Endogenous heparin decreases the thrombotic response to hemorrhagic shock in rabbits

Endogenous Heparin Decreases the Thrombotic Response to Hemorrhagic Shock in Rabbits Vance G. Nielsen Purpose: The purpose of this study was to determine if endogenous heparin release would modulate the hemostatic response to hemorrhagic shock in rabbits. Materials and Methods: Anesthetized rabbits (n ⴝ 13) underwent hemorrhagic shock (MAP 30-40 mm Hg) for 60 minutes. Blood samples obtained before and 60 minutes after hemorrhagic shock had thrombelastographic variables (R, reaction time [min]; angle, ␣ [°]; and G [dynes/cm2]) determined. Hemostatic function was assessed by modified thrombelastography under four conditions: (1) unmodified sample; (2) platelet inhibition with cytochalasin D; (3) heparinase I exposure; and (4) platelet inhibition and heparinase I exposure. Results: Thrombelastographic variable values in samples without platelet inhibition or heparinase exposure did not significantly change after hemorrhage

(before hemorrhage: R ⴝ 22.01 ⴞ 0.7 min, ␣ ⴝ 43.6 ⴞ 1.3°, G ⴝ 7,089 ⴞ 379 dyne/cm2; after hemorrhage: R ⴝ 22.1 ⴞ 2.4, ␣ ⴝ 41.6 ⴞ 3.9, G ⴝ 5,662 ⴞ 564; mean ⴞ SEM). However, blood samples exposed to heparinase after hemorrhage demonstrated enhanced hemostatic function with thrombelastographic values (R ⴝ 13.4 ⴞ 1.5, ␣ ⴝ 56.0 ⴞ 3.4, G ⴝ 7012 ⴞ 565) significantly different (P ⬍ .05) from samples not exposed to heparinase. Samples with platelet inhibition demonstrated a similar pattern. Conclusion: Hemorrhagic shock significantly increased circulating endogenous heparin activity, attenuating the thrombotic response to hemorrhage in rabbits. Heparin-mediated regulation of hemostasis may serve as a protective mechanism in shock states. Copyright © 2001 by W.B. Saunders Company

A

Although post-reperfusion thrombotic states may threaten organ function, endogenous antithrombotic mechanisms may modulate this response to ischemia. The release of endogenous heparinoids and consequent significant coagulopathy has been documented in the setting of orthotopic liver transplantation,11-13 with heparinoids presumably released from mast cells in the hepatoenteric vasculature. Similarly, the decrease in the speed of clot formation and clot strength in a rabbit model of hepatoenteric ischemia-reperfusion14 has been determined to be mediated by an increase in circulating endogenous heparin activity.15 Although these marked decreases in hemostatic function occurred following ischemic insults not survivable outside of the operating room, the modulation of circulating of heparinoid activity may be an unrecognized mechanism that prevents microcirculatory thrombi formation. In view of the aforementioned data, it is critically important to determine if endogenous heparin is released during shock states before determining if endogenous heparin modulates reperfusion injury. Thus, I hypothesized that endogenous heparin activity may decrease the hemostatic response to hemorrhagic shock in rabbits. The presence of and hemostatic consequences of heparin activity was determined by thrombelastography.

MONG THE MANY functions of blood, hemostasis is of great clinical interest after major vascular and trauma surgery, as hemorrhage and thrombosis are major sources of morbidity and mortality.1-6 Although multiple clinical factors may impact on hemostasis in the perioperative period (eg, volume of blood loss, hemodilution, temperature), the elaboration of several thrombotic mediators may adversely affect organ function and injury. Cytokines, such as interleukin 6,7 have been identified as positive modulators of hemostasis in the setting of hemorrhagic shock. Also of interest, the inhibition of tissue factor-mediated thrombin generation has been demonstrated to decrease hepatic reperfusion injury.8,9 Additionally, enhancement of the antithrombin III (AT III) system via exogenous heparin or AT III administration significantly decreased hepatic reperfusion injury.10 Taken as a whole, these data suggest that a thrombotic response occurs after an ischemic insult, possibly contributing to reperfusion injury.

