Direct Thrombin Inhibitors, but Not Factor Xa Inhibitors, Enhance Thrombin Formation in Human Plasma by Interfering with the Thrombin–Thrombomodulin–Protein C System

Blood ◽  
2008 ◽  
Vol 112 (11) ◽  
pp. 528-528 ◽  
Author(s):  
Elisabeth Perzborn ◽  
Michaela Harwardt

Abstract Activation of protein C is dependent on thrombin complexed with thrombomodulin (TM). Activated protein C (APC), together with its cofactor protein S, degrades coagulation Factors Va and VIIIa, thereby limiting further thrombin formation. Thus, in addition to suppressing the procoagulant effects of thrombin, direct thrombin inhibitors (DTIs) may also downregulate anticoagulant effects of thrombin-mediated feedback mechanisms. By contrast, direct Factor Xa (FXa) inhibitors block the formation of thrombin, but not its actions. The objective of this study was to investigate whether the direct FXa inhibitor, rivaroxaban, and the DTIs, dabigatran and melagatran, inhibit the negative-feedback reaction of the thrombin–TM complex/APC (thrombin–TM/APC) system and thereby increase thrombin formation. Experiments were conducted in plateletpoor plasma from healthy donors (normal plasma) and in pooled protein C-deficient plasma, both substituted with 1.33 μM phospholipids, in the presence or absence of 10 nM TM with increasing concentrations of rivaroxaban, dabigatran, melagatran, or the appropriate vehicles. Thrombin formation was initiated by adding 1.67 pM tissue factor (TF) and assessed by measuring the cleavage of the fluorogenic substrate Z-Gly-Gly-Arg-AMC (Bachem) using the Calibrated Automated Thrombogram (CAT, Thrombinoscope® BV) method. The parameters assessed were lag time, time to peak thrombin generation (tmax), peak thrombin generation (Cmax), and endogenous thrombin potential (ETP). In addition, formation of prothrombin fragments 1+2 (F1+2) was determined by ELISA (Enzygnost® F1+2 monoclonal [Dade Behring]). Rivaroxaban potently inhibited thrombin formation in the absence and presence of TM across all parameters in a concentration-dependent manner in both normal plasma and protein C-deficient plasma (see Table). In the absence of TM, melagatran and dabigatran also inhibited thrombin formation in a concentration-dependent manner, both in normal plasma and protein C-deficient plasma. In the presence of TM, DTIs prolonged lag time and tmax in a concentration-dependent manner. However, only high concentrations of the DTIs reduced ETP, Cmax, and F1+2, In normal plasma,lower concentrations even increased ETP, Cmax, and F1+2. Increased thrombin formation was observed with melagatran 119–474 nM or dabigatran 68–545 nM. DTIs did not increase thrombin formation in protein C-deficient plasma, suggesting that both protein C and TM are needed for the DTI-mediated increase in thrombin formation. The results suggest that low concentrations of DTIs suppress the anticoagulant effects of the thrombin–TM/APC system by inhibiting activation of protein C by the thrombin–TM complex, and thereby enhance thrombin formation. Conversely, rivaroxaban does not increase thrombin formation, suggesting that it does not suppress the negative-feedback reaction by inhibition of protein C activation. This hypothesis is supported by the absence of enhanced thrombin formation in protein C-deficient plasma. Enhanced thrombin formation might explain the hypercoagulation observed with DTIs in a rat model of TF-induced intravascular coagulation (Furugohri, et al. 2005; Morishima, et al. 2005; Perzborn, et al. 2008) and suggests that DTIs could cause activation of coagulation at lower plasma concentrations. Table. Effect of rivaroxaban, dabigatran, and melagatran on peak thrombin formation (Cmax [nM thrombin]) in the absence or presence of thrombomodulin (TM) in normal plasma (NP) from healthy volunteers (n=8–12), and in pooled protein C-deficient plasma (PPC) in the presence of TM (n=3). Results were obtained by the CAT method and are a mean of n plasma samples. Prothrombin fragments F1+2 (nM F1+2) results were obtained in the presence of TM (mean results; n=3 [NP]). Thrombin (nM) in normal and in protein C-deficient plasma n.d:, no data. Rivaroxaban (nM) 0 18 91 182 363 1,090 Cmax: − TM in NP 273 222 113 71 43 15 Cmax: + TM in NP 100 67 32 18 10 3 Cmax: + TM in PPC 290 249 168 117 80 38 F1+2: + TM (in NP) 184 89 43 18 n.d. 2 Dabigatran (nM) 0 68 136 273 545 1,090 Cmax: − TM in NP 261 287 289 282 239 102 Cmax: + TM in NP 81 156 203 240 218 96 Cmax: + TM in PPC 298 301 300 292 253 116 F1+2: + TM in NP 252 n.d. 362 422 351 98 Melagatran(nM) 0 24 119 237 474 948 Cmax: − TM in NP 276 290 297 289 251 127 Cmax: + TM in NP 101 133 215 251 237 113 Cmax: + TM in PPC 294 296 299 292 256 132 F1+2: + TM in NP 213 n.d. 389 427 393 123

