| JOURNAL HOME | CME HOME | THIS MONTH | PAST ISSUES | ETOC | COLLECTIONS |
| AUTHORS | REVIEWERS | EDITORIAL BOARD | FEEDBACK | RSS | HELP |
| ||||||||||||||
| ||||||||||||||
|
|
|
|
||||||||||||
|
||||||||||||||
From the Departments of *Anesthesiology, Pathology and Hematology/Oncology, Emory University School of Medicine, Atlanta, Georgia.
Address correspondence and reprint requests to Kenichi A. Tanaka, MD, MSc, Associate Professor of Anesthesiology, Division of Cardiothoracic Anesthesia, Department of Anesthesiology, Emory University, School of Medicine, 1364 Clifton Rd., NE, Atlanta, GA 30322. Address e-mail to kenichi.tanaka{at}emoryhealthcare.org.
Abstract
BACKGROUND: Thromboembolic complications can occur with recombinant activated factor VII (rFVIIa) treatment in trauma and surgical patients but they are infrequent in hemophiliacs. Bleeding diathesis in these conditions is often attributed to reduced thrombin generation, which may be improved with rFVIIa. Normally, thrombin that diffuses from local vascular injury sites is quickly inactivated by antithrombin (AT). Evaluating the influence of AT levels on thrombin generation in hypocoagulable FVIII-deficient plasma would be a simple approach to better understand how procoagulant stimuli, such as rFVIIa, might result in postoperative thrombotic complications. We hypothesize that reduced AT concentrations would increase the procoagulant effects of rFVIIa in FVIII-deficient plasma.
METHODS: Thrombin generation was evaluated in vitro in FVIII-deficient and AT/FVIII-deficient plasma using thrombelastography and a thrombin generation assay (ThrombinoscopeTM). The effect of added rFVIIa on these variables was evaluated.
RESULTS: Delayed thrombus formation based on thrombelastography in FVIII-deficient plasma was predictably reversed by rFVIIa. Improved thrombus formation and responses to rFVIIa were observed when AT levels were 20%–50% of normal. Thrombin generation in FVIII-deficient plasma increased in an inverse relationship to AT levels. Supplemental rFVIIa decreased the lag time of thrombin generation but not the amount of thrombin generated.
CONCLUSIONS: Using FVIII-deficient plasma as a model of reduced thrombin generation, we demonstrate that low AT levels enhance in vitro hemostatic responses to rFVIIa. Reduced AT levels in trauma and surgical patients with normal or increased FVIII levels may be considered potentially prothrombotic. Monitoring of AT levels during rFVIIa therapy may thus reduce thrombotic complications in nonhemophiliacs.
Thrombin generation is a pivotal component of hemostasis after vascular injury (Fig. 1) with the coagulation process starting when factor VIIa binds to exposed tissue factor (TF) forming TF-FVIIa complex. In patients with hemophilia A, initial formation of trace amounts of thrombin (2–5 nM) via the TF/factor VIIa-factor Xa (TF/FVIIa-FXa) pathway occurs normally, but sustained thrombin generation (propagation process) is very slow due to the lack of intrinsic tenase (FVIIIa-FIXa-Ca2+) complex formation.1 This diminished "thrombin burst" results in smaller thrombi that are more susceptible to fibrinolysis due to decreased activation of tissue activatable fibrinolysis inhibitor.1 On the other hand, during the initiation of coagulation in both healthy individuals and hemophiliacs, the initially formed FXa and thrombin are susceptible to plasma protease inhibitors such as tissue factor pathway inhibitor (TFPI) and antithrombin (AT) (Fig. 1).2,3 Rapid inhibition of trace amounts of FXa and thrombin are regulatory mechanisms to limit hemostatic responses to areas of vascular injury (i.e., TF exposure), hence preventing uncontrolled thrombin generation. This safety mechanism poses major challenges in establishing local hemostasis in patients with hemophilia A in tissues with lesser TF expression, such as muscles and joints.4 The efficacy of recombinant activated factor VII (rFVIIa) to abort bleeding in patients with hemophilia A suggests that additional FXa formation, via the TF-FVIIa pathway, may partially overcome TFPI-mediated inhibition leading to improved hemostasis.2,5,6
|
Thromboembolic complications are infrequent in hemophiliacs receiving rFVIIa, but they may occur in trauma and cardiac surgical patients receiving this therapy for life-threatening hemorrhage.7 In hemophilia, thrombin generation is reduced by an isolated factor deficiency (factor VIII or IX) with normal anticoagulant levels, whereas reduced thrombin generation is accompanied by decreased procoagulant as well as anticoagulant elements (e.g., AT) in patients after trauma and surgery.8–10 Evaluating the influence of AT levels on thrombin generation in FVIII-deficient plasma would be a simple but useful approach to better understand how procoagulant stimuli, such as rFVIIa, might result in postoperative thrombotic complications. In this study, we hypothesize that reduced AT concentrations would increase hemostatic responses to rFVIIa because AT-mediated inhibition of FXa and thrombin is decreased (Fig. 1). Therefore, the purpose of this in vitro study was to evaluate the effects of rFVIIa on thrombus formation and thrombin generation in FVIII-deficient plasma at different AT levels.
