Coagulation Abnormalities in the Trauma Patient: the Role of Point-Of-Care Thromboelastography

Coagulation Abnormalities in the Trauma Patient: The Role of Point-of-Care Thromboelastography

Eduardo Gonzalez, M.D.,1 Fredric M. Pieracci, M.D., M.P.H.,1 Ernest E. Moore, M.D.,1 and Jeffry L. Kashuk, M.D.2

ABSTRACT

Current recommendations for resuscitation of the critically injured patient are limited by a lack of point-of-care (POC) assessment of coagulation status. Accordingly, the potential exists for indiscriminant blood component administration. Furthermore, although thromboembolic events have been described shortly after injury, the time sequence of post-injury coagulation changes is unknown. Our current understanding of hemostasis has shifted from a classic view, in which coagulation was considered a chain of catalytic enzyme reactions, to the cell-based model (CBM), representing the interplay between the cellular and plasma components of clot formation. Thromboe- lastography (TEG), a time-sensitive dynamic assay of the viscoelastic properties of blood, closely parallels the CBM, permitting timely, goal-directed restoration of hemostasis via POC monitoring of coagulation status. TEG-based therapy allows for goal-directed blood product administration in trauma, with potential avoidance of the complications resulting from overzealous component administration, as well as the ability to monitor post-injury coagulation status and thromboprophylaxis. This overview addresses coagulation status and thromboprophylaxis management in the trauma patient and the emerging role of POC TEG.

KEYWORDS: Coagulopathy, hypercoagulability, trauma, cell-based model, thromboelastography, massive transfusion, transfusion ratio Downloaded by: University of Pennsylvania Library. Copyrighted material.

Restoration and subsequent maintenance of nor- a hypercoagulable state, demanding accurate risk strat- mal coagulation function is essential for survival of the ification and chemoprophylaxis for prevention of highly severely injured bleeding patient. During the acute phase morbid thromboembolic events (TEs). of trauma, efforts are focused on resuscitation and The coagulation system involves a fine balance hemorrhage control, with attempts at restoration of between the procoagulant (potentially prothrombotic) normal blood clotting. Furthermore, after initial stabili- and fibrinolytic systems. Such a homeostatic state pre- zation, trauma patients paradoxically face the dangers of vents catastrophic hemorrhage at the site of injury, with

1Department of Surgery, Denver Health Medical Center and Univer- (e-mail: [email protected]). sity of Colorado School of Medicine, Denver, Colorado; 2Department Global Hemostasis: New Approaches to Patient Diagnosis and of Surgery, Division of Acute Care Surgery, Penn State-Hershey Treatment Monitoring; Guest Editor, Maha Othman, M.D., Ph.D. College of Medicine, Hershey, Pennsylvania. Semin Thromb Hemost 2010;36:723–737. Copyright # 2010 by Address for correspondence and reprint requests: Jeffry L. Kashuk, Thieme Medical Publishers, Inc., 333 Seventh Avenue, New York, NY M.D., Department of Surgery, Division of Trauma, Acute Care, 10001, USA. Tel: +1(212) 584-4662. Critical Care Surgery, Penn State – Hershey Medical Center, 500 DOI: http://dx.doi.org/10.1055/s-0030-1265289. University Drive, MC H075, Room C5520, Hershey, PA 17033-0850 ISSN 0094-6176. 723 724 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 36, NUMBER 7 2010

simultaneous control of indiscriminant thrombosis else- activated protein C, leading to impaired thrombin gen- where. Massive tissue injury and shock overwhelm the eration and increased fibrinolysis. Such a process may be body’s endogenous regulatory mechanisms, placing the protective teleologically by inducing an ‘‘auto-anticoagu- critically injured patient at risk for both excessive bleed- lation’’ state that protects the perfusion of tissue beds ing and unwanted thrombosis. Early recognition of such from thrombosis in the milieu of an activated coagulation dysregulation is a prerequisite toward the evolution of system. These authors suggest that fibrinolysis results effective management strategies. from activated protein C’s inactivation of plasminogen Thrombelastography (TEG), a point-of-care activator inhibitor (PAI)-1. However, an established (POC) assay of the viscoelastic properties of blood, mechanistic link of fibrinolysis to the acute coagulopathy provides a comprehensive real-time analysis of hemo- of trauma has not been shown, and its occurrence in stasis, from initial thrombin burst to fibrinolysis, permit- trauma patients has been variably reported, from 15% to ting improved transfusion strategies resulting in the 20%,14 but is likely underestimated due to a lack of potential for goal-directed therapy of coagulation ab- accurate and timely laboratory tests for this entity. Others normalities following injury. This review describes the believe that acute coagulopathy after injury is a variant of coagulation abnormalities associated with trauma, disseminated intravascular coagulation, with accelerated underscores the limitations of the traditional approach fibrinolysis mediated by PAI-1 consumption.15 to post-injury coagulopathy, and highlights the emerg- Fig. 1 depicts our current view of the ‘‘bloody ing role of POC TEG in addressing these limitations. vicious cycle,’’ which emphasizes that acute endogenous coagulopathy exists immediately after injury, is unrelated to clotting factor deficiency, and is thus resistant to factor COAGULATION DISTURBANCES IN THE replacement. Recent evidence from our group12,16 and TRAUMA PATIENT others8,17 supports this theory, suggesting that more than a third of multiply injured patients are coagulo- Hypocoagulability pathic within 30 minutes from injury. This subset of Hemorrhage accounts for up to 40% of all trauma-related patients has an increased incidence of multiple organ deaths1,2 and is compounded by the presence of coagul- failure (MOF) and death. Of note, although the acute opathy.3,4 Furthermore, many trauma patients presenting coagulopathy of trauma predominates in the immediate with massive hemorrhage succumb to refractory coagul- post-injury phase, subsequent uncontrolled hemorrhage opathy despite surgical control of bleeding.3,5,6 In fact, and disrupted tissue consumption of clotting factors continuing hemorrhage due to coagulopathy remains the eventuates in factor depletion, leading to a state of most common indication for damage control procedures,2 systemic coagulopathy. In contrast to the acute endog- but the acute coagulopathy of trauma remains poorly enous coagulopathy of trauma, the resulting systemic characterized.7,8 Traditionally, post-injury coagulopathy coagulopathy requires prompt repletion of factors to was thought to originate as a result of the depletion and restore hemostatic integrity. dilution of coagulation factors resulting from hemorrhage Understanding how these factors contribute to and subsequent fluid resuscitation, termed the ‘‘bloody the development and perpetuation of coagulopathy is vicious cycle’’ by our institution in 1982.5 In the face of essential to designing optimal treatment strategies. The ongoing hemorrhage, hypothermia and acidosis com- complex interactions of thrombin, fibrinogen, platelets, pound this coagulopathy, leading to a worsening coagul- protein clotting factors, calcium ions, inflammatory Downloaded by: University of Pennsylvania Library. Copyrighted material. opathy culminating in death.9,10 mediators, and the endothelium18 create a challenge The coagulopathy of trauma is a complex process, for designing a comprehensive approach to the manage- with tissue injury and hypoperfusion initiating a cascade ment of acute coagulopathy of trauma. of events resulting in a hypocoagulable state.8,11,12 Although clotting factor depletion, dilutional effects from crystalloid resuscitation, hypothermia, and acidosis Hypercoagulability have significant roles, current evidence suggests that The oscillation in coagulation status after injury has tissue injury, hypoperfusion, accelerated fibrinolysis, been recognized since Walter Cannon’s studies in and inflammatory mechanisms may be more pivotal 1917.19 During World War I, Cannon et al studied inciting events. the physiology of traumatic shock, reporting brief Immunoactivation is an important component of hypercoagulability immediately after injury, followed this coagulopathic process, particularly as it relates to by hypocoagulability, ultimately returning to a hyper- endothelial cell activation, neutrophil priming, and sub- coagulable state. Current clinical evidence confirms sequent release of inflammatory mediators from the that a transition from a hypocoagulable to a hyper- innate immune system via activation of the complement coagulable state occurs in patients who survive the acute cascade.13 Brohi et al8 suggest that the acute coagulop- phase of trauma.20 Our work21 suggests that the de- athy of trauma is mediated by thrombomodulin via layed hypercoagulable state predisposes trauma patients POINT-OF-CARE THROMBOELASTOGRAPHY/GONZALEZ ET AL 725