From the Department of Anesthesiology, Division of Cardiothoracic Anesthesia, The University of Alabama at Birmingham, Birmingham, AL. This study was supported in part by a grant from BioTime, Inc., Berkeley, CA, and the Department of Anesthesiology. Address reprint requests to Vance G. Nielsen, MD, Department of Anesthesiology, The University of Alabama at Birmingham, 619 South 19th St, Birmingham, AL 35249. Copyright © 2001 by W.B. Saunders Company 0883-9441/01/1602-0004$35.00/0 doi: 10.1053/jcrc.2001.26292 64

MATERIALS AND METHODS The study was approved by our animal review committee. New Zealand White male rabbits (n ⫽ 13, 2 to 3 kg) were anesthetized with 10 mg/kg intravenously (IV) ketamine via a Journal of Critical Care, Vol 16, No 2 (June), 2001: pp 64-68

HEPARIN AND HEMORRHAGIC SHOCK

65

marginal ear vein and subsequently administered inhaled 1% isoflurane carried in 99% oxygen. After tracheotomy and placement of a 3.5-mm OD endotracheal tube, mechanical ventilation was performed with PaCO2 maintained at 32 to 45 mm Hg. Pancuronium was administered 0.3 mg/kg/h IV to facilitate mechanical ventilation. Mean arterial blood pressure (MAP) and heart rate (HR) were monitored by placement of a right femoral arterial catheter and data were recorded with a Grass Model 7D polygraph (Grass Instruments, Quincy, MA). All rabbits received a maintenance infusion of lactated Ringer’s at 4 ml/kg/hr, and esophageal temperatures were maintained at 38°C-39°C with a heating pad. A 15-minute equilibration period followed completion of the surgical preparation. After equilibration, rabbits were bled from the arterial catheter for 2 to 3 minutes until a MAP of 30 to 40 mm Hg was obtained. The hemorrhaged blood was withdrawn into syringes containing citrate-phosphate-dextrose solution with adenine (CPD-A; 9 volumes blood to 1 volume CPD-A; Sigma, St. Louis, MO). Blood was intermittently removed or returned to maintain the 30 to 40 mm Hg MAP for 60 minutes. After 60 minutes, samples subsequently described were removed before euthanasia with an injection of KCl. Arterial blood samples were obtained after 15 minutes of equilibration and after 60 minutes of hemorrhagic shock for blood gas analysis (Model 1640, Instrumentation Laboratory, Lexington, MA) with K⫹, Ca2⫹, and hematocrit determination. Platelet concentrations were concurrently determined with a Sysmex K-800 (TOA Medical Electronics Co., LTD, Japan). Modified thrombelastographic analyses were performed after equilibration and after hemorrhagic shock with two computercontrolled thrombelastographs (Model 5000, Haemoscope Corp., Niles, IL), each with two channels for a total of four thrombelastograms generated per time point. The four thrombelastographic conditions were as follows: (1) 350 ␮L of blood with 10 ␮L of 0.9% NaCl; (2) 350 ␮L of blood with 10 ␮L of cytochalasin D (final concentration 10 ␮mol/L); (3) 350 ␮L of blood with 10 ␮L of 0.9% NaCl in the presence of heparinase I (from Flavobacterium heparinum, 2.0 IU per cup); and (4) 350 ␮L of blood with 10 ␮L of cytochalasin D in the presence of heparinase I. Cytochalasin D inhibits microtubule formation (and glycoprotein IIb/IIIa activation) in platelets, resulting in a thrombelastographic signature due to coagulation proteins in whole blood.16,17 Cups containing 2 IU of heparinase will digest up to 6 IU/mL of heparin activity, which is more than sufficient to neutralize the circulating heparin activity encountered in the rabbit.15 The proper functioning of the thrombelastograph was confirmed daily with quality control standards purchased from Haemoscope (Niles, IL). The following thrombelastographic variables were measured for each sample over a 1-hour

