2005.BAY 59-7939 ...phase II dose ranging study by sitebay


									Journal of Thrombosis and Haemostasis, 3: 2479–2486


BAY 59-7939: an oral, direct Factor Xa inhibitor for the
prevention of venous thromboembolism in patients after total
knee replacement. A phase II dose-ranging study
A. G. G. TURPIE,* W. D. FISHER,  K. A. BAUER,à L. M. KWONG,§ M. W. IRWIN,– P. KALEBO,**               ¨
F . M I S S E L W I T Z ,     and M . G E N T * F O R T H E O D I X a - K N E E S T U D Y G R O U P
*HHS-General Hospital, Hamilton, Canada;  McGill University Hospital Centre, Montreal, Canada; àBeth Israel Deaconess Medical Center,
Boston, MA, USA; §Harbor-UCLA Medical Center, Torrance, CA, USA; –Bayer Inc., Toronto, Canada; **Sahlgrenska University Hospital/Ostra,
Goteborg, Sweden; and   Bayer HealthCare AG, Wuppertal, Germany

To cite this article: Turpie AGG, Fisher WD, Bauer KA, Kwong LM, Irwin MW, Kalebo P, Misselwitz F, Gent M, for the ODIXa-KNEE Study Group.
BAY 59-7939: an oral, direct Factor Xa inhibitor for the prevention of venous thromboembolism in patients after total knee replacement. A phase II
dose-ranging study. J Thromb Haemost 2005; 3: 2479–86.

                                                                           with enoxaparin. Bleeding endpoints were lower for the
Summary. Background: BAY 59-7939, a novel, oral, direct                    2.5–10 mg b.i.d. doses compared with higher doses of
factor Xa inhibitor, is in clinical development for the prevention         BAY 59-7939. Conclusions: Oral administration of 2.5–10 mg
of venous thromboembolism (VTE), a frequent complication                   b.i.d. of BAY 59-7939, early in the postoperative period,
following orthopaedic surgery. Methods: In a multicenter,                  showed potential efficacy and an acceptable safety profile,
parallel-group, double-blind, double-dummy study, 621                      similar to enoxaparin, for the prevention of VTE in patients
patients undergoing elective total knee replacement were                   undergoing elective total knee replacement.
randomly assigned to oral BAY 59-7939 (2.5, 5, 10, 20, and
30 mg b.i.d., initiated 6–8 h postsurgery), or subcutaneous                Keywords: direct Factor Xa inhibitor, oral anticoagulant,
enoxaparin (30 mg b.i.d., initiated 12–24 h postsurgery). Treat-           prophylaxis, total knee replacement, venous thromboembo-
ment was continued until mandatory bilateral venography                    lism.
5–9 days after surgery. The primary efficacy endpoint was a
composite of any deep vein thrombosis (proximal and/or                     Introduction
distal), confirmed non-fatal pulmonary embolism and all-cause
mortality during treatment. The primary safety endpoint was                Venous thromboembolism (VTE) after elective total knee
major, postoperative bleeding during treatment. Results: Of                replacement is well characterized; without thromboprophyl-
the 613 patients treated, 366 (59.7%) were evaluable for the               axis, venographic deep vein thrombosis (DVT) at 7–14 days
primary efficacy analysis. The primary efficacy endpoint                       occurs in 41–85% of patients, with proximal DVT rates of
occurred in 31.7%, 40.4%, 23.3%, 35.1%, and 25.4% of                       about 22% [1]. Because of the high risk of VTE, and the need
patients receiving 2.5, 5, 10, 20 and 30 mg b.i.d. doses of                for active prevention, total joint replacement is an ideal clinical
BAY 59-7939, respectively (test for trend, P ¼ 0.29), com-                 environment in which to assess the efficacy and safety of new
pared with 44.3% in the enoxaparin group. The frequency                    anticoagulant agents. Seeking new and improved agents is a
of major, postoperative bleeding increased with increasing                 continuing clinical challenge. Active agents are available,
doses of BAY 59-7939 (test for trend, P ¼ 0.0007), with no                 including low molecular weight heparins (LMWHs), the
significant difference between any dose group compared                       indirect Factor Xa (FXa) inhibitor fondaparinux, and vitamin
                                                                           K antagonists (e.g. warfarin), but these have limitations. In
Correspondence: A. G. G. Turpie, HHS-General Hospital, 237 Barton          particular, LMWHs and fondaparinux require parenteral
Street East, Hamilton, Ontario, L8L 2X2, Canada.                           administration, while vitamin K antagonists have poor
Tel.: +1 905 528 9946; fax: +1 905 521 1551; e-mail:                       pharmacokinetic and pharmacodynamic profiles that may
turpiea@mcmaster.ca                                                        result in unpredictable anticoagulation.
                                                                              BAY 59-7939, a novel, oral, direct FXa inhibitor, is in
Presented as an oral communication at the XXth International Society in
Thrombosis and Haemostasis, Sydney, Australia, 6–12 August 2005.
                                                                           clinical development for the prevention and treatment of
                                                                           thromboembolic disorders. It is a small-molecule, active-site-
Study participants are listed at the end of the paper in Appendix.         directed FXa inhibitor with antithrombotic potential that has
                                                                           been well documented in animal models of venous and arterial
Received 20 May 2005, accepted 22 July 2005                                thrombosis [2]. The pharmacokinetic profile of BAY 59-7939 in