METHODS
This study was conducted after the approval from the IRB of Emory University, Atlanta, GA. Recombinant human thrombin was kindly provided by Zymogenetics (Seattle, WA). The lyophilized product was dissolved in 5 mL of normal saline to obtain stock solution containing 1000 U/mL (or 0.32 mg/mL). Pooled plasma was purchased from George King Biomedical (Overland Park, KS). Antithrombin deficient plasma, FVIII-deficient (AT+/FVIII–), and AT-FVIII double deficient (AT–/FVIII–) plasma were obtained from Enzyme Research Laboratories (South Bend, IN). The fibrinogen level for each plasma was specified by the manufacturer as within the normal range (323–328 mg/dL). Corn trypsin inhibitor was purchased from Hematologic Technologies Inc. (Essex Junction, VT). rFVIIa (FVIIa, NovoSeven®) was kindly provided by Novo-Nordisk A/S (Bagsvaerd, Denmark). Each vial containing 1.2 mg of rFVIIa was reconstituted with 2.2 mL of sterile water, resulting in a final concentration of 0.545 mg/mL of rFVIIa.
Thrombelastography (TEG®)
To obtain plasma samples with different AT levels, but deficient in FVIII, AT+/FVIII– plasma was mixed with an incremental portion of plasma deficient in both factors, AT–/FVIII–, to achieve AT activity of 20% or 50% of normal (AT20/FVIII– and AT50/FVIII–). To prevent XII (contact) activation, corn trypsin inhibitor (100 µg/mL, final concentration) was added to the plasma samples. Samples were incubated for 5 min at 37°C before commencement of TEG analyses. To allow thrombus formation to be triggered only by exogenous thrombin, plasma samples were recalcified with 10 µL of 0.4 M CaCl2 without additional activators (i.e., kaolin, TF) of coagulation. For the study, four TEG-5000 analyzers with two channels each (Hemoscope, Niles, IL) were simultaneously used. The following TEG variables were collected: 1) lag time (R, min), which represents the elapsed time from the start of the analysis to the initiation of clotting, which is a measure comparable to the whole blood thrombin time; 2) the angle (°) of thrombus formation, which reflects the rate of fibrin polymerization and the speed of thrombin generation, and 3) maximum amplitude (MA, mm), which represents the peak of viscoelastic strength of the thrombus.
