- Email: [email protected]
Comparison of antithrombotic and haemorrhagic effects of edoxaban, an oral direct factor Xa inhibitor, with warfarin and enoxaparin in rats
Thrombosis Research 130 (2012) 514–519
Contents lists available at SciVerse ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Comparison of antithrombotic and haemorrhagic effects of edoxaban, an oral direct factor Xa inhibitor, with warfarin and enoxaparin in rats Yoshiyuki Morishima ⁎, Yuko Honda, Chikako Kamisato, Naoki Tsuji, Akemi Kita, Naoko Edo, Toshiro Shibano Biological Research Laboratories, R & D Division, Daiichi Sankyo Co., Ltd., 1-2-58 Hiromachi, Shinagawa-ku, Tokyo 140–8710, Japan
a r t i c l e
i n f o
Article history: Received 1 March 2012 Received in revised form 11 April 2012 Accepted 7 May 2012 Available online 29 May 2012 Keywords: Oral anticoagulant ED50 Bleeding time prolongation dose
a b s t r a c t Introduction: Factor Xa (FXa) is a key serine protease in the coagulation cascade and a promising target for a new antithrombotic agent. Edoxaban is an oral, selective and direct FXa inhibitor. The objective of this study was to compare the antithrombotic and haemorrhagic effects of edoxaban with clinically available anticoagulants, warfarin and enoxaparin, in rat models of thrombosis and haemorrhage. Methods: Rats were treated with single oral administration of edoxaban, repeated oral dosing of warfarin for 4 days and single subcutaneous administration of enoxaparin before thrombosis or haemorrhage induction. Thrombosis was induced by the insertion of a platinum wire into the inferior vena cava for 60 min. Tail template bleeding time was measured after making an incision on the tail. Results: Edoxaban at 0.3, 1 and 3 mg/kg exerted dose-dependent and significant inhibition of venous thrombus formation. The 50% thrombus inhibition dose (ED50) was 1.9 mg/kg. At supra-therapeutic doses (10 and 20 mg/kg), edoxaban significantly but moderately (less than 2-fold) prolonged bleeding time. Warfarin and enoxaparin also dose-dependently inhibited venous thrombosis and prolonged bleeding time. The ED50 values of warfarin and enoxaparin were 0.12 mg/kg and 500 IU/kg, and the 2-fold bleeding time prolongation doses (BT2) were 0.16 mg/kg and 1700 IU/kg, respectively. The safety margin (ratio of BT2 to ED50) of edoxaban (> 10.5) was greater than those of warfarin (1.3) and enoxaparin (3.4). Conclusions: Edoxaban inhibited venous thrombosis comparably to warfarin and enoxaparin, and the attendant bleeding risk of edoxaban was lower than that of warfarin and enoxaparin in rats. © 2012 Elsevier Ltd. All rights reserved.
Introduction Anticoagulant therapy has been dominated by oral vitamin K antagonists (e.g. warfarin), parenteral unfractionated heparin and low molecular weight heparins (LMWH) [1–3]. Recently novel oral anticoagulants, including a direct thrombin inhibitor (dabigatran) and factor Xa (FXa) inhibitors (rivaroxaban and apixaban) have been introduced into the anticoagulant market [4]. These new drugs have advantages over warfarin such as lessened interindividual variation of action, faster onset of action, fewer drug-drug or drug-food interactions and wider therapeutic windows. In turn, they offer the principal advantage over unfractionated heparin and LMWH of being orally administered agents and thus are suitable for long-term outpatient therapy.
Abbreviations: FXa, factor Xa; ED50, dose required for 50% inhibition of thrombus formation; BT2, dose required to double bleeding time; LMWH, low molecular weight heparin; PT, prothrombin time; APTT, activated partial thromboplastin time; SEM, standard error of mean. ⁎ Corresponding author. Tel.: + 81 3 3494 3131; fax: + 81 3 5436 8587. E-mail address: [email protected] (Y. Morishima). 0049-3848/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2012.05.008
Edoxaban [the free form of edoxaban tosilate hydrate (Japanese Accepted Names for Pharmaceuticals)] is a novel oral direct factor Xa inhibitor with Ki value of 0.561 nM for human FXa [5]. This compound effectively prevents thrombosis in rat and rabbit models of thrombosis after oral administration [5] and is expected to be a potential replacement for warfarin. In a single ascending dose study in healthy males, edoxaban (10 – 150 mg) is well tolerated and the exposure is proportional to doses. Pharmacokinetic profiles are consistent across doses with rapid absorption, biphasic elimination, and terminal elimination half-life of 5.8 to 10.7 hr. The plasma edoxaban concentrations are linearly correlated with coagulation parameters [6]. Edoxaban tosilate hydrate is currently marketed for the prophylaxis of venous thromboembolism after orthopedic surgery [7–9] in Japan as a once daily oral anticoagulant and is undergoing global phase III studies for the prevention of stroke in patients with atrial fibrillation (ENGAGE AF-TIMI 48, NCT00986154) [10] and treatment and prevention of recurrent venous thromboembolism (HOKUSAI VTE, NCT00781391). The aim of the present study was to compare the antithrombotic effect of edoxaban in a rat model of venous thrombosis with those of warfarin and enoxaparin. We also determined the effects of edoxaban, warfarin and enoxaparin on bleeding time in a rat tail template bleeding model and explored potential safety windows between the antithrombotic effects and haemorrhagic complications.
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
515
Materials and Methods
Effects on Bleeding Time
Materials
Edoxaban (3, 10 and 20 mg/kg) and vehicle (0.5% MC) were orally administered to fasted rats (n = 8 in each group) 30 min before the induction of bleeding. The doses of warfarin were the same as those in the venous thrombosis model: loading doses (0.30, 0.45 and 0.60 mg/kg) on the first day and maintenance doses (0.10, 0.15 and 0.20 mg/kg) on subsequent 3 days (n = 8 in each group). Enoxaparin (800, 1600 and 3200 IU/kg) and vehicle (saline) were subcutaneously administered to fasted rats (n = 8 in each group) 30 min before the induction of bleeding. These doses of edoxaban and enoxaparin were selected because they represent the fully effective and supramaximal doses in the rat venous thrombosis model. A rat tail incision bleeding model was used [5]. Rats were anesthetized with thiopental sodium (100 mg/kg i.p.) 15 min before the incision. Blood samples were collected from the jugular vein for the measurement of PT and APTT as previously described. A 1 mm deep incision was made with a blade on the artery of the ventral aspect of the tail 4 cm from the tip. The blood was blotted every 15 sec with filter paper. Cessation of bleeding was confirmed as the absence of any detectable blood stain on the filter paper for two consecutive intervals of blotting. The bleeding time was defined as the interval from the tail incision to the last detectable blood stain. When bleeding did not stop within 30 min, the bleeding time was defined as 30 min. The bleeding time was measured by investigators blinded to study treatment.