Ó 2005 International Society on Thrombosis and Haemostasis
2480 A. G. G. Turpie et al

healthy subjects is characterized by a time to peak plasma          enoxaparin (sanofi-aventis, Bridgewater, NJ, USA) was given
concentration of 2.5–4 h, and a terminal half-life of 9 h after     by subcutaneous injection every 12 h with the first dose given
multiple doses [3].                                                 on the morning after surgery to maintain blinding. Treatment
   This dose-finding study was performed to evaluate the             was continued until mandatory bilateral venography was
efficacy and safety of a 12-fold dose range of BAY 59-7939 in        performed 5–9 days after surgery. All patients received match-
the prevention of VTE after major orthopaedic surgery (total        ing placebo injections or tablets. Continued VTE prophylaxis
knee replacement), relative to the LMWH, enoxaparin, given          was at the discretion of the treating physician. Intermittent
postoperatively at a dose of 30 mg b.i.d.                           pneumatic compression devices were prohibited, but elastic
                                                                    compression stockings were allowed.
                                                                    Study endpoints
Study design
                                                                    The primary efficacy endpoint was the composite of the
This was a randomized, double-blind, double-dummy, active           incidence of proximal and/or distal DVT (at screening or
comparator controlled, parallel-group, dose-ranging study           confirmed symptomatic events), confirmed non-fatal PE, and
conducted at 26 centers in Canada, and 17 in the USA.               all-cause mortality during the treatment period. Secondary
Patients were randomly assigned to six treatment groups, using      endpoints included the incidence of individual components of
a computer-generated randomization list and interactive voice       DVT (total, proximal, and distal), major VTE (composite of
response system. Treatment was continued until mandatory            proximal DVT, PE and VTE-related death), and symptomatic
bilateral venography was performed 5–9 days after surgery.          VTE during the treatment period and within 30 days after
Blood samples for clinical chemistry, hematology, and coagu-        stopping treatment. Systematic evaluation of DVT was
lation were taken during screening, after surgery (before the       performed 5–9 days after surgery (or sooner if symptoms were
first dose of study treatment), 2–4 h after administration of        present). The assessment of the efficacy endpoints was based
treatment (day 2 and day 4), and directly before the last intake    solely on the analysis made by two independent central
of study treatment. A clinical assessment of patients, which        adjudication committees (Venography and VTE) blinded to
included recording of adverse events, symptoms and signs of         the treatment allocation.
DVT or pulmonary embolism (PE) and bleeding events, as well            The last dose of oral medication was given on the morning of
as confirmed diagnoses of VTE, was undertaken at a follow-up         bilateral venography. Venography was performed according to
visit 30 days (up to 60 days in some patients) after stopping       a modification of the Rabinov and Paulin technique [4–6]. The
treatment. The study protocol and other study documentation         technique was standardized to ensure complete visualization of
were approved by the institutional review boards at each site.      the deep veins, including the three paired calf veins, with at least
The study was conducted in accordance with accepted clinical        one fully opacified and visualized calf muscular vein, popliteal,
guidelines.                                                         superficial and common femoral, external and common iliac
                                                                    veins. Proximal DVT was defined as DVT in or above the
                                                                    popliteal vein. Criteria for an adequate examination included
                                                                    either a bilateral venogram with full visualization of all veins
Male patients aged 18 years or above and postmenopausal             listed above without DVT or findings of an intraluminal filling
female patients scheduled for elective total knee replacement,      defect of same shape in at least two images (the sole DVT
and who signed an informed consent, were eligible for the study.    criterion) irrespective of findings in the other veins. Sympto-
Exclusion criteria included: any bleeding disorder; current use     matic DVT during hospitalization required confirmation by
of drugs that may affect study outcome such as anticoagulants,      venography. PE was diagnosed by ventilation–perfusion scin-
platelet aggregation inhibitors (e.g. aspirin, clopidogrel, dipy-   tigraphy, spiral CT or pulmonary angiography, depending on
ridamole), or any other drug influencing coagulation (except         local center preference and accepted criteria for diagnosis [7].
non-steroidal anti-inflammatory drugs with half-life <17 h);         Symptomatic DVT during follow-up was confirmed by ultra-
DVT, PE, myocardial infarction, transient ischemic attack, or       sound or venography. All tests for VTE during the treatment
ischemic stroke within the last 6 months; severe hypertension;      period were first evaluated locally and subsequently by the
severe liver or renal impairment; body-weight <45 kg; and           independent adjudication committees.
alcohol or drug abuse. There was no upper weight limitation on         The primary safety endpoint was major, postoperative
patient inclusion.                                                  bleeding observed after the first administration of study drug
                                                                    and not later than 2 days after the last administration of study
                                                                    drug. Major postoperative bleeding was defined as clinically
Treatment schedule
                                                                    overt bleeding associated with a ‡2 g dL)1 fall in hemoglobin;
BAY 59-7939 (Bayer HealthCare AG, Wuppertal, Germany)               clinically overt bleeding leading to transfusion of ‡2 units of
was administered orally at 2.5, 5, 10, 20, or 30 mg b.i.d. every    blood; fatal bleeding, bleeding into a critical organ (e.g.
12 ± 1 h at meal times or within 2 h of food. The first dose         retroperitoneal, intracranial, or intraocular); or bleeding war-
was given 6–8 h after surgical wound closure. A 30-mg dose of       ranting treatment cessation or leading to re-operation.