Figure 1 The updated bloody vicious cycle. This schematic depicts an updated view of the multiple factors contributing to ongoing post-injury coagulopathy. Recent observations suggest that acute endogenous coagulopathy occurs early after traumatic injury. The subsequent development of progressive systemic coagulopathy is time dependent, resulting from persistent hypotension, acidosis, and cellular shock. FFP, fresh-frozen plasma; RBC, red blood cells. Reproduced with permission from Kashuk et al.12 to subsequent TE events. Currently, little is known erythrocytes, mostly located in proximity of the endo- about the timing of this transition. The etiology of the thelium, as well as areas dominated by a platelet-fibrin hypercoagulable state is likely multifactorial, involving network, located more distally in the growing throm- endothelial injury, circulatory stasis, platelet activation, bus.26 Of note, scanning electron microscopy studies decreased levels of endogenous anticoagulants, and show aggregated platelets attached to venous endothe- impaired fibrinolysis. Deep vein thrombosis has been lium.27 Moreover, inhibition of P-selectin on the surface Downloaded by: University of Pennsylvania Library. Copyrighted material. reported to occur in 10 to 80% of patients following of an activated platelet results in impaired thrombus trauma.22 Pulmonary embolism has been observed in 2 formation both in an experimental model of venous to 22% of injured patients, and fatal pulmonary emb- thrombosis and in vivo.28 Accordingly, platelets likely olisms are the third most common cause of death in contribute significantly to the formation of venous those who survive the first 24 hours.23 thrombi. Finally, proinflammatory activation of endo- Although endothelial injury and platelet activa- thelium has been implicated in the genesis of venous tion are known contributors to arterial thromboem- thrombi.13 Risk stratification models for post-injury boli,18 the sequence of events leading to venous hypercoagulability should therefore incorporate both thrombosis is not as clear. According to the long-stand- enzymatic and platelet components of hemostasis. ing triad of Virchow, three major factors contribute to the development of venous TE: stasis of blood flow, hypercoagulability, and damaged endothelium. In addi- LIMITATIONS OF CURRENT CONCEPTS tion, several lines of evidence suggest that activation of AND TESTING platelets contributes to the development and propagation of venous thrombi.24,25 In an analysis of 50 venous Classic versus Cell-Based Model of Coagulation thrombi obtained from femoral valve pockets, Sevitt et al Our current understanding of the scientific basis of the identified areas that were characterized by fibrin and acute coagulopathy of trauma is limited by the fact that 726 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 36, NUMBER 7 2010

the classic laboratory tests of coagulation function, such procedure that can take >180 minutes; hence it is not a as prothrombin time/international normalized ratio (PT/ practical test in the acutely bleeding patient. Other INR) and activated partial thromboplastin time (aPTT), techniques used to diagnose hyperfibrinolysis, such as are largely designed for assessment of anticoagulation plasmin-antiplasmin complex, PAI-1, thrombin-acti- therapy, and their applicability to the acutely injured vatable fibrinolytic inhibitor, and D-dimers, suffer patient has never been proven. Consequently, no con- from similar limitations. Gando et al failed to validate sensus exists regarding the definition of either coagulop- D-dimer, plasmin-antiplasmin complexes, fibrinopep- athy or subsequent hypercoagulable states, and currently tide-B b 15–42, or PAI-1 as indicators for increased there is a regrettable lack of scientific evidence to support fibrinolysis after trauma.15 preventive diagnostic and therapeutic strategies. The limitations of classic measurements of coag- The classic model of hemostasis suggests a central ulation among trauma patients have been documented in role of protein coagulation factors directing the hemo- several clinical studies. Over 30 years ago, Counts et al,35 static process, with the cells surface serving primarily to from the Puget Sound Blood Center, performed exten- provide an anionic phospholipid region for procoagulant sive prospective analyses of coagulation function in complex assembly.29 Accordingly, the ‘‘cascade’’ model patients receiving massive transfusions (MTs), where of coagulation (extrinsic and intrinsic pathways) is struc- 69% of packed red blood cells (PRBCs) were adminis- tured as a series of proteolytic reactions. Although this tered within 24 hours of hospitalization. In this cohort, model supports traditional laboratory tests of coagula- the predictive value of the PT, aPTT, and bleeding time tion by plasma-based algorithms, it does not correlate was poor for assessment of generalized bleeding. Fur- with current concepts of hemostasis occurring in vivo.18 ther, based on animal studies of hypovolemic shock and The two laboratory tests performed traditionally resuscitation, Lucas and Ledgerwood36 observed that to assess coagulation are PT/INR and aPTT. These PT and aPTT changes were infrequent unless trans- plasma-based assays are performed by recalcifying a fusions in excess of an equivalent of 15 units of whole sample of citrated plasma in the presence of animal- blood occurred. derived or recombinant thromboplastin (PT) or a neg- The classic view of the intrinsic and extrinsic atively charged substrate (aPTT), which results in factor coagulation cascades has been replaced recently by a cell activation, thereby initiating coagulation via the extrinsic based model (CBM) of coagulation.29 This model recog- (PT/INR) or intrinsic (aPTT) clotting pathway.30 The nizes the important interactions of the cellular and plasma end point for these tests is the time (in seconds) until the components to clot formation.9 The CBM suggests that earliest formation of a fibrin clot is detected by visual, procoagulant properties result from expression of a variety optical, or electromechanical means. of cell-based features, including protein receptors, which The applicability of these laboratory tests in the activate components of the coagulation system at specific trauma setting has not been validated. Furthermore, the cell surfaces. This model also facilitates our understanding time required to conduct these assays in most trauma of potential mechanistic links of cross-talk among inflam- centers is 45 minutes, limiting usefulness in the setting mation and coagulation components because interaction of hemorrhagic shock.31 Indeed, the entire blood volume between platelet receptors, endothelial cells, and cytokines of the bleeding trauma patient may have been exchanged have important roles in coagulation.37 This model also several times during this time interval, rendering the considers the red blood cells (RBC) and their interactions results of these laboratory tests obsolete by the time with the hemostatic process.38 For example, the provision Downloaded by: University of Pennsylvania Library. Copyrighted material. they are available. Furthermore, both the PT/INR and of adenosine diphosphate from RBCs is known to con- aPTT are performed on platelet-poor plasma, and these tribute to platelet activation.39 Furthermore, the rheologic tests are sensitive only to the earliest initiation of clot margination of platelets toward the vessel’s perimeter formation. However, >95% of thrombin generation (areas with greatest shear stress) and subsequent adhesion occurs after the initial polymerization of fibrinogen.32 after endothelial injury appears dependent on the quantity Finally, these tests are performed in an artificial environ- and size of RBCs.40 ment (pH ¼ 7.40, temperature ¼ 388C), irrespective of In sum, the CBM represents a major paradigm the patient’s core body temperature and pH. Impairment shift from a theory that views coagulation as being of coagulation by acidosis and hypothermia has been controlled by concentrations and kinetics of coagulation thoroughly documented both in vitro and in vivo.33,34 proteins to one that considers the process to be con- Thus coagulation testing performed at a normal pH and trolled by diverse cellular interactions. temperature will not reflect in vivo derangements responsible for coagulopathy.9 Diagnosis of fibrinolysis is also problematic. The Presumptive Blood Component Therapy euglobulin lysis time (ELT) has been used for estimating for Post-Injury Coagulopathy the functional fibrinolytic capacity of plasma.14 Unfortu- Consumption and dilution of clotting factors via crystal- nately, ELT remains a complex and time-consuming loid resuscitation and other factor-poor blood products POINT-OF-CARE THROMBOELASTOGRAPHY/GONZALEZ ET AL 727 perpetuates the acute coagulopathy of trauma. Coagula- opathy. Furthermore, they noted that the absence of time- tion factors present in plasma contained in PRBCs have varying assessment of blood component administration in minimal activity due to prolonged storage and associated trauma likely results in a survivor bias in studies assessing short coagulation factor half-lives.41 Isolated administra- cumulative transfusion ratios. These reports further ques- tion of PRBCs in the absence of plasma will therefore tion the accuracy of prior studies using a cumulative ratio potentiate the acute coagulopathy of trauma. Accord- calculated at 24 hours. Pending further data, most trauma ingly, most MT protocols advocate early replacement of surgeons recommend a preemptive product transfusion of factors and platelets. However, the definition of MT, and the critically injured patient in hemorrhagic shock using a the timing and ratio of specific factor replacement, RBC-to-FFP ratio of 1:2. remains widely debated, due at least in part to differences The role of early platelet transfusion in the setting in protocols as well as inherent flaws in retrospective data of hemorrhagic shock also remains debated. As with analysis. FFP, recent military reports51 have suggested routine A valid definition of MT is lacking. MT has been administration of apheresis platelets to the injured pa- defined as administration of 10 units of PRBCs to an tient. However, a similar survival bias has been suggested individual or the transfusion of more than one blood to explain the apparent benefit of early platelet admin- volume (70 mL/kg) in 24 hours. Other dynamic defi- istration. Furthermore, studies from more than 2 deca- nitions of MT are currently used, particularly to design des ago31,52 evaluating clotting factor and platelet counts appropriate MT protocols. Such definitions include in massively transfused patients concluded that a platelet transfusion >10 units of PRBCs in a 24-, 12- or 6- count of 100,000/mm3 is the threshold for diffuse hour period. Our group recently reviewed transfusion bleeding, and that thrombocytopenia was not a clinically practices in severely injured patients at risk for post-injury significant problem until transfusions exceeded 15 to 20 coagulopathy.42 We noted that >85% of transfusions units of blood. Specifically, patients with a platelet count were accomplished within 6 hours post-injury, suggesting >50,000/mm3 had only a 4% chance of developing this is the critical period to assess the impact of pre- diffuse bleeding. The authors of these reports concluded emptive factor replacement, rather than the 24-hour time that routine platelet administration was not warranted in period others have emphasized.43 this group of patients.31,52 Hiippala and colleagues53 Current clinical MT protocols promoting ‘‘dam- assessed the changes in platelet and fibrinogen concen- age control resuscitation’’ (i.e., preemptive transfusion of trations after major surgical bleeding and noted that a plasma, platelets, and fibrinogen) assume that patients platelet count of 50,000/mm3 was only reached late in presenting with life-threatening hemorrhage at risk for the course of blood loss, and the changes in individual post-injury traumatic coagulopathy should receive com- patients were unpredictable. ponent therapy in amounts approximating those found Although the classic threshold for platelet trans- in whole blood during the first 24 hours.43–46 The U.S. fusion has been 50,000/mm3, a higher target level of military experience in Iraq suggesting improved survival 100,000/mm3 has been suggested for multiply injured based on a 1:1:1 fresh-frozen plasma (FFP)-to-RBC-to patients and patients with massive hemorrhage.53 platelet ratio44 has led to recommendations of fixed However, the relationship of platelet count to hemo- ratios of these blood products during the first 24 hours stasis and the contribution of platelets to formation of a post-injury in civilian trauma centers.43,45,46 stable clot in the injured patient remain largely un- In contrast, our results12 as well as those of known. Furthermore, platelet function, irrespective of Downloaded by: University of Pennsylvania Library. Copyrighted material. others47 suggest that the optimal survival ratio appears number, is also of crucial importance. The complex to be in the range of a 1:2 to 1:3 FFP-to-RBC ratio. Of relationship of thrombin generation to platelet activa- note, a prospective study by Scalea et al48 showed no tion requires dynamic evaluation of clot function,54,55 benefit to a standardized 1:1 RBC-to-FFP transfusion as opposed to isolated measurements of platelet count, protocol for critically injured trauma patients, and their or older methods of clot assessment, such as the bleed- study did not specifically address the early (<6 hour) ing time, which is of no use in the trauma setting. Thus time frame. We believe the reported benefits from a 1:1 the significance of a decreased platelet count should be strategy likely represent a surrogate marker of survival. interpreted in the patient’s specific clinical context and Specifically, those patients who survive injury are simply with knowledge of platelet function. At this time, no able to receive more plasma transfusions, as opposed to evidence supports an absolute trigger for platelet trans- those who die from acute hemorrhagic shock early after fusions in trauma. injury. Two recent reports provide further evidence of Concerns over high ratios of blood component this survival bias. Snyder et al49 as well as Magnotti therapy stem in large part from a growing body of et al50 analyzed sequential timing of transfusions ad- evidence documenting the adverse effects of transfu- ministered for post-injury traumatic resuscitation and sion. The association of massive transfusion of PRBCs found evidence that early (<6 hour) assessment of bleed- with nosocomial pneumonia, acute lung injury, and ing is critical for analysis of post-injury traumatic coagul- acute respiratory distress syndrome (ARDS) has been 728 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 36, NUMBER 7 2010