2011 ◽  
Vol 106 (12) ◽  
pp. 1076-1083 ◽  
Author(s):  
Nobutoshi Sugiyama ◽  
Yoshiyuki Morishima ◽  
Toshiro Shibano ◽  
Taketoshi Furugohri

SummaryThere is increasing concern that some anticoagulants can paradoxically increase thrombogenesis under certain circumstances. Previously, we demonstrated that at certain doses a direct thrombin inhibitor, melag-atran, worsens the coagulation status induced by tissue factor (TF) in-jection in a rat model. We utilised an in vitro thrombin generation (TG) assay to determine if direct thrombin inhibitors could enhance TG in human plasma, and whether inhibition of the negative-feedback sys-tem [thrombin-thrombomodulin (TM)-protein C] contributed to the TG enhancement. TG in human plasma was assayed by means of the cali-brated automated thrombography. In this assay, direct factor Xa (FXa) inhibitors such as edoxaban and antithrombin (AT)-dependent anti-coagulants such as heparin did not increase, but simply suppressed TG. AT-independent thrombin inhibitors (melagatran, lepirudin, and active site blocked thrombin (IIai)) increased peak levels of TG (2.0, 1.6, and 2.2-fold, respectively) in the presence of 12 nM recombinant human soluble TM (rhsTM). Melagatran and lepirudin at higher concentrations began to suppress TG. In the absence of rhsTM, the enhancement of peak TG by melagatran decreased to 1.2-fold. Furthermore, in protein C-deficient plasma, AT-independent thrombin inhibitors failed to enhance TG. In addition, a human protein C neutralising antibody increased the peak height of TG in the presence of rhsTM. These results suggest that AT-independent thrombin inhibitors may activate throm-bogenesis by suppression of the thrombin-induced negative-feedback system through inhibition of protein C activation. In contrast, direct FXa inhibitors are more useful than AT-independent thrombin inhibitors in terms of lower possibility of activation of the coagulation pathway.


2009 ◽  
Vol 101 (01) ◽  
pp. 185-196 ◽  
Author(s):  
Katrien Devreese ◽  
Kathelijne Peerlinck ◽  
Jef Arnout ◽  
Marc Hoylaerts

SummaryLupus anticoagulants (LAC) consist of antiphospholipid antibodies, detected via their anticoagulant properties in vitro. Strong LAC relate to thromboembolic events, a hallmark of the anti-phospholipid syndrome. We have analyzed whether detection of this syndrome would benefit from thrombin generation measurements. Therefore, calibrated automated thrombography was done in normal plasma (n=30) and LAC patient plasma (n=48 non-anticoagulated, n=12 on oral anticoagulants), diluted 1:1 with a normal plasma pool. The anti-β2-glycoprotein I monoclonal antibody 23H9, with known LAC properties, delayed the lag time and reduced the peak height during thrombin generation induction in normal plasma dose-dependently (0–150 μg/ml). At variance, LAC patient 1:1 plasma mixtures manifested variable lag time prolongations and/or peak height reductions. Coupling these two most informative thrombin generation parameters in a peak height/lag time ratio, and upon normalization versus the normal plasma pool, this ratio distributed normally and was reduced in the plasma mixtures, for 59/60 known LAC plasmas. The normalized peak height/lag time ratio correlated well with the normalized dilute prothrombin time, diluted Russell’s viper venom time and silica clotting time, measured in 1:1 plasma mixtures (correlation coefficients 0.59–0.72). The anticoagulant effects of activated protein C (0–7.5 nM) or 23H9 (0–150 μg/ml), spiked in the 1:1 LAC plasma mixtures were reduced for the majority of patients, compatible with functional competition between patient LAC and activated protein C and LAC and 23H9, respectively. Hence, the normalized thrombin generation-derived peak height/lag time ratio identifies LAC in plasma with high sensitivity in a single assay, irrespective of the patient’s treatment with oral anticoagulants.