period at 39° C (the normal temperature of the rabbit): reaction time (R, minutes), angle (␣, degrees), maximum amplitude (MA, mm) and shear elastic modulus (G, dyne/cm2). A detailed description of the method of thrombelastography has been presented in great detail elsewhere.17,18 In brief, R is defined as the time from when the blood sample is placed into the thrombelastograph cup until initial fibrin formation occurs as noted by a signal of 2-mm amplitude. ␣ is the angle formed from R to the inflection point of the thrombelastographic signal as clot strength stabilizes; it is a measure of the kinetics of clot formation. MA is the largest amplitude of the thrombelastographic signal and is a measure of clot strength. Finally, G is a measure of clot strength17 calculated from MA as follows: G ⫽ (5,000 ⫻ MA)/(100 ⫺ MA). Although MA was determined, G was reported. The contribution of platelets to G (GP) was defined by the total G of whole blood not exposed to cytochalasin D (GT) minus the G of blood exposed to cytochalasin D, which is attributable to the soluble components of the coagulation pathway (GSC).14-17 All variables are expressed as mean ⫾ SEM. Analyses of the effects of hemorrhagic shock on thrombelastographic variables were conducted with two-way analysis of variance (ANOVA) with repeated measures. Post-hoc analysis was performed with the Student-Newman-Keuls test. The effects of hemorrhagic shock on all other physiologic data were determined with a twotailed paired t test. An alpha error of ⱕ 0.05 was considered significant.

RESULTS

Hemodynamic and arterial blood gas is displayed in Table 1. All rabbits in the hemorrhagic shock group survived the 60 minutes of hemorrhagic shock, with MAP maintained between 30 and 40 mm Hg. Heart rate significantly decreased during the shock period. A total of 35 ⫾ 1 mL/kg of blood was removed by the end of the shock period, without any rabbit requiring any significant reinfusion of hemorrhaged blood (eg, no more than 1 to 3 mL during the shock period). Hemorrhagic shock resulted in a significant decrease in arterial pH, PaCO2, base excess, hematocrit, and platelet concentration. In contrast, K⫹ concentration and PaO2 significantly increased over time. Thrombelastographic data are depicted in Figures 1 and 2. Hemorrhagic shock did not significantly change R, ␣, or G values in blood samples

Table 1. Arterial Blood Gas and Hemodynamic Data 15 minutes of pHa 7.45 ⫾ 0.01 60 minutes of pHa 7.25 ⫾ 0.02*

Equilibration PaCO2 39.9 ⫾ 1.0 Shock PaCO2 30.6 ⫾ 1.4*

PaO2 492 ⫾ 18

BE 4.7 ⫾ 0.4

Kⴙ 3.9 ⫾ 0.1

Ca2ⴙ 1.51 ⫾ 0.03

HCT 34 ⫾ 1

Platelets 381 ⫾ 19

MAP 88 ⫾ 3

HR 321 ⫾ 5

PaO2 548 ⫾ 7*

BE ⫺11.6 ⫾ 1.1

Kⴙ 4.7 ⫾ 0.2*

Ca2ⴙ 1.54 ⫾ 0.06

HCT 18 ⫾ 2*

Platelets 233 ⫾ 15*

MAP 35 ⫾ 1*

HR 293 ⫾ 5*

Note. All values are expressed as the mean ⫾ SEM. PaCO2 and PaO2 expressed as mm Hg, base excess (BE), K⫹ and Ca2⫹ expressed as mmol/L, hematocrit (HCT) as %, platelets as 103/mm3, mean arterial pressure (MAP) as mm Hg, and heart rate (HR) as beats/min. *P ⬍ .05 vs. 15 minutes of equilibration.