documented.56,57 Regarding transfusion of plasma-rich Table 1 Independent Risk Factors for Venous components after injury, Watson et al58 evaluated Thromboembolism after Trauma (Multivariate Logistic independent risks associated with administration of Regression) FFP, platelets, and cryoprecipitate in a multicenter Risk Factor OR (95% CI) p Value cohort study. Although early transfusion of these prod- Age 40 years OR ¼ 2.01 (1.74–2.32) <0.0001 ucts was associated with decreased mortality within the Lower extremity OR ¼ 1.92 (1.64–2.26) <0.0001 first 48 hours, it was noted that every unit of FFP given fracture (AIS 3) was independently associated with a 2.1% and 2.5% Head injury (AIS 3) OR ¼ 1.24 (1.05–1.46) 0.0125 increased risk of MOF and ARDS, respectively. Our Ventilator days >3OR¼ 8.08 (6.86–9.52) <0.0001 group and others have found a similar association Vascular injury OR ¼ 3.56 (2.25–5.72) <0.0001 between transfusion of FFP and organ failure.59,60 Major operative OR ¼ 1.53 (1.30–1.80) <0.0001 Similar to FFP and PRBC administration, platelet procedure transfusion is also associated with immunological com- OR, odds ratio; CI, confidence interval; AIS, abbreviated injury scale. plications, with a reported incidence of >200 per Reproduced with permission from Knudson et al.101 100,000 transfused patients.61 Additionally, blood products are an expensive resource and of limited tinely given recent controversy regarding the lack of supply. correlation with TE events. A major limiting factor of current MT protocols is Another current limitation of chemoprophylaxis the lack of a real-time assessment of coagulation function. is the failure to assess platelet hypercoagulability. Cur- Temporal and logistical limitations of conventional rent thromboprophylactic regimens using UH or measurements of coagulation have precluded true goal- LMWH decrease thrombin levels through antithrombin directed therapy of hemorrhagic shock, with current potentiation and by direct factor Xa inhibition.65 treatment algorithms advocating preemptive replacement Although thrombin is the main platelet activator, given of products. Accordingly, the role of goal-directed ther- the multiplicity of factors known to activate platelets and apy specifically addressing coagulation abnormalities re- the direct interactions between the platelet and the mains to be established. Furthermore, current treatment endothelium, solely targeting thrombin production via algorithms fail to consider the complex temporal and current regimens may not adequately prevent platelet pathophysiological dynamics of coagulation. Preemptive activation.18 Accordingly, a hypercoagulable state may coagulation factor replacement is essential for the initial persist despite ‘‘therapeutic’’ anticoagulation. The addi- resuscitation of the acutely bleeding patient. We remain tion of pharmacological inhibition of platelet activation concerned, however, regarding the continued adminis- and aggregation could further reduce the incidence and tration of FFP and platelets via protocols that lack POC propagation of thrombus growth and lead to reduction assessment of ongoing blood component requirements. of subsequent TEs, although the use of platelet antag- onists in the acutely injured patient requires precise monitoring of hemostasis due to the serious and poten- Diagnosis and Treatment of Hypercoagulability tially devastating consequences of recurrent hemorrhage. of Traumatic Critical Illness Given the limitations just described, it is not Thromboprophylaxis remains a challenge because objec- surprising that several studies have reported that the tive measurements of hypercoagulability are lacking. frequency of TEs has not changed despite adoption of Downloaded by: University of Pennsylvania Library. Copyrighted material. Designed originally to assess hypocoagulable states, current thromboprophylactic guidelines.66–68 In a multi- classic coagulation tests can neither identify hypercoa- center study involving 12 medical and surgical intensive gulability nor predict TEs.62,63 Current treatment care units (ICUs), Patel et al concluded that most venous guidelines are thus generally based on clinical risk thromboembolisms (VTEs) occurred due to prophylaxis stratification (Table 1). Most protocols advocate a com- failure rather than failure to provide prophylaxis.66 In bination of subcutaneous heparinoids and sequential their study, 65% of VTEs occurred in patients receiving compression devices.64 Unfortunately, generic dosing standard pharmacological thromboprophylaxis. of both unfractionated heparin (UH) and low molecular weight heparin (LMWH) does not account for variable pharmacokinetics between patients, which in turn de- POINT-OF-CARE pend on several clinical parameters, including weight, THROMBOELASTOGRAPHY mobility, and degree of inflammation. Furthermore, response to thromboprophylaxis is not routinely moni- Historical Perspective tored. Currently, the test used for monitoring LMWH Hellmut Hartert conceived TEG in Germany, at the effect is the anti-factor Xa activity level (anti-Xa) in Heidelberg University School of Medicine in 1948.69 plasma. However, anti-Xa measurement is expensive and Introduced initially for the evaluation of clotting factor labor intensive, and it has not been implemented rou- deficiencies, TEG was applied increasingly throughout POINT-OF-CARE THROMBOELASTOGRAPHY/GONZALEZ ET AL 729