2009 ◽  
Vol 101 (05) ◽  
pp. 827-833 ◽  
Author(s):  
Kenichi Tanaka ◽  
Fania Szlam ◽  
Christopher Rusconi ◽  
Jerrold Levy

SummaryThe REG1 system consists of factor IXa inhibitor, RB006, an ap-tamer-based anticoagulant and its antidote, RB007. The optimal use of RB006 can be facilitated by understanding its effect on the formation of thrombin and fibrin, and other standard tests of coagulation. Blood from consented volunteers was drawn into 3.2% citrate (9:1 v/v) and either used immediately or centrifuged to obtain platelet-poor plasma. Increasing concentrations of ap-tamer (6–24 μg/ml) alone or in combination with heparin (0.1 U/ml) or lepirudin (0.2 μg/ml) were added to blood and plasma samples. Activated clotting times (ACT+, low range-ACT), thrombelastometry (ROTEM™) or thrombelastography (TEG®) were performed in recalcified whole blood samples. Thrombin generation, prothrombin time (PT) and activated partial thromboplastin time (aPTT) were performed in plasma samples. To some samples the antidote RB007 was added to neutralise the anticoagulation activity of RB006. In all experiment the ratio of RB006 to RB007 was kept 1:2. RB006 dose-dependently prolonged aPTT and low range-ACT, but, as expected, had no effect on PT. RB006 prolonged the lag time and decreased the peak of Actin-triggered thrombin generation. Thrombin-activated TEG demonstrated that RB006 decreases the rate of clot formation. These effects were potentiated when RB006 was combined with heparin or lepirudin. In all experiments RB007 reversed the effects of RB006 back to baseline. In conclusion, RB006 inhibits thrombin generation and clot formation in a concentration-dependent manner. It is feasible to monitor RB006 and its reversal with RB007 using aPTT, low range-ACT, and thrombin-activated TEG.


2012 ◽  
Vol 32 (suppl_1) ◽  
Author(s):  
Yiming Xu ◽  
Weizhen Wu ◽  
Andrew S Plump ◽  
Madhu Chintala ◽  
Martin L Ogletree ◽  
...  

Thrombin is a central enzyme in haemostasis and thrombosis, and a proven target for anticoagulant therapies. Different classes of thrombin inhibitors, while exerting therapeutic benefits in most clinical trials, have different indications, dosing regimens, and bleeding complications. To gain more insight into the underlying mechanisms for their differential clinical profiles, we compared four marketed and representative agents, including dabigatran, hirudin, bivalirudin (direct thrombin inhibitors, DTIs), and heparin (an indirect thrombin inhibitor), in two in vitro spike-in assays with concentration titrations covering their therapeutic ranges. The two assays were the Thrombinoscope TGA with plasma, triggered by low tissue factor (1 nM TF), and TEG with whole blood, triggered by 1:8000 Recombiplastin (equivalent to low TF), with or without a threshold level of tPA to induce fibrinolysis. In TGA, the largest effect was prolongation of lag time, with the potency of the three DTIs rank-ordered as hirudin>dabigatran>bivalirudin; regarding peak, slope, and ETP, while complete inhibition was achieved with 1-2 μM dabigatran or hirudin, bivalirudin had no effect even at 4 μM, possibly due to its short half life in plasma. In TEG, the three DTIs prolonged clotting time (R) in the same rank order as TGA; for clot strength (MA), while all four agents reduced MA in synergy with tPA, only hirudin reduced MA without tPA, likely due to its highest potency. With tPA-induced fibrinolytic activity (Ly30), dabigatran and bivalirudin enhanced Ly30 (dabigatran>bivalirudin), but hirudin and heparin did not. This contrast might involve differential access to clot-bound thrombin. Heparin had a steep dose-response curve for both lag time in TGA and R in TEG, which is in line with its very narrow therapeutic index. All three DTIs, but not heparin, displayed the previously reported paradoxical increase in peak and slope in TGA in the low concentration range, suggesting this is indeed a class effect of DTI. In summary, our observations highlight the distinct features of each agent in thrombin generation, coagulation, and fibrinolysis. These results in combination with known clinical properties are informative on efforts to define the optimal profiles of new anticoagulants.


