Use of Viscoelastography in Malignancy-Associated Coagulopathy and Thrombosis: A Review

Abstract

The relationship between malignancy and coagulopathy is one that is well documented yet incompletely understood. Clinicians have attempted to quantify the hypercoagulable state produced in various malignancies using common coagulation tests such as prothrombin time, activated partial thromboplastin time, and platelet count; however, due to these tests’ focus on individual aspects of coagulation during one specific time point, they have failed to provide clinicians the complete picture of malignancy-associated coagulopathy (MAC). Viscoelastic tests (VETs), such as thromboelastography (TEG) and rotational thromboelastometry (ROTEM), are whole blood analyses that have the advantage of providing information related to the cumulative effects of plasma clotting factors, platelets, leukocytes, and red cells during all stages of the coagulation and fibrinolytic processes. VETs have gained popularity in the care of trauma patients to objectively measure trauma-induced coagulopathy (TIC), but the utility of VETs remains yet unrealized in many other medical specialties. The authors discuss the similarities and differences between TIC and MAC, and propose a mechanism for the hypercoagulable state of MAC that revolves around the thrombomodulin–thrombin complex as it switches between activating the protein C anticoagulation pathway or the thrombin activatable fibrinolysis inhibitor coagulation pathway. Additionally, they review the current literature on the use of TEG and ROTEM in patients with various malignancies. Although limited research is currently available, early results demonstrate the utility of both TEG and ROTEM in the prediction of hypercoagulable states and thromboembolic complications in oncologic patients.

“Omnis cellula e cellula” [All cells from pre-existing cells] R. Virchow, 1855.^1,2

Despite increased awareness of venous thromboembolic prophylaxis, thromboemboli occur with increasing frequency in the United States, particularly in patients with malignancy.³ The association between thrombosis and malignancy was first described by Trousseau over a century ago.^4,5 At minimum, an estimated 10 to 20% of patients with carcinoma of the pancreas, lung, or stomach also present with concomitant thromboembolic disease, which is associated with higher staging and increased thrombotic risk.^5,6 Moreover, development of venous thrombosis within 1 year of diagnosis of a primary carcinoma increases the probability of detecting advanced disease.^5,7,8 Among patients with solid malignant tumors and hematopoietic malignancies, activation of coagulation occurs despite the absence of clinically apparent thromboembolic disease.^6,7,9–11 The process of ongoing activation of the coagulation system in malignancy is not well understood, but may involve tumor-induced activation of platelets, increased expression of tumor-related intravascular tissue factor (TF) in concert with circulating mononuclear phagocytes, and upregulation of endogenous inhibitors of fibrinolysis. These dysregulated host responses occur within the milieu of endothelial cell activation that exists as a principal component of a chronic, low-grade systemic inflammatory response.^10,11 We refer to this dysregulated coagulation as malignancy-associated coagulopathy (MAC).

Common coagulation tests (CCTs), including PT/INR (prothrombin time/international normalized ratio) and aPTT (activated partial thromboplastin time), have traditionally been used for identifying hypocoagulable states. Although recently a shortened aPTT has been associated with thrombosis risk, there is still concern that such results are variably attributable to collection artifacts, reagent disparities, or other preanalytic issues.¹² Further, CCTs are plasma-based assays that analyze a static snapshot of factor deficiencies in coagulation.³ Accordingly, they are poor predictors of the evolving clotting dynamics during a hypercoagulable state.³ Thus, it is critical to identify patients at risk for developing tumor-associated thrombosis using readily available biomarkers such as platelet count and plasma D-dimer levels, as well as research tools that identify circulating TF and P-selectin.^13–21 These thrombotic biomarkers are only part of the story of MAC. Thus, recent American Society of Clinical Oncology guidelines counsel against assessing risk based on a single test, but rather emphasize scoring systems that consider multiple factors such as Khorana’s risk assessment model for assessing risk of venous thromboembolism in cancer patients.^{15–18,20–22}

Over a century after Virchow foreshadowed the idea in the 1850s, literature regarding trauma resuscitation, liver transplantation, and cardiac surgery has brought to light the possibility of using whole blood viscoelastic tests (VETs) such as thromboelastography (TEG; Haemonetics) and rotational thromboelastometry (ROTEM; TEM International) to risk stratify cancer patients for venous thromboembolism.^3,23,24 Recent literature has indicated that in noncancerous states, such as critically ill, septic, and postoperative states, VETs are useful in identifying a hypercoagulable state and guiding resuscitation.^23–29 The use of VETs to guide hemostatic resuscitation of patients with trauma-induced coagulopathy (TIC) has also led to these point-of-care tests being used as predictors of thromboembolism and multiorgan failure (MOF) in trauma patients during the postresuscitation period.^30–34 The utility of these assays has now been extended to surgical oncology patients with colon, rectal, and breast cancer who present with seemingly normal fibrinogen and platelets.³⁵ High rates of thrombotic complications are observed with advanced pancreatic cancer and VETs have been able to accurately predict which patients will develop postoperative pulmonary emboli after pancreatic resection.^35,36

Mechanisms of Hypercoagulable State in Cancer

Of interest for this review are the similarities of these derangements and hypercoagulability during the posttrauma resuscitation periods and the more prolonged cancer period. In cancer patients, the gradual and subclinical presentation of hemostatic activation and dysregulated coagulation results from structural, hemostatic, and biochemical derangements of coagulation factors, platelets, and endothelium. Thus, this complicated interplay of MAC is illustrated in ►Fig. 1. In TIC, there are several theories that explain the coagulopathy derangements as illustrated in ►Fig. 2. Then, as illustrated in ►Fig. 3, MAC is compared to TIC. The comparison with trauma is a useful starting point for discussing the utility of VETs in diagnosing and monitoring cancer patients at risk for thromboembolism.

Fig. 1.

Proposed molecules and cells involved in the hypercoagulable/hypofibrinolytic phenotype present in malignancy-associated coagulopathy (MAC). ADP, adenosine diphosphate; NETs, neutrophil extracellular traps; PAI-1, plasminogen activator inhibitor-1; TNF, tissue necrosis factor.

Fig. 2.

Proposed mechanisms for trauma-induced coagulopathy (TIC). This teeter-totter spectrum shows the balance and molecular interplay between hypocoagulable/hyperfibrinolytic and hypercoagulable/hypofibrinolytic phenotypes in TIC. There are four current hypotheses to explain TIC which are each centered around the balance of key molecules as follows: (A)DIC–fibrinolysis, (B) glycocalyx, (C) activated protein C, and (D) fibrinogen. These hypotheses are not mutually exclusive, but instead explain the mechanisms at play in TIC within minutes to hours following trauma. At the center, the TM–thrombin complex is the fulcrum of coagulation regulation. Structural and/or posttranslational covalent modifications of the TM–thrombin complex, along with certain cofactors and receptors, allow the TM–thrombin complex to switch between activating the protein C anticoagulation pathway or the TAFI coagulation pathway. (A) The DIC–fibrinolysis hypothesis centers around the consumption of clotting factors and platelets following hypoperfusion and endothelial injury; this consumption leads to a hypocoagulable/hyperfibrinolytic phenotype where fibrinolytic activity exceeds clot formation resulting in hemorrhage. (B) The glycocalyx hypothesis centers around endothelium injury resulting in glycocalyx shedding and systemic autoheparinization. (C) According to the protein C activation hypothesis, endothelium injury causes the activation of protein C by the TM–thrombin complex at the EPCR; protein C activation favors anticoagulation by inactivating PAI-1 and increasing levels of tPA. (D) The fibrinogen-centric hypothesis refers to the inverse relationship between fibrinogen levels and the activation of protein C. Accordingly, when fibrinolysis predominates, fibrinogen levels are low and protein C is activated on the TM–thrombin complex. DIC, disseminated intravascular coagulation; EGF, epidermal growth factor; EPCR, endothelial protein C receptor; PAI-1, plasminogen activator inhibitor 1; Prtn C, protein C; TAFI, thrombin activatable fibrinolysis inhibitor; TM, thrombomodulin; tPA, tissue plasminogen activator.