Europe during the 1950s, validating its use as a reliable Fig. 2. In earlier iterations, this tracing was recorded on assessment of anticoagulant effect, the presence of heat-sensitive paper moving at a rate of 2 mm/minute. thrombocytopenia, and fibrinolysis.70 The availability More recent computer technology has allowed for auto- of TEG in the United States was limited until the matic calculation of TEG variables. At our institution, a 1980s, at which time it was found to be useful during dynamic TEG tracing is transmitted directly to the liver transplantation due to the frequency of fibrinolysis operating room or ICU via computer within minutes, during the anhepatic phase.71 From this initial experi- enabling prompt interpretation. ence, TEG evolved for use in monitoring patients Based on time-resistance correlations, the follow- undergoing cardiopulmonary bypass. In this group, it ing seven TEG parameters have been defined: (1) split was primarily useful for monitoring intraoperative hep- point (SP, minutes), (2) reaction time (R, minutes), (3) arinization and was also found beneficial for predicting coagulation time (K, minutes), (4) a angle (a, degrees), postoperative bleeding.72 TEG gained additional ac- (5) maximum amplitude (MA, mm), (6) clot strength ceptance as a tool for assessment of coagulation in the (G, dynes/cm2), and (7) estimated percentage lysis general surgical patient, permitting differentiation of (EPL, %). The various components of the TEG tracing coagulopathy from surgical bleeding.73,74 are depicted in Fig. 2. The SP time is a measure of the time to initial clot formation, interpreted from the earliest resistance detected by the TEG analyzer causing the Technique and Interpretation tracing to split; this is the point at which all other The TEG analyzer is composed of two mechanical parts platelet-poor plasma clotting assays (e.g., PT and separated by a blood specimen: a plastic cup or cuvette, aPTT) fail to progress. The R value is defined as the into which a 0.36-mL blood specimen is pipetted, and a time elapsed from the initiation of the test until the point plastic pin attached to a torsion wire and suspended where the onset of clotting provides enough resistance to within the specimen (Fig. 2). Once the sample within produce a 2-mm amplitude reading on the TEG tracing. the cuvette is placed on the TEG analyzer, the temper- Of note, in the rapid-TEG (R-TEG) assay (discussed ature is adjusted to that of the patient. The cup then later, due to the acceleration of clotting initiation, the R oscillates slowly through an angle of 48450. Initially, time is represented by a TEG-derived activated clotting movement of the cuvette does not affect the pin, but as time (TEG-ACT). The R time and TEG-ACT are clot develops, resistance from the developing fibrin most representative of the initiation phase of enzymatic strands couples the pin to the motion of the cuvette. clotting factors. Whereas a prolonged R time or TEG- In turn, the torsion wire generates a signal that is ACT is diagnostic of hypocoagulability, decreased values amplified and records the characteristic tracing seen in suggest hypercoagulability. The K time is a measurement Downloaded by: University of Pennsylvania Library. Copyrighted material.

Figure 2 Thrombelastography instrument and tracing. The instrument diagram depicts the cuvette where a whole blood sample is placed and the pin attached to a torsion wire. Once the assay is initiated, a tracing is produced and an initial linear segment (zone of precoagulation) extends from the beginning of the test to the formation of the first fibrin strand, causing the tracing to split (split point). The progressive divergence of the tracing reflects the formation of the clot. The R time is reached when the onset of clotting provides enough resistance to produce a 2-mm amplitude reading. The K time is a measurement of the time interval from the R time to the point where fibrin cross-linking provides enough clot resistance to produce a 20-mm amplitude reading. The a angle is the angle formed by the slope of a tangent line traced from the R to the K time measured in degrees. The maximum amplitude, MA, indicates the point at which clot strength reaches its maximum measure in millimeters. Estimated percentage lysis, EPL, is a measure of fibrinolysis and reflects the slow yet progressive decrease in clot strength once MA is reached. Reproduced with permission from Haemonetics Corporation. 730 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 36, NUMBER 7 2010

Table 2 Thromboelastrography Parameters Parameter Significance SP time Earliest activity of enzymatic factors, causing tracing to split R time Initiation phase of enzymatic factor activity K time Potentiation phase of enzymatic factors yielding clot strengthening Alpha-angle Rate of clot strengthening through polymerization of available fibrinogen MA Platelet functional contribution to clot strength through GP IIb-IIIa receptor interaction with fibrin G Overall total clot strength resulting from all coagulation interactions, calculated from amplitude (A), G ¼ (5000 A)/(100 A). EPL Fibrinolytic activity, derived from percentage decrease in clot strength after MA is reached. SP, split point, minutes; R, reaction time, minutes; K, coagulation time, minutes; a angle, degrees; MA, maximum amplitude, millimeters; G, clot strength, dynes/cm2; EPL, estimated percent agelysis, %.

of the time interval from the R time to the point where (normal K time and a angle) and platelet function fibrin cross-linking provides enough clot resistance to (normal MA) (Fig. 3B). Platelet dysfunction is noted produce a 20-mm amplitude reading. The a angle is the by a tracing with a normal R time and a primarily angle formed by the slope of a tangent line traced from decreased MA, as seen in Fig. 3C. We have defined the R to the K time measured in degrees. Both the K clinically significant fibrinolysis when EPL values exceed time and the a-angle denote the rate at which the clot 15%7 resulting in a characteristic tapering of the TEG strengthens and are most representative of thrombin’s tracing immediately after the MA is reached (Fig. 3D). cleaving of available fibrinogen into fibrin. The MA Depicted in Fig. 3E is a characteristic tracing of hyper- indicates the point at which clot strength reaches its coagulability, identifiable by decreased R and K times maximum measure in millimeters on the TEG tracing, along with an elevated a angle and MA. and it reflects the end result of maximal platelet-fibrin Coagulopathic tracings seen in the severely in- interaction via glycoprotein (GP) IIb-IIIa receptors.75 jured patient involve a combination of the examples just Finally, G is a calculated measure of total clot strength described. A prolonged TEG-ACT and R time derived from amplitude (A, mm); G ¼ (5000 A)/ early after injury typically characterize an evolving (100 A). Due to its exponential relationship with A, G is a more realistic representation of overall clot strength.76 Table 2 summarizes the various TEG components and their significance. Blood coagulation in TEG is initiated by addition of an activating solution consisting of kaolin, phospho- lipids, and buffered stabilizers, which requires an activation phase of several minutes before coagulation starts. To expedite time to generate results (e.g., in the setting of hemorrhagic shock), clotting initiation can be further prompted by the addition of tissue factor. This permits the earliest tracings via R-TEG to be viewed within 10 Downloaded by: University of Pennsylvania Library. Copyrighted material. minutes. Furthermore, when heparinized patients are evaluated, the effects of heparin on the TEG tracing may be delineated via a simultaneous analysis of the heparinized sample on a conventional cuvette and another containing heparinase. This strategy is particularly useful for assessment of the efficacy of chemoprophylaxis. Similarly, TEG platelet mapping assesses the patient’s percentage platelet inhibition against their own maximum platelet function, as measured by changes in the MA secondary to anti-platelet therapy (e.g., acetylsali- Figure 3 Normal and abnormal thrombelastographic patterns. (A) Normal tracing. (B) Prolonged clotting time seen cylic acid, dipyridamole, clopidogrel).77 with coagulation factor deficiency or inhibition by anticoagu- Specific coagulation derangements produce char- lation. (C) Decreased maximum amplitude (MA), seen during acteristic TEG patterns. Fig. 3 depicts these deranged platelet dysfunction or pharmacological inhibition. (D) In- profiles along with a normal TEG tracing. Anticoagu- creased percentage lysis in a fibrinolytic tracing. (E) De- lation causes enzymatic inhibition of coagulation factors creased clotting and clotting formation time, elevated a and produces a prolonged R time due to delayed ini- angle, and increased MA represent a hypercoagulable state. tiation of clot formation, along with normal fibrinogen Reproduced with permission from Haemonetics Corporation. POINT-OF-CARE THROMBOELASTOGRAPHY/GONZALEZ ET AL 731 coagulopathy. Further resuscitation with crystalloid maximum rate of thrombus generation’’ (TMRTG), and PRBCs may lead to decreased platelet function calculated from the time of clot initiation to the (reduced MA), as seen in Fig. 3C. point where maximum velocity of thrombus growth The basic TEG tracing has been further studied to is reached, correlated with thrombin-antithrombin generate novel TEG parameters that may be useful levels,80 validating TMRTG as a potential marker of markers of coagulation abnormalities.78–81 Using the thrombin generation. CBM,29 Nielsen described a clot lifespan model We recently studied V-curve parameters in a (CLSM) of coagulation.78 Employing a TEG-based cohort of surgical intensive care unit patients at risk for approach with a standardized clotting and fibrinolytic TE complications and demonstrated a linear correlation stimulus, he assessed parametric, elastic modulus-based (r ¼ 0.94) between a novel parameter delta and parameters of clot formation. The model provides mech- TMRTG.81 Delta (depicted in Fig. 4) is calculated anistic insight into hemostasis through clotting and from the same time interval as TMRTG by subtracting fibrinolytic stimuli that would not otherwise be explained SP from the R time. A decreased delta indicates increased by coagulation factor activity alone. The CLSM was able thrombin generation, as seen in a hypercoagulable state, to differentiate enzymatic inhibition from cellular-con- whereas an increased delta indicates decreased thrombin trolled noncatalytic inhibition of protein–protein inter- generation, as seen in hypocoagulability or enzymatic actions responsible for changes in both clot growth and inhibition of coagulation factors. In our pilot study, disintegration during hemodilution. Observing this 30% of patients receiving prophylactic LMWH had an CLSM of coagulation, Sorensen et al noted that throm- appropriately increased delta but remained hypercoagu- belastographic coagulation kinetics correlated with lable according to increased total clot strength (G) and thrombin generation in vivo.79 This correlation resulted MA. Delta may thus aid in differentiation of enzymatic from calculating the first derivative of changes in clot from platelet hypercoagulability. resistance expressed as a change in amplitude per unit of The various TEG values, derived from a single time (mm 100/second), representing the maximum measurement of whole blood coagulation, are not in- velocity of clot formation. Based on these derivatives, dependent parameters but rather a continuum of blood clot-velocity curves (V curve), depicted in Fig. 4, may coagulation with interactions between all components. be plotted using standard TEG software. Rivard et al For instance, thrombin liberates fibrinopeptides from demonstrated that the V-curve parameter ‘‘time to fibrinogen, allowing association with other fibrinogen Downloaded by: University of Pennsylvania Library. Copyrighted material.