Edoxaban tosilate hydrate (Japanese Accepted Names for Pharmaceuticals) was synthesized at Daiichi Sankyo (Tokyo, Japan). Warfarin sodium and 0.5 w/v% methyl cellulose 400 solution (0.5% MC) were purchased from Wako Pure Chemical Industries (Osaka, Japan). Enoxaparin sodium was obtained from Sanofi-Aventis (Tokyo, Japan). The doses of edoxaban are expressed as the amount of the anhydrous freebase. The doses of warfarin are expressed as the amount of the salt form. Platinum wires (diameter 0.5 mm, length 2.5 cm) were obtained from Eikoukagaku (Tokyo, Japan). Thiopental sodium was obtained from Mitsubishi Tanabe Pharma (Osaka, Japan). Coomassie Plus – the better Bradford assay reagent was obtained from Thermo Fisher Scientific (Waltham, MA). Thromboplastin C Plus and Platelin LS II were purchased from Sysmex (Kobe, Japan) and Kyowa Medex (Tokyo, Japan), respectively. Animals All experimental procedures were performed in accordance with the in-house guideline of the Institutional Animal Care and Use Committee of Daiichi Sankyo. Male Slc:Wistar rats weighing 209 – 258 g were purchased from Japan SLC (Hamamatsu, Japan) and maintained on a 7:00 am/7:00 pm light/dark schedule. Rats were housed 4 – 6 per cage and food and water were available ad libitum. Antithrombotic Effects Edoxaban and warfarin were dissolved in 0.5% MC. Enoxaparin was diluted with saline. Edoxaban (0.3, 1 and 3 mg/kg) and vehicle (0.5% MC) were orally administered to fasted rats (n = 8 in each group) 30 min before the induction of thrombosis. Warfarin and vehicle (0.5% MC) was orally administered to rats with loading doses (0.30, 0.45 and 0.60 mg/kg) on the first day and maintenance doses (0.10, 0.15 and 0.20 mg/kg) on subsequent 3 days (n = 8 in each group). Thrombosis was induced 24 hr after the last warfarin dose. Enoxaparin (100, 300 and 1000 IU/kg) and vehicle (saline) were subcutaneously administered to fasted rats (n = 8 in each group) 30 min before the induction of thrombosis. The rats were anesthetized with thiopental sodium (100 mg/kg, i.p.) 15 min before the thrombus induction, and placed on heating mats at approximately 37 °C to maintain body temperature. The inferior vena cava was exposed through a midline incision. Blood samples were collected from the jugular vein for the measurement of prothrombin time (PT) and activated partial thromboplastin time (APTT) 3 min before the thrombus induction. Blood was collected into a syringe containing 1/10 volume of 3.13% sodium citrate tribasic dihydrate. Venous thrombosis was induced as follows [5]. A platinum wire was inserted 2 cm into the inferior vena cava just below the left renal vein caudally, towards the lower limbs, and remained in place for 60 min. Immediately before removing the thrombus, 1 mL of 1% glutaraldehyde was injected into the inferior vena cava to fix the thrombus in situ, and 1 mL of 3.13% sodium citrate tribasic dihydrate was injected to the site to block any subsequent thrombus formation. The wire with the thrombus was immediately dissected free and washed with saline. The protein content of the thrombus was measured by the Bradford method with Coomassie Plus – the better Bradford assay reagent and a spectrophotometer SpectraMax 190 or SpectraMax PLUS384 and the analysis software SoftMax Pro 5.2 or 5.4 (Molecular Devices, Sunnyvale, CA). Bovine serum albumin was used as a standard.
Measurement of PT and APTT PT was measured with a microcoagulometer (Amelung KC-10A micro, Trinity Biotech, Bray, Ireland) as follows. Plasma (50 μL) and 50 μL of saline were added to a cuvette and pre-incubated at 37 °C for 1 min. Coagulation was started by the addition of 100 μL of Thromboplastin C Plus to the mixture, and the clotting time was measured. APTT was measured with the microcoagulometer as follows. Plasma (50 μL), 25 μL of saline and 50 μL of Platelin LS II were added to a cuvette and pre-incubated at 37 °C for 5 min. Coagulation was started by the addition of 25 μL of 50 mM CaCl2 solution, and the clotting time was measured. Statistical Analysis All data are expressed as mean ± standard error of mean (SEM). Statistical significance between the control and treatment groups was analyzed by the Dunnett test. The dose–response relationships were evaluated by Spearman's rank correlation coefficient. The statistical significance level was p b 0.05. Doses required for 50% inhibition of thrombus formation (ED50) and doses required to double bleeding time (BT2) were calculated as follows. Two doses at which the means of inhibition (%) of thrombus formation were adjacent to 50% were selected. Then, the ED50 was estimated by linear regression analysis [11]. To calculate BT2, the dose group at which the mean of prolongation (fold) of bleeding time was above 2-fold and the next lower dose group were selected. Then, the BT2 was estimated by linear regression analysis. These statistical analyses were performed using SAS System Release 8.2 (SAS Institute Inc., Cary, NC). Results Antithrombotic Effects The placement of a platinum wire in the rat inferior vena cava induced thrombus formation on the surface of the wire. The protein content of the thrombi in the control groups of edoxaban, warfarin and enoxaparin were 128 ± 12, 186 ± 33 and 167 ± 26 μg, respectively (Figs. 1A, 2A and 3A).
516
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
Fig. 1. Antithrombotic effect of edoxaban in a rat platinum wire-induced venous thrombosis model. (A) Thrombus protein content, (B) Prothrombin time (PT), (C) activated partial thromboplastin time (APTT). Edoxaban was orally administered 30 min before the platinum wire placement in the inferior vena cava. Blood samples were drawn 3 min before the thrombus induction. Data represent the mean ± SEM (n = 8). ** P b 0.01, *** P b 0.001 vs control.
Edoxaban (0.30 to 3.0 mg/kg, p.o.) dose-dependently and statistically significantly inhibited thrombus formation (P b 0.01 or 0.001, Fig. 1A). The ED50 value of edoxaban was 1.9 mg/kg (Table 1). Edoxaban at 3.0 mg/kg significantly prolonged PT (P b 0.01), but had no effect on APTT (Fig. 1B and C).
Fig. 2. Antithrombotic effect of warfarin in a rat platinum-wire-induced venous thrombosis model. (A) Thrombus protein content, (B) Prothrombin time (PT), (C) activated partial thromboplastin time (APTT). Warfarin was orally administered for 4 days (1-day loading dose and following 3-day maintenance doses). The doses of warfarin cited are the maintenance doses. Thrombosis was induced 24 hr after the last dosing. Blood samples were drawn 3 min before the thrombus induction. Data represent the mean ± SEM (n = 8). ** P b 0.01, *** P b 0.001 vs control.