66

VANCE G. NIELSEN

nificantly increased the contribution of platelets (GP) and decreased the role of soluble coagulation proteins (GSC) to final clot strength (GT) as depicted in Table 2. DISCUSSION

The hemostatic response to hemorrhagic shock in the rabbit appears to involve the simultaneous release of thrombotic mediators and endogenous heparin, with no significant change in overall clotting function noted. In contrast, after aortic occlusion-reperfusion, endogenous heparin overwhelms the coagulation system, resulting in a sig-

Fig 1. Thrombelastographic variables R, ␣, and G without platelet inhibition. Hemorrhagic shock did not significantly change R, ␣, or G values in blood samples not exposed to heparinase (closed circles). In contrast, samples exposed to heparinase (open circles) demonstrated a significant decrease in R, increase in ␣, and greater G values after hemorrhage compared with samples without heparinase. 15E ⴝ 15-minute equilibration, 60H ⴝ minutes of hemorrhagic shock. *P ⬍ .05 vs. 15E, †P ⬍ .05 vs. samples without heparinase.

not exposed to heparinase (Fig 1). In contrast, samples exposed to heparinase demonstrated a significant decrease in R, increase in ␣ and greater G values after hemorrhage compared with samples without heparinase. Similarly, hemorrhagic shock did not significantly change R or ␣ values in blood samples not exposed to heparinase after platelet inhibition with cytochalasin D (Fig 2). In contrast, samples exposed to heparinase with platelet inhibition demonstrated a significant decrease in R and increase in ␣ values after hemorrhage compared with samples without heparinase. G significantly decreased after hemorrhagic shock without any differences between the groups noted in samples exposed to cytochalasin D. Hemorrhagic shock sig-

Fig 2. Thrombelastographic variables R, ␣, and G with platelet inhibition. Hemorrhagic shock did not significantly change R or ␣ values in blood samples not exposed to heparinase (closed circles). In contrast, samples exposed to heparinase (open circles) demonstrated a significant decrease in R and increase in ␣ values after hemorrhage compared with samples without heparinase. G significantly decreased after hemorrhagic shock without any differences between the groups noted. 15E ⫽ 15-minute equilibration, 60H ⴝ 60 minutes of hemorrhagic shock. *P ⬍ .05 vs. 15E, †P ⬍ .05 vs. samples without heparinase.

HEPARIN AND HEMORRHAGIC SHOCK

67

Table 2. Relative Contributions of Platelets (GP) and Soluble Coagulation Proteins (GSC) to Total Clot Strength (GT) 15 minutes of Equilibration Heparinase GP (%GT) ⫺ 85.9 ⫾ 0.6% ⫹ 86.4 ⫾ 1.2% 60 minutes of Shock Heparinase GP (%GT) ⫺ 90.5 ⫾ 1.2%* ⫹ 89.7 ⫾ 1.1%*

GSC (%GT) 14.1 ⫾ 0.6% 13.6 ⫾ 1.2% GSC (%GT) 9.5 ⫾ 1.2%* 10.3 ⫾ 1.1%*

Note. All values are expressed as the mean ⫾ SEM. *P ⬍ .05 vs. 15 E.

nificant decrease in hemostatic function.15 One could posit that greater quantities of heparin are released after aortic occlusion-reperfusion than during hemorrhagic shock. The quantity of heparin released after aortic occlusion reperfusion in rabbits was determined15 by comparing R time changes (samples without platelet inhibition) after exposure to various concentrations of exogenous heparin. It was determined that R increased 1 minute for every 3.6 mIU/mL increase in heparin activity (R2 ⫽ 0.79, P ⬍ .0001).15 Using this relationship, it was determined that 30 minutes of aortic occlusion increased circulating heparin activity by approximately 32 mIU/mL after 30 minutes of reperfusion.15 Using this same relationship to analyze data from this study, hemorrhagic shock increased circulating heparin activity by 23 mIU/mL. Teleologically, the degree of hepatoenteric ischemia (or low-flow versus no-flow state) may dictate whether the quantity of heparin released favorably modulates hemostasis or pathologically decreases clotting function. Given that reduction of thrombin formation or activity has reduced ischemia-reperfusion injury,8-10 one may speculate that the release of endogenous heparin may serve to protect vasculature particularly vulnerable to reperfusion injury (eg, the intestine). Indeed, mast cells are a putative source of endogenous heparin release, and the intestine is an organ resplendent with mast cells.19,20 Taken as a whole, the regulated release of endogenous heparin during lowflow states to counter the hemostatic effects of the release of thrombotic mediators and microcirculatory thrombi formation may represent a previously unknown defense mechanism against hemorrhagic shock-mediated organ dysfunction. The heparinoid released by man after liver transplantation11-13 and by the rabbit after aortic occlusion-reperfusion15 and hemorrhagic shock is most likely heparin, as heparinase I does not degrade heparan sulfate to any great extent.21,22 Hep-