Figure 4 Thrombus velocity curve. Sample of a thrombus velocity curve (V curve, in gray), calculated from the ﬁrst derivative of changes in clot resistance, depicted here over a standard thromboelastographic tracing, showing relationship of delta (R to SP) with thrombus generation parameters. MRTG, maximum rate of thrombus generation; R, reaction time; SP, split point; TMRTG, time to maximum rate of thrombus generation; TTG, total thrombus generation; D, delta. 732 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 36, NUMBER 7 2010

molecules for soluble fibrin and subsequently thrombin-activated factor XIII converts soluble into cross- linked fibrin. This impacts the rate of clot strength as represented by the a angle and K time. Furthermore, thrombin exerts known effects on platelet function due to interactions with factor VIII and von Willebrand factor; accordingly, such changes mainly impact MA and G. Recent data suggest the superiority of TEG as compared with both aPTT and PT/INR for assessment of the acute coagulopathy of trauma. Kheirabadi et al82 showed in a rabbit model that TEG is a more sensitive indicator of dilutional hypothermic coagulopathy than Figure 5 Coagulopathic tracing after injury and subsequent PT. A recent clinical study of trauma patients surviving correction via goal-directed therapy. The black thromboelas- the first 24 hours reported that TEG detected hyper- tographic tracing was obtained upon admission and shows coagulability (by an increased MA and G) whereas the delayed clot initiation, decreased rate of clot formation (decreased angle), and poor platelet contribution to clot strength PT and aPTT did not.83 In a retrospective review of 84 (decreased maximum amplitude). Administration of fresh- penetrating injuries, Plotkin et al recently demon- frozen plasma, cryoprecipitate, and platelets gradually strated hypocoagulation based on delayed propagation achieved a normal profile, as seen by the subsequent tra- of the clot (increased K time and reduced a angle) and cings in gray. decreased clot strength (reduced MA), where MA correlated with blood product use (r ¼ 0.57, p < 0.01). shows a clinical example of this stepwise goal-directed These findings emphasize the limitations of classic approach to post-injury coagulopathy. coagulation tests and their lack of effectiveness, partic- With results available within 10 minutes, an ularly in post-injury coagulopathy. initial hemostatic assessment with R-TEG identifies POC testing has also proved useful for the de- patients at risk for post-injury coagulopathy upon arrival. tection of fibrinolysis. In a cohort of trauma patients, Blood component therapy is then tailored to address Scho¨chl et al85 determined fibrinolysis as detected by each deranged phase of clotting in a specific manner, and ROTEM as a better predictor of mortality than injury subsequent reassessment allows the evaluation of re- severity score. The primary difference between ROTEM sponse until a set threshold is reached. This strategy and TEG relates to hardware; whereas TEG operates by also permits improved communication with the blood oscillating a cup filled with a sample that engages a pin/ bank; based on initial assessment and response to com- wire transduction system as clot formation occurs, the ponent therapy, more accurate estimations of component ROTEM involves an immobile cup in which the pin/ requirements can be made. wire transduction system slowly oscillates.86 Parameters Fig. 6 depicts our current goal-directed approach and diagnostic nomenclature among these two assays are to coagulopathy. Reflecting the initiation phase of en- similar. In our recent experience with trauma patients zymatic factor activity, a prolonged TEG-ACT value is requiring MT, fibrinolysis was identified by TEG in the earliest indicator of coagulopathy; when the value is 18% of patients and occurred <1 hour post-injury.7 above threshold, FFP is administered. K time and Downloaded by: University of Pennsylvania Library. Copyrighted material. Among patients requiring MT, 65% had fibrinolysis, a angle follow and are most dependent on the avail- with the risk of death correlating significantly with the ability of fibrinogen to be cleaved into fibrin while in the presence of fibrinolysis (p ¼ 0.027). presence of thrombin. If indicated by K and a angle, cryoprecipitate is administered, providing a concentrated form of fibrinogen (150 to 250 mg/10 mL).88 MA is Goal-Directed Resuscitation of Post-Injury then noted, considering the relationship between fibrin Coagulopathy generated during the initial phases of hemostasis and Goal-directed transfusion therapy guided by TEG tai- platelets via GP IIb-IIIa receptor interaction.75 Platelets lors blood product administration to the pathophysio- are administered based on an MA <54 mm, which logical state of the individual patient. At our institution, reflects the platelets’ functional contribution to clot this approach has become an integral part of resuscita- formation. Of note, in this protocol, platelet transfusion tion.87 Using this technology, a variety of coagulation is not based on platelet count but on functional contri- abnormalities have been noted that in the past would bution to clotting, reflected by the MA. have been overlooked. Our current protocol of compo- Antifibrinolytic drugs such as aprotinin or amino- nent transfusion therapy emphasizes a goal-directed caproic acid have proven effective in hemorrhage during approach, permitting stepwise correction of coagulation cardiac surgery and hepatic transplantation.89,90 How- dysfunction by repeated assessment via TEG. Fig. 5 ever, both cost and morbidity associated with indiscrim- POINT-OF-CARE THROMBOELASTOGRAPHY/GONZALEZ ET AL 733

Figure 6 Algorithm of goal-directed approach to post-traumatic coagulopathy. a angle, rate of clot formation; ACA, aminocaproic acid; ACT, activated clotting time; EPL, estimated percentage lysis; FFP, fresh-frozen plasma; G, clot strength; MA, maximum amplitude; r-TEG, rapid thromboelastography; TEG, thromboelastography. inant use mandate an accurate diagnosis of fibrinolysis. potential benefits: (1) reduction of transfusion volumes Specifically, these drugs have been associated with an via specific goal-directed treatment of identifiable coag- increased risk of renal failure and thrombotic events such ulation abnormalities, (2) earlier correction of coagulation as stroke and myocardial infarction.91 Of note, TEG is abnormalities with more efficient restoration of physio- the only current test able to establish a diagnosis of logical hemostasis, (3) improved survival in the acute Downloaded by: University of Pennsylvania Library. Copyrighted material. fibrinolysis rapidly and reliably in the acutely bleeding hemorrhagic phase due to improved hemostasis from patient.92,93 After the tracing has reached MA, an EPL index is obtained based on the decreasing rate of clot strength. Epsilon-aminocaproic acid is indicated in the presence of significant fibrinolysis (EPL >15%). Fig. 7 shows a clinical example of correction of a fibrinolytic TEG tracing in a hemorrhaging patient following administration of aminocaproic acid. Furthermore, we have observed a post-fibrinolysis consumptive coagulopathy, which represents diffuse clotting factor deficiency secondary to massive consumption after fibrinolysis. The resulting severe thrombin deficiency may be an indication for recombinant factor VIIa, and we have noted rapid improvement with normalization of TEG patterns Figure 7 Fibrinolytic tracing. Profound fibrinolysis (tracing after such treatment. in black) seen in hemorrhaging patient after injury. Subse- In summary, implementation of a goal-directed quent correction of coagulation profile is achieved (tracing in approach to post-injury coagulopathy offers the following gray) after administration of aminocaproic acid. 734 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 36, NUMBER 7 2010