Fig. 3.

A double teeter-totter (or scale) spectrum showing the balance and molecular interplay between hypocoagulable/hyperfibrinolytic and hypercoagulable/hypofibrinolytic phenotypes in malignancy-associated coagulopathy (MAC) as compared with trauma-induced coagulopathy (TIC). This figure depicts the proposed mechanisms, endothelial interactions, and TEG/ROTEM tracings of both MAC and TIC; both MAC and TIC are centered around the thrombomodulin–thrombin complex as it switches between activating the protein C anticoagulation pathway and the TAFI coagulation pathway. While the majority of MAC patients favor coagulation and many TIC patients favor anticoagulation, they both exist within a fluid spectrum where the “switch” can be flipped back and forth from a procoagulant to an anticoagulant state when the conditions are favorable (i.e., protein C vs. TAFI activation and PAI-1 vs. tPA activity). The relative importance of PAI-1 and tPA ratios creates a milieu which favors thrombosis in patients with longstanding cancer and in trauma patients with persistent fibrinolytic shutdown following resuscitation. During the acute stage of severe trauma, the hyperfibrinolytic/anticoagulant phenotype gives way in minutes to hours to a fibrinolytic shutdown/procoagulant phenotype with successful resuscitation; this is similar to the hyperfibrinolytic phenotype of patients with early untreated acute promyelocytic leukemia. The four TIC hypotheses (right image) were included for the purposes of comparison with MAC. As explained in ►Fig. 2 and in-text, these hypotheses center around (A) DIC–fibrinolysis, (B) glycocalyx, (C) activated protein C, and (D) fibrinogen.⁴⁹ ADP, adenosine diphosphate; DIC, disseminated intravascular coagulation; EGF, epidermal growth factor; EPCR, endothelial protein C receptor; FDPs, fibrin degradation products; G-CSF, granulocyte colony-stimulating factor; ICAM, intercellular adhesion molecule 1; IL-1, interleukin-1; NETs, neutrophil extracellular traps; PAI-1, plasminogen activator inhibitor-1; Prtn C, protein C; ROTEM, rotational thromboelastometry; TAFI, thrombin activatable fibrinolysis inhibitor; TEG, thromboelastography; TM, thrombomodulin; TNF, tumor necrosis factor; tPA, tissue plasminogen activator; VCAM, vascular cell adhesion molecule 1; VEGF, vascular endothelial growth factor; VTE, venous thromboembolism.

Proposed Molecules and Cells Involved in MAC

Malignancy-associated coagulopathy encompasses a spectrum of hemostatic derangements, platelet dysfunction, and vascular disturbances in patients with cancer. There are many proposed molecules and cells involved in the hypercoagulable/hypofibrinolytic phenotype of malignancy. Falanga et al categorize these molecules into different groups labeled as procoagulant factors, soluble mediators, and adhesion particles.¹⁰ ►Fig. 1 illustrates how the endothelial cells, tumor cells, free-floating cells, and particles interact with these procoagulant factors, soluble mediators, and adhesion particles.

Procoagulant factors include TF, TF-bearing microparticles, heparainase, and cancer procoagulant. TF is a transmembrane protein expressed by various cell types including some cancer cells. As shown in ►Fig. 1, TF stimulates factor X and factor IX by forming a complex with factor VII; this activates the coagulation cascade. Endothelial cells release tissue necrosis factor (TNF), interleukin-1 (IL-1), and P-selectin which induce expression of TF.¹⁰ TNF has multiple effects on coagulability including thrombomodulin (TM) downregulation, fibrinolytic activity suppression, and further expression of IL-1.¹⁰

Similar to TF, but without the transmembrane domain, TF-bearing microparticles are cancer-associated procoagulant proteins that free float in the plasma and also interact with factor X to activate the coagulation cascade.^7,10,37 And, as shown in ►Fig. 1, monocytes can be stimulated to make TF and direct factor X activators.¹⁰ And, neutrophil cell-death is thought to play a role in hypercoagulability because their death releases a variety of proteins that create a mesh of chromatin called neutrophil extracellular traps (NETs).^10,38 NETs react with surrounding structures and further the process of thrombin generation.¹⁰

Malignant cells can produce a variety of soluble mediators (i.e., adenosine diphosphate [ADP], IL-1, TNF, vascular endothelial growth factor, granulocyte-colony stimulating factor, thromboxane A2, and podoplanin) and adhesion molecules (i.e., vascular cell adhesion molecule, intercellular adhesion molecule) which have the ability to induce procoagulant states in other neighboring cell types.^10,38,39 Cancer procoagulant is a molecule found only in malignant and fetal tissue and, as indicated in ►Fig. 1, directly activates factor X; the relative contribution of cancer procoagulant to malignancy-associated hypercoagulability is unknown at this time.¹⁰ MET oncogene is found to be expressed in multiple types of malignancies and, as demonstrated in ►Fig. 1, has been shown to increase plasminogen activator inhibitor-1 (PAI-1) and cyclooxygenase-2 production.^7,40,41 Polyphosphate activates the intrinsic pathway. As indicated in ►Fig. 1, vesicles with high amounts of polyphosphate have been found in prostate cancer and shown to activate factor XII.⁴² Heparanase is expressed by a variety of cancer cells and its nonenzymatic domain enhances TF activity through phosphorylation and by acting as a cofactor for TF.¹⁰

Further, the release of activators from malignant cells stimulates platelets. Platelets can be activated via increased ADP expression by malignant cells, increased von Willebrand factor, tumor-induced thrombin creation, and/or increased expression of TF via release of P-selectin.^10,38,43 For example, mucins released by mucinous adenocarcinomas activate platelets in the presence of P- and L-selectin.⁴⁴ In myeloproliferative disorders, increased platelet clonality could also contribute to hypercoagulability.^45,46 Furthermore, in malignancy-causing liver metastases or membranous nephropathy, for example, decreased circulating antithrombin could result in increased thrombin creation contributing to hypercoagulability.^47,48

Proposed Mechanisms for TIC

For decades, the literature has described trauma patients—including those experiencing shock, hemorrhage, burns, cardiac arrest, and major surgery—with hemostatic disorders that we now call TIC.⁴⁹ The endothelial TM–thrombin complex is the fulcrum in the in vivo regulation of the coagulation spectrum. The TM–thrombin complex can activate the protein C anticoagulation pathway as depicted on the right-hand side of ►Fig. 2. Or, the TM–thrombin complex can activate the thrombin activatable fibrinolysis inhibitor (TAFI) coagulation pathway as shown on the left-hand side of ►Fig. 2. The switch between the two pathways occurs within seconds to minutes after trauma and involves either structural or posttranslational covalent modifications of different sites on the TM–thrombin complex. Then, in the presence of their respective cofactors and/or receptors, the modified TM–thrombin complex binds and activates protein C or TAFI. The activation of protein C by the TM–thrombin complex is accelerated by endothelial protein C receptor (EPCR) binding.

