Activated coagulation time vs. intrinsically activated modified rotational thromboelastometry in assessment of hemostatic disturbances and blood loss after protamine administration in elective cardiac surgery: analysis from the clinical trial (NCT01281397)

Excessive bleeding after CPB continues to be an important cause of morbidity and mortality [1]. Dixon B et al. reported chest tube output (CTO) as the strongest independent predictor of mortality [2]. However, the question how to predict and prevent excessive bleeding due to hemostatic disturbances remains challenging. Therapeutic approach to surgically or coagulopathically-induced bleeding is quite different. In addition to, conventional coagulation tests are not able to predict postoperative bleeding [3, 4]. Thus, bedside suitable devices capable to identify hemostatic disturbances after weaning from cardiopulmonary bypass (CPB) are desirable. Identifying hemostatic disorder after CPB may make the important differentiation between coagulopathic (nonsurgical) bleeding requiring procoagulant blood components transfusion and surgical bleeding requiring surgical intervention.

The Activated Coagulation Time (ACT) was first described by Hattersley in 1966 and is essentially a point-of-care test of coagulation that is used to monitor the anticoagulant effect of unfractionated heparin (UFH) in patients on bypass surgery [5]. The ACT first came into clinical use in the mid-1970s to guide the administration and reversal of heparin during CPB procedures. Although originally proposed as a routine pre-operative screening test - it is now used almost exclusively for monitoring patients on CPB. The test is now more commonly performed using a fully automated technique with several devices available in which the end point i.e. clot formation is recorded electronically – the principle, however, remains the same as described by Hattersley many years ago [5]. ACT is the most common intraoperative point-of-care hemostatic test to confirm profound anticoagulation during CPB and adequacy of heparin-protamine neutralization after weaning from CPB [6]. Although the most commonly used hemostatic point-of-care device in cardiosurgical procedures, the ACT is the least sensitive hemostatic test to detect residual heparin anticoagulation [6], as well as other hemostatic disorders which may lead to excessive bleeding after CPB. Although ACT is regularly used for assessment of heparine-protamine neutralization management adequacy and assesses intrinsic pathway of coagulation, its reliability in predicting hemostatic disorder resulting in excessive bleeding remains elusive. The best point-of-care test to diagnose hemostatic disorder immediately after CPB, is still unclear. Rotation thromboelastometry (TEM) performed on whole blood samples provides information on the contribution of fibrinogen and platelets to clot formation, however, the precise relationship between TEM values and postoperative bleeding still remains unclear.

This study sought to compare and determine the accuracy of intrinsically activated rotational TEM test (InTEM test) variables and ACT in detection of excessive bleeding and blood products transfusion requirements after CPB. This study has been presented during 23rd World Congress of the World Society of Cardio-Thoracic Surgeons and published in supplement issue of the Journal of Cardiothoracic Surgery [7].

This study presents retrospective exploratory analysis of prospectively collected data within research project approved by the Ethics Committee of the University Hospital Center Zagreb and registered at the Clinical-trials.gov website (Identifier NCT01281397). The purpose of this study was prospectively defined. Clinical trial NCT01281397 was designed in prospective observational fashion. The aim of trial was to assess possibility of point-of-care hemostatic devices (rotational thromboelastometry and multiple electrode aggregometry) to predict excessive bleeding in elective cardiac surgery. Parameters of rotational thromboelastometry and multiple electrode aggregometry were obtained perioperatively at three time points and respective values were correlated with observed key endpoints such as CTO and transfusion requirements. Initially, a trial was supposed to recruit 400 patients. During study period, interim statistical analysis was performed after approximately every 50 patients. After 148 patients enrolled, interim analysis revealed positive results in regard to primary hypothesis that point-of-care tests for assessment of platelet function and viscoelastic blood properties predict bleeding in cardiac surgery patients. Thus, due to fact that positive results confirmed primary hypothesis after 148 patients were recruited, we decided to terminate study earlier.

All patients from sample were discharged from hospital. Study cohort demographic, surgical, and baseline routine laboratory findings are presented in Tables 1 and 2. Total amount of CTO value of 1507,50 mL presented 75th percentile of distribution, thus a cut-off value for bleeder category. No surgical reexploration was performed in study cohort. Patients in “Bleeder” category had longer CPB time (median, 116 vs. 95 min, p = 0.023) and significantly lower value of the lowest body temperature during CPB (median, 30 vs. 32 degrees of Celsius, p = 0.012). Patients with total amount of CTO exceeding 75th percentile were more frequently transfused with fresh frozen plasma (51.4% vs. 9.9%, p < 0.001), fibrinogen concentrate (21.6% vs. 2.7%, p = 0.001) and platelet concentrate (13.5% vs. 0.9%, p = 0.004). Differences in ACT and InTEM parameters of patient groups according to bleeding tendency are presented in Table 3. InTEM parameters, but not ACT correlated to total amount of CTO (mL) (Table 4). InTEM parameters with the strongest correlation to CTO were tested for accuracy in predicting excessive postoperative bleeding using ROC analysis. Accuracies of the InTEM A 10 (AUC 0.636, p = 0.02), InTEM A 20 (AUC 0.632, p = 0.02) and InTEM A 30 (AUC 0.630, p = 0.02) are shown in Figure 1.

The best test to detect hemostatic disturbances and optimize hemostatic properties after CPB remains unclear. Conventional laboratory tests are unable to decompose multifactorial coagulopathy pathogenesis as described by Paparella et al. [11]. Considering the fact that immediate and precise diagnosis of the underlying cause of hemostatic disturbances, which result in excessive bleeding, is essential, we decided to compare ACT with another bedside suitable test, TEM with InTEM parameter. Our study showed that InTEM parameters, but not ACT, significantly correlated to bleeding extent following CPB. There are several advantages of InTEM test over ACT. (1) ACT fails to predict excessive bleeding. Observed correlations between ACT and CTO were not significant in our study (p = ns) which is in line with results recently published by Galeone et al. [12]. The role of the ACT as a monitor of profound heparin anticoagulation has also been questioned. Notably, Metz and coworkers found that maintaining a minimum ACT during CPB was not necessary if a 300 U/kg intravenous heparin dose was administered before CPB [13]. Worthwhile, a profound heparin anticoagulation can be easily confirmed with TEM using InTEM and HepTEM tests concurrently. (2) While Murray et coworkers [6] showed that ACT is not very sensitive to the low heparin concentrations that might be present after protamine reversal, Mittermayr et coworkers [14] described TEM as a valuable tool for heparin-protamine management. Using CT parameter, authors showed that TEM can be used to distinguish the effects of heparin excess from those of protamine excess [14]. Heparin rebound and incomplete heparin reversal are very common phenomena after cardiac surgery with CPB [12]. Recently, Galeone et al. [12] showed ACT not to be able to detect residual heparin activity, whereas thromboelastography analysis with and without heparinase allowed the diagnosis of heparin rebound [12]. However, no association was observed between heparin concentration, ACT, and thromboelastography parameters with postoperative bleeding and need for blood and blood components transfusions [12]. Contrary to, in our study InTEM CT parameter significantly correlated to the total amount of CTO. Although statistically significant, this correlation is relatively weak (p = 0.045), focusing CT parameter only to heparin-protamine management in cases of high InTEM CT values. (3) Hemostatic disturbances after CPB have multifactorial etiology due to many components of the hemostatic system involved in. TEM, but not ACT provides comprehensive insight into hemostatic properties, therefore allows possibility to detect the most dominant underlying cause of hemostatic disorder. (4). Spiess et al. reported thromboelastography guided hemostatic management to significantly reduce incidence of overall transfusion and mediastinal re-exploration due to excessive bleeding [15]. We believe that similar results could be obtained using TEM. However, such a hemostatic management should be evaluated through a prospective randomized controlled trial. In present study, observed InTEM parameters significantly correlated to total amount of CTO. Aside from heparin-protamine neutralization management, InTEM parameters provide possibility for targeted procoagulant blood component administration. Such a transfusion algorithm deems reasonable especially if InTEM parameters fall below cut-of values which delineate bleeding tendency according to the ROC curves. According to our results we propose following hemostatic management. In patients with high InTEM CT values we propose concurrent use of InTEM and HepTEM test with heparin-protamine management as proposed by Mittermayr et coworkers [14]. HepTEM test can be used to identify the presence of heparin and its effects on coagulation by comparison with the InTEM test. If bleeding occurs after InTEM-HepTEM heparin neutralization management and if InTEM parameters fall below cut-of values defined with ROC curves, targeted procoagulant blood components administration should be considered. MCF parameters significantly correlate to amount of CTO. Thus, ROC derived cut-off values that correspond to excessive bleeding may be used as triggers for targeted administration of procoagulant blood components.

In our study, patients in “Bleeder” category had longer CPB time (median, 116 vs. 95 min, p = 0.023) and significantly lower value of the lowest body temperature during CPB (median, 30 vs. 32 degrees of Celsius, p = 0.012). It is well known that CPB alters the hemostatic balance and predisposes cardiac surgery patients to increased risk of excessive bleeding [16]. Pathophysiology of excessive bleeding after CPB has been described by Green et al. [16]. There are several factors related to CPB that contribute to onset of hemostatic disorder such as: foreign surface contact [17], consumption of clotting factors [18], platelet activation and dysfunction [19, 20] and fibrinolysis [21]. In addition to, hemostatic impairment during CPB arises in some extent from systemic hypothermia that induces kinetic slowing of coagulation, kinin and kalikrein activation, platelet function and fibrinolysis [22]. The fact that hypothermia tends to increase bleeding in the cardiosurgical patients has been intuitively recognized by cardiac surgeons despite scarce evidence available at the beginning. Canine studies have shown that hypothermia causes thrombocytopenia [23, 24] and activates fibrinolytic system [24]. Those results were confirmed in normal volunteers both in vitro and in vivo[25] suggesting that adequate rewarming strategy may reduce the need for less safe alternative such as transfusion of procoagulant blood components.

This study has few limitations. First of all, study was prospective observational, so transfusion management was performed according to regular clinical practice. In our study the attending clinicians, both the anesthesiologists and surgeons were blinded to InTEM results, therefore administering procoagulant blood products mainly on the clinical grounds. The use of procoagulant blood products certainly affected the amount of bleeding. This decrease in amount of bleeding would reduce the degree of correlation between blood loss and both ACT and InTEM parameters by reducing the sensitivity of each parameter. Administration of procoagulant blood components certainly attenuated correlation coefficients between hemostatic test values and extent of CTO. However, it would not be ethically acceptable to evaluate the relationship between point-of-care test values and values of CTO volume amount in setting without use of procoagulant blood components administered. In our study all patients were treated in line with the best available hemostatic management. However, even if administration of procoagulant blood components distorted and attenuated correlations, the fact is that correlations were equally distorted for each test at the same patient as correlations were analyzed between CTO extent and values of respective tests. The same explanation applies for tranexaminic acid which was regularly used for all patients at two time points, (1) at the induction of anesthesia, and (2) after protamine administration. Since all patients received tranexaminic acid in the same dose and at the same time, we assume that all patients were well balanced in respect to possible effects of tranexaminic acid. In conclusion, although the transfusion of procoagulant blood components and antifibrinolytic agents certainly reduced the correlation coefficients between point-of-care test results and amount of CTO, the fact is that all patients had the same hemostatic management. Further, as previously reported within “Methods” section, patients requiring surgical reexploration for excessive bleeding were supposed to be excluded from study if bleeding vessel would be identified during reexploration. During the study period, no surgical reexploration for bleeding was performed. It is, however, obvious that CTO is consisted of both “surgical” and “hemostatic disorder” bleeding. Without surgical reexploration performed it is impossible to detect whether some proportion of patients was actually bleeding predominantly due to surgical cause. It is impossible to extract surgical bleeding from total amount of CTO. Unfortunately, it is not feasible to differentiate chest tubes bleeding volume in respect to surgical or coagulopathic origin. At our center, surgical reexploration is performed in each case with suspicion to surgical cause of bleeding based on consensus between consultant surgeon and anesthesiologist. We do not have any specific reexploration criteria based on dynamics of CTO over time as well as on amount of CTO. Without measurement of hourly CTO dynamics and clear criteria that suggest surgical bleeding it is theoretically possible that some proportion of patients had surgical bleeding even though they were not reexplorated. That may explain relatively high cut-of value for excessive bleeding in our study.

In conclusion, we suggest that patients with InTEM test parameters below the cut-off values are at increased risk of excessive bleeding and use of TEM guided hemostatic management is required. Excessive bleeding with normal TEM values implies surgical bleeding and administration of procoagulant blood components such as fresh frozen plasma, fibrinogen concentrate and platelet concentrate should not simply be used empirically. Values above the cut-off values do not imply the fact that the patients will not bleed, however they may advise to pay attention to surgical cause of bleeding. Concomitant use of InTEM and HepTEM tests enables precise detection of low to moderate heparin concentration [14]. Therefore, TEM test also enables physician to detect or prevent protamine excess resulting from empirical or ACT-based additional protamine administration. Hemostatic disturbances after adequate heparin-protamine neutralization management may be easily detected with additional use of ExTEM and FibTEM assays. This may be very useful, especially in patients who developed the coagulopathy due to other CPB associated factors. In such cases, detection of coagulation factor depletion, platelets and/or fibrinogen dysfunction may lead to appropriate, “targeted” hemostatic therapy and more efficient hemostatic management. Such an approach may help to improve clinical outcome with lower incidence of excessive bleeding, lower transfusion requirements with lower incidence of transfusion related adverse outcomes.