correction of coagulopathy, and (4) improved outcomes linear correlation with TMRTG may allow for identifi- in the later phase due to attenuation of immuno-inflam- cation of thrombin’s important contribution to clot for- matory complications, including ARDS and MOF. mation.81 Using delta as a measure of enzymatic activity Current retrospective data and pilot studies sup- may allow for differentiation of the enzymatic contribu- port a reduction in MT rates, decreased need for multi- tion to hypercoagulability from that of platelet reactivity. ple and repeated classic coagulation tests, and decreased morbidity and mortality after implementation of TEG in trauma care.21,81,85,94 We recently compared pre- LIMITATIONS AND FUTURE DIRECTIONS TEG to post-TEG outcomes of patients at risk of Introduction of TEG into clinical practice has multiple post-injury coagulopathy admitted to our trauma cen- challenges. Although the routine operational cost of a ter.95 The TEG G value was significantly associated TEG assay is comparable with that of a single classic with survival (p ¼ 0.03), whereas PT/INR and PTT did coagulation test (PT/INR, aPTT), the cost of the TEG not discriminate between survivors and nonsurvivors. analyzer is significantly higher than that of the equip- Further validation studies are ongoing at our institution ment required for classic coagulation testing. However, as well as other trauma centers. reductions in financial and administrative costs have been documented at institutions using TEG in cardiac surgery patients because decreased blood transfusion and Post-Injury Hypercoagulability: Risk surgical reexploration rates have decreased after imple- Stratification mentation of TEG.99,100 Despite this, no documented Because POC TEG characterizes dynamic hypercoagul- financial benefit has yet been established prospectively ability and simultaneously reflects the antithrombotic comparing TEG with routine clinical practice when effect of chemoprophylaxis, it may serve as a template for caring for trauma patients. The potential benefits of designing tailored thromboprophylactic regimens. We TEG are contingent on accurate interpretation of re- anticipate that TEG will prove useful for optimizing sults. This comes with a significant learning curve for the prophylactic LMWH dosing and monitoring, and for clinician to translate these potential benefits into clinical identifying those patients requiring antiplatelet therapy practice. Our initial experience with goal-directed ther- in addition to, or in lieu of, LMWH therapy for the apy via TEG and our current preliminary studies have prevention of TE events. Additionally, recent experience led us to initiate a randomized prospective clinical trial to gained in cardiovascular medicine with TEG platelet evaluate more accurately the usefulness of TEG as a mapping for dosing and monitoring of platelet inhib- screening test for hypercoagulability and monitoring of itors96–98 suggests its usefulness for guiding prophylactic thromboprophylaxis. If confirmed, an improved risk antiplatelet therapy in this subgroup. stratification of surgical patients based on the presence, Our recent work has validated that TEG hyper- degree, and differentiation of the hypercoagulable state coagulable G values were predictive of TE events in a could facilitate changes to our current thromboprophy- cohort of critically ill surgical patients receiving standard laxis regimens. We have also initiated a randomized chemoprophylaxis.21 Despite chemoprophylaxis, 60% of prospective clinical trial evaluating the efficacy of our patients remained hypercoagulable as measured by G.In current TEG protocol in managing post-injury coagul- a multiple regression analysis, for every 1-point increase opathy in the context of MT. These studies will help in G there was a 25% increased risk of TE complications, elucidate the usefulness of this promising technology. Downloaded by: University of Pennsylvania Library. Copyrighted material. when controlling for the presence of chemoprophylaxis (odds ratio: 1.25, 95% confidence interval, 1.12–1.39). None of the patients with a normal coagulation TEG CONCLUSION profile developed a TE event (p < 0.001). Successful management of the severely injured patient Recently, an in vivo study evaluating the effect of requires knowledge of the dynamic process of coagula- LMWH on TEG parameters reported that TEG pa- tion. Impaired platelet function and coagulation factor rameters correlated significantly with anti-Xa levels. activity are not predicted by dilution or hemorrhagic Among these parameters, the TEG R time had the shock alone. Rather, the coagulopathy of trauma is a best correlation with anti-Xa levels with the most favor- complex multifactorial event. Furthermore, complica- able combination of sensitivity and specificity for the tions resulting from hypercoagulable states developing therapeutic range of anti-Xa levels (r ¼ 0.82; p < 0.0001; after post injury remain significant continuing challenges sensitivity 68%, specificity 100%).62 after successful resuscitation has been accomplished. As mentioned previously, one prominent limita- Current thromboprophylactic regimens incompletely tion of current thromboprophylactic regimens is the address the pathophysiology of venous thrombosis, absence of platelet inhibition; no differentiation between which likely reflects multiple complex interactions in- enzymatic and platelet contribution to hypercoagulability volving coagulation proteins, platelets, endothelial cells, can be made based on classic coagulation tests. Delta’s and inflammatory mediators. Recent experience with POINT-OF-CARE THROMBOELASTOGRAPHY/GONZALEZ ET AL 735

TEG suggests that this technology could serve as a 17. Maegele M, Lefering R, Yucel N, et al; AG Polytrauma of template for clinical applications of the CBM of coag- the German Trauma Society (DGU). Early coagulopathy in ulation in the injured patient. Subsequent treatment multiple injury: an analysis from the German Trauma protocols are now being tailored based on specific Registry on 8724 patients. Injury 2007;38(3):298–304 18. Furie B, Furie BC. Mechanisms of thrombus formation. evaluation of clot formation as an ex vivo representative N Engl J Med 2008;359(9):938–949 assay of the coagulation process. 19. Cannon WB, Fraser J, Cowell E. The preventive treatment of wound shock. JAMA 1918;70:618–621 20. Schreiber MA, Differding JS, Thorborg P, Mayberry JC, REFERENCES Mullins RJ. Hypercoagulability is most prevalent early after injury and in female patients. J Trauma 2005;58(3):475– 1. Sauaia A, Moore FA, Moore EE, et al. Epidemiology 480; discussion 480–481 of trauma deaths: a reassessment. J Trauma 1995;38: 21. Kashuk JL, Moore EE, Sabel A, et al. Rapid thrombelastog- 185–193 raphy (r-TEG) identifies hypercoagulability and predicts 2. Kauvar DS, Lefering R, Wade CE. Impact of hemorrhage thromboembolic events in surgical patients. Surgery 2009; on trauma outcome: an overview of epidemiology, clinical 146(4):764–772; discussion 772–774 presentations, and therapeutic considerations. J Trauma 22. Attia J, Ray JG, Cook DJ, Douketis J, Ginsberg JS, Geerts 2006;60(6, Suppl):S3–S11 WH. Deep vein thrombosis and its prevention in critically 3. Cosgriff N, Moore EE, Sauaia A, Kenny-Moynihan M, ill adults. Arch Intern Med 2001;161(10):1268–1279 Burch JM, Galloway B. Predicting life-threatening 23. O’Malley KF, Ross SE. Pulmonary embolism in major coagulopathy in the massively transfused trauma patient: trauma patients. J Trauma 1990;30(6):748–750 hypothermia and acidoses revisited. J Trauma 1997; 42(5): 24. Ramacciotti E, Myers DD Jr, Wrobleski SK, et al. P- 857–861; discussion 861–862 selectin/ PSGL-1 inhibitors versus enoxaparin in the 4. MacLeod JB, Lynn M, McKenney MG, Cohn SM, Murtha resolution of venous thrombosis: a meta-analysis. Thromb M. Early coagulopathy predicts mortality in trauma. Res 2010;125(4):e138–e142 J Trauma 2003;55(1):39–44 25. Lo´pez JA, Kearon C, Lee AY. Deep venous thrombosis. 5. Kashuk JL, Moore EE, Millikan JS, Moore JB. Major Hematology Am Soc Hematol Educ Program 2004: abdominal vascular trauma—a unified approach. J Trauma 439–456 1982;22(8):672–679 26. Sevitt S. The structure and growth of valve-pocket thrombi 6. Tieu BH, Holcomb JB, Schreiber MA. Coagulopathy: its in femoral veins. J Clin Pathol 1974;27(7):517–528 pathophysiology and treatment in the injured patient. 27. Kim Y, Nakase H, Nagata K, Sakaki T, Maeda M, World J Surg 2007;31(5):1055–1064 Yamamoto K. Observation of arterial and venous thrombus 7. Kashuk JL, Moore EE, Sawyer Met al. Primary fibrinolysis formation by scanning and transmission electron micro- is integral in the pathogenesis of acute coagulopathy of scopy. Acta Neurochir (Wien) 2004;146(1):45–51; discus- trauma. Paper presented at: American Surgical Association sion 51 130th Annual Meeting, 2010; Chicago, IL 28. Myers DD Jr, Rectenwald JE, Bedard PW, et al. Decreased 8. Brohi K, Cohen MJ, Ganter MT, Matthay MA, Mackersie venous thrombosis with an oral inhibitor of P selectin. J RC, Pittet JF. Acute traumatic coagulopathy: initiated by Vasc Surg 2005;42(2):329–336 hypoperfusion: modulated through the protein C pathway? 29. Hoffman M, Monroe DM III. A cell-based model of Ann Surg 2007;245(5):812–818 hemostasis. Thromb Haemost 2001;85(6):958–965 9. Rossaint R, Cerny V, Coats TJ, et al. Key issues in advanced 30. McPherson MA, Pincus MR. Henry’s Clinical Diagnosis bleeding care in trauma. Shock 2006;26(4):322–331 and Management by Laboratory Methods. 21st ed. 10. Spivey M, Parr MJ. Therapeutic approaches in trauma- Philadelphia, PA: W.B. Saunders; 2006 induced coagulopathy. Minerva Anestesiol 2005;71(6): 31. Lucas CE, Ledgerwood AM. Clinical significance of altered