There are four main theories of TIC in relation to the TM–thrombin activated protein C pathway. As shown in ►Figs. 2 and 3, current hypotheses to explain TIC include (A) the disseminated intravascular coagulation (DIC)–fibrinolysis hypothesis, (B) the glycocalyx hypothesis, (C) the protein C activation hypothesis, and (D) the fibrinogen hypothesis (correspondingly labeled in ►Fig. 3). (A) In DIC–fibrinolysis, the theory proposes that the anticoagulant state results from hypoperfusion and endothelial injury that leads to a consumption of clotting factors and platelets. This consumption results in hyperfibrinolysis as evidenced by increased fibrinogen degradation products (FDPs) and high FDP/D-dimer ratios. Following the DIC-fibrinolytic phase, there can be a DIC-thrombotic state that results from high levels of PAI-1 and inhibition of fibrinolysis.⁴⁹ Thus, DIC is not synonymous with TIC because DIC’s anticoagulant state is the result of diffuse fibrin deposition leading to a state in which fibrinolytic activity exceeds clot formation and hemorrhage results.⁴⁹ (B) The glycocalyx injury hypothesis also proposes that bleeding occurs secondary to endothelium injury from hypoperfusion resulting in a shedding of the glycocalyx layer on the luminal side of the endothelium. This leads to what has been called by Johansson et al as SHock INduced Endotheliopathy (SHINE).^27,50 This glycocalyx breakdown signals further endothelial activation and results in systemic autoheparinization as evidenced by increased expression of syndecan-1.⁴⁹ (C) Similar to the glycocalyx hypothesis’s focus on endothelium injury from hypoperfusion, the protein C activation hypothesis also proposes that endothelium injury and/or hypoperfusion results in an anticoagulant state. However, according to the activated protein C hypothesis, the anticoagulant state is the result of protein C activation by the TM–thrombin complex at the EPCR. Increased levels of activated protein C inhibit thrombin generation and favor hyperfibrinolysis by depleting PAI-1 levels which leads to increased tissue plasminogen activator (tPA) activity.^49,51 (D) The fibrinogen hypothesis centers on the concentration of fibrinogen. The hypothesis is that the anticoagulant state is driven by the decrease in fibrinogen which reduces viscoelastic clot amplitudes. The hypofibrinogenemia occurs when fibrinolysis occurs at a greater rate than fibrinogen synthesis. Without sufficient fibrinogen, platelet function is altered and protein C activation on the TM–thrombin complex is increased leading to an anticoagulant state.⁴⁹ The relationship between fibrinogen levels and activated protein C has been proposed as an inverse relationship. Accordingly, when fibrinolysis predominates, fibrinogen levels are low and protein C is activated on the TM–thrombin complex resulting in an anticoagulant state. Conversely, at high levels of fibrinogen, the TM–thrombin complex is inhibited from activating protein C.^49,52

Comparison between MAC and TIC: Pathophysiologic Mechanisms Correlated to TEG/ROTEM Patterns

In both TIC and MAC, TF and its “activation” of the TM–thrombin complex are critical for maintaining normal hemostasis.^10,49 The TM–thrombin activated protein C pathway is present on approximately 7,000 m² of the endothelial surface in both trauma and cancer patients and can become dysregulated. Compared to trauma patients in whom dysregulation occurs over hours, the dysregulation in cancer patients occurs over a much longer period.^6,10,31,49 However, whether over hours or weeks, dysregulated coagulopathy is driven by an imbalance of a number of tumor cells and factors. The balance between the anticoagulant/hypocoagulable and procoagulant/hypercoagulable states in cancer is a function of many interrelated factors as shown in ►Fig. 1 and many of these factors contribute to the increased incidence of thrombosis in cancer. As depicted in ►Fig. 2, these factors are similar to the factors in TIC. For example, the relative importance of PAI-1 and tPA ratios creates a milieu favoring thrombosis in patients with cancer and patients with persistent fibrinolytic “shutdown” following trauma resuscitation. As defined by Moore et al, a fibrinolytic “shutdown” occurs when a physiologic dysregulation of the plasminogen–plasmin system causes a relative resistance to tPA.³¹ During the acute stage of severe trauma, the hyperfibrinolytic phenotype may give way in minutes to hours to a fibrinolytic shutdown phenotype which is similar in many ways to the hyperfibrinolytic phenotype of patients with early untreated acute promyelocytic leukemia (APL).^31,53–59 And, just as increased polyphosphate in prostate cancer has been shown to increase hypercoagulability, in the event of trauma, polyphosphate is released from platelets and strengthens fibrin aggregates while reducing sites for tPA and plasminogen.^31,42,60,61

A classic comparative example of the spectrum of coagulopathies between TIC and MAC can be seen in ►Fig. 3. The hyperfibrinolytic phenotype in the severely injured patient with TIC is the hyperfibrinolysis-induced anticoagulant phenotype, while for hematologic malignancy the similar hyperfibrinolytic phenotype is the initial pretreatment anticoagulant state characteristic of APL. This APL-associated hyperfibrinolytic phenotype is dependent on increased expression of tPA, annexin A2, and urokinase-type plasminogen activator by the leukemic promyelocytes.⁵³ Similar to the hyperfibrinolysis of TIC, the APL-associated hyperfibrinolytic phenotype may be quickly reversed and rapidly become a hypercoagulable state when treated with all-trans retinoic acid (ATRA) and other therapeutic agents.^49,53–59 Inhibitors of the TF–FVIIa–FXa complex, such as TF pathway inhibitor are present in APL patients. Also, PAI-1 is complexed with tPA reducing both their activities in APL patients.^53,54 Severe hyperfibrinolytic TIC bears similarities to the MAC present in APL because an early anticoagulant state is a function of hyperfibrinolysis but, following treatment in APL or resuscitation in TIC, patients become hypercoagulopathic. Therefore, whether caused by malignancy or trauma, TEG and the ROTEM allow for evaluation of the patient’s phenotype along the spectrum of hypocoagulable/hyperfibrinolytic to hypercoagulable/hypofibrinolytic.

TEG and ROTEM

Thromboelastography

The first documented use of TEG was in 1948 by Hellmut Hartert where it was used to demonstrate hemostatic function for whole blood samples.^62,63 Since then, the use of TEG within the clinical setting has continued to grow.^62,64–68

Thromboelastography is a method to graph and evaluate the entire process of blood coagulation from the beginning of clot formation to fibrinolysis.^69,70 By measuring the viscoelastic properties of the blood clot, TEG has the advantage of providing information related to the cumulative effects of plasma clotting factors, platelets, leukocytes, and red cells during all stages of the coagulation and fibrinolytic processes.⁶⁹

The standard kaolin TEG is a point-of-care whole blood test whereby citrated blood is placed in a warmed cup into which a pin descends. As the cup rotates and the clot forms, the connection between the rotating cup and the pin induces a torque on the pin, which is registered by a transducer that provides a characteristic curve that looks like a shovel. The TEG setup is illustrated on the right-hand side of ►Fig. 4A. A normal or physiologic TEG tracing (►Fig. 4A) resembles a wide flat (non-functional) “shovel” with a short handle. As enumerated in ►Table 1, the TEG parameters are abbreviated as follows: reaction time (R) for the transducer to be displaced 2 mm, clot formation/kinetics (K) or time to accomplish a specific level of clot strength, α-angle or rate of clot formation, maximum amplitude (MA) or overall clot strength, and lysis at 30 minutes (LY30) or percent decrease in amplitude 30 minutes after achieving MA.⁷¹

Fig. 4.

