Validation of rotation thrombelastography in a model of systemic activation of fibrinolysis and coagulation in humans

Authors


Bernd Jilma, Department of Clinical Pharmacology, Medical University of Vienna, Waehringer Guertel 18-20, 1090 Vienna, Austria.
Tel.: ++43 1 40400 2981; fax: ++43 1 40400 2998; e-mail: bernd.jilma@meduniwien.ac.at

Abstract

Summary. Background: Thrombelastography (TEG) is a whole blood assay to evaluate the viscoelastic properties during blood clot formation and clot lysis. Rotation thrombelastography (e.g. ROTEM®) has overcome some of the limitations of classical TEG and is used as a point-of-care device in several clinical settings of coagulation disorders. Endotoxemia leads to systemic activation of the coagulation system and fibrinolysis in humans. Objectives: We validated whether ROTEM® is sensitive to endotoxin induced, tissue factor-triggered coagulation and fibrinolysis and if its measures correlate with biohumoral markers of coagulation and fibrinolysis. Patients and methods: Twenty healthy male volunteers participated in this randomized placebo-controlled trial. Volunteers received either 2 ng kg−1 National Reference Endotoxin or saline. Results: Endotoxemia significantly shortened ROTEM® clotting time (CT) by 36% (CI 0.26–0.46; P < 0.05) with a strong inverse correlation with the peak plasma levels of prothrombin fragments (F1 + 2) (r = −0.83, P < 0.05). Additionally, endotoxin infusion enhanced maximal lysis (ML) 3.9-fold (CI: 2.5–5.2) compared with placebo or baseline after 2 h (P < 0.05). Peak ML and peak tissue plasminogen activator (t-PA) values correlated excellently (r = 0.82, P < 0.05). ROTEM® parameters clot formation time and maximal clot firmness were not affected by LPS infusion, whereas platelet function analyzer (PFA-100) closure times decreased. Conclusions: Rotation thrombelastography (ROTEM®) detects systemic changes of in vivo coagulation activation, and importantly it is a point of care device, which is sensitive to changes in fibrinolysis in humans. The ex vivo measures CT and ML correlate very well with established in vivo markers of coagulation activation (F1 + 2) and fibrinolysis (t-PA), respectively.

Introduction

Thrombelastography (TEG) is a method to evaluate the viscoelastic properties during blood clot formation and clot lysis. By using whole blood, it is able to yield information relating to the cumulative effects of various parameters (plasma factors, platelets and leukocytes) of all phases of the coagulation process and fibrinolysis [1,2]. Therefore, application of TEG could have several advantages compared with standard coagulation analysis, which reflect only a part of the hemostatic process and additionally include partially unphysiologic activators.

Recently, the modified rotation thrombelastogram analyzer (ROTEM®; Pentapharm, Munich, Germany), has overcome some of the limitations of classical TEG. The ROTEM® is very robust and not susceptible to vibrations or mechanical shocks. By using an electronic pipette, reproducibility and performance has increased. Also, depending on the parameters measured, ROTEM® results are available as early as 15 min up to 1 h. Therefore, ROTEM® application may lead to accelerated and more appropriate clinical decision making.

Thrombelastography is already utilized in several clinical settings, e.g. for monitoring of hemostasis during liver transplantation and cardiac surgery and has been shown to be a point of care device for rapid diagnosis and differentiation of hypercoagulable and hyperfibrinolytic situations [3,4].

Endotoxemia leads to systemic activation of the coagulation system and fibrinolysis in humans [5,6]. Hence, it is a good model to test anticoagulants [7–9], as well as diagnostic coagulation tests and fibrinolytic markers [10–12].

It is currently unknown whether ROTEM® is sensitive to systemic changes of in vivo coagulation or if its measures of coagulation and fibrinolysis correlate with endogenous changes in thrombin formation and tissue plasminogen activator (t-PA) or plasminogen activator inhibitor (PAI-1) levels, respectively. Thus, the current study aimed to characterize the functional properties of ROTEM® in a well-defined human model of endotoxin induced, tissue factor (TF) triggered coagulation [8,13] and tumor necrosis factor (TNF-α) triggered fibrinolysis [14]. We hypothesized that ROTEM® may be a whole blood assay, which not only detects changes in coagulation like two other assays [11,12], but may also detect changes in fibrinolysis in human endotoxemia.

Patients and methods

Study design

The study was approved by the Ethics Committee of the Medical University of Vienna and all participants gave written informed consent. Twenty healthy male volunteers [aged 26 years (95% CI 22–29)] participated in a placebo-controlled trial. The control group consisted of four volunteers (4:1 randomization). Medical screening included medical history, physical examination, laboratory parameters, and virologic and standard drug screening. In addition, study subjects were tested for hereditary thrombophilia i.e. factor V Leiden, protein C and S deficiency, to minimize potential risks of endotoxin-induced coagulation activation [8]. Exclusion criteria were regular or recent intake of medication including over the counter (OTC) drugs, and clinically relevant abnormal findings in medical history or laboratory parameters.

The experimental procedures of our endotoxin model have been described in detail in other trials [7,8]. Briefly, volunteers were admitted to the study ward at 8:00 am after an overnight fast. Throughout the entire study period, participants were confined to bed rest and kept fasting for 8.5 h after LPS infusion. In the current study, volunteers received either a bolus of 2 ng kg−1 National Reference Endotoxin (LPS, Escherichia coli; USP, Rockville, MD, USA) or 0.9% NaCl as placebo.

Physiologic saline solution (200 mL h−1) was administered to maintain adequate hydration for all subjects, who were kept fasting during the first 6 h.

Sampling and analysis

Sampling times were selected based on the kinetics of coagulation and fibrinolysis seen in subjects challenged with LPS in previous trials [8,15–17] (F1 + 2: 0, 2, 4, 6, 8, 24; t-PA: 0, 2, 4, 24; PAI-1: 0, 2, 3, 4, 24; plasmin antiplasmin complexes (PAP): 0, 1, 2, 4, 8, 24; CEPI-CT: 0, 2, 4, 8, 24; VWF-Ag: 0, 4, 24, platelet count: 0, 1, 2, 3, 4, 6, 8, 24; ROTEM® parameters: 0, 1, 2, 3, 4, 6, 8 and 24 hours after LPS-infusion). Blood samples were collected by repeated venipunctures into citrated or ethylenediaminetetraacetic acid (EDTA) – anticoagulated vacutainer tubes (Becton Dickinson, Vienna, Austria). Plasma was obtained by centrifugation at 2000 g (15 min at 4°C) and stored in 0.5 mL aliquots at −80°C until analysis.

Most coagulation and inflammatory parameters were measured by enzyme immunoassays (EIA). Plasma levels of prothrombin fragment (Enzygnost®F1 + 2 micro; Dade Behring, Marburg, Germany) were used as markers of in vivo thrombin generation and plasma levels of t-PA (t-PA Kit; Technoclone, Vienna, Austria), plasmin activator inhibitor (PAI-1, Actibind® Kit; Technoclone, Vienna, Austria) and PAP (PAP micro; DRG International, East Mountainside, NJ, USA) were used as markers for endogenous fibrinolytic capacity. Plasma levels of von Willebrand factor were measured using an enzyme-linked immunosorbent assay (REAADS VWF Ag Test Kit; Corgenix, CO, USA).

The ROTEM® Modified Rotation Thrombelastogram Analyzer

The method and the parameters of TEG and the ROTEM® Coagulation Analyzer (Pentapharm, Munich, Germany) have been described in detail previously [1,18]. Briefly, TEG measures shear elastic modulus during clot formation and subsequent fibrinolysis. The ROTEM® uses a ball-bearing system for power transduction, which makes it less susceptible to mechanical stress, movement and vibration.

Whole blood samples were collected into 3.8% sodium citrate tubes. In the literature, inconsistent references exist on the stability and reproducibility of TEG measurements after sample storage at room temperature. Depending on the references [19–21], TEG measurements produced accurate and reproducible results within 30 min up to 4 h. We performed TEG measurements between 1 and 2 h after blood storage at room temperature. Just before running the assay, citrated blood samples were recalcified with 20 μL of CaCl2 0.2 m (Start-TEG; Nobis, Endingen, Germany) and the test was started. To adjust as much as possible to physiologic conditions and to quantify the intrinsic changes in TF-triggered coagulation we did not add activators to the test system [so-called non-activated TEM (NATEM)]. The following ROTEM® parameters were analyzed: the clotting time (CT), the clot formation time (CFT), the maximum clot firmness (MCF) and the maximum lysis (ML).

PFA-100 assay

Peripheral venous blood was collected into tubes containing 3.8% sodium citrate. The functional properties of the PFA-100 (Dade Behring, Deerfield, IL, USA) have been described in detail elsewhere [22,23]. In brief, the system consists of a disposable test cartridge where a platelet plug occludes a microscope aperture cut into a membrane coated with collagen and epinephrine (CEPI) or collagen and ADP (CADP). The plug formation occurs under high shear flow conditions produced by a constant vacuum and controlled by a capillary. The time required for occlusion (closure time, CT) is indicative of platelet function and primary haemostasis capacity.

Data analysis

Data are expressed as mean and the 95% confidence intervals for description in the text. Non-parametric statistics were applied. All statistical comparisons were performed with the Friedman anova and the Wilcoxon signed rank test for post hoc comparisons. A two-tailed P-value of <0.05 was considered significant. The Spearman ranks correlation test was used for computations of associations. All statistical calculations were performed using commercially available statistical software (Statistica Vers. 5.0; Stat Soft, Tulsa, OK, USA).

Results

No severe, serious, or unexpected adverse events were observed after LPS infusion. There was no difference in ROTEM® baseline parameters between groups (Table 1).

Table 1.  Baseline values of ROTEM® parameters and markers of coagulation and fibrinolysis
 Placebo, mean (±95% CI); n = 4Endotoxin, mean (±95% CI); n = 16P-value
  1. CT, clotting time; CFT, clot formation time; MCF, maximum clot formation; VWF-Ag, von Willebrand factor-antigen; t-PA, tissue plasminogen activator; PAI-1; plasminogen activator inhibitor; PAP, plasmin antiplasmin complexes.

CT (s)729 (484–972)689 (625–753)ns
CFT (s)260 (125–396)293 (245–341)ns
MCF (mm)54 (46–61)53 (50–56)ns
CT plus CFT (s)990 (609–1370)998 (891–1104)ns
ML (% of MCF)14 (8–21)14 (12–16)ns
Prothrombin fragment 1 + 2 (nmol L−1)0.41 (0.34–0.48)0.61 (0.42–0.80)ns
PFA-100, CEPI-CT (s)165 (130–200)161 (120–202)ns
VWF-Ag levels (IU dL−1)81 (68–94)83 (64–102)ns
Platelet count (G L−1)200 (177–224)194 (135–253)ns
t-PA (ng mL−1)1.7 (−1.0–4.6)1.1 (−0.5–2.7)ns
PAI-1 (ng mL−1)8.1 (4.2–12.0)7.3 (−2.6–16.1)ns
PAP (μg L−1)476 (338–616)468 (307–627)ns

Effect of endotoxin on ROTEM® CT and CFT in their relation to coagulation activation (F1 + 2)

LPS-infusion shortened CT by 36% (95%CI 0.26–0.46) after 6 h [CT: 441 s (95%CI 373–509)] compared with placebo and baseline (P < 0.05 between group and time) (Fig. 1). After 24 h these LPS induced changes diminished and the clotting time returned to baseline values. CFT showed no significant changes in the time course, as well as compared with placebo.

Figure 1.

Effect of endotoxin on ROTEM® clotting time (CT) and clot formation time (CFT) in their relation to coagulation activation (F1 + 2). Healthy male volunteers received either an endotoxin (LPS) bolus infusion (2 ng kg−1) (bsl00001, n = 16) or placebo (bsl00043, n = 4). Endotoxemia led to a significant transient shortening of ROTEM® CT by 36% (CI 0.26–0.46) with an inverse correlation to the peak plasma levels of F1 + 2. Data represent the mean value ±95% CI; *P < 0.05 vs. placebo.

The maximum shortening of CT coincides with the maximum coagulation activation as measured with prothrombin fragments (F1 + 2), which showed an approximately sevenfold increase compared with baseline values after 4 h [(F1 + 2: 3.4 nmol L−1 [95% CI 2.4–4.4)] (P < 0.05 between time). Nadir values of CT showed a strong inverse correlation with peak F1 + 2 levels in the endotoxemia group (r = −0.83, P < 0.05).

Effect of endotoxin on ROTEM® MCF in relation to CEPI closure time (CEPI-CT) measured with the PFA-100, VWF-antigen levels and platelet counts

LPS-infusion had no obvious effect on the strength of the blood clot, which was measured as MCF, and which is considered a measure of platelet function [24,25] (Fig. 2). There was neither a change in the time course nor compared with placebo.

Figure 2.

Effect of endotoxin on ROTEM® maximal clot firmness (MCF) in relation to CEPI closure time (CEPI-CT) measured with the PFA-100, von Willebrand factor (VWF)-antigen levels and platelet counts. Healthy male volunteers received either an endotoxin (LPS) bolus infusion (2 ng kg−1) (bsl00001, n = 16) or placebo (bsl00043, n = 4). MCF was not affected by endotoxemia. data represent the mean value ±95%CI; *P < 0.05 vs. placebo.

These results are in contrast to the effects of endotoxemia on the closure time measured with the PFA-100 device [26]. LPS-infusion significantly decreased CEPI-closure time by approximately 50% to reach minimum levels after 4 h [80 s (95%CI 73–88)]; P < 0.05 vs. baseline and placebo).

Von Willebrand factor-Ag levels increased from baseline levels almost threefold (229 IU dL−1; 95% CI 187–271) at 4 h after endotoxin challenge. As previously published [26–28], peak VWF levels inversely correlated with nadir values of CEPI-CT (r = −0.56, P < 0.05).

The LPS-infusion induced a transient decrease in platelet count by 16% to reach minimum levels after 3 h [platelet count: 168 × 109 L−1 (95%CI 144–192); P < 0.05 vs. baseline]. In the placebo period we could not observe changes in the time course.

Effect of endotoxin on ROTEM® ML in relation to markers of fibrinolysis (t-PA, PAI-1, PAP)

Endotoxin infusion caused an obvious short-term activation of ML with a 3.9-fold (CI: 2.5–5.2) increase to a peak value of 53% (95% CI: 35–72) of MCF compared with placebo after 2 h (P > 0.05 vs. baseline and placebo) (Fig. 3). Activation of lysis diminished 2 h later.

Figure 3.

Effect of endotoxin on ROTEM® maximal lysis (ML) [% of maximum clot formation (MCF)] in relation to markers of fibrinolysis [tissue plasminogen activator (t-PA), plasminogen activator inhibitor (PAI-1), plasmin antiplasmin complexes (PAP)]. Healthy male volunteers received either an endotoxin (LPS) bolus infusion (2 ng kg−1) (bsl00001, n = 16) or placebo (bsl00043, n = 4). Endotoxin infusion caused an obvious short-term increase in ML with an 3.9-fold (CI: 2.5–5.2) increase compared with placebo after 2 h. Peak ML and peak t-PA values correlated excellently (r = 0.82, P < 0.05). Data represent the mean value ±95% CI; *P < 0.05 vs. placebo.

These results are consistent with the effects of endotoxemia on t-PA plasma levels, which increased transiently 13-fold compared with baseline after 2 h [t-PA: 23.3 ng mL−1 (95% CI 14.0–32.6)]. Peak t-PA levels correlated very well with peak maximal lysis (ML) (r = 0.82, P < 0.05).

Plasmin-antiplasmin complexes (PAP) showed a similar time course, with peak values and a 4.4-fold increase compared with baseline after 2 h [PAP: 2095 ng mL−1 (95% CI 1480–2711)].

Plasminogen activator inhibitor increased with an approximately 2 h delay compared with t-PA and reached peak values after 4 h [PAI: 60 ng mL−1 (95% CI 43.5–76.5)] with an sevenfold increase (95%CI 5.3–9.4) vs. baseline.

Discussion

This study demonstrates that rotation TEG (ROTEM®) not only detects in vivo changes of LPS induced, TF-triggered coagulation in vivo but also changes in fibrinolysis. Furthermore, the ROTEM® parameters CT and ML correlate with markers of coagulation activation (prothrombin fragments, F1 + 2) and markers of the fibrinolytic system (t-PA, PAP, PAI-1) respectively.

Endotoxemia led to a significant transient shortening of CT (Fig. 1). We observed an excellent correlation between the nadir of CT values and the peak plasma levels of prothrombin fragments (F1 + 2) (r = −0.83), which are a marker of LPS induced coagulation activation. Similarly, Rivard et al. [29] recently described an excellent correlation between thrombin/antithrombin (TAT) complex generation and the calculated TEG parameter total thrombus generation (TTG). Additionally, our results confirm and extend the separate observations of two previous studies that validated two different point-of-care devices measuring the TF clotting time and the clotting onset time in the same human endotoxemia model [11,12]. However, TEG has the advantage of measuring all parts of the coagulation process including fibrinolysis.

In contrast to CT, CFT was not affected by endotoxemia. Therefore, the time to onset (CT), i.e. the initiation of coagulation, and not its propagation (CFT) seems to be sensitive to low dose endotoxemia.

Of particular interest, we observed a short lasting activation of lysis (ML) after 2 h in the endotoxin model, which was paralleled by activation of fibrinolysis and consequent inhibition of fibrinolysis as measured by t-PA, PAI-1 and PAP (Fig. 3). Peak ML and peak t-PA values correlated excellently (r = 0.82), indicating that ML is a measure of circulating t-PA activity. To our knowledge, this is the first report of monitoring the fibrinolytic system with a global bedside assay in experimental human endotoxemia, and its cross-validation with an established fibrinolysis marker (t-PA). So far, monitoring of hyperfibrinolytic states by TEG has been performed in other clinical settings, e.g. orthotopic liver transplantation [30–32], and TEG has been shown to be useful in the early detection of hyperfibrinolysis [33,34].

Furthermore, MCF, which primarily depends on platelet function and platelet count [24,25], was not affected although platelet counts decreased slightly, and VWF-levels increased substantially. These results are in contrast to the effects of endotoxemia on the PFA-100 closure times (CEPI-CT) [26] (Fig. 2). There are likely explanations for these results. First, the platelet count only slightly decreased compared with baseline (minus 16%), which may not be sufficient to have an impact on MCF. Secondly, the activation of platelets, which occurs in human endotoxemia [35], might counteract the slight decrease in platelet counts. Finally, as TEG is not a high shear system, VWF-GPIb interactions are unlikely to be measurable. However, a modified TEG method may still be a valuable tool to assess the effects of anti-platelet drugs including non-steroidal anti-inflammatory drugs, clopidogrel or GPIIb/IIIa inhibitors [24,36,37].

One limitation of this study is that the human low grade endotoxemia model is not a sepsis model, as discussed previously [9]. Therefore, it is difficult to extrapolate whether TEG might be a useful tool in monitoring sepsis and disseminated intravascular coagulation in the clinical setting. In addition, we only validated the ROTEM® without any activators for the above described reasons, and use of TF-activated TEG may show lower scattering of data as compared with the non-activated test [20,24].

In conclusion, rotation TEG (ROTEM®) not only detects systemic changes of in vivo coagulation activation, but also changes in fibrinolysis in human endotoxemia. Ex vivo clotting time correlated very well with prothrombin fragment levels, a marker of in vivo thrombin generation, and maximal lysis with the in vivo fibrinolysis marker t-PA.

Ancillary