aran sulfate is degraded by heparinase III.21,22 Unlike heparin, which is contained in primarily pulmonary and hepatoenteric mast cells,23-26 heparan is ubiquitously distributed on cells surfaces and in basement membranes of nearly all tissues.23-26 After aortic occlusion-reperfusion, hemostasis was essentially maintained after ex vivo exposure to heparinase I, the implication being that if significant quantities of heparan sulfate were released, restoration of hemostasis should have been incomplete.15 Consequently, heparin most likely mediates the hemostatic effects observed in this study and other investigations.11-13,15 The hemorrhagic shock-mediated changes in the relative contributions of GP and GSC to final clot strength are also of interest. In absolute terms, final clot strength in samples with intact platelet function was not significantly different after hemorrhagic shock compared with GT values at equilibration (although the heparin-mediated differences between samples with or without heparinase exposure did increase). However, GSC significantly decreased after hemorrhage in both absolute (dyne/cm2) and relative (%GT) terms. These changes in the contribution of GP and GSC to GT were heparin-independent and are reasonably interpreted as a simultaneous increase of platelet function and decrease of coagulation protein function mediated by hemorrhagic shock. Although loss of coagulation protein function may, in part, be explained by hemorrhageinduced loss of protein, the mediators responsible for enhanced platelet function cannot be determined by this investigation. In conclusion, hemorrhagic shock in the rabbit elicits both thrombotic and anticoagulant hemostatic responses. Endogenous heparin release plays a pivotal role in the maintenance of normal hemostasis during hemorrhagic shock in the rabbit. Additionally, the release of heparin may represent a mechanism by which the microcirculation is protected from thrombus formation and ischemia/infarction. Further study is warranted to determine the mechanisms responsible for heparin release in low-flow and no-flow states. The effect of resuscitation (eg, crystalloid vs. colloid vs. whole blood) on hemorrhage-induced hemostatic abnormalities also remains as a future line of investigation. Finally, this study serves as the rational basis for the determination of the role endogenous heparinoids may play in clinical settings involving hemorrhagic shock and other forms of ischemiareperfusion.