Effects of rFVIIa on Thrombus Formation Measured with TEG
This portion of the study consisted of two parts. In the first part, the TEG variables were obtained using commercial pooled plasma samples activated with increasing concentrations of thrombin (1, 2, 5, and 10 nM). In the second part of the study, thrombin (1 or 2 nM final concentration) was added to the TEG cuvettes containing 0, 30, or 60 nM of rFVIIa (final concentration 0, 1.5, and 3.0 µg/mL) followed by the addition of 340 µL of AT+/FVIII–, AT20/FVIII–, and AT50/FVIII– plasma samples, as described above. TEG measurements were allowed to run for 60 min before the recording was terminated. The concentrations of factor VIIa chosen for this study (30–60 nM) encompass the clinically relevant plasma concentrations (i.e., 26 nM after 90 µg/kg dose).11
Thrombin Generation Assay (ThrombinoscopeTM)
The fluorogenic substrate for this assay, benzyloxycarbonyl-Gly-Gly-Arg-7-amido-4-methylcoumarin (Z-Gly-Gly-Arg-AMC) was purchased from Bachem Bioscience Inc. (King of Prussia, PA). HEPES (N-2-hydroxyethyl piperazine-N'-2-ethanesulfonic acid), CaCl2, bovine serum albumin, and dimethyl sulfoxide were from purchased form Sigma-Aldrich (St. Louis, MO). For the preparation of the calcium buffer solution, 1.75 mL of 20 mM HEPES buffer (pH 7.35) containing 60 mg/mL of bovine serum albumin was added to 0.2 mL of 1 M CaCl2 in a glass test tube, mixed, and warmed at 37°C for 5 min. Just before use, 50 µL of the 100 mM fluorogenic substrate in dimethyl sulfoxide was added, yielding 2.5 mM substrate and 100 mM CaCl2 solution.
This aspect of the study was performed with a Thrombinoscope (Synapse B.V., Maastricht, the Netherlands), which allows for the measurement of thrombin generation in activated platelet-poor (PPP) or platelet-rich plasma based on the signal generated by the cleavage of a thrombin specific fluorogenic substrate. All experiments were performed in PPP using a commercially available TF-based activator (PPP reagent low, Diagnostica Stago, Parsippany, NJ).
Because our study involved FVIII-deficient plasma, we used low concentrations of TF activator (1 pM) since higher concentrations (standard for normal PPP is 5 pM) would blunt the difference in thrombin generation between hemophilia and normal plasma.12 For these measurements, 80 µL of PPP followed by 20 µL of an activator was added to round-bottom microtiter plate wells and the plate incubated for 2–3 min at 37°C after which 20 µL of the substrate buffer was added. The reaction was monitored by a fluorescence reader (Fluoroscan FL Ascent, Thermo Labsystems, Franklin, MA) set at 390 nm excitation wavelength and 460 nm emission wavelength. Fluorescence was recorded for 60–90 min, and the acquired data were processed to measure the thrombin generation variables. A thrombin calibrator with known thrombin-like activity was monitored in parallel sample wells to eliminate the signal differences of different plasmas and to allow for calculation of generated thrombin in nanomoles.12
Statistical Analysis
Based on previous studies with Thrombinoscope and TEG, a sample size of six experiments is needed to detect 20% change from the control samples in peak thrombin generation or TEG amplitude with a βs 
0.8 and an 
<0.05.13,14 Statistical analyses were conducted with the paired t-test or the repeated-measures analysis of variance, followed by the paired t-test with the Bonferroni correction. All data were expressed as mean value or mean ± sd. P 
0.05 was considered significant. All statistical analyses were performed using SPSS 15.0 (SPSS Inc., Chicago, IL) software.
RESULTS
Effects of Thrombin Concentration on TEG Variables
With increasing concentrations of thrombin (1, 2, 5, and 10 nM final concentration) in normal pooled plasma samples, there was a progressive decrease in R time and a simultaneous increase in angle and MA (Table 1). On the basis of these results, we chose a thrombin concentration of 1 and 2 nM as a coagulation trigger (activator) since higher concentrations of added thrombin resulted in near immediate clotting. In addition, the highest dose of thrombin (10 nM) led to a lower MA.