Blood ◽  
2005 ◽  
Vol 106 (11) ◽  
pp. 1877-1877
Author(s):  
Kathleen M. Welch ◽  
Kamlai Saiya-Cork ◽  
Weston R. Gould ◽  
Robert J. Leadley

Abstract Inhibition of either coagulation factor Xa (FXa) or thrombin (FIIa) alters ex vivo biomarkers of coagulation and decreases thrombus size in animal models of experimental thrombosis. The objective of this study was to determine if combining FXa and FIIa inhibitors would synergistically reduce the magnitude of IIa generated without elevating markers of bleeding. A synergistic combination may translate into lower doses, more effective anticoagulation and better safety. To predict the FXa and FIIa inhibitor combinations with the maximum potential for synergy, PD 0313052, a potent, selective FXa inhibitor, and argatroban, a potent FIIa inhibitor, were each tested independently and in combination using an in vitro thrombin generation assay. Individually, PD 0313052 and argatroban reduced total thrombin generation (TG) in a concentration dependent manner with IC50’s of 497±148 and 882±193 nM, respectively. Subsequently, PD 0313052 and argatroban were combined in 96 well plates at concentrations ranging from 0.125x to 8x their respective IC50 concentrations. Analysis using the Bliss Independence Model identified statistically significant synergistic activity, with the greatest increase (33%) over simple additivity at 249 and 441 nM FXa to FIIa, respectively, both below their respective IC50 concentrations. Furthermore, combinations of PD 0313052 and argatroban were evaluated in an assay measuring activated clotting time (ACT). Although both PD 0313052 and argatroban dose-dependently elongated the ACT, the combination of 0.5x the TG IC50 concentrations, which demonstrated the greatest synergy in the TG assay, showed the smallest increase in the ACT, prolonging clotting time by only 15%. These data demonstrate that the combination of a specific factor Xa inhibitor and a specific IIa inhibitor can synergistically reduce thrombin generation without appreciable elevation of ACT, suggesting that dual inhibition of FXa and FIIa, using relatively low doses of each, may provide efficacious and safe treatment for thromboembolic diseases.


Blood ◽  
2008 ◽  
Vol 112 (11) ◽  
pp. 22-22
Author(s):  
Fionnuala Ni Ainle ◽  
Roger JS Preston ◽  
Vince P Jenkins ◽  
Jennifer A Johnson ◽  
Shona Harmon ◽  
...  

