The Role of TEG and ROTEM in Damage Control Resuscitation.

ABSTRACT: Trauma-induced coagulopathy is associated with very high mortality, and hemorrhage remains the leading preventable cause of death after injury. Directed methods to combat coagulopathy and attain hemostasis are needed. The available literature regarding viscoelastic testing, including thrombelastography (TEG) and rotational thromboelastometry (ROTEM), was reviewed to provide clinically relevant guidance for emergency resuscitation. These tests predict massive transfusion and developing coagulopathy earlier than conventional coagulation testing, within 15 min using rapid testing. They can guide resuscitation after trauma, as well. TEG and ROTEM direct early transfusion of fresh frozen plasma when clinical gestalt has not activated a massive transfusion protocol. Reaction time and clotting time via these tests can also detect clinically significant levels of direct oral anticoagulants. Slowed clot kinetics suggest the need for transfusion of fibrinogen via concentrates or cryoprecipitate. Lowered clot strength can be corrected with platelets and fibrinogen. Finally, viscoelastic tests identify fibrinolysis, a finding associated with significantly increased mortality yet one that no conventional coagulation test can reliably detect. Using these parameters, guided resuscitation begins within minutes of a patient's arrival. A growing body of evidence suggests this approach may improve survival while reducing volumes of blood products transfused.

INTRODUCTION

Hemorrhage is the leading preventable cause of trauma-related death. Despite advances in care through the last four decades, mortality rates remain consistently high (1–5). Incompressible hemorrhage, anticoagulants, and other coagulopathies can add complexity to an already challenging problem for any individual patient. A particularly deadly coagulopathy directly related to traumatic hemorrhage has been titled “trauma-induced coagulopathy” (TIC), “acute traumatic coagulopathy,” “acute coagulopathy of trauma,” “acute coagulopathy of trauma-shock,” or “early coagulopathy of trauma” (6–9). These various terms for coagulopathy following major trauma all portend death. Present in a third of both civilian and military trauma patients, TIC corresponds to nearly 50% mortality rates (6–8, 10).

Expedient control of ongoing hemorrhage in its myriad forms remains beyond the scope of a single review. Rather, combating TIC with early recognition, identification of compounding factors, and appropriate resuscitation will be discussed here. The basic tenets of damage control resuscitation are transfusion of whole blood (or product ratios approximating whole blood), limited crystalloid use, and permissive hypotension. Beyond this, achieving optimal outcomes in bleeding patients benefits from viscoelastic coagulation testing, which is the focus of this article. Both thrombelastography (TEG) (Haemonetics Corporation, Braintree, Mass) and rotational thromboelastometry (ROTEM) (TEM International GmbH, Munich, Germany) will be reviewed with attention to their clinical uses in the initial resuscitation of the severely injured patient. Topics of review will also include history of viscoelastic testing, interpretation of its results, and guided versus fixed-ratio resuscitation.

METHODS

A systematic search was undertaken to find peer-reviewed literature describing TEG and ROTEM employed for resuscitation of trauma patients. Specific search terms in varying Boolean strings were “thrombelastography,” “TEG,” “rotational thromboelastometry,” “ROTEM,” “viscoelastic,” “trauma” or “traumatic,” “shock,” “injury,” “hemorrhage” or “hemorrhagic,” and “resuscitation.” A combination of PubMed, Ovid MEDLINE, and Google Scholar database searches was employed, and the end date limit was September 15, 2019. Patent descriptions, animal studies, non-clinical in vitro, and elective surgical studies were excluded. Only full-text articles available in the English language were reviewed. Additionally, product manuals and technical information sheets were requested from manufacturers of TEG and ROTEM testing equipment.