Warfarin (0.15 and 0.20 mg/kg, p.o. as maintenance doses) and enoxaparin (1000 IU/kg, s.c.) statistically significantly inhibited thrombus formation (P b 0.001, Figs. 2A and 3A) with the ED50 values of 0.12 mg/kg and 500 IU/kg, respectively (Table 1). Warfarin at 0.15 and
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
517
Table 1 ED50, BT2 and therapeutic indexes (BT2/ED50) values of edoxaban, warfarin and enoxaparin. Agent
Antithrombotic effect (ED50)
Bleeding effect (BT2)
Therapeutic index (BT2/ED50)
Edoxaban Warfarin Enoxaparin
1.9 mg/kg 0.12 mg/kg 500 IU/kg
>20 mg/kg 0.16 mg/kg 1700 IU/kg
>10.5 1.3 3.4
ED50: dose required for 50% inhibition of thrombus formation. BT2: dose required to double bleeding time.
Effects on Bleeding Time In the control groups of edoxaban, warfarin, and enoxaparin, reproducible bleeding times were observed with mean values of 3.16 ± 0.26 min, 4.13 ± 0.30 min and 3.31 ± 0.39 min, respectively (Fig. 4). All three anticoagulants used in this study dose-dependently prolonged bleeding time (Fig. 4). Edoxaban at 10 and 20 mg/kg, p.o. significantly prolonged bleeding time 1.72-fold (P b 0.01) and 1.67fold (P b 0.05), respectively (Fig. 4A). The BT2 value of edoxaban was more than 20 mg/kg (Table 1). Warfarin at 0.20 mg/kg, p.o. (as a maintenance dose) statistically significantly prolonged bleeding time 4.19-fold (P b 0.001, Fig. 4B). The BT2 value was 0.16 mg/kg (Table 1). Enoxaparin at 3200 IU/kg, s.c. exerted a significant 6.52fold prolongation of bleeding time (P b 0.001, Fig. 4C) with the BT2 value of 1700 IU/kg (Table 1). The dose response curves of the antithrombotic effects and bleedingtime prolongation of these anticoagulants are shown in Fig. 5. The therapeutic index (calculated as the BT2 divided by the antithrombotic ED50) of edoxaban was> 10.5, whereas that of warfarin and enoxaparin were 1.3 and 3.4, respectively (Table 1), indicating the wider therapeutic window of edoxaban than warfarin and enoxaparin. Discussion
Fig. 3. Antithrombotic effect of enoxaparin in a rat platinum wire-induced venous thrombosis model. (A) Thrombus protein content, (B) Prothrombin time (PT), (C) activated partial thromboplastin time (APTT). Enoxaparin was subcutaneously administered 30 min before the platinum-wire placement in the inferior vena cava. Blood samples were drawn 3 min before the thrombus induction. Data represent the mean± SEM (n= 8). ** P b 0.01, *** P b 0.001 vs control.
0.20 mg/kg statistically significantly prolonged PT and APTT (P b 0.01 or 0.001, Fig. 2B). Enoxaparin prolonged APTT at 300 and 1000 IU/kg (P b 0.01 or 0.001), whereas PT prolongation by 1000 IU/kg enoxaparin (P b 0.001) was less pronounced than APTT (Fig. 3B and C).
In phase-III studies of edoxaban for the prevention of stroke in patients with atrial fibrillation [10] and for the prevention of venous thromboembolism in patients undergoing orthopedic surgeries [7–9], warfarin and enoxaparin are used as active comparators. In the present study, we determined the antithrombotic and bleeding time prolongation effects of these anticoagulants. We chose rat models of venous thrombosis and tail bleeding because they have been frequently used to characterize the antithrombotic efficacy and haemorrhagic complications of anticoagulants [5,12,13]. In a rat venous thrombosis model, pretreatment with a single oral dose of edoxaban resulted in a dose-dependent inhibition of thrombosis similar to 4-day repeated dose of warfarin and a single subcutaneous injection of enoxaparin. Thrombosis in patients with atrial fibrillation and patients undergoing orthopedic surgeries is caused by stagnant blood and/or vascular damage. The prevention of venous thromboembolism by edoxaban was demonstrated in patients undergoing orthopedic surgery in completed phase III studies. And edoxaban can be expected to prevent thrombosis in patients with atrial fibrillation. Warfarin blocks the blood coagulation cascade by inhibiting the synthesis of vitamin K-dependent coagulation factors, factors II, VII, IX, and X [1]. The onset of action of warfarin is slow because of the need to deplete these coagulation factors from the circulation. In contrast, edoxaban is a direct inhibitor of FXa and exhibits a rapid onset of action, exerting an antithrombotic effect from as early as 30 min after a single oral dose in rat thrombosis models [5]. Furthermore, in a rat venous thrombosis model, warfarin prolonged the PT (19.7 sec in control to 40.3 sec with warfarin) at an antithrombotic dose (77.5% inhibition), whereas edoxaban caused minimal PT prolongation (18.4 sec in control to 23.1 sec with edoxaban) at an antithrombotic dose (61.4% inhibition). These results indicate that the
518
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
Fig. 4. Effects of edoxaban, warfarin and enoxaparin on tail-bleeding time in rats. (A) Edoxaban, (B) warfarin, (C) enoxaparin. Bleeding time was measured in a tail template bleeding model in rats. Data is expressed as mean ± SEM (n = 8). * P b 0.05, ** P b 0.01, *** P b 0.001 vs control.
relationships between the antithrombotic activity and PT prolongation of these anticoagulants differ. Therefore, the warfarin-based concept of a therapeutic international normalized ratio (INR) with a target range of 2.0 to 3.0 cannot be applied to edoxaban. Significant inhibition of thrombosis at doses that cause minimal prolongation of the clotting time may be a common property of direct factor Xa inhibitors [14].
Fig. 5. Dose-relationship of antithrombotic effects and bleeding-time prolongation of edoxaban, warfarin and enoxaparin in rats. (A) Edoxaban, (B) warfarin, (C) enoxaparin. Antithrombotic effect (% control of thrombus formation) and bleeding-time prolongation (ratio to control) were plotted against doses. Data represent the mean ± SEM (n = 8).