68

VANCE G. NIELSEN

REFERENCES 1. Jackson MR, Olson DW, Beckett WC, et al: Abdominal vascular trauma: A review of 106 injuries. Am Surg 58:622-626, 1992 2. Reilly LM, Ramos TK, Murray SP, et al: Optimal exposure of the proximal abdominal aorta: A critical appraisal of transabdominal medial visceral rotation. J Vasc Surg 19:375390, 1994 3. Gibbs NM, Crawford GPM, Michalopoulos N: Thrombelastographic patterns following abdominal aortic surgery. Anaesth Intensive Care 22:534-538, 1994 4. Money SR, Ballinger BA, Crockett DE, et al: The effects of supraceliac aortic clamping on the anticoagulant effects of heparin. Am J Surg 172:155-157, 1996 5. Richardson JD, Bergamini TM, Spain DA, et al: Operative strategies for management of abdominal aortic gunshot wounds. Surgery 120:667-671, 1996 6. Gibbs NM, Crawford GPM, Michalopoulos N: A comparison of postoperative thrombotic potential following abdominal aortic surgery, carotid endarterectomy, and femoro-popliteal bypass. Anaesth Intensive Care 24:11-14, 1996 7. Katsuyama I, Mayumi T, Kohnawa M, et al: Bleeding induced interleukin-6 decreases blood loss via activation of coagulation. Shock 11:87-92, 1999 8. Kobayashi Y, Yoshimura N, Yamagishi H, et al: Role of tissue factor in ischemic reperfusion injury: (I) Tissue factor levels of liver tissue and serum after hepatic injury in rats. Transplant Proc 30:3726-3727, 1998 9. Yoshimura N, Kobayashi Y, Nakamura K, et al: The effect of tissue factor pathway inhibitor on hepatic ischemic reperfusion injury of the rat. Transplantation 67:45-53, 1999 10. Hisama N, Yamaguchi Y, Okajima K, et al: Anticoagulant pretreatment attenuates production of cytokine-induced neutrophil chemoattractant following ischemia-reperfusion of rat liver. Dig Dis Sci 41:1481-1486, 1996 11. Harding SA, Mallet SV, Peachey TD, et al: Use of heparinase modified thrombelastography in liver transplantation. Br J Anaesth 78:175-179, 1997 12. Pivalizza EG, Abramson DC, King FS: Thrombelastography with heparinase in orthotopic liver transplantation. J Cardiothorac Vasc Anesth 12:305-308, 1998 13. Kettner SC, Gonano C, Seebach F, et al: Endogenous

heparin-like substances significantly impair coagulation in patients undergoing orthotopic liver transplantation. Anesth Analg 86:691-695, 1998 14. Nielsen VG, Geary BT: Thoracic aorta occlusionreperfusion decreases hemostasis as assessed by thrombelastography in rabbits. Anesth Analg 91:517-521, 2000 15. Nielsen VG, Geary BT: Hepatoenteric ischemiareperfusion increases circulating heparinoid activity in rabbits. J Crit Care 15:142-146, 2000 16. Nielsen VG, Geary BT, Baird MS: Evaluation of the contribution of platelets to clot strength in rabbits: Role of tissue factor and cytochalasin D. Anesth Analg 91:35-39, 2000 17. Khurana S, Mattson JC, Westley S, et al: Monitoring platelet glycoprotein IIb/IIIa-fibrin interaction with tissue factor-activated thrombelastography. J Lab Clin Med 130:401411, 1997 18. Hartert H, Schaeder JA: The physical and biologic constants of thrombelastography. Biorheology 1:31-39, 1962 19. Kanwar S, Kubes P: Mast cells contribute to ischemiareperfusion-induced granulocyte infiltration and intestinal dysfunction. Am J Physiol (Gastrointest Liver Physiol 30) 267:G316-G321, 1994 20. Szabo A, Boros M, Kaszaki J, et al: The role of mast cells in mucosal permeability changes during ischemia-reperfusion injury of the small intestine. Shock 8:284-291, 1997 21. Linhardt RJ, Turnbull JE, Wang HM, et al: Examination of the substrate specificity of heparin and heparan sulfate lyases. Biochemistry 29:2611-2617, 1990 22. Desai UR, Wang H, Linhardt RJ: Substrate specificity of the heparin lyases from Flavobacterium heparinum. Arch Biochem Biophys 306:461-468, 1993 23. Lindahl U, Kjellen L: Heparin or heparan sulfate—what is the difference? Thromb Haemostasis 66:44-48, 1991 24. Jackson RL, Busch SJ, Cardin AD: Glycosaminoglycans: Molecular properties, protein interactions, and role in physiological processes. Physiol Rev 71:481-522, 1991 25. Gunay NS, Linhardt RJ: Heparinoids: Structure, biological activities and therapeutic applications. Planta Med 65:301306, 1999 26. Kjellen L, Lindahl U: Proteoglycans: Structures and interactions. Ann Rev Biochem 60:443-475, 1991