281–289 coagulation tests after massive transfusion for trauma. Am Downloaded by: University of Pennsylvania Library. Copyrighted material. 11. MacLeod JB. Trauma and coagulopathy: a new paradigm to Surg 1981;47(3):125–130 consider. Arch Surg 2008;143(8):797–801 32. Brummel KE, Paradis SG, Butenas S, Mann KG. 12. Kashuk JL, Moore EE, Johnson JL, et al. Postinjury life Thrombin functions during tissue factor-induced blood threatening coagulopathy: is 1:1 fresh frozen plasma:packed coagulation. Blood 2002;100(1):148–152 red blood cells the answer? J Trauma 2008;65(2):261–270, 33. Martini WZ. Coagulopathy by hypothermia and acidosis: discussion 270–271 mechanisms of thrombin generation and fibrinogen 13. Amara U, Rittirsch D, Flierl M, et al. Interaction between availability. J Trauma 2009;67(1):202–208; discussion the coagulation and complement system. Adv Exp Med Biol 208–209 2008;632:71–79 34. Ferrara A, MacArthur JD, Wright HK, Modlin IM, 14. Kapsch DN, Metzler M, Harrington M, Mitchell FL, Silver McMillen MA. Hypothermia and acidosis worsen coagul- D. Fibrinolytic response to trauma. Surgery 1984;95(4): opathy in the patient requiring massive transfusion. Am J 473–478 Surg 1990;160(5):515–518 15. Gando S, Tedo I, Kubota M. Posttrauma coagulation and 35. Counts RB, Haisch C, Simon TL, Maxwell NG, Heimbach fibrinolysis. Crit Care Med 1992;20(5):594–600 DM, Carrico CJ. Hemostasis in massively transfused trauma 16. Moore EE, Moore FA, Fabian TC, et al; PolyHeme Study patients. Ann Surg 1979;190(1):91–99 Group. Human polymerized hemoglobin for the treatment 36. Lucas CE, Ledgerwood AM. Clinical significance of altered of hemorrhagic shock when blood is unavailable: the USA coagulation tests after massive transfusion for trauma. Am multicenter trial. J Am Coll Surg 2009;208(1):1–13 Surg 1981;47(3):125–130 736 SEMINARS IN THROMBOSIS AND HEMOSTASIS/VOLUME 36, NUMBER 7 2010

37. Opal SM, Esmon CT. Bench-to-bedside review: functional 55. Ellis TC, Nielsen VG, Marques MB, et al. Thrombelasto- relationships between coagulation and the innate immune graphic measures of clot propagation: a comparison of alpha response and their respective roles in the pathogenesis of with the maximum rate of thrombus generation. Blood sepsis. Crit Care 2003;7(1):23–38 Coagul Fibrinolysis 2007;18:45–48 38. Lier H, Krep H, Schroeder S, Stuber F. Preconditions of 56. Marik PE, Corwin HL. Efficacy of red blood cell hemostasis in trauma: a review. The influence of acidosis, transfusion in the critically ill: a systematic review of the hypocalcemia, anemia, and hypothermia on functional literature. Crit Care Med 2008;36(9):2667–2674 hemostasis in trauma. J Trauma 2008;65(4):951–960 57. Moore FA, Moore EE, Sauaia A. Blood transfusion. An 39. Quaknine-Orlando B, Samama CM, Riou B, et al. Role of independent risk factor for postinjury multiple organ failure. the hematocrit in a rabbit model of arterial thrombosis and Arch Surg 1997;132(6):620–624; discussion 624–625 bleeding. Anesthesiology 1999;90(5):1454–1461 58. Watson GA, Sperry JL, Rosengart MR, et al; Inflamma- 40. Turitto VT, Weiss HJ. Red blood cells: their dual role in tion and Host Response to Injury Investigators. Fresh thrombus formation. Science 1980;207(4430):541–543 frozen plasma is independently associated with a higher 41. Aucar JA, Isaak E, Anthony D. The effect of red blood risk of multiple organ failure and acute respiratory distress cell age on coagulation. Am J Surg 2009;198(6):900–904 syndrome. J Trauma 2009;67(2):221–227; discussion 42. Kashuk JL, Moore EE, Johnson JL, et al. Postinjury life 228–230 threatening coagulopathy: is 1:1 fresh frozen plasma:packed 59. Johnson JL, Moore EE, Kashuk JL. Early transfusion of red blood cells the answer? J Trauma 2008;65(2):261–270; FFP is independently associated with postinjury multiple discussion 270–271 organ failure. Arch Surg 2010; In press 43. Holcomb JB, Wade CE, Michalek JE, et al. Increased 60. Khan H, Belsher J, Yilmaz M, et al. Fresh-frozen plasma plasma and platelet to red blood cell ratios improves and platelet transfusions are associated with development of outcome in 466 massively transfused civilian trauma acute lung injury in critically ill medical patients. Chest patients. Ann Surg 2008;248(3):447–458 2007;131(5):1308–1314 44. Borgman MA, Spinella PC, Perkins JG, et al. The ratio of 61. Flesland O. A comparison of complication rates based on blood products transfused affects mortality in patients published haemovigilance data. Intensive Care Med receiving massive transfusions at a combat support hospital. 2007;33(Suppl 1):S17–S21 J Trauma 2007;63(4):805–813 62. Artang R, Frandsen NJ, Nielsen JD. Application of basic 45. Gonzalez EA, Moore FA, Holcomb JB, et al. Fresh frozen and composite thrombelastography parameters in monitor- plasma should be given earlier to patients requiring massive ing of the antithrombotic effect of the low molecular weight transfusion. J Trauma 2007;62(1):112–119 heparin dalteparin: an in vivo study. Thromb J 2009; 46. Gunter OL Jr, Au BK, Isbell JM, Mowery NT, Young PP, 10(7):14 Cotton BA. Optimizing outcomes in damage control 63. Van PY, Cho SD, Underwood SJ, Morris MS, Watters JM, resuscitation: identifying blood product ratios associated Schreiber MA. Thrombelastography versus AntiFactor Xa with improved survival. J Trauma 2008;65(3):527–534 levels in the assessment of prophylactic-dose enoxaparin in 47. Davenport R, De’Ath H, Manson J, et al. Coagulation critically ill patients. J Trauma 2009;66(6):1509–1515; response to different plasma:red cell ratios in trauma trans- discussion 1515–1517 fusion. Paper presented at: Annual Meeting of the Eastern 64. Geerts WH, Bergqvist D, Pineo GF, et al. Prevention of Association for the Surgery of Trauma; 2010; Naples, FL venous thromboembolism: American College of Chest 48. Scalea TM, Bochicchio KM, Lumpkins K, et al. Early Physicians Evidence-Based Clinical Practice Guidelines. aggressive use of fresh frozen plasma does not improve Chest 2008;133(6 Suppl):381S–453S outcome in critically injured trauma patients. Ann Surg 65. Weitz JI. Low-molecular-weight heparins. N Engl J Med 2008;248(4):578–584 1997;337(10):688–698 49. Snyder CW, Weinberg JA, McGwin G Jr, et al. The 66. Patel R, Cook DJ, Meade MO, et al; Burden of Illness in relationship of blood product ratio to mortality: survival venous ThromboEmbolism in Critical care (BITEC) Study