RESULTS

An initial screening search of titles, abstracts, and keyword searches for “‘thromboelastography,’ ‘thrombelastography,’ or ‘rotational thromboelastometry,”’ their abbreviations, or “viscoelastic” yielded 4,370 results. Eligibility searches of those results revealed 883 individual available texts, after removal of duplicates, in the English language with keywords, titles, or abstracts including the terms “trauma/traumatic,” “shock,” “injury,” “hemorrhage/hemorrhagic,” or “resuscitation” with truncation wildcards where appropriate. Full text search followed. Of these texts, 176 described animal studies. A further 371 were excluded as primarily elective surgical studies or purely in vitro/ex vivo studies. The search, therefore, yielded 336 studies comprising the available peer-reviewed literature of TEG and ROTEM in clinical trauma settings on which to base this narrative review (Fig. 1). This body of literature was categorized and summarized by topic heading.

Fig. 1

Flowchart of the review design and results of literature search.

Coagulopathy following injury

Bleeding begets more bleeding. After injury and major hemorrhage, multiple mechanisms contribute to ongoing hemorrhage, beyond the classic triad of acidosis, hypothermia, and coagulopathy. Acidosis, hypothermia, endothelial injury, hypocalcemia, dilutional coagulopathy, pre-injury medication-related coagulopathy, and TIC can act synergistically and often appear in combinations (11, 12). Most of the enzymes critical to the coagulation cascade, as well as platelets, function most efficiently at physiologic to slightly alkalotic pH (13). Hypothermia, besides its direct effect on decreasing protease and platelet activity (14, 15), negatively alters calcium ion receptor binding in calcium-dependent pro-coagulants (16). Component blood transfusion, the most common approach to replace blood volume and combat coagulopathy, carries its own deleterious effects including diluted coagulation factors, calcium chelation, and immunoinflammatory effects (17–22).

TIC is defined as an independent contributor to hemorrhage after trauma via activated protein C (aPC) (23–25), thrombomodulin (23, 26), plasminogen activator inhibitor-1 (PAI-1), tissue plasminogen activator (tPA), hyperfibrinolysis (27, 28), qualitative platelet dysfunction (29), and release of thrombogenic microparticles (30). In other words, correction of acidosis, hypothermia, and other contributors to coagulopathy assist with hemorrhage control, but TIC can persist despite their control. Moreover, TIC is apparent early and is highly lethal. In the civilian setting, at least a quarter of trauma patients arriving from the scene are coagulopathic and the presence of coagulopathy increases mortality 4-fold (23). Among military casualties in Iraq and Afghanistan, Niles et al. noted that almost 40% presented with evidence of TIC. Those with TIC on admission had a 6-fold higher mortality than their non-coagulopathic cohorts (10).

Because of its lethality, a critical point in combating TIC becomes apparent: it must be predicted or identified early. Several retrospective studies from the civilian arena identified that the median time to death from hemorrhage is within 3 h of injury (31, 32). This time to achieving hemostasis and hemorrhage control, or not, was the same regardless of penetrating or blunt mechanism of injury. In a randomized trial of civilian transfusion ratios, the time to death or hemostasis was the same at 2.6 h after injury (31). The inflection point beyond which bleeding patients suffer significantly worse outcomes has been repeatedly under 3 h in multiple large studies (33, 34). Even brief delays in acting upon hemorrhage carry significant mortality. In a multicenter study including 12 Level I US trauma centers, every minute in delay of activation of a massive transfusion protocol was associated with a 5% incremental increase in mortality (35).

The detection of TIC, therefore, depends on rapid or point of care testing. Even experienced clinicians predict poorly the need for massive transfusion based on judgment alone (36), and clinical scoring systems for predicting transfusion requirements are not designed to predict coagulopathy. Initially, TIC was often defined using conventional coagulation testing (CCT) such as prothrombin time (PT) > 18 s (6, 37), international normalized ratio (INR) > 1.5–1.6 (38–40), active partial thromboplastin time (PTT) > 60 s (6, 37), and decreased platelet (typically < 50,000–100,000/μL) and fibrinogen counts (such as < 100 mg/dL) (39, 41, 42). Unfortunately, CCT carries several disadvantages making it less than ideal for diagnosing TIC. Tests are run on spun-down plasma samples (rather than whole blood specimens), focusing on single parts of the complex clotting cascade. The tests themselves were never designed to diagnose coagulopathy in trauma, but rather, were developed to follow isolated lesions seen in disease such as hemophilia. Moreover, PT and PTT tests are concluded when approximately 5% of the total available thrombin is generated, providing an incomplete representation of fibrin formation (43). Fibrinogen counts have no ability to account for functional versus nonfunctional proteins. CCT also take time: as an example, a multicenter prospective study including actively bleeding patients during major surgical procedures reported a median turnaround time of 88 min, with a range that extended up to 235 min (44).