Enoxaparin forms a complex with antithrombin, a physiological anticoagulant factor in the blood, and enhances the ability of antithrombin to inhibit primarily FXa and to a lesser degree thrombin in an irreversible manner [3]. However, enoxaparin is a parenteral agent, which imposes a considerable burden upon both patients and caregivers. Edoxaban is orally active and thus has an important clinical advantage in this regard over enoxaparin. In the current rat-tail bleeding model, edoxaban did not affect bleeding time at a therapeutic dose (3 mg/kg), but did so at higher
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
doses (10 and 20 mg/kg). The bleeding time prolongation by edoxaban was less pronounced than warfarin and enoxaparin. Edoxaban prolonged the bleeding time by less than 2-fold even at a dose 10 times higher than its antithrombotic ED50, whereas warfarin and enoxaparin prolonged the bleeding time 4.19-fold and 6.52-fold at doses 1.7 times and 6.4 times higher than their antithrombotic ED50, respectively. The therapeutic index of edoxaban (> 10.5) was greater than warfarin (1.3) and enoxaparin (3.4) (Table 1), suggesting that edoxaban may demonstrate a clinical benefit relative to warfarin and enoxaparin in terms of a lower bleeding risk. Completed phase III studies conducted in Japan for the prevention of thromboembolic events following orthopedic surgery demonstrated that edoxaban 30 mg qd has superior efficacy to enoxaparin 2,000 IU bid with a similar or lower incidence of bleeding [7–9]. Thus it appears that the safety margin of edoxaban is greater than that of enoxaparin in humans. The reason why the risk/benefit ratio of edoxaban appears to be better than that of warfarin and enoxaparin is not clear. One critical factor might be a reversible and specific inhibition of the single coagulation factor, FXa, by edoxaban, whereas warfarin inhibits the synthesis of multiple coagulation factors and enoxaparin irreversibly inhibits FXa and thrombin. This reversible and specific inhibition of FXa by edoxaban may result in efficient inhibition of coagulation at the site of thrombus formation without systemic hyperanticoagulation. Another element in favor of edoxaban might be its lack of effect on platelet aggregation. Edoxaban does not inhibit platelet aggregation induced by adenosine diphosphate, collagen and a thromboxane A2 agonist [5]. Bleeding is a common complication of anticoagulant therapy. Therefore, antidotes or reversal agents for anticoagulants are highly desirable [15]. In terms of reversal agents for edoxaban, we have demonstrated that recombinant factor VIIa significantly reduces bleeding induced by a supratherapeutic dose of edoxaban in a rat model and recombinant factor VIIa, Feiba (activated prothrombin complex concentrate), and PPSB-HT (prothrombin complex concentrate) effectively reverse the anticoagulant effect of edoxaban in vitro [16]. Thus, these haemostatic agents may prove useful in managing bleeding in the setting of edoxaban therapy. In conclusion, edoxaban inhibited venous thrombosis comparably to warfarin and enoxaparin. The attendant bleeding risk of edoxaban may be lower than that of warfarin and enoxaparin as evidenced by a lesser effect on bleeding time in a rat-tail bleeding model. Conflict of Interest Statement All authors are employees of Daiichi Sankyo Co., Ltd.
519
Acknowledgments We thank Ms. Chikako Nishimoto for her technical assistance.
References [1] Van Aken H, Bode C, Darius H, Diehm C, Encke A, Gulba DC, et al. Anticoagulation: the present and future. Clin Appl Thromb Hemost 2001;7:195–204. [2] Heparin Hirsh J. N Engl J Med 1991;324:1565–74. [3] Weitz JI. Low-molecular-weight-heparins. N Engl J Med 1997;337:688–98. [4] Bauer KA. Recent progress in anticoagulant therapy: oral direct inhibitors of thrombin and factor Xa. J Thromb Haemost 2011;9(Suppl. 1):12–9. [5] Furugohri T, Isobe K, Honda Y, Kamisato-Matsumoto C, Sugiyama N, Nagahara T, et al. DU-176b, a potent and orally active factor Xa inhibitor: in vitro and in vivo pharmacological profiles. J Thromb Haemost 2008;6:1542–9. [6] Ogata K, Mendell-Harary J, Tachibana M, Masumoto H, Oguma T, Kojima M, et al. Clinical safety, tolerability, pharmacokinetics, and pharmacodynamics of the novel factor Xa inhibitor edoxaban in healthy volunteers. J Clin Pharmacol 2010;50: 743–53. [7] Fuji T, Wang CJ, Fujita S, Tachibana S, Kawai Y, Koretsune Y, et al. Edoxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty: the STARS E-3 Trial. Pathophysiol Haemost Thromb 2010;37:A20 [OC297]. [8] Fujita S, Fuji T, Tachibana S, Nakamura, Kawai Y. Safety and efficacy of edoxaban in patients undergoing hip fracture surgery. Pathophysiol Haemost Thromb 2010;37: A95 [P366]. [9] Fuji T, Fujita S, Tachibana S, Kawai Y, Koretsune Y, Yamashita T, et al. Efficacy and Safety of Edoxaban Versus Enoxaparin for the Prevention of Venous Thromboembolism Following Total Hip Arthroplasty: STARS J-V trial. Blood 2010;116 [Abstract 3320]. [10] Ruff CT, Giugliano RP, Antman EM, Crugnale SE, Bocanegra T, Mercuri M, et al. Evaluation of the novel factor Xa inhibitor edoxaban compared with warfarin in patients with atrial fibrillation: design and rationale for the Effective aNticoaGulation with factor xA next GEneration in Atrial Fibrillation-Thrombolysis In Myocardial Infarction study 48 (ENGAGE AF-TIMI 48). Am Heart J 2010;160: 635–41. [11] Wienen W, Stassen JM, Priepke H, Ries UJ, Hauel N. Effects of the direct thrombin inhibitor dabigatran and its orally active prodrug, dabigatran etexilate, on thrombus formation and bleeding time in rats. Thromb Haemost 2007;98:333–8. [12] Lavelle SM, Iomhair MM. The reliability of thrombotic tendency measured by intravascular wires in the rat. Thromb Res 1997;87:353–7. [13] Morishima Y, Tanabe K, Terada Y, Hara T, Kunitada S. Antithrombotic and hemorrhagic effects of DX-9065a, a direct and selective factor Xa inhibitor: comparison with a direct thrombin inhibitor and antithrombin III-dependent anticoagulants. Thromb Haemost 1997;78:1366–71. [14] Leadley Jr RJ. Coagulation factor Xa inhibition: biological background and rationale. Curr Top Med Chem 2001;1:151–9. [15] Levi MM, Eerenberg E, Löwenberg E, Kamphuisen PW. Bleeding in patients using new anticoagulants or antiplatelet agents: risk factors and management. Neth J Med 2010;68:68–76. [16] Fukuda T, Honda Y, Kamisato C, Morishima Y, Shibano T. Reversal of anticoagulant effects of edoxaban, an oral, direct factor Xa inhibitor, with haemostatic agents. Thromb Haemost 2011;107:253–9.