(A) Depiction of a TEG.²⁸ TEG and ROTEM use independently labeled parameters that are equivalent. Reaction time (R) is the clotting time to cause a 2-mmdisplacement during initial clot formation. Kinetics (K) is the clot formation time. α-Angle is the rate of clot formation. Maximum amplitude (MA) is the maximum clot strength. Lysis at 30 minutes (LY30) is the percent decrease in amplitude 30minutes after achieving MA. As illustrated on the right-hand side, citrated blood is placed in a warmed cup into which a pin descends. As the cup rotates and the clot forms, the connection between the rotating cup and the pin induces torque on the pin, which is registered by a transducer as a characteristic shovel-shaped curve. Abbreviations: TEG, thromboelastography; ROTEM, rotational thromboelastometry; R, reaction time; K, kinetics; α-angle, rate of clot formation; MA, maximal amplitude; LY30, lysis at 30 minutes. (B)Depiction of a ROTEM tracing.²⁸ TEG and ROTEM use independently labeled parameters that are equivalent. Clotting time (CT) is the reaction time to cause a 2-mm displacement during initial clot formation. Clot formation time (CFT) measures clot kinetics. α-Angle is the rate of clot formation. Maximum clot firmness (MCF) measures maximum clot strength. Lysis at 30 minutes (LI30) is the percent decrease in amplitude 30 minutes after MCF. As illustrated on the right-hand side, citrated blood is placed in a warmed cup with a pin. Unlike the TEG, the ROTEM pin revolves within a fixed cup. The resultant force of the clot on the pin is transduced and then traced as a characteristic shovel-shaped curve. TEG, thromboelastography; ROTEM, rotational thromboelastometry.

Table 1.

TEG/ROTEM normal parameters

Parameter	TEG	ROTEM
Clotting time (2 mm amplitude)	R (reaction time)Normal (citrate/kaolin = 3–8 min	CT (clotting time)Normal (EXTEMa) = 42–74 s Normal (INTEMb) = 137–246 s
Clot formation/kinetics (20 mm amplitude)	K (kinetics)Normal(citrate/kaolin) = 1–3 min	CFT (clot formation time) Normal (EXTEM) = 46–148 s Normal (INTEM) 40–100 s
Clot strengthening (angle of clot formation)	α-Angle (slope between R and k points) Normal(citrate/kaolin) = 55–78°	α-Angle (slope of tangent at 2 mm amplitude) Normal (EXTEM) = 63–81° Normal (INTEM) = 71–82°
Amplitude/maximal firmness	MA (maximal amplitude)Normal (citrate/kaolin) = 51–69 mm	MCF (maximum clot firmness)Normal (EXTEM) = 49–71 mm Normal (INTEM) = 52–72 mm Normal (FIBTEMc) = 9–25 mm A5, A10, etc. – amplitudes at dedicated time points predicting the final clot firmness
Lysis	LY30, CL30, CL60, CL	LI30, LI60, ML

Abbreviations: CFT, clot formation time; CT, clotting time; EXTEM, extrinsic activator thromboelastometry; FIBTEM, fibrin-based thromboelastometry; INTEM, intrinsic activator thromboelastometry; K, kinetics; LI, lysis; LY, lysis; MA, maximal amplitude; MCF, maximum clot firmness; R, reaction time; ROTEM, rotational thromboelastometry; TEG, thromboelastography; TF, tissue factor.

^aThe EXTEM uses TF as the activator and is a test of the extrinsic pathway.

^bThe INTEM uses ellagic acid and is a test of the intrinsic pathway.

^cFIBTEM uses both TF and platelet inhibition so it is a test of fibrin polymerization after inactivation of the platelets.

Rotational Thromboelastometry

Thromboelastography and ROTEM utilize identical principles with the exception that in the ROTEM the pin revolves within a fixed cup as illustrated on the right-hand side of ►Fig. 4B.⁷² The resulting ROTEM curve is identical to the TEG, but each assay uses different nomenclature (►Fig. 4B). As shown in ►Table 1, clot time (CT) corresponds to TEG’s R, clot formation time (CFT) corresponds to TEG’s K, maximum clot firmness (MCF) corresponds to TEG’s MA, and lysis after 30 minutes (LI30) corresponds to TEG’s LY30.^{64–68,72,73}

Tissue factor is used to initiate the rapid TEG for the TEG and the extrinsic activator thromboelastometry (EXTEM) for ROTEM in assessing the hypocoagulopathic patient while kaolin TEG is more commonly employed to assess hypercoagulability for the TEG.^{3,29,72,74,75}

Analysis of TEG or ROTEM

The authors suggest that the shapes of the TEG/ROTEM tracings can be analogized to the shape of a shovel. With hemostatic competence or normal hemostasis, when whole blood is neither hypocoagulable nor hypercoagulable, the shovel-shaped tracing has an ideal handle length, blade slope, blade width, and blade tip. This ideal shovel shape is shown in ►Fig. 5 as a black shovel. The shovel attributes correspond to TEG parameters as follows: the R is the handle, the α-angle is the slope of the blade, MA is the width of the blade, and LY30 is the point of the blade. The ends of the coagulation spectrum are represented by different shovels where the ease of tilling and moving the earth corresponds to the ease of moving blood. Continuing the analogy, the shovel tracing in hyperfibrinolysis has a long handle with a narrow and pointed blade (red shovel in ►Fig. 5) which makes it easy for the earth to be tilled and transported; similarly, in hyperfibrinolysis, the blood is easily transported because it is hypocoagulable. On the other end of the spectrum, the shovel tracing in a hypercoagulable state has a short handle and a wide blade with an absent tip (blue shovel in ►Fig. 5) which makes it very difficult for the earth to be broken up; by analogy, the blood in the hypercoagulable state has difficulty being broken up and is not easily transported. In summary, in ►Fig. 5, the black shovel represents hemostatic equilibrium, the red shovel represents hypocoagulable disequilibrium, and the blue shovel represents hypercoagulopathic disequilibrium.

Fig. 5.

Schematic of simplified TEG tracing.²⁸ The TEG shovel-shaped tracing in black, with a normal handle length and moderately wide shovel blade with slight narrowing of the tip of the blade, represents physiologic hemostasis with normal R, α-angle, MA, and LY30. The superimposed TEG shovel tracing in red depicts a tracing with a prolonged R, narrow α-angle, a narrow MA, and pointed and increased LY30; these parameters are indicative of a hypocoagulopathic/hyperfibrinolytic phenotype. The superimposed TEG shovel tracing in blue with a very short handle and wide blade with very little tapering of the tip of the blade depicts a tracing with a decreased R and a wide α-angle, MA, and very low LY30. The blue shovel tracing is indicative of a hypercoagulable/fibrinolytic shutdown phenotype. α-Angle, rate of clot formation; K, kinetics; LY30, lysis at 30 minutes; MA, maximal amplitude; R, reaction time; TEG, thromboelastography.