This is precisely where the advantages of viscoelastic testing have been shown to identify TIC with increased sensitivity and decreased time to interpretation (45–47). In a large prospective cohort, rapid TEG predicted substantial bleeding and red blood cell transfusion requirements better than PT or PTT, α-angle was superior to fibrinogen for predicting plasma transfusion, and maximum amplitude (MA) was superior to platelet count for predicting platelet transfusion (48). Multiple other authors have validated the utility of TEG and ROTEM for predicting coagulopathy and massive transfusion requirement in a wide variety of civilian and military settings (49–60). Most importantly, the figures of clot activation, kinetics, and strength are available within 15 min using rapid testing (61).

History, clot parameter generation, and interpretation of viscoelastic testing

Hartert (62) first reported a method to evaluate whole blood coagulation in 1948. This early version of TEG was relegated to research purposes only, until clinical use was developed for liver transplantation in the 1960s (63). Hyperfibrinolysis after implantation of the donor liver was accurately identified by TEG, a feat not achievable by CCT. It should be noted that viscoelastic testing still provides early detection of fibrinolysis better than CCT, one of the persisting justifications of its use (64, 65). In the 1980s, TEG showed promise in reducing individual blood component transfusion in both liver transplant and later cardiac surgery (66). Kaufmann et al. (67) reported TEG's predictive use in blunt trauma in 1997, one of the first descriptions of viscoelastic testing following trauma. Although a small series, hypocoagulable results predicted early transfusion.

TEG has become the primary version of viscoelastic testing in North America, while ROTEM is seen more commonly in Europe. While they are similar in concept, they rely on different reagents and are not equivalent (68–72). Both provide information on clot initiation, progression, strength, and lysis from whole blood samples run at 37°C in a plastic cylindrical cup with a central wire or pin suspended in the sample. In TEG, the cup rotates while the central pin senses torque via an electromagnetic inducer as the sample coagulates and eventually lyses to varying degrees. In ROTEM, the pin is the rotating piece.

A conventional TEG tracing from a TEG 5000 machine plots time on the X axis with resistance to oscillation on the Y axis, resulting in a symmetric set of diverging curves as the cup rotates clockwise then counterclockwise. Samples are typically run in paired chambers to minimize measurement errors. Citrated blood samples are first reversed with calcium. Kaolin (a type of clay primarily containing hydrated aluminum silicate) and cephalins (a class of phospholipids) start the clotting process. The entire process can be accelerated by adding both kaolin and tissue factor, a process trademarked as Rapid TEG, which reveals early parameters of coagulation initiation in under 5 min. Additional subtypes of TEG include PlateletMapping, which describes the inhibitory effects of antiplatelet agents, and hTEG, in which heparinase is added to the sample. A functional fibrinogren assay is also available (73).

ROTEM produces five tracings, many of which correspond to TEG subtypes: intrinsic activation (INTEM), extrinsic activation (EXTEM), fibrin component of the clot (FIBTEM), aprotinin addition (APTEM) to show the effect of fibrinolysis inhibition, and heparinase addition (HEPTEM) to remove the effect of heparin on the sample (74). The similarities in interpretation of TEG and EXTEM tracings are shown in Figure 2. Normal values for each component of the tests are shown in Table 1, along with the primary contributors of the coagulation and clot formation web reflected in the individual components of the tracings.

Fig. 2

Comparison of TEG and ROTEM tracings to demonstrate their similarities. TEG indicates thrombelastography; ROTEM, rotational thromboelastometry; R, reaction time; CT, clotting time; k, k-time or kinetics; CFT, clot formation time; α, alpha-angle; MA, maximum amplitude; MCF, maximum clot firmness; LY30, lysis at 30 min; CL30, clot lysis at 30 min.