Contents lists available at SciVerse ScienceDirect
Thrombosis Research journal homepage: www.elsevier.com/locate/thromres
Regular Article
Comparison of antithrombotic and haemorrhagic effects of edoxaban, an oral direct factor Xa inhibitor, with warfarin and enoxaparin in rats Yoshiyuki Morishima ⁎, Yuko Honda, Chikako Kamisato, Naoki Tsuji, Akemi Kita, Naoko Edo, Toshiro Shibano Biological Research Laboratories, R & D Division, Daiichi Sankyo Co., Ltd., 1-2-58 Hiromachi, Shinagawa-ku, Tokyo 140–8710, Japan
a r t i c l e
i n f o
Article history: Received 1 March 2012 Received in revised form 11 April 2012 Accepted 7 May 2012 Available online 29 May 2012 Keywords: Oral anticoagulant ED50 Bleeding time prolongation dose
a b s t r a c t Introduction: Factor Xa (FXa) is a key serine protease in the coagulation cascade and a promising target for a new antithrombotic agent. Edoxaban is an oral, selective and direct FXa inhibitor. The objective of this study was to compare the antithrombotic and haemorrhagic effects of edoxaban with clinically available anticoagulants, warfarin and enoxaparin, in rat models of thrombosis and haemorrhage. Methods: Rats were treated with single oral administration of edoxaban, repeated oral dosing of warfarin for 4 days and single subcutaneous administration of enoxaparin before thrombosis or haemorrhage induction. Thrombosis was induced by the insertion of a platinum wire into the inferior vena cava for 60 min. Tail template bleeding time was measured after making an incision on the tail. Results: Edoxaban at 0.3, 1 and 3 mg/kg exerted dose-dependent and significant inhibition of venous thrombus formation. The 50% thrombus inhibition dose (ED50) was 1.9 mg/kg. At supra-therapeutic doses (10 and 20 mg/kg), edoxaban significantly but moderately (less than 2-fold) prolonged bleeding time. Warfarin and enoxaparin also dose-dependently inhibited venous thrombosis and prolonged bleeding time. The ED50 values of warfarin and enoxaparin were 0.12 mg/kg and 500 IU/kg, and the 2-fold bleeding time prolongation doses (BT2) were 0.16 mg/kg and 1700 IU/kg, respectively. The safety margin (ratio of BT2 to ED50) of edoxaban (> 10.5) was greater than those of warfarin (1.3) and enoxaparin (3.4). Conclusions: Edoxaban inhibited venous thrombosis comparably to warfarin and enoxaparin, and the attendant bleeding risk of edoxaban was lower than that of warfarin and enoxaparin in rats. © 2012 Elsevier Ltd. All rights reserved.
Introduction Anticoagulant therapy has been dominated by oral vitamin K antagonists (e.g. warfarin), parenteral unfractionated heparin and low molecular weight heparins (LMWH) [1–3]. Recently novel oral anticoagulants, including a direct thrombin inhibitor (dabigatran) and factor Xa (FXa) inhibitors (rivaroxaban and apixaban) have been introduced into the anticoagulant market [4]. These new drugs have advantages over warfarin such as lessened interindividual variation of action, faster onset of action, fewer drug-drug or drug-food interactions and wider therapeutic windows. In turn, they offer the principal advantage over unfractionated heparin and LMWH of being orally administered agents and thus are suitable for long-term outpatient therapy.
Abbreviations: FXa, factor Xa; ED50, dose required for 50% inhibition of thrombus formation; BT2, dose required to double bleeding time; LMWH, low molecular weight heparin; PT, prothrombin time; APTT, activated partial thromboplastin time; SEM, standard error of mean. ⁎ Corresponding author. Tel.: + 81 3 3494 3131; fax: + 81 3 5436 8587. E-mail address: [email protected] (Y. Morishima). 0049-3848/$ – see front matter © 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.thromres.2012.05.008
Edoxaban [the free form of edoxaban tosilate hydrate (Japanese Accepted Names for Pharmaceuticals)] is a novel oral direct factor Xa inhibitor with Ki value of 0.561 nM for human FXa [5]. This compound effectively prevents thrombosis in rat and rabbit models of thrombosis after oral administration [5] and is expected to be a potential replacement for warfarin. In a single ascending dose study in healthy males, edoxaban (10 – 150 mg) is well tolerated and the exposure is proportional to doses. Pharmacokinetic profiles are consistent across doses with rapid absorption, biphasic elimination, and terminal elimination half-life of 5.8 to 10.7 hr. The plasma edoxaban concentrations are linearly correlated with coagulation parameters [6]. Edoxaban tosilate hydrate is currently marketed for the prophylaxis of venous thromboembolism after orthopedic surgery [7–9] in Japan as a once daily oral anticoagulant and is undergoing global phase III studies for the prevention of stroke in patients with atrial fibrillation (ENGAGE AF-TIMI 48, NCT00986154) [10] and treatment and prevention of recurrent venous thromboembolism (HOKUSAI VTE, NCT00781391). The aim of the present study was to compare the antithrombotic effect of edoxaban in a rat model of venous thrombosis with those of warfarin and enoxaparin. We also determined the effects of edoxaban, warfarin and enoxaparin on bleeding time in a rat tail template bleeding model and explored potential safety windows between the antithrombotic effects and haemorrhagic complications.
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
515
Materials and Methods
Effects on Bleeding Time
Materials
Edoxaban (3, 10 and 20 mg/kg) and vehicle (0.5% MC) were orally administered to fasted rats (n = 8 in each group) 30 min before the induction of bleeding. The doses of warfarin were the same as those in the venous thrombosis model: loading doses (0.30, 0.45 and 0.60 mg/kg) on the first day and maintenance doses (0.10, 0.15 and 0.20 mg/kg) on subsequent 3 days (n = 8 in each group). Enoxaparin (800, 1600 and 3200 IU/kg) and vehicle (saline) were subcutaneously administered to fasted rats (n = 8 in each group) 30 min before the induction of bleeding. These doses of edoxaban and enoxaparin were selected because they represent the fully effective and supramaximal doses in the rat venous thrombosis model. A rat tail incision bleeding model was used [5]. Rats were anesthetized with thiopental sodium (100 mg/kg i.p.) 15 min before the incision. Blood samples were collected from the jugular vein for the measurement of PT and APTT as previously described. A 1 mm deep incision was made with a blade on the artery of the ventral aspect of the tail 4 cm from the tip. The blood was blotted every 15 sec with filter paper. Cessation of bleeding was confirmed as the absence of any detectable blood stain on the filter paper for two consecutive intervals of blotting. The bleeding time was defined as the interval from the tail incision to the last detectable blood stain. When bleeding did not stop within 30 min, the bleeding time was defined as 30 min. The bleeding time was measured by investigators blinded to study treatment.