The hyperfibrinolytic TEG tracings at the bottom left and right corners of ►Fig. 3 demonstrate a prolonged reaction time (R or CT), decreased kinetics (K or CFT) or time to accomplish a specific level of clot formation, flat α-angle or rate of clot formation, small MA (MCF) or overall clot strength/firmness, and increased LY30 (LI30) or percent decrease in amplitude 30 minutes after achieving MA. These values indicate a systemic coagulopathy with hyperfibrinolysis.^64–68,73 In response to TEG/ROTEM values, appropriate therapy can be planned. Fresh frozen plasma or prothrombin complex concentrate is given for a prolonged R (CT), cryoprecipitate or fibrinogen concentrate is given for a flat α-angle, platelets are given for a small MA (MCF), and tranexamic acid (TXA) can be given for an increased LY30 (LI30).^{28,29,64–70,73,76}

Coagulation index is another way to interpret the data from TEG. It is the sum of the R, K, MA, and α-angle values. The resulting value represents an individual’s hemostatic balance and has been suggested as a way to assess hemorrhagic or thrombotic risk.⁷⁷

Efficacy of VETs to Define the Hypercoagulable State in Patients

Recently, it has been suggested that VETs such as TEG should replace CCTs for the guidance of blood component therapy (BCT) in the trauma population.⁷⁸ Whereas VETs evaluate whole blood samples, CCTs are plasma-based assays that assess only a part of the story.³ CCTs not only fail to identify deficiencies in coagulation factors, fibrinogen, and fibrin, but they also do not evaluate blood cellular contribution to clotting function and fibrinolysis.^78–82 VETs provide a rapid and accurate assessment of clot formation, stability, and firmness which allows for individualized treatment of patients with TIC.^3,29,78,83

Sepsis-induced coagulopathy and MAC have been determined by ROTEM.⁸⁴ Reductions in the MA, CFT, α-angles have been associated with increased mortality at 28 days in the critical care unit in septic patients with and withoutDIC.⁸⁵ The ROTEM parameters for hypocoagulability have demonstrated reduced fibrinolysis or fibrinolytic shutdown in patients with overt DIC defined by the International Society on Thrombosis and Haemostasis or Japanese Association for Acute Medicine criteria alone.⁸⁶ At the other end of the coagulopathy spectrum, VETs were a better indicator of the postinjury hypercoagulable state than CCTs.⁸⁷ Postoperative patients with and without cancer in the intensive care unit have been found to be at risk for venous thromboembolism and this risk can be predicted by ROTEM and TEG.^23,34 Specifically, the MA value of TEG has been found to be predictive of increased risk of deep vein thrombosis and pulmonary embolism in patients with major orthopedic trauma.³⁴

In patients with malignancy, both TEG and ROTEM parameters can predict hypercoagulable states. Specifically in patients with malignancy, TEG parameters have shown decreased R and K, and increased MA.^37,88,89 ROTEM parameters demonstrated hypercoagulability in patients with solid tumors by identifying the contribution of fibrinogen and platelets to clot strength.²⁴ MCF of the ROTEM tracings on all assays—the intrinsic, extrinsic, fibrin, and aprotinin activation thromboelastometry tests (INTEM [intrinsic activator thromboelastometry], EXTEM, FIBTEM [fibrin-based thromboelastometry], APTEM [aprotinin thromboelastometry])—has revealed a statistically significant shortening of CFT and increasing of MCF in cancer patients compared to routine laboratory tests.^23,24,90,91 Also, venous thromboembolism and arterial thromboembolismmay also precede the diagnosis of cancer by months or years and, therefore, ROTEM and TEG data may function as a marker for occult malignancy.^10,37,92,93

VETs and the Hypercoagulable State in the Noncancer Setting

Multiorgan failure and thrombosis result from severe inflammation and hemostatic derangement in the emergent and intensive care unit populations.^26,75,94,95 Recently, in an attempt to further categorize the hemostatic competence of critically ill patients, VETs have been proposed as an adjunctive point-of-care methodology to further delineate the hypercoagulable states of these patients.^{25,34,87,96–102}

The first studies of VETs in critically ill patients were performed by Starzl’s liver transplantation group in Colorado in the 1960s. They first identified hyperfibrinolysis during the anhepatic phase of liver transplantation and began using routine epsilon aminocaproic acid (EACA) to treat hyperfibrinolysis.^103,104 However, in their subsequent experience, this led to fatal pulmonaryemboli. This led to the recommendation of EACA for select patients with posttransplant hyperfibrinolysis as documented by TEG.^64,105 Another antifibrinolytic agent, aprotinin, was known to be associated with both venous thromboemboli and MOF following cardiac surgery.¹⁰⁶

Also, CCTs do not adequately predict thrombosis because of “rebalanced hemostasis” associated with chronic liver failure.¹⁰⁵ “Rebalanced hemostasis” in patients with chronic liver disease refers to deficits in the production of proteins C and S resulting in abnormalities in prohemostatic drivers that are then compensated for by commensurate changes in antihemostatic drivers. For example, the clot dissolution phase of hemostasis is defective in patients with cirrhosis due to diminished levels of profibrinolytic proteins and depressed levels of protein C from decreased hepatic biosynthesis. These patients maintain a fragile balance of hemostasis due to such deficiencies and compensation measures. Standard CCTs, such as INR, do not adequately assess the fragility of this state of balance, but VETs are much more informative. On the other side of the coagulation spectrum, VETs reveal the presence of hypercoagulability and a prothrombotic state and also provide information about the presence of endogenous heparinoids with vascular endothelial injury associated with liver failure, sepsis, or acute and chronic inflammation.^{105,107–109}

Since the early Starzl studies, extensive research has been conducted on the use of VETs in the acute setting. Moore et al have proposed that hypofibrinolysis in the postoperative period was the “missing link” that caused Starzl to use the TEG in order to identify those patients at greatest risk for post-liver transplant venous thromboembolism.⁷⁵ Recent studies have confirmed Starzl’s prescient utilization of the TEG by identifying a group of patients with severe trauma and shock who developed MOF and venous thromboembolism following resuscitation and TXA administration. TXA has been recently demonstrated in a large, retrospective study of more than 21,000 severely injured patients from the University of Pittsburgh to be associated with a significantly increased risk of thromboemboli with no survival benefit in patients with multiple trauma.¹¹⁰ In this study of 21,931 patients, 189 pairs werewellmatched across propensity score variables and had a median injury severity score of 19. TXA was not significantly associated with survival (adjusted odds ratio [aOR]: 0.86; 95% confidence interval [CI]: 0.23–3.25; p = 0.83) and was associated with more than threefold increase in the odds of venous thromboembolism (aOR: 3.3; 95% CI: 1.3–9.1; p = 0.02). Risk of venous thromboembolism remained elevated in the TXA cohort despite accounting for mortality. In other studies, patients have demonstrated a similar hypofibrinolysis in the postresuscitation and post-TXA period which has been called a fibrinolytic shutdown.^{30,31,110–112} As defined by Moore et al, a fibrinolytic shutdown occurs when a physiologic dysregulation of the plasminogen–plasmin system causes a relative resistance to tPA and thus, a hypercoagulable state.³¹ MAC exhibits a similar state with a decrease in tPA activity.¹¹³

The pathophysiologic causes that drive the hypercoagulability and fibrinolytic shutdown after trauma are a matter of speculation. In trauma, the early fibrinolytic response is believed to be the release of tPA from the vasculature as a response to hypotension and shock; this shock is absent in cancer.^75,114 Rather, in the absence of shock and with tissue injury such as malignancy, a hypercoagulable state exists. For example, tissue injury in the absence of shock in animal models promotes fibrinolysis resistance and a hypercoagulable state.^115,116 However, this translation to the human scenarios remains a challenge because most patients have variable combinations of tissue injury and shock after experiencing trauma. With a cancer patient, there is less incidence of acute shock and therefore a greater incidence of a hypercoagulopathic state. It is appropriate then to apply the knowledge learned from more than a half century use of VETs, such as in liver transplantation and trauma, to predict coagulopathy at both ends of the coagulation spectrum in the cancer patient.