beneﬁt or survival bias? J Trauma 2009;66(2):358–362; Investigators; Canadian Critical Care Trials Group. Burden Downloaded by: University of Pennsylvania Library. Copyrighted material. discussion 362–364 of illness in venous thromboembolism in critical care: a 50. Magnotti LJ, Zarzaur BL, Croce MA, et al. Improved multicenter observational study. J Crit Care 2005;20(4): survival after hemostatic resuscitation: does the emperor 341–347 have no clothes? J Trauma 2010; In press 67. Ibrahim EH, Iregui M, Prentice D, Sherman G, Kollef 51. Perkins JG, Cap AP, Andrew CP, et al. An evaluation of MH, Shannon W. Deep vein thrombosis during prolonged the impact of apheresis platelets used in the setting of mechanical ventilation despite prophylaxis. Crit Care Med massively transfused trauma patients. J Trauma 2009;66 2002;30(4):771–774 (4, Suppl):S77–S84; discussion S84–S85 68. Geerts WH, Jay RM, Code KI, et al. A comparison of low- 52. Ciavarella D, Reed RL, Counts RB, et al. Clotting factor dose heparin with low-molecular-weight heparin as pro- levels and the risk of diffuse microvascular bleeding in phylaxis against venous thromboembolism after major the massively transfused patient. Br J Haematol 1987;67: trauma. N Engl J Med 1996;335(10):701–707 365–368 69. Hartert H. Blutgerinnungsstudien mit der Thrombelas- 53. Hiippala ST, Myllyla¨ GJ, Vahtera EM. Hemostatic factors tographie, einem neuen Untersuchungsverfahren. Klin and replacement of major blood loss with plasma-poor red Wochenschr 1948;16(37–38):577–583 cell concentrates. Anesth Analg 1995;81(2):360–365 70. De Nicola P. Thromboelastography. Oxford, United 54. Zuckerman L, Cohen E, Vagher JP, Woodward E, Caprini Kingdom: Blackwell Scientiﬁc; 1957 JA. Comparison of thrombelastography with common 71. Kang YG, Martin DJ, Marquez J, et al. Intraoperative coagulation tests. Thromb Haemost 1981;46(4):752–756 changes in blood coagulation and thrombelastographic POINT-OF-CARE THROMBOELASTOGRAPHY/GONZALEZ ET AL 737

monitoring in liver transplantation. Anesth Analg 1985; 87. Stahel PF, Moore EE, Schreier SL, Flierl MA, Kashuk JL. 64(9):888–896 Transfusion strategies in postinjury coagulopathy. Curr 72. Tuman KJ, Spiess BD, McCarthy RJ, Ivankovich AD. Opin Anaesthesiol 2009;22(2):289–298 Comparison of viscoelastic measures of coagulation after 88. Bolliger D, Szlam F, Molinaro RJ, Rahe-Meyer N, Levy cardiopulmonary bypass. Anesth Analg 1989;69(1):69–75 JH, Tanaka KA. Finding the optimal concentration range 73. Gibbs NM, Crawford GP, Michalopoulos N. Thrombelas- for fibrinogen replacement after severe haemodilution: an tographic patterns following abdominal aortic surgery. in vitro model. Br J Anaesth 2009;102(6):793–799 Anaesth Intensive Care 1994;22(5):534–538 89. Vorweg M, Hartmann B, Knu¨ttgen D, Jahn MC, Doehn 74. Chen A, Teruya J. Global hemostasis testing thromboelas- M. Management of fulminant fibrinolysis during abdomi- tography: old technology, new applications. Clin Lab Med nal aortic surgery. J Cardiothorac Vasc Anesth 2001;15(6): 2009;29(2):391–407 764–767 75. Khurana S, Mattson JC, Westley S, O’Neill WW, Timmis 90. Kang Y. Thromboelastography in liver transplantation. GC, Safian RD. Monitoring platelet glycoprotein IIb/IIIa- Semin Thromb Hemost 1995;21(Suppl 4):34–44 fibrin interaction with tissue factor-activated thromboelas- 91. Ray WA, Stein CM. The aprotinin story—is BART the tography. J Lab Clin Med 1997;130(4):401–411 final chapter? N Engl J Med 2008;358(22):2398–2400 76. Nielsen VG, Geary BT, Baird MS. Evaluation of the 92. Ide M, Bolliger D, Taketomi T, Tanaka KA. Lessons from contribution of platelets to clot strength by thromboelastog- the aprotinin saga: current perspective on antifibrinolytic raphy in rabbits: the role of tissue factor and cytochalasin D. therapy in cardiac surgery. J Anesth 2010;24(1):96–106 Anesth Analg 2000;91(1):35–39 93. Stamou SC, Reames MK, Skipper E, et al. Aprotinin in 77. Michelson AD. Methods for the measurement of platelet cardiac surgery patients: is the risk worth the benefit?. function. Am J Cardiol 2009;103(3, Suppl):20A–26A Presented at: The Society of Critical Care Medicine; 78. Nielsen VG. Beyond cell based models of coagulation: January 2008; Nashville, TN analyses of coagulation with clot ‘‘lifespan’’ resistance-time 94. Zumwinkle L, Kashuk J, Moore EE. Rapid thrombelastog- relationships. Thromb Res 2008;122(2):145–152 raphy differentiates primary from secondary fibrinolysis. 79. Sørensen B, Johansen P, Christiansen K, Woelke M, Crit Care Med 2008;36:187–190 Ingerslev J. Whole blood coagulation thrombelastographic 95. Kashuk JL, Moore EE, Wohlauer M, et al. Point of care profiles employing minimal tissue factor activation. rapid thrombelastography improves management of life J Thromb Haemost 2003;1(3):551–558 threatening post-injury coagulopathy. Paper presented at: 80. Rivard GE, Brummel-Ziedins KE, Mann KG, Fan L, American Association for the Surgery of Trauma 68th Hofer A, Cohen E. Evaluation of the profile of thrombin Annual Meeting; October 2009; Pittsburgh, PA generation during the process of whole blood clotting as 96. Kaur J, Jones N, Mallett S. Thrombelastography Platelet assessed by thrombelastography. J Thromb Haemost 2005; Mapping is a useful preoperative tool in surgical patients 3(9):2039–2043 taking antiplatelet medication. Br J Anaesth 2009;103(2):304; 81. Gonzalez E, Kahuk JL, Moore EE, Silliman C. Differ- author reply 304–305 entiation of enzymatic from platelet hypercoagulability 97. Agarwal S, Coakley M, Coakely M, Reddy K, Riddell A, using the novel thrombelastography parameter delta. J Surg Mallett S. Quantifying the effect of antiplatelet therapy: a Res 2010;April 21 (Epub ahead of print) comparison of the platelet function analyzer (PFA-100) and 82. Kheirabadi BS, Crissey JM, Deguzman R, Holcomb JB. modified thromboelastography (mTEG) with light trans- In vivo bleeding time and in vitro thrombelastography mission platelet aggregometry. Anesthesiology 2006;105(4): measurements are better indicators of dilutional hypo- 676–683 thermic coagulopathy than prothrombin time. J Trauma 98. Tantry US, Bliden KP, Gurbel PA. Overestimation of 2007;62(6):1352–1359; discussion 1359–1361 platelet aspirin resistance detection by thrombelastograph 83. Park MS, Martini WZ, Dubick MA, et al. Thromboelas- platelet mapping and validation by conventional aggregom- tography as a better indicator of hypercoagulable state after etry using arachidonic acid stimulation. J Am Coll Cardiol

injury than prothrombin time or activated partial throm- 2005;46(9):1705–1709 Downloaded by: University of Pennsylvania Library. Copyrighted material. boplastin time. J Trauma 2009;67(2):266–275; discussion 99. Spalding GJ, Hartrumpf M, Sierig T, Oesberg N, Kirschke 275–276 CG, Albes JM. Bedside thrombelastography. Cost reduc- 84. Plotkin AJ, Wade CE, Jenkins DH, et al. A reduction in clot tion in cardiac surgery [in German]. Anaesthesist 2007; formation rate and strength assessed by thromboelastography 56(8):765–771 is indicative of transfusion requirements in patients with 100. Spiess BD, Gillies BS, Chandler W, Verrier E. Changes in penetrating injuries. J Trauma 2008;64(2 Suppl):S64–S68 transfusion therapy and reexploration rate after institution 85. Scho¨chl H, Frietsch T, Pavelka M, Ja´mbor C. Hyper- of a blood management program in cardiac surgical patients. ﬁbrinolysis after major trauma: differential diagnosis of lysis J Cardiothorac Vasc Anesth 1995;9(2):168–173 patterns and prognostic value of thrombelastometry. 101. Knudson MM, Ikossi DG, Khaw L, Morabito D, Speetzen J Trauma 2009;67(1):125–131 LS. Thromboembolism after trauma: an analysis of 1602 86. Nielsen VG. A comparison of the Thrombelastograph and episodes from the American College of Surgeons National the ROTEM. Blood Coagul Fibrinolysis 2007;18(3): Trauma Data Bank. Ann Surg 2004;240(3):490–496; 247–252 discussion 496–498