Edoxaban tosilate hydrate (Japanese Accepted Names for Pharmaceuticals) was synthesized at Daiichi Sankyo (Tokyo, Japan). Warfarin sodium and 0.5 w/v% methyl cellulose 400 solution (0.5% MC) were purchased from Wako Pure Chemical Industries (Osaka, Japan). Enoxaparin sodium was obtained from Sanofi-Aventis (Tokyo, Japan). The doses of edoxaban are expressed as the amount of the anhydrous freebase. The doses of warfarin are expressed as the amount of the salt form. Platinum wires (diameter 0.5 mm, length 2.5 cm) were obtained from Eikoukagaku (Tokyo, Japan). Thiopental sodium was obtained from Mitsubishi Tanabe Pharma (Osaka, Japan). Coomassie Plus – the better Bradford assay reagent was obtained from Thermo Fisher Scientific (Waltham, MA). Thromboplastin C Plus and Platelin LS II were purchased from Sysmex (Kobe, Japan) and Kyowa Medex (Tokyo, Japan), respectively. Animals All experimental procedures were performed in accordance with the in-house guideline of the Institutional Animal Care and Use Committee of Daiichi Sankyo. Male Slc:Wistar rats weighing 209 – 258 g were purchased from Japan SLC (Hamamatsu, Japan) and maintained on a 7:00 am/7:00 pm light/dark schedule. Rats were housed 4 – 6 per cage and food and water were available ad libitum. Antithrombotic Effects Edoxaban and warfarin were dissolved in 0.5% MC. Enoxaparin was diluted with saline. Edoxaban (0.3, 1 and 3 mg/kg) and vehicle (0.5% MC) were orally administered to fasted rats (n = 8 in each group) 30 min before the induction of thrombosis. Warfarin and vehicle (0.5% MC) was orally administered to rats with loading doses (0.30, 0.45 and 0.60 mg/kg) on the first day and maintenance doses (0.10, 0.15 and 0.20 mg/kg) on subsequent 3 days (n = 8 in each group). Thrombosis was induced 24 hr after the last warfarin dose. Enoxaparin (100, 300 and 1000 IU/kg) and vehicle (saline) were subcutaneously administered to fasted rats (n = 8 in each group) 30 min before the induction of thrombosis. The rats were anesthetized with thiopental sodium (100 mg/kg, i.p.) 15 min before the thrombus induction, and placed on heating mats at approximately 37 °C to maintain body temperature. The inferior vena cava was exposed through a midline incision. Blood samples were collected from the jugular vein for the measurement of prothrombin time (PT) and activated partial thromboplastin time (APTT) 3 min before the thrombus induction. Blood was collected into a syringe containing 1/10 volume of 3.13% sodium citrate tribasic dihydrate. Venous thrombosis was induced as follows [5]. A platinum wire was inserted 2 cm into the inferior vena cava just below the left renal vein caudally, towards the lower limbs, and remained in place for 60 min. Immediately before removing the thrombus, 1 mL of 1% glutaraldehyde was injected into the inferior vena cava to fix the thrombus in situ, and 1 mL of 3.13% sodium citrate tribasic dihydrate was injected to the site to block any subsequent thrombus formation. The wire with the thrombus was immediately dissected free and washed with saline. The protein content of the thrombus was measured by the Bradford method with Coomassie Plus – the better Bradford assay reagent and a spectrophotometer SpectraMax 190 or SpectraMax PLUS384 and the analysis software SoftMax Pro 5.2 or 5.4 (Molecular Devices, Sunnyvale, CA). Bovine serum albumin was used as a standard.
Measurement of PT and APTT PT was measured with a microcoagulometer (Amelung KC-10A micro, Trinity Biotech, Bray, Ireland) as follows. Plasma (50 μL) and 50 μL of saline were added to a cuvette and pre-incubated at 37 °C for 1 min. Coagulation was started by the addition of 100 μL of Thromboplastin C Plus to the mixture, and the clotting time was measured. APTT was measured with the microcoagulometer as follows. Plasma (50 μL), 25 μL of saline and 50 μL of Platelin LS II were added to a cuvette and pre-incubated at 37 °C for 5 min. Coagulation was started by the addition of 25 μL of 50 mM CaCl2 solution, and the clotting time was measured. Statistical Analysis All data are expressed as mean ± standard error of mean (SEM). Statistical significance between the control and treatment groups was analyzed by the Dunnett test. The dose–response relationships were evaluated by Spearman's rank correlation coefficient. The statistical significance level was p b 0.05. Doses required for 50% inhibition of thrombus formation (ED50) and doses required to double bleeding time (BT2) were calculated as follows. Two doses at which the means of inhibition (%) of thrombus formation were adjacent to 50% were selected. Then, the ED50 was estimated by linear regression analysis [11]. To calculate BT2, the dose group at which the mean of prolongation (fold) of bleeding time was above 2-fold and the next lower dose group were selected. Then, the BT2 was estimated by linear regression analysis. These statistical analyses were performed using SAS System Release 8.2 (SAS Institute Inc., Cary, NC). Results Antithrombotic Effects The placement of a platinum wire in the rat inferior vena cava induced thrombus formation on the surface of the wire. The protein content of the thrombi in the control groups of edoxaban, warfarin and enoxaparin were 128 ± 12, 186 ± 33 and 167 ± 26 μg, respectively (Figs. 1A, 2A and 3A).
516
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
Fig. 1. Antithrombotic effect of edoxaban in a rat platinum wire-induced venous thrombosis model. (A) Thrombus protein content, (B) Prothrombin time (PT), (C) activated partial thromboplastin time (APTT). Edoxaban was orally administered 30 min before the platinum wire placement in the inferior vena cava. Blood samples were drawn 3 min before the thrombus induction. Data represent the mean ± SEM (n = 8). ** P b 0.01, *** P b 0.001 vs control.
Edoxaban (0.30 to 3.0 mg/kg, p.o.) dose-dependently and statistically significantly inhibited thrombus formation (P b 0.01 or 0.001, Fig. 1A). The ED50 value of edoxaban was 1.9 mg/kg (Table 1). Edoxaban at 3.0 mg/kg significantly prolonged PT (P b 0.01), but had no effect on APTT (Fig. 1B and C).
Fig. 2. Antithrombotic effect of warfarin in a rat platinum-wire-induced venous thrombosis model. (A) Thrombus protein content, (B) Prothrombin time (PT), (C) activated partial thromboplastin time (APTT). Warfarin was orally administered for 4 days (1-day loading dose and following 3-day maintenance doses). The doses of warfarin cited are the maintenance doses. Thrombosis was induced 24 hr after the last dosing. Blood samples were drawn 3 min before the thrombus induction. Data represent the mean ± SEM (n = 8). ** P b 0.01, *** P b 0.001 vs control.