Thromboelastography-identified postoperative hypercoagulability has been demonstrated following hepatic and intra-abdominal surgery, and neurosurgery.^117–120 In addition, TEG-identified hypercoagulability has been assessed in coronary heart disease, dialysis patients, cut-down intravenous catheters, oral contraceptives, and burn patients.^{102,121–124} And, TEG-identified hypercoagulability has been used in major orthopedic trauma injuries to predict venous thromboemboli during hospitalization.³⁴

VETs and Specific Malignancies

Solid Tumors

Liver and Pancreaticobiliary Malignancies

Blasi et al and Moore et al have described a positive correlation Between ROTEM and TEG parameters and venous thromboembolism (►Table 2).^35,125 Specifically, cholangiocarcinoma has demonstrated correlation between increased CFT and MCF in patients with venous thromboembolism.¹²⁵ Based on their study, they recommended addition of ROTEM abnormalities to an expanded Khorana risk model for venous thromboembolism.¹²⁵ Moore et al noted increased α-angle and low LY30 in 11% of patients with pancreatic neoplasms who developed venous thromboembolism.³⁵ De Pietri et al noticed prolonged PT/INR despite normal TEGs and a reduction in antithrombin and platelet count which reflected the rebalanced hemostasis in patients with liver cancer.¹²⁶

Table 2.

Solid tumors

Study	Cancer/number	Test	Findings
Blasi et al, 2018125	Cholangiocarcinoma	ROTEM	Recommended addition of ROTEM to expanded Khorana risk model for VTE after identifying ↑CFT, ↑MCF, and a 22% incidence of VTE.
	Patients: 27
Scărlătescu et al, 201884	Solid tumors/sepsis	ROTEM	Organ dysfunction associated with reduced fibrinolysis. Standard coagulation tests and plasmatic indicators of fibrinolysis are inaccurate predictors of bleeding.
	Patients: 35
Moore et al, 201835	Pancreatic neoplasms	Native TEG	Pulmonary embolism with ↓ R, ↑α-angle, and ↓ LY30. 11% incidence of VTE.
	Patients: 100
Jansohn et al, 2017133	Glioma	ROTEM	↓ CT with concomitant local fibrinolysis in tissue-spiked tumor samples.
	Patients: 21
Wang and Ye, 201689	Nonsmall cell lung cancer	TEG	↓ R, ↓ K, ↑MA, ↑α-angle, ↑fibrinogen, and ↑d-dimer in 15 of 48 lung cancer acute cerebral infarction patients.
	Patients: 48
Wikner et al, 2015129	Free flap in head/neck	ROTEM	Heparin guided by ROTEM was noninferior to aPTT. ROTEM may augment traditional coagulation tests in microvascular tissue surgery.
	Patients: 25
Toukh et al, 201437	Prostate cancer	TEG	68% patients were hypercoagulable based on ↓ R, ↑α-angle, ↑MA, and ↑CI, including ↑microparticles.
	Patients: 32
Thorson et a, 201491	Intra-abdominal malignancies	ROTEM	In resectable malignancies, ↓ CFT, ↑α-angle, and ↑MCF in pancreatic compared to esophageal/liver cancers, and in all cancers compared to benign disease. Pancreatic neoplasms have greatest risk for hypercoagulability.
	Patients: 72
Hincker et al, 201423	Solid tumor surgery	ROTEM	Preoperative EXTEM and INTEM indicated patients (147/313 or 47% had cancer) had significantly ↓ CFT, ↑α-angle, and ↑MCF associated with increased thromboembolic complications.
	Patients: 147
Akay et al, 2014132	Gynecologic malignancies	ROTEM	Hetastarch causes hypocoagulopathic state in patients with gynecologic cancers as shown by ↑CT, ↑CFT, and ↓ MCF. HES may reduce VTE incidence in gynecologic cancer surgery.
	Patients: 22
Thorson et al, 201390	Intra-abdominal malignancies	ROTEM	Postsurgical hypercoagulability with ↓ CFT, ↑α-angle, and ↑MCF within 1 wk of surgery. Patients were hypercoagulable at 3–4 wk. Supports postdischarge thromboprophylaxis in high-risk cancer patients.
	Patients: 35
Attaran et al, 2010131	Lung tumor	TEG	Advocated TEG use in patients with hypercoagulable states to monitor LMWH.
	Patients: 60
De Pietri et al, 2010126	Liver and pancreatic cancer	TEG	In postoperative period of liver patients, TEG parameters remained normal while PT/INR remained prolonged despite ↓ antithrombin III, ↓ platelets, normal aPTT, and normal fibrinogen. Pancreatic patients revealed transient ↑R and ↑K.
	Patients: 56
Liang et al, 201088	Colon cancer Patients: 35	TEG	TEG parameters in patients with colon cancer indicated ↓ R, ↓ K, and ↑MA. TEG parameters indicated that presurgical loading with HES inhibited platelet activation.
Akay et al, 200924	Solid tumors	ROTEM	Hypercoagulability reflected by ROTEM ↓ CFT and ↑MCF. Contribution of fibrinogen and platelets to clot strength was determined by ROTEM.
	Patients: 78
Papa et al, 2007127	Digestive tract cancers	ROTEM	↑Maximum velocity and ↓ time to maximum velocity reflect hypercoagulability.
	Patients: 50
Wen et al, 1997130	Solid tumors	TEG	Hypercoagulable states in cancer patients do not have direct contribution as determined by TEG.
	Patients: 76
Goh et al, 1997118	Brain tumors	TEG	In postoperative patients without hematoma, TEG demonstrated ↓ K, ↑α-angle, and ↑MA. In patients with hematoma, there was increased fibrinolysis and ↑R.
	Patients: 50
Francis et al, 1994128	Solid tumors	TEG	TEG demonstrated ↓ R in patients with colon cancer, breast cancer, and miscellaneous cancers. Patients with breast cancer and colon cancer also had ↑MA and ↑α-angle. Sonoclot Analyzer demonstrated similar coagulopathy.
	Patients: 17

Abbreviations: aPTT, activated partial thromboplastin time; CFT, clot formation time; CI, coagulation index; CT, clotting time; EXTEM, extrinsic activator thromboelastometry; HES, hetastarch; INR, international normalized ratio; INTEM, intrinsic activator thromboelastometry; K, kinetics; LI, lysis; LMWH, low molecular weight heparin; LY, lysis; MA, maximal amplitude; MCF, maximum clot firmness; PT, prothrombin time; R, reaction time; ROTEM, rotational thromboelastometry; TEG, thromboelastography; VTE, venous thromboembolism.