Table 1

Viscoelastic testing parameters and their reference ranges for citrated samples according to manufacturer labeling

Kaolin TEG	ROTEM EXTEM	Description	Main contributor
R: 5–10 min rTEG ACT: 86–118 s	CT: 43–82 s	Activation: time from start of test to first detectable clot formation, defined as an amplitude of 2 mm	Coagulation factors initiating thrombin generation
k: 1–3 min rTEG k: 34–138 s	CFT: 48–127 s	Amplification: time to 20 mm clot strength due to fibrin deposition and cross-linking	Concentration of fibrinogen and activation by thrombin
α: 53–72° rTEG α: 64–80°	α: 65–80°	Propagation: slope of tracing indicating speed of thrombin generation and fibrin deposition and cross-linking	Concentration of functioning fibrinogen, and to a lesser extent platelets
MA: 50–70 mm rTEG MA: 52–71 mm	MCF: 52–70 mm	Termination: maximal amplitude of clot strength	Platelet count and activity, and to a lesser extent fibrinogen
G: 4,500–11,000 kilodynes/cm2	MCE: no provided range (dimensionless value)	Calculated value (logarithmic computation based on MA and MCF)	MA and MCF
LY30: -2.3–5.77% rTEG LY30: < 7.5%	LY30 or CL30: < 18%	Fibrinolysis: percentage decrease in amplitude at 30 min after MA, indicating clot lysis	Speed of fibrinolysis (concentration of plasminogen and its activators)

α indicates alpha-angle; ACT, activated clotting time; CFT, clot formation time; CL30, clot lysis at 30 min; CT, clotting time; EXTEM, extrinsic thromboelastometry; G, G-value; k, k-time or kinetics; LY30, lysis at 30 min; MA, maximum amplitude; MCE, maximum clot elasticity; MCF, maximum clot firmness; R, reaction time; ROTEM, rotational thromboelastometry; rTEG, rapid thrombelastography; TEG, thrombelastography.

Correct interpretation of these laboratory tests relies on cognizance of their limitations as much as their strengths. While TEG and ROTEM are relatively rapid global assessments in comparison to CCT, they suffer from clinically significant shortcomings. Plastic cylinders do not mimic the complex endovascular environment of coagulation, a contributor to the coagulation cascade gaining increasing attention (75, 76). Hypothermia, especially core body temperature below 34°C, is common after major injury and significantly reduces platelet and factor activity (11, 77, 78). However, standard protocol runs the sample at 37°C, introducing mismatch between in vivo and in vitro coagulability profiles. Addition of reagents requires a high degree of precision in viscoelastic testing, introducing inter-technician and inter-facility variability (79). As a non-citrated whole blood sample must be run in under 4 min, citrated samples are more widely employed. Storage and stability time, pipetting of reagents, vibration artifacts, and frequent quality control checks are all sources of concern (73, 80–82). United Kingdom National External Quality Assessment Scheme (NEQAS) data have revealed coefficients of variances of 7.1% to 39.9% for TEG and 7.0% to 83.6% for ROTEM (81). It should also be noted that a newer generation of both TEG (6s, employing resonance-based measurements rather than rotation resistance) and ROTEM (Sigma) equipment is now available, both of which are cartridge-based. These more automated devices may reduce inter-test variability and increase correlation between the two technologies (83, 84). Finally, as with all laboratory testing, TEG and ROTEM require clinical judgment for safe and appropriate interpretation. Moreover, if you can “hear the bleeding,” the patient does not need viscoelastic testing to guide the need for blood products, suture, and a surgeon.

Employing viscoelastic testing to guide resuscitation

As discussed previously, TEG and ROTEM, guided by sound clinical judgment, can efficiently identify patients likely to require massive transfusion. Taken as a whole, a hypocoagulable tracing should prompt immediate efforts to correct the modifiable contributors to TIC: obtaining early control of massive hemorrhage, addressing acidosis, preventing hypothermia, treating hypocalcemia, and avoiding dilutional coagulopathy. Individual components of the TEG or ROTEM tracing can further guide resuscitation strategies. A summary of the measures describes below is shown in Table 2.