                                                                          Ó 2005 International Society on Thrombosis and Haemostasis
                                                                                                     BAY 59-7939 for prevention of VTE 2481

   Clinically relevant, non-major bleeding events were defined                 Armitage test for trend using the total daily dose of
as: skin hematoma >25 cm2; wound hematoma >100 cm2; a                         BAY 59-7939 as a covariate and center as a fixed effect.
multiple site or extrasurgical site bleed; epistaxis >5 min, and
macroscopic hematuria >24 h. Minor bleeding events were
defined as those not fulfilling the criteria for major, postoper-
ative bleeding or clinically relevant, non-major bleeding. Blood
loss (perioperative and postoperative) and transfusion volumes
were measured during the treatment period. All bleeding events                Of the 621 patients randomized to either BAY 59-7939 (n ¼
were assessed centrally by a blinded independent bleeding event               516) or enoxaparin (n ¼ 105) between February and Novem-
committee.                                                                    ber 2004 (Fig. 1), eight patients were not treated with study
                                                                              drug and were not included in the analysis. The safety
                                                                              population comprised 613 patients who received at least one
Statistical analyses
                                                                              dose of study drug; 43 of these patients did not complete the
The primary statistical analysis was an assessment of the                     entire treatment period, 26 because of adverse events. The
dose effect of BAY 59-7939 on the primary efficacy                             intention-to-treat population comprised 404 patients; 209 were
endpoint, using logistic-regression. The sample size deter-                   excluded because either bilateral venography was not per-
mination was based on the assumption that the incidence of                    formed or the venograms were adjudicated to be indeterminate.
the primary efficacy endpoint for the BAY 59-7939 groups                       The per-protocol population consisted of 366 patients.
would range from 9% (highest dose) to 25% (lowest dose).                         The six treatment groups were well matched with regard to
Assuming a linear trend in the dose–response relationship                     baseline and surgical characteristics (Table 1). The age range
between BAY 59-7939 and the primary efficacy endpoint,                         was 39–92 years, the majority of patients were female and
randomization of 552 patients (92 per treatment group) was                    the BMI ranged between 19.5 to 61.3 kg m)2. General and
required to detect a dose trend with 80% power (at a 5%                       regional anesthesia was used in similar numbers of patients.
level of significance). This sample size assumed 22% of                        The mean time interval between surgery and first dose of BAY
patients would not have an evaluable venogram and was                         59-7939 was 7.2 h, with 83% of patients receiving study drug
calculated using nQuery Advisor, Version 4, Module PGT                        6–8 h after surgery. The mean duration of treatment was
1-1 (Cork, Ireland). The safety population comprised all                      7 days. After the treatment period, three (0.5%) patients were
randomized patients who received at least one dose of study                   lost to follow-up.
drug. The intention-to-treat population included those
patients who received study drug and had adequate
                                                                              Efficacy endpoints
assessment of VTE. The primary efficacy analysis was
performed on the per-protocol population (those patients                      Among the five BAY 59-7939 dose groups tested, the incidence
valid for the intention-to-treat analysis who had assessment                  of the primary efficacy endpoint ranged from 23.3% (10 mg
of VTE no more than 1 day after stopping study drug and                       b.i.d.) to 40.4% (5 mg b.i.d.), with no significant dose trend
showed no major protocol violations). The presence or                         according to the total daily dose (P ¼ 0.29) (Table 2 and
absence of a dose trend was estimated using the Cochran–                      Fig. 2). The incidence of the primary endpoint with enoxaparin

Fig. 1. Trial profile. ITT, intention-to-treat; PP, per-protocol. *Includes withdrawal of consent, failure to obtain adequate cannulation of both legs,
unilateral venography not demonstrating thrombus.  Adjudicated to be indeterminate by Venography Adjudication Committee. àPatients could have
more than one protocol violation.

Ó 2005 International Society on Thrombosis and Haemostasis
2482 A. G. G. Turpie et al

                                                                                                                                                                                                                                                                                                                                                             was 44.3%. Similar findings were obtained using the intention-

                                                                                                                                                        b.i.d., twice daily; VTE, venous