Abstract Protamine sulphate is a positively-charged polypeptide widely used to reverse heparin-induced anticoagulation. Paradoxically, protamine also possesses intrinsic anticoagulant properties. Furthermore, administration of excess protamine in the neutralization of UFH is associated with increased bleeding, particularly following cardiothoracic surgery. In this study we have investigated the molecular mechanisms underlying the anticoagulant properties of protamine. In pooled normal plasma, we observed a dose-dependent prolongation of both PT and APTT assays with increasing protamine concentrations (0–30μg/ml). The anticoagulant effects of protamine in normal plasma were also examined using a tissue factor-initiated thrombin generation assay. 30μg/ml protamine resulted in a two-fold prolongation of lag-time, a two-fold reduction in peak thrombin generation, and a 41±17% (p=0.047) decrease in endogenous thrombin potential (ETP). In heparinised plasma (0.3U/ml), addition of increasing protamine concentration initially reversed the anticoagulant effect of heparin, resulting in a progressive increase in ETP to that of normal plasma. However, further increases in protamine resulted in a dose-dependent reduction in ETP to a minimum of 61±16%. Recent studies have shown that platelet factor 4 (PF4), another cationic protein used to reverse heparin, can bind to the anionic Gla domain of protein C, thereby enhancing APC generation (up to 25-fold). Consequently, we investigated potential interaction(s) between protamine and the protein C anticoagulant pathway. As expected, in normal plasma, APC (0–20nM) caused a concentration-dependent prolongation in APTT to 180±8%, and a parallel reduction in ETP. However in the presence of 30μg/ml protamine, the effects of APC on both the APTT and ETP were markedly enhanced compared to the effect of either substance alone. As APC down-regulates thrombin generation by inactivating FVa and FVIIIa, we used a phospholipid-dependent FVa proteolysis assay to elucidate the mechanism responsible for the synergistic interaction between APC and protamine. The ability of APC to reduce FVa cofactor activity in this assay (in the presence or absence of protein S) was not significantly affected by the presence of protamine. However, a potent synergistic anticoagulant interaction between APC and protamine was also observed in plasma from patients with homozygous FV Leiden, suggesting protamine enhances APC cleavage of FVa at position Arg-306. To determine whether protamine influences the rate of FVIIIa proteolysis by APC, we expressed and purified an APC-resistant FVIII variant (R336Q/R562Q). FVIII-deficient plasma was spiked with physiological concentrations of wild type or variant FVIII, and the anticoagulant effects of protamine ± APC studied using plasma thrombin generation assays. Similar synergistic anticoagulant effects of APC in combination with protamine were observed for both wildtype and variant FVIII. To assess whether protamine also influences procoagulant processes, the rate of factor V activation by thrombin was analysed by SDS-PAGE. In the presence of protamine (30μg/ml), FVa generation was significantly reduced. In addition to inhibiting the rate of FVa generation, we also observed that protamine significantly impaired the functional activity of the prothrombinase complex in a concentration dependent manner. In contrast, cationic polybrene had no significant effect on either rate of FVa generation or prothrombinase complex activity. In conclusion, we have shown a novel and profound anticoagulant synergy between protamine sulphate and APC. Moreover, we demonstrate that this synergistic effect is mediated by independent effects on FVa generation and proteolysis respectively. These novel findings provide further insights into the molecular mechanism underlying the anticoagulant effect of excess protamine in human plasma.


2010 ◽  
Vol 30 (04) ◽  
pp. 212-216 ◽  
Author(s):  
R. Jovic ◽  
M. Hollenstein ◽  
P. Degiacomi ◽  
M. Gautschi ◽  
A. Ferrández ◽  
...  

SummaryThe activated partial thromboplastin time test (aPTT) represents one of the most commonly used diagnostic tools in order to monitor patients undergoing heparin therapy. Expression of aPTT coagulation time in seconds represents common practice in order to evaluate the integrity of the coagulation cascade. The prolongation of the aPTT thus can indicate whether or not the heparin level is likely to be within therapeutic range. Unfortunately aPTT results are highly variable depending on patient properties, manufacturer, different reagents and instruments among others but most importantly aPTT’s dose response curve to heparin often lacks linearity. Furthermore, aPTT assays are insensitive to drugs such as, for example, low molecular weight heparin (LMWH) and direct factor Xa (FXa) inhibitors among others. On the other hand, the protrombinase-induced clotting time assay (PiCT®) has been show to be a reliable functional assay sensitive to all heparinoids as well as direct thrombin inhibitors (DTIs). So far, the commercially available PiCT assay (Pefakit®-PiCT®, DSM Nutritional Products Ltd. Branch Pentapharm, Basel, Switzerland) is designed to express results in terms of units with the help of specific calibrators, while aPTT results are most commonly expressed as coagulation time in seconds. In this report, we describe the results of a pilot study indicating that the Pefakit PiCT UC assay is superior to the aPTT for the efficient monitoring of patients undergoing UFH therapy; it is also suitable to determine and quantitate the effect of LMWH therapy. This indicates a distinct benefit when using this new approach over the use of aPPT for heparin monitoring.