Table 2

A protocol for guided resuscitation based on rTEG values with normal and abnormal tracings

Prolonged reaction time (R) in TEG and clotting time (CT) in ROTEM indicates delayed initiation of the coagulation web. A surrogate in rTEG is the activated clotting time (ACT). As plasma contains the greatest concentration of clotting factors among fractionated blood products, many protocols start with infusion of plasma while starting red blood cells (RBC) during the initial phases of a resuscitation. While plasma use in patients who eventually required fewer than 10 units of RBC in 24 h may be considered undesirable, early administration of plasma in a massive transfusion event of more than 10 units improves survival (32, 85–87). R, ACT, and CT are the first values to return in viscoelastic testing. Thus, they sound the alarm for early activation of massive transfusion protocols (if not already ordered) and early transfusion of plasma.

Anticoagulant-associated hemorrhage can also often be identified via early viscoelastic measures, a critical finding in a patient with unknown medication history. R, ACT, and CT are usually prolonged in the presence of clinically significant bleeding due to direct thrombin inhibitors and Xa inhibitors such as dabigatran, rivaroxaban, and apixaban use (88, 89). R-time in the presence of dabigatran correlates linearly to gold-standard assays (90). It should be noted that warfarin does not similarly affect these values, and INR remains the standard for monitoring this anticoagulant. Although consensus is lacking regarding plasma and factor concentrate use, targeted methods do exist to treat anticoagulant-associated hemorrhage. Dabigatran can be reversed with idarucizumab (91), and andexanet alfa is now available for reversal of rivaroxaban and apixaban (92). Retrospective data have associated ROTEM-guided prothrombin complex concentrate use with increased survival and decreased transfusion requirements (93, 94), although prospective trials are still needed to validate these results.

The next resulting values in TEG and ROTEM describe clot kinetics and indicate the need for additional plasma, cryoprecipitate, and/or fibrinogen concentrate. The alpha angle in both TEG and ROTEM represents the early rate of fibrin polymerization and can be thought of as the slope of the curve as clot forms. The kinetic or “k-time” for TEG or clot formation time (CFT) for ROTEM measure the time from initiation of clotting (at the point of separation of the resistance curves) to a resistance level of 20 mm. The clot amplitude at 10 min (CA10) for ROTEM EXTEM, more specific for fibrinogen with FIBTEM, also describes early polymerization. Cryoprecipitate and fibrinogen concentrates, and FFP to a lesser degree, can maintain or increase the circulating fibrinogen to increase the rate of polymerization (95, 96). A clinical trial is underway to compare fibrinogen concentrate versus cryoprecipitate for this purpose (97).

Fibrinogen and platelets contribute to maximum amplitude (MA) for TEG and maximum clot firmness (MCF) for ROTEM. These values occur at the peak of the resistance curve and indicate the clot strength. Fibrinogen contributes between approximately 20% and 30% of the amplitude of MA and MCF, the rest belonging to platelet count and activity (98–100). While platelet transfusions are the obvious choice to correct a low MA or MCF, prospective data show that platelets may not restore function despite rise in platelet counts (22). Correction of MA and MCF should therefore take into consideration both platelet and fibrinogen replacement, especially if derangements in clot kinetics are apparent.