Warfarin (0.15 and 0.20 mg/kg, p.o. as maintenance doses) and enoxaparin (1000 IU/kg, s.c.) statistically significantly inhibited thrombus formation (P b 0.001, Figs. 2A and 3A) with the ED50 values of 0.12 mg/kg and 500 IU/kg, respectively (Table 1). Warfarin at 0.15 and
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
517
Table 1 ED50, BT2 and therapeutic indexes (BT2/ED50) values of edoxaban, warfarin and enoxaparin. Agent
Antithrombotic effect (ED50)
Bleeding effect (BT2)
Therapeutic index (BT2/ED50)
Edoxaban Warfarin Enoxaparin
1.9 mg/kg 0.12 mg/kg 500 IU/kg
>20 mg/kg 0.16 mg/kg 1700 IU/kg
>10.5 1.3 3.4
ED50: dose required for 50% inhibition of thrombus formation. BT2: dose required to double bleeding time.
Effects on Bleeding Time In the control groups of edoxaban, warfarin, and enoxaparin, reproducible bleeding times were observed with mean values of 3.16 ± 0.26 min, 4.13 ± 0.30 min and 3.31 ± 0.39 min, respectively (Fig. 4). All three anticoagulants used in this study dose-dependently prolonged bleeding time (Fig. 4). Edoxaban at 10 and 20 mg/kg, p.o. significantly prolonged bleeding time 1.72-fold (P b 0.01) and 1.67fold (P b 0.05), respectively (Fig. 4A). The BT2 value of edoxaban was more than 20 mg/kg (Table 1). Warfarin at 0.20 mg/kg, p.o. (as a maintenance dose) statistically significantly prolonged bleeding time 4.19-fold (P b 0.001, Fig. 4B). The BT2 value was 0.16 mg/kg (Table 1). Enoxaparin at 3200 IU/kg, s.c. exerted a significant 6.52fold prolongation of bleeding time (P b 0.001, Fig. 4C) with the BT2 value of 1700 IU/kg (Table 1). The dose response curves of the antithrombotic effects and bleedingtime prolongation of these anticoagulants are shown in Fig. 5. The therapeutic index (calculated as the BT2 divided by the antithrombotic ED50) of edoxaban was> 10.5, whereas that of warfarin and enoxaparin were 1.3 and 3.4, respectively (Table 1), indicating the wider therapeutic window of edoxaban than warfarin and enoxaparin. Discussion
Fig. 3. Antithrombotic effect of enoxaparin in a rat platinum wire-induced venous thrombosis model. (A) Thrombus protein content, (B) Prothrombin time (PT), (C) activated partial thromboplastin time (APTT). Enoxaparin was subcutaneously administered 30 min before the platinum-wire placement in the inferior vena cava. Blood samples were drawn 3 min before the thrombus induction. Data represent the mean± SEM (n= 8). ** P b 0.01, *** P b 0.001 vs control.
0.20 mg/kg statistically significantly prolonged PT and APTT (P b 0.01 or 0.001, Fig. 2B). Enoxaparin prolonged APTT at 300 and 1000 IU/kg (P b 0.01 or 0.001), whereas PT prolongation by 1000 IU/kg enoxaparin (P b 0.001) was less pronounced than APTT (Fig. 3B and C).
In phase-III studies of edoxaban for the prevention of stroke in patients with atrial fibrillation [10] and for the prevention of venous thromboembolism in patients undergoing orthopedic surgeries [7–9], warfarin and enoxaparin are used as active comparators. In the present study, we determined the antithrombotic and bleeding time prolongation effects of these anticoagulants. We chose rat models of venous thrombosis and tail bleeding because they have been frequently used to characterize the antithrombotic efficacy and haemorrhagic complications of anticoagulants [5,12,13]. In a rat venous thrombosis model, pretreatment with a single oral dose of edoxaban resulted in a dose-dependent inhibition of thrombosis similar to 4-day repeated dose of warfarin and a single subcutaneous injection of enoxaparin. Thrombosis in patients with atrial fibrillation and patients undergoing orthopedic surgeries is caused by stagnant blood and/or vascular damage. The prevention of venous thromboembolism by edoxaban was demonstrated in patients undergoing orthopedic surgery in completed phase III studies. And edoxaban can be expected to prevent thrombosis in patients with atrial fibrillation. Warfarin blocks the blood coagulation cascade by inhibiting the synthesis of vitamin K-dependent coagulation factors, factors II, VII, IX, and X [1]. The onset of action of warfarin is slow because of the need to deplete these coagulation factors from the circulation. In contrast, edoxaban is a direct inhibitor of FXa and exhibits a rapid onset of action, exerting an antithrombotic effect from as early as 30 min after a single oral dose in rat thrombosis models [5]. Furthermore, in a rat venous thrombosis model, warfarin prolonged the PT (19.7 sec in control to 40.3 sec with warfarin) at an antithrombotic dose (77.5% inhibition), whereas edoxaban caused minimal PT prolongation (18.4 sec in control to 23.1 sec with edoxaban) at an antithrombotic dose (61.4% inhibition). These results indicate that the
518
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
Fig. 4. Effects of edoxaban, warfarin and enoxaparin on tail-bleeding time in rats. (A) Edoxaban, (B) warfarin, (C) enoxaparin. Bleeding time was measured in a tail template bleeding model in rats. Data is expressed as mean ± SEM (n = 8). * P b 0.05, ** P b 0.01, *** P b 0.001 vs control.
relationships between the antithrombotic activity and PT prolongation of these anticoagulants differ. Therefore, the warfarin-based concept of a therapeutic international normalized ratio (INR) with a target range of 2.0 to 3.0 cannot be applied to edoxaban. Significant inhibition of thrombosis at doses that cause minimal prolongation of the clotting time may be a common property of direct factor Xa inhibitors [14].
Fig. 5. Dose-relationship of antithrombotic effects and bleeding-time prolongation of edoxaban, warfarin and enoxaparin in rats. (A) Edoxaban, (B) warfarin, (C) enoxaparin. Antithrombotic effect (% control of thrombus formation) and bleeding-time prolongation (ratio to control) were plotted against doses. Data represent the mean ± SEM (n = 8).