1997 ◽  
Vol 78 (04) ◽  
pp. 1215-1220 ◽  
Author(s):  
D Prasa ◽  
L Svendsen ◽  
J Stürzebecher

SummaryA series of inhibitors of factor Xa (FXa) were investigated using the thrombin generation assay to evaluate the potency and specificity needed to efficiently block thrombin generation in activated human plasma. By inhibiting FXa the generation of thrombin in plasma is delayed and decreased. Inhibitor concentrations which cause 50 percent inhibition of thrombin generation (IC50) correlate in principle with the Ki values for inhibition of free FXa. Recombinant tick anticoagulant peptide (r-TAP) is able to inhibit thrombin generation with considerably low IC50 values of 49 nM and 37 nM for extrinsic and intrinsic activation, respectively. However, the potent synthetic, low molecular weight inhibitors of FXa (Ki values of about 20 nM) are less effective in inhibiting the generation of thrombin with IC50 values at micromolar concentrations.The overall effect of inhibitors of FXa in the thrombin generation assay was compared to that of thrombin inhibitors. On the basis of similar Ki values for the inhibition of the respective enzyme, synthetic FXa inhibitors are less effective than thrombin inhibitors. In contrast, the highly potent FXa inhibitor r-TAP causes a stronger reduction of the thrombin activity in plasma than the most potent thrombin inhibitor hirudin.


1992 ◽  
Vol 68 (03) ◽  
pp. 297-300 ◽  
Author(s):  
Monica Galli ◽  
Paul Comfurius ◽  
Tiziano Barbui ◽  
Robert F A Zwaal ◽  
Edouard M Bevers

SummaryPlasmas of 16 patients positive for both IgG anticardiolipin (aCL) antibodies and lupus anticoagulant (LA) antibodies were subjected to adsorption with liposomes containing cardiolipin. In 5 of these plasmas both the anticardiolipin and the anticoagulant activities were co-sedimented with the liposomes in a dose-dependent manner, whereas in the remaining cases only the anticardiolipin activity could be removed by the liposomes, leaving the anticoagulant activity (LA) in the supernatant plasma. aCL antibodies purified from the first 5 plasmas were defined as aCL-type A, while the term aCL-type B was used for antibodies in the other 11 plasmas, from which 2 were selected for this study.Prolongation of the dRVVT was produced by affinity-purified aCL-type A antibodies in plasma of human as well as animal (bovine, rat and goat) origin. aCL-type B antibodies were found to be devoid of anticoagulant activity, while the corresponding supernatants containing LA IgG produced prolongation of the dRVVT only in human plasma.These anticoagulant activities of aCL-type A and of LA IgG's were subsequently evaluated in human plasma depleted of β2-glycoprotein I (β2-GPI), a protein which was previously shown to be essential in the binding of aCL antibodies to anionic phospholipids. Prolongation of the dRVVT by aCL-type A antibodies was abolished using β2-GPI deficient plasma, but could be restored upon addition of β2-GPI. In contrast, LA IgG caused prolongation of the dRVVT irrespective of the presence or absence of β2-GPI.Since β2-GPI binds to negatively-charged phospholipids and impedes the conversion of prothrombin by the factor Xa/Va enzyme complex (Nimpf et al., Biochim Biophys Acta 1986; 884: 142–9), comparison was made of the effect of aCL-type A and aCL-type B antibodies on the rate of thrombin formation in the presence and absence of β2-GPI. This was measured in a system containing highly purified coagulation factors Xa, Va and prothrombin and lipid vesicles composed of 20 mole% phosphatidylserine and 80 mole% phosphatidylcholine. No inhibition on the rate of thrombin formation was observed with both types of aCL antibodies when either β2-GPI or the lipid vesicles were omitted. Addition of β2-GPI to the prothrombinase assay in the presence of lipid vesicles causes a time-dependent inhibition which was not affected by the presence of aCL-type B or non-specific IgG. In contrast, the presence of aCL-type A antibodies dramatically increased the anticoagulant effect of β2-GPI. These data indicate that the anticoagulant activity of aCL-type A antibodies in plasma is mediated by β2-GPI.


Sign in / Sign up

Export Citation Format

Share Document