Fibrinolysis, based upon the percentage lysis at 30 mi (LY30) or clot lysis at 30 min (CL-30) after maximum amplitude (MA), has gained the most attention perhaps because viscoelastic testing identifies this issue whereas no other CCT reliably can. Many centers use a threshold of 3% lysis as the upper limit of normal, given the 10-fold increase in hemorrhagic death associated with LY30 > 3% (64), although consensus on the exact percentage is lacking. The Clinical Randomisation of Anti-fibrinolytic in Severe Hemorrhage trial (CRASH-2) (101) and Military Application of Tranexamic Acid in Trauma Emergency Resuscitation Study (MATTERs) (102) showed decrease in mortality with tranexamic acid (TXA) as an antifibrinolytic, but concerns abound regarding its use in certain settings (103), particularly given the increase in mortality and thromboembolic events seen if TXA was administered after 3 h postinjury. More selective use of TXA in only those patients demonstrating fibrinolysis has gained increasing support (104–106). The other end of the fibrinolytic spectrum also carries increased mortality risk. Often defined as LY30 < 0.8%, fibrinolysis shutdown was seen in 46% of severely injured patients in a large multicenter observational study. This phenotype, while not associated with as high a mortality as hyperfibrinolysis, shutdown remained an independent risk factor for mortality (OR 1.6 when compared to physiologic LY30 of 0.8–2.9%) (107). Given its high incidence, shutdown has also been used as an argument against nonselective use of TXA (108). It should be noted that this topic continues to be controversial, with ongoing debate regarding implications of shutdown, such as overdiagnosis, differences between TEG and ROTEM lysis values, and whether it actually represents a pathophysiologic phenomenon (109).

Guided versus fixed ratio resuscitation

Multiple algorithms have been published describing viscoelastic testing-guided resuscitation (57, 61, 110–115). Military experience, where judicious use of limited resources such as blood products becomes particularly essential, has shown ROTEM-guided resuscitation to be feasible (116, 117). A remaining question is whether this type of guided resuscitation is superior to CCT-guided or fixed ratio transfusion of red blood cells, plasma, and platelets, a standard fairly well established in clinical trials (33, 118). Three systematic reviews found a paucity of studies with low risk of bias regarding viscoelastic-guided identification and management of TIC (119–121), although some findings were encouraging. Veigas et al. found 13 observational studies after trauma, including nine prospective, showing ROTEM CA and MCF consistently predicted coagulopathy, massive transfusion, and mortality. Da Luz et al., with more relaxed search criteria, identified 55 studies and concluded viscoelastic testing “may also be used to direct blood and blood-product transfusion; effects on patient-important outcomes are uncertain.” Whiting et al. (122) pooled 31 studies to demonstrate substantial cost savings for viscoelastic testing of trauma patients. More recently, Wikkelsø et al. (123) performed a meta-analysis on 17 trials covering bleeding patients in both elective and trauma settings, finding that blood product use and mortality were decreased with TEG and ROTEM use, compared to historical controls. Unfortunately, risk of bias was generally high and the majority of patients underwent elective cardiac procedures. The most recent meta-analysis by Dias et al. (124) confirmed these findings in elective settings. The Denver group has recently published the only randomized trial in trauma patients comparing TEG to CCT-guided resuscitation in patients predicted to receive a massive transfusion. The Goal-Directed Hemostatic Resuscitation of Trauma-Induced Coagulopathy Trial showed a survival benefit at 28 days postinjury, as well as less transfused plasma and platelets, in patients receiving TEG-guided resuscitation (125). Additional clinical trials are underway to compare directly patient outcomes and transfusions with viscoelastic testing versus conventional management (126, 127).

CONCLUSION

Trauma causes 973 million people a year to seek care for injury worldwide (128). Among these casualties, hemorrhage remains a major cause of mortality, both in the immediate phase of care as well as later contributing to multi-organ system failure. TIC is associated with a 10-fold increase in mortality for the bleeding patient. Viscoelastic testing, including TEG and ROTEM, has been shown to predict massive transfusion requirement and identify TIC with greater sensitivity than clinical judgment or CCT. Further, TEG and ROTEM can guide resuscitation to combat specific derangements of physiologic coagulation within minutes of a patient's arrival. When minutes count, immediate control of hemorrhage and prevention of fulminant coagulopathy are keys to increasing survival. Serial reassessment of coagulation can then proceed. While viscoelastic testing has its limitations, it expediently provides a global assessment of clotting. In elective surgical settings, viscoelastic testing decreases blood product consumption, and the available trauma literature suggests the same may be true in emergency resuscitation. Ongoing trials may reinforce the growing evidence promoting guided resuscitation.