Enoxaparin forms a complex with antithrombin, a physiological anticoagulant factor in the blood, and enhances the ability of antithrombin to inhibit primarily FXa and to a lesser degree thrombin in an irreversible manner [3]. However, enoxaparin is a parenteral agent, which imposes a considerable burden upon both patients and caregivers. Edoxaban is orally active and thus has an important clinical advantage in this regard over enoxaparin. In the current rat-tail bleeding model, edoxaban did not affect bleeding time at a therapeutic dose (3 mg/kg), but did so at higher
Y. Morishima et al. / Thrombosis Research 130 (2012) 514–519
doses (10 and 20 mg/kg). The bleeding time prolongation by edoxaban was less pronounced than warfarin and enoxaparin. Edoxaban prolonged the bleeding time by less than 2-fold even at a dose 10 times higher than its antithrombotic ED50, whereas warfarin and enoxaparin prolonged the bleeding time 4.19-fold and 6.52-fold at doses 1.7 times and 6.4 times higher than their antithrombotic ED50, respectively. The therapeutic index of edoxaban (> 10.5) was greater than warfarin (1.3) and enoxaparin (3.4) (Table 1), suggesting that edoxaban may demonstrate a clinical benefit relative to warfarin and enoxaparin in terms of a lower bleeding risk. Completed phase III studies conducted in Japan for the prevention of thromboembolic events following orthopedic surgery demonstrated that edoxaban 30 mg qd has superior efficacy to enoxaparin 2,000 IU bid with a similar or lower incidence of bleeding [7–9]. Thus it appears that the safety margin of edoxaban is greater than that of enoxaparin in humans. The reason why the risk/benefit ratio of edoxaban appears to be better than that of warfarin and enoxaparin is not clear. One critical factor might be a reversible and specific inhibition of the single coagulation factor, FXa, by edoxaban, whereas warfarin inhibits the synthesis of multiple coagulation factors and enoxaparin irreversibly inhibits FXa and thrombin. This reversible and specific inhibition of FXa by edoxaban may result in efficient inhibition of coagulation at the site of thrombus formation without systemic hyperanticoagulation. Another element in favor of edoxaban might be its lack of effect on platelet aggregation. Edoxaban does not inhibit platelet aggregation induced by adenosine diphosphate, collagen and a thromboxane A2 agonist [5]. Bleeding is a common complication of anticoagulant therapy. Therefore, antidotes or reversal agents for anticoagulants are highly desirable [15]. In terms of reversal agents for edoxaban, we have demonstrated that recombinant factor VIIa significantly reduces bleeding induced by a supratherapeutic dose of edoxaban in a rat model and recombinant factor VIIa, Feiba (activated prothrombin complex concentrate), and PPSB-HT (prothrombin complex concentrate) effectively reverse the anticoagulant effect of edoxaban in vitro [16]. Thus, these haemostatic agents may prove useful in managing bleeding in the setting of edoxaban therapy. In conclusion, edoxaban inhibited venous thrombosis comparably to warfarin and enoxaparin. The attendant bleeding risk of edoxaban may be lower than that of warfarin and enoxaparin as evidenced by a lesser effect on bleeding time in a rat-tail bleeding model. Conflict of Interest Statement All authors are employees of Daiichi Sankyo Co., Ltd.
519
Acknowledgments We thank Ms. Chikako Nishimoto for her technical assistance.
References [1] Van Aken H, Bode C, Darius H, Diehm C, Encke A, Gulba DC, et al. Anticoagulation: the present and future. Clin Appl Thromb Hemost 2001;7:195–204. [2] Heparin Hirsh J. N Engl J Med 1991;324:1565–74. [3] Weitz JI. Low-molecular-weight-heparins. N Engl J Med 1997;337:688–98. [4] Bauer KA. Recent progress in anticoagulant therapy: oral direct inhibitors of thrombin and factor Xa. J Thromb Haemost 2011;9(Suppl. 1):12–9. [5] Furugohri T, Isobe K, Honda Y, Kamisato-Matsumoto C, Sugiyama N, Nagahara T, et al. DU-176b, a potent and orally active factor Xa inhibitor: in vitro and in vivo pharmacological profiles. J Thromb Haemost 2008;6:1542–9. [6] Ogata K, Mendell-Harary J, Tachibana M, Masumoto H, Oguma T, Kojima M, et al. Clinical safety, tolerability, pharmacokinetics, and pharmacodynamics of the novel factor Xa inhibitor edoxaban in healthy volunteers. J Clin Pharmacol 2010;50: 743–53. [7] Fuji T, Wang CJ, Fujita S, Tachibana S, Kawai Y, Koretsune Y, et al. Edoxaban versus enoxaparin for thromboprophylaxis after total knee arthroplasty: the STARS E-3 Trial. Pathophysiol Haemost Thromb 2010;37:A20 [OC297]. [8] Fujita S, Fuji T, Tachibana S, Nakamura, Kawai Y. Safety and efficacy of edoxaban in patients undergoing hip fracture surgery. Pathophysiol Haemost Thromb 2010;37: A95 [P366]. [9] Fuji T, Fujita S, Tachibana S, Kawai Y, Koretsune Y, Yamashita T, et al. Efficacy and Safety of Edoxaban Versus Enoxaparin for the Prevention of Venous Thromboembolism Following Total Hip Arthroplasty: STARS J-V trial. Blood 2010;116 [Abstract 3320]. [10] Ruff CT, Giugliano RP, Antman EM, Crugnale SE, Bocanegra T, Mercuri M, et al. Evaluation of the novel factor Xa inhibitor edoxaban compared with warfarin in patients with atrial fibrillation: design and rationale for the Effective aNticoaGulation with factor xA next GEneration in Atrial Fibrillation-Thrombolysis In Myocardial Infarction study 48 (ENGAGE AF-TIMI 48). Am Heart J 2010;160: 635–41. [11] Wienen W, Stassen JM, Priepke H, Ries UJ, Hauel N. Effects of the direct thrombin inhibitor dabigatran and its orally active prodrug, dabigatran etexilate, on thrombus formation and bleeding time in rats. Thromb Haemost 2007;98:333–8. [12] Lavelle SM, Iomhair MM. The reliability of thrombotic tendency measured by intravascular wires in the rat. Thromb Res 1997;87:353–7. [13] Morishima Y, Tanabe K, Terada Y, Hara T, Kunitada S. Antithrombotic and hemorrhagic effects of DX-9065a, a direct and selective factor Xa inhibitor: comparison with a direct thrombin inhibitor and antithrombin III-dependent anticoagulants. Thromb Haemost 1997;78:1366–71. [14] Leadley Jr RJ. Coagulation factor Xa inhibition: biological background and rationale. Curr Top Med Chem 2001;1:151–9. [15] Levi MM, Eerenberg E, Löwenberg E, Kamphuisen PW. Bleeding in patients using new anticoagulants or antiplatelet agents: risk factors and management. Neth J Med 2010;68:68–76. [16] Fukuda T, Honda Y, Kamisato C, Morishima Y, Shibano T. Reversal of anticoagulant effects of edoxaban, an oral, direct factor Xa inhibitor, with haemostatic agents. Thromb Haemost 2011;107:253–9.