Intra-abdominal Malignancies

In 2013, Thorson et al described postoperative abnormalities of the ROTEM which predicted a hypercoagulable state at 1 and 4 weeks postoperatively in patients with intra-abdominal malignancies.⁹⁰ At 1 week, 86% had more rapid CFT, high α-angles, and elevated MCF; all of these patients remained hypercoagulable at 4 weeks. In total, 81% of patients with abnormal ROTEM values demonstrating hypercoagulability had normal coagulation profiles preoperatively. Thorson et al also published a follow-up study which confirmed their previous study showing the following hypercoagulable parameters associated with intra-abdominal malignancies: low CFT, high α-angle, and elevated MCF.⁹¹ They noted that pancreatic neoplasms have the greatest risk for hypercoagulability, and that pancreas, esophageal, and liver cancers had more hypercoagulability measured by ROTEM than benign diseases. Papa et al similarly described a shortened time to maximum velocity (i.e., low CFT) in patients with “digestive tract” cancers.¹²⁷ Specifically for colon cancer patients, Francis et al and Liang et al noted that TEG parameters demonstrated a hypercoagulable state with decreased R and an increased MA.^88,128 To address the hypercoagulable state and increased platelet activation, Liang et al found that presurgical loading with HES 200/0.5 inhibited platelet activation as evidenced by TEG parameter changes.⁸⁸

Head and Neck Malignancies

Wikner et al observed that ROTEM was noninferior to aPTT in guiding heparin use in patients with free flaps following head and neck cancer surgery.¹²⁹ Once it becomes standardized, ROTEM may augment traditional coagulation tests in microvascular tissue surgery.

Solid Tumors (Not Otherwise Specified)

Scărlătescu et al used ROTEM to associate fibrinolytic shutdown with MOF.⁸⁴ In those cases, standard coagulation tests and plasma indicators of fibrinolysis were inaccurate indicators of bleeding and thrombosis. Hincker et al described patients undergoing noncardiac surgery including 147 patients (147/313 or 47% of patients) with cancer. In all patients with thromboembolic complications, preoperative EXTEM and INTEM revealed reduced CFT, elevated α-angle, and elevated MCF in patients with thromboembolic complications.²³ There was no difference using FIBTEM, which excludes platelets. In 2009, Akay et al noted significant correlation between ROTEM parameters and hypercoagulability including decreased CFT and increased MCF.²⁴ The strongest correlations were in the INTEM, EXTEM, FIBTEM, and APTEM for CFT, MCF, and α-angle. Similarly in 1994, Francis et al noted hypercoagulable TEGs in patients with solid tumors as evidenced by a short R and an elevated MA.¹²⁸ In comparison, in 1997Wen et al found no correlation between TEG parameters and thrombosis of implantable venous catheters in patients with solid tumors.¹³⁰

Lung Cancer

Wang et al in 2016 noted that patients with nonsmall cell carcinoma of the lung have hypercoagulability as demonstrated by increased MA, α-angle, and fibrinogen.⁸⁹ Attaran et al advocated the use of TEG in patients with hypercoagulable states to monitor low molecular weight heparin (LMWH) thromboprophylaxis.¹³¹

Gynecologic Malignancies

In 2014, Akay et al performed a similar study of gynecologic malignancies and demonstrated that hetastarch caused a hypocoagulable state in patients with gynecologic cancers.¹³² Using ROTEM, they identified an increased CT and CFT, and a reduced MCF compared to controls. This study was done to demonstrate that hetastarch could potentially reduce the incidence of venous thromboembolism in gynecologic surgery.

Prostate Malignancies

Toukh et al in 2014 noted that 68% of patients with prostate malignancies showed hypercoagulable TEG tracings as demonstrated by short R, elevated α-angle, increased MA, and increased coagulation index.³⁷ They also noticed increased microparticles in the patients who were hypercoagulable.

Central Nervous System Malignancies

In one study, Jansohn et al noted decreased clotting times with maintained local fibrinolysis in tissue-spiked tumor samples of patientswithglioblastomas.¹³³ In another study by Gohet al of patients with primary brain tumors, TEG tracings demonstrated both hyper- and hypocoagulable patients in the pre- and postoperative periods based on changes in R, α-angle, MA, and LY30.¹¹⁸ Specifically, postoperative patients without hematoma had decreased K, increased α-angle, and increased MA. Comparatively, patients with hematoma demonstrated a fibrinolysis phenotype with an increased R.

Hematopoietic Malignancies

As noted above, a spectrum of coagulopathy from the hypocoagulopathic and hyperfibrinolytic to the hypercoagulopathic and hypofibrinolytic states exists in MAC, much as it does in TIC (►Table 3). We have seen that the paradigm for hypocoagulability and hyperfibrinolytic hematologic malignancy is observed in APL. Patients with this form of leukemia are often hypocoagulable at presentation and the hypocoagulability progressed with chemotherapy in the pre-ATRA era; these patients’ coagulopathy is mostly abated by 7 to 10 days with ATRA treatment.^{53,54,56–58,134} To date there are no studies of VETs in this group of patients. However, there have been reported cases of the utilization of TEG/ROTEM to predict thromboembolic events and hypercoagulability in patients with hematologic malignancies.

Table 3.

Hematopoietic malignancies

Study	Cancer/number	Test	Findings
Crowley et al, 2017137	Multiple myeloma	TEG	↓ K and ↑α-angle after induction. Hypercoagulability markers directly proportional to myeloma burden.
	Patients: 16
Giaccherini et al, 201645	Myeloproliferative neoplasms Patients: 39	ROTEM	↓ CFT and ↑MCF. MCF values indicated lower platelet reactivity in MPN patients.
Gracheva et al, 2015135	Multiple myeloma	TEG	Thrombin generation test and TEG parameters (R, K, and MA) reveal hypercoagulability in patients with multiple myeloma.
	Patients: 20
Crowley et al, 2015138	Myeloma/MGUS Patients: 24	TEG	No difference in TEG parameters between multiple myeloma and MGUS.
Ko et al, 2015136	Acute lymphoblastic leukemia	TEG	↓ R and ↑MA during induction.
	Patients: 80
Akay et al, 2015140	Hematologic malignancies	ROTEM	No difference in TEG parameters in low-dose or high-dose random or single-donor platelet transfusions.
	Patients: 100
Flisberg et al, 2009139	Thrombocytopenia due to hematologic malignancy	ROTEM	Platelet transfusion in patients with hematologic malignancies is associated with ↓ CFT and ↑MCF.
	Patients: 20

Abbreviations: CFT, clot formation time; K, kinetics; MA, maximal amplitude; MCF, maximum clot firmness; MGUS, monoclonal gammopathy of undetermined significance; MPN, myeloproliferative neoplasm; R, reaction time; ROTEM, rotational thromboelastometry; TEG, thromboelastography.

Giaccherini et al found a hypercoagulable profile in patients with myeloproliferative neoplasms and ROTEM analysis demonstrated increased MCF and shorter CFT in this population.⁴⁵ When MCF values were corrected for platelet counts, the parameters supported the hypothesis that platelet exhaustion occurred from clotting activation. Gracheva et al in 2015 described 20 patients utilizing TEG (specifically the R, K, and MA parameters), thrombin generation test, and thrombodynamics which revealed hypercoagulability in patients with multiple myeloma.¹³⁵ Ko et al in 2017 demonstrated a short R and an increased MA in 80 patients with acute lymphoblastic leukemia during induction.¹³⁶ Crowley et al in 2017 noted borderline shortened K and increased α-angle of the TEG in patients with multiple myeloma that reflected the burden of disease in the postinduction period.¹³⁷ However, previously in 2015 Crowley et al found no difference in TEG parameters between multiple myeloma, monoclonal gammopathy of undetermined significance, and control.¹³⁸

Flisberg et al in 2009 demonstrated that patients with hematologic malignancies who were given platelets prior to central venous port placement responded with shortened CFT and increased MCF.¹³⁹ The implication was that TEG measurements are a useful parameter for the necessity of platelet transfusion in patients with hematologic malignancies needing this procedure. Akay et al noted in 2015 that ROTEM measurements had no statistical difference between low-dose and high-dose, random or single-donor platelet transfusion in patients with hematologic malignancies.¹⁴⁰ Therefore, they proposed that based on ROTEM, low-dose platelet transfusion should be considered because of its reduced likelihood of transfusion reactions and economic burden.

The Spectrum of MAC

The recent application of the description of the spectrum of fibrinolytic phenotypes in trauma may allow adaptation of this phenotype classification to patients with MAC. The trauma patient can evolve in seconds to minutes from one phenotype to another in what has been called “phenotype switching.”^114,141 Patients who remain in a prolonged state of hypofibrinolysis or fibrinolytic shutdown are at increased risk for venous thromboembolism and MOF.³⁰ In posttrauma resuscitation hypercoagulability, this shutdown occurs in the days following trauma and is comparable to the hypercoagulability of cancer that evolves over months to years.

The importance of the fibrinolytic shutdown in the pathophysiology of trauma and cancer-related hypercoagulability has long been a matter of intense investigation. The parameters for the definition of fibrinolytic shutdown as determined by VETs and used to guide BCT in bleeding patients recently have been refined and are matters of continued investigation.^25,142–147

As has been studied in trauma and liver transplant patients, it would be reasonable to infer that a similar fibrinolytic spectrum exists in cancer patients; there are theoretical and clinical studies that confirm this inference. In fact, there has been a significant body of work regarding the fibrinolytic system and cancer. This work has demonstrated a similar event to fibrinolytic shutdown occurring in the 24-hour postoperative period following cancer surgeries. For example, Galster et al and others have noted that differences in the activation of coagulation depend on the cancer type, which suggests that this process is mainly related to procoagulants released from the neoplastic cells.^148,149 This clinical observation has fueled research into the fibrinolytic spectrum including fibrinolytic shutdown, whether in trauma or cancer, as a reflection of the nascent state of hypercoagulability.^{35,75,150–152}

We have seen that there are similarities between the mechanisms of coagulopathies of TIC and MAC. While TIC evolves over a much more rapid period and has fewer variables causing the hemostatic derangements than MAC, the basic influences of endothelial dysfunction, the TF effects on the TM–thrombin activated protein C pathway, and the balances between PAI-1, antifibrinolytics, and anticoagulants are similar between TIC and MAC. In both TIC and MAC, these changes result in an eventual hypercoagulable state for trauma patients who survive the initial insult and for cancer patients who do not succumb to initial hypocoagulable complications. The switch from bleeding due to thrombocytopenia, DIC, or immediate fibrinolysis to a hypercoagulable state has been documented in successfully treated APL patients.^{53,54,56–59,134}

In trauma and in malignancy, PAI-1 is one of the main factors responsible for hypercoagulability, increased rates of venous thromboemboli, and MOF.¹¹² It has been shown that elevated levels of PAI-1 are associated with increased risk of venous thromboembolism in orthopedic surgery, and in particular, in patients with acute hip fractures who have been given TXA.^153,154 Specific ratios of PAIs to activators may shut down fibrinolysis after certain surgeries. It has been noted that following postoperative hip replacement, endothelial cells cultured in the plasma of these patients have high PAI-1 levels.^154,155 Antifibrinolytics are not recommended in the treatment of acute hip fractures since an increased rate of venous thromboembolism has been noted in those patients treated with TXA confirming the above-mentioned recent large study demonstrating increased risk of thromboemboli without mortality benefits in patients given TXA.^110,153,155 In MAC, elevation of PAI-1 and reduced levels of tPA and antithrombin have been noted and associated with a hypofibrinolysis phenotype.^38,47,113 The use of more modern assays revealed similar elevation of PAI-1 and other natural antifibrinolytics, as well as reductions in fibrinolytic proteins.^38,156

Potential use of Precision-Based Medicine in MAC and TIC

Large randomized controlled trials regarding the empiric management of bleeding trauma and obstetrical patients have been met with differing levels of acceptances as the histories of the CRASH-2 and WOMAN trials have shown.^27,157–163 Because of the unintended consequence of “one-size-fits-all” randomized controlled trial philosophy touted for the advancement of evidence-based clinical science, it is more difficult to engage in smaller proof-of-concept studies that evaluate the influence of specific pathophysiological derangements and phenotypes on clinical outcomes.^164,165

The concept of precision-based medicine has been developed to allow for the integration of epidemiologic findings from randomized controlled trials with select patient-group-specific clinical findings.^165,166 For example, the field of proteomics can help identify patients with specific fibrinolytic phenotypes and these subsets of patients can be studied in a focused manner with attention to precision-based medicine.^158,167

Conclusion

Virchow’s prescient anticipation of the cell-based theory of hemostasis in 1855 with the words: “Omnis cellulae cellula” [All cells from pre-existing cells], sets the stage for the next phase in delineating hemostatic competence of patients along the spectra of coagulopathies.^1,2 Currently, the VETs, TEG, and ROTEM are the most commonly used whole-blood assays that avail themselves of the cell-based theory of hemostasis in order to predict bleeding and coagulation, and to guide BCT in cancer patients. However, these point-of-care tests have barriers because there is no universal system that is agreed upon resulting in interobserver variability and institution-specific procedures. The TEG and ROTEM parameters are not directly interchangeable and separate expertise is required in order to utilize each test.¹⁶⁸ These tests are not yet at the level of reproducibility that the plasma-based tests are and therefore, these tests are in their relative infancy in the areas of trauma resuscitation and assessment of MAC.¹⁶⁹ Yet, as seen in ►Tables 2 and 3, there is substantial literature that describes the utilization of the TEG and ROTEM to guide decisions regarding coagulopathic patients with MAC. The concern regarding interobserver reproducibility and institutional variability has now been addressed with cartridge system VETs such as the TEG6s and the ROTEM Sigma. These cartridge VETs facilitate a more standardized performance of assays that detect the lifespan of the clot from initiation through amplification, propagation, and termination via fibrinolysis. These cartridge VETs also provide a reproducible quantification of the efficacy of anti-Xa and direct thrombin inhibitors. These new cartridge VETs such as the TEG6s and the ROTEM Sigma may hold the future to a more vigorous and refined utilization of VETs to define and guide therapy and prevention of thrombosis in cancer patients.^170,171 Moreover, recent development of multichannel microfluidic assays that include endothelium may allow us soon to advance confidence in quantifying hemostasis based on plasma, platelets, and endothelium. Thus, emerging endothelium-based assays may be used to further refine the assessment of hemostatic integrity, not just in trauma patients, but also in patients with hypercoagulopathic states in cancer.^172–174 The future of assaying coagulopathy in cancer, but also in all areas of surgery, transplantation, and trauma, remains an exciting, open field that promises to bear fruit with the expansion of these whole-blood-based technologies. The history of the utilization of VETs to guide BCT and to predict venous and arterial thrombosis in transplant patients and in trauma patients has formed a pathophysiologic foundation that will usher in an era where VETs may be used in order to risk stratify patients with MAC. This new era will allow for more refined prevention and treatment of thrombosis in cancer patients. Guidelines concerning the dosage and timing of administration of venous thromboembolism prophylaxis in cancer patients are not clear. Recently, direct oral anticoagulants (DOACs) have been shown to be noninferior to LMWH in the treatment of venous thromboembolism from MAC.¹⁷⁵ Because TEG has been validated in monitoring the dosage of LMWH and may be validated to monitor DOACs, the role of TEG as a resource for guiding prevention and treatment of MAC in cancer patients is likely to continue to evolve.^131,176,177