|
Effects of AT and FVIII Deficiency on TEG Variables
The TEG variables obtained from plasma samples deficient in FVIII and with varying AT levels are summarized in Table 2. There was no thrombus formation in FVIII-deficient plasma samples with normal AT even at 60 min when activated with 1 nM thrombin. The lag time was decreased to 40.6 ± 21.8 min when the thrombin concentration was increased to 2 nM. Decreasing AT concentrations (AT50/FVIII– and AT20/FVIII– plasma samples) significantly reduced R time and the rate of thrombus formation (larger angle and MA) when compared with AT+/ FVIII– plasma samples. However, there was no statistical difference in any of the TEG variables between AT50/FVIII– and AT20/FVIII– (AT 50% vs AT 20%) samples suggesting that the AT effect was maximized at 50% of normal and further decreases in AT concentration had no additional effects on any of the variables. The addition of rFVIIa (30 or 60 nM) to the FVIII– plasma samples reduced R time, increased the angle, and increased MA (Table 2, Fig. 2) regardless of whether blood samples were activated with 1 or 2 nM of thrombin, and regardless of AT concentration. Decreasing the AT level to 50% of normal further improved the TEG variables, especially by decreasing the R time and increasing the angles shown in Table 2 and Figure 2.
|
|
Thrombin Generation Assay (Thrombinoscope)
A representative tracing of thrombin generation data is shown in Figure 3. Thrombin generation was significantly dependent on normal concentrations of AT and FVIII. The absence of AT (AT–/FVIII+) resulted in more than a three-fold increase in peak thrombin generation when compared with the control (AT+/FVIII+) samples (peak thrombin 450 nM vs 141 nM in control, P = 0.03). In AT+/FVIII– plasma, peak thrombin generation was decreased by approximately 55% to 64.3 nM from 141 nM in controls (Fig. 3). Decreasing the AT level was a dominant factor in determining the corresponding increase in peak thrombin generation, as demonstrated by the lack of statistical difference in peak thrombin generation between AT–/FVIII+ and AT–/FVIII– deficient factor plasma samples (450 vs 465 nM). Decreasing the AT level in the FVIII-deficient plasma samples to 50% of normal increased thrombin generation to a level similar to or higher than that observed with the control samples (178 vs 141 nM, P = 0.8, Fig. 3). Addition of rFVIIa (data not shown) had no effect on peak thrombin generation but it shortened the of R time of thrombin generation.
|
DISCUSSION
In this investigation, we measured the effects of AT and rFVIIa on thrombus formation and thrombin generation in factor VIII-deficient plasma. We found that in factor VIII-deficient plasma decreasing the AT level to between 20% and 50% of normal significantly decreased the R time and increased the angle and MA of thrombus formation as measured with TEG. Modulation of thrombus formation by AT was enhanced by the addition of 30 or 60 nM rFVIIa (Fig. 2, Table 2). The results from our thrombin generation experiments clearly showed that AT activity inversely affected thrombin generation in FVIII-deficient plasma and that, by decreasing the AT concentration to 50% of normal, the peak thrombin generation in factor VIII-deficient plasma was normalized (178 vs141 nM). Therefore, it can be inferred that the normalization of the TEG variables by AT in FVIII-deficient plasma is due to an increased thrombin generation resulting in a faster and stronger thrombus formation.
Normally two protease inhibitors, TFPI and AT, separately regulate the activity of TF/FVIIa complex, FXa, and thrombin.15 It is believed that these processes allow a rapid but regulated hemostatic response that is limited to the sites of vascular injury since it is initiated in relation to the extent of TF exposure (Fig. 1). In hemophiliacs, muscles and joints are often subjected to acute or repeated acute hemorrhage due to less TF expression compared with vital organs.4 In this study, we examined thrombin generated by the TF/FVIIa pathway and whether associated thrombus formation could be enhanced by decreased AT levels, since AT neutralizes FXa and thrombin. In vitro models of severe hemophilia A combined with factor V Leiden demonstrated that thrombin generation is increased by three- to seven-fold depending on the factor V Leiden concentration in the system.16 Factor V Leiden is an inherited prothrombotic condition where Factor V cannot be efficiently inactivated by activated protein C resulting in sustained thrombin generation. In neonatal plasma depleted of FVIII, thrombin generation is nearly normal because TF pathway inhibitor and AT levels are approximately 50% of adult levels.17 Our results extend these findings showing that administration of rFVIIa in an in vitro model of hemophilia A is affected by levels of natural anticoagulant factors such as AT.
The use of rFVIIa is rapidly expanding outside of its approved indication for hemostatic therapy in hemophilic patients with inhibitors to recombinant FVIII. In this situation, neutralizing antibodies develop in response to replacement therapy with the missing factor.18 Large dose rFVIIa (150–300 µg/kg) generates FXa by both TF-dependent and independent mechanisms.19 The resulting increased levels of FXa are more likely to not be inactivated by TFPI or AT (Fig. 1). It was recently recognized that the rate of adverse events from rFVIIa therapy is significantly higher in surgical patients than in hemophiliacs. Among 185 thromboembolic events reported to the Food and Drug Administration's (FDA) Adverse Event Reporting System after rFVIIa administration, 75 events occurred in patients given the drug for surgical bleeding and 17 events occurred in hemophiliacs.7 It is not uncommon to observe levels of AT that are decreased by 40%–60% from normal and/or protein C levels approximately 60% of normal after cardiac surgery with cardiopulmonary bypass.9,10 Further, as acute phase reactants, elevated levels of FVIII-von Willebrand factor are common in nonhemophilia surgical patients due to the stress response.8 This combination of reduced levels of antithrombotic factors and increased procoagulants might thus contribute to a greater response to rFVIIa leading to an enhanced rate of thromboembolic complications. Under low or absent TF-rFVIIa stimulation, thrombin generation is highly dependent on feedback activation of FVIII and FV by trace levels of thrombin (Fig. 1).4,20 In normal plasma, exogenously added thrombin at 2 nM, but not 1 nM, was sufficient to promote thrombin generation and thrombus formation, but this was not the case in FVIII-deficient plasma (Table 1). Lack of factor VIII not only prolongs the time to "thrombin burst" but it also decreases the total amount of generated thrombin, therefore 2 nM exogenously added thrombin is insufficient to promote efficient thrombin and thrombus formation in FVIII-deficient plasma with normal AT level. TEG activated by small amounts of thrombin (2–5 nM) should better reflect thrombin-mediated feedback activation of FVIII and FV than TF activation particularly in the presence of rFVIIa. Wielders et al. have demonstrated the importance of feedback activation using Thrombinoscope,21 and our work essentially extends the same concept to the TEG system.
Our study demonstrated that a 50% reduction of AT enhanced hemostatic effects of rFVIIa; thus, these findings maybe clinically pertinent as to the possible cause of the increased incidence of rFVIIa-related thromboembolic complications in perioperative patients with moderate AT deficiency.7 Although perioperative patients have multifactorial disturbances in coagulation relative to a single factor defect in the hemophilia patient, both demonstrate the common feature of delayed and reduced thrombin burst.1,17,22 The hemostatic response to rFVIIa in nonhemophilia patients may be increased in the presence of reduced AT and stress-induced increase of FVIII-von Willebrand factor.8 On the other hand, moderate reduction of AT may be favorable for achieving hemostasis in hemophiliac patients with FVIII inhibitors undergoing surgery. In either case, an appropriate adjustment of FVIIa dosing and monitoring of AT levels seem prudent to minimize thrombotic complications.
Our current data are limited to in vitro plasma models, which lack major cellular components of hemostasis (e.g., monocyte, platelet). TEG and Thrombinoscope assay are conducted under static conditions, and therefore our results exclude rheological elements. Additional studies are currently in progress to evaluate the modulation of hemophilia phenotype by AT levels in animal models.
Footnotes
Accepted for publication November 2, 2007.
Supported by Department of Anesthesiology, Emory University School of Medicine, Atlanta, Georgia.
Dr. Jerrold H. Levy, Section Editor for Hemostasis and Transfusion Medicine, was recused from all editorial decisions related to this manuscript.
REFERENCES
This article has been cited by other articles:
|
|
 |
|
  K. A. Tanaka, N. S. Key, and J. H. Levy Blood Coagulation: Hemostasis and Thrombin Regulation Anesth. Analg., May 1, 2009; 108(5): 1433 - 1446. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
|
