time remained unchanged in A group after hemodilution. Thrombin generation potential and peak concentration of thrombin were unchanged in A samples but significantly reduced in H and D diluted samples. Subsequent splenic injury led to almost equal blood losses ( approximately 54 +/- 1 mL/kg) in H and D groups, which were higher (p < 0.01) than in A rabbits (37 +/- 4 mL/kg). This resulted in death of 100% (H), 75% (D), and 50% (A) of the rabbits with significant difference in survival time among the groups. CONCLUSION: : TEG and thrombin generation assays identified more severe coagulopathy development with H and D than A dilution, whereas plasma PT and activated partial thromboplastin time measurements did not differentiate between these colloids. These results suggest that resuscitation with albumin maintained coagulation function, decreased blood loss, and improved survival time compared with the synthetic colloids.

A novel swine model for evaluation of potential intravascular hemostatic agents.

Because uncontrolled hemorrhage is a leading cause of battlefield mortality, finding an intravenous treatment that could assist endogenous clotting mechanisms is a major mission for military researchers. Evaluation of potential intravenous hemostatic agents requires both in vitro and in vivo tests. For in vivo evaluation, we have developed a novel swine model in which 1) bleeding times (BT) and coagulation function could be ascertained after multiple doses of hemostatic drug administration and 2) a subsequent exsanguinating injury could be performed in the same animal, yielding screening information regarding the effects of drug pretreatment on blood loss and survival. Transection of small mesenteric arteries and veins allowed for multiple and reproducible BT measures that correlated with coagulation function. Subsequent excision of defined areas of the liver produced bleeding predominantly from small vessels (diameter, less than 2 mm) and parenchyma while resulting in 62% mortality without the use of either heparinization or aggressive fluid infusion. This swine model allows for multiple, repeatable BT measures in the same animal in experiments already involving laparotomy. Such a model is well suited for terminal studies to test effects of multiple doses of the same drug or multiple drugs on BT and allows for multiple, easily visualized measures that permit enhanced repeatability. The liver injury provides for numerous small vessel lesions that could be amenable to closure by coagulation. Therefore, drugs or mechanisms that enhance coagulation and concomitantly decrease blood loss and increase survival time may be accurately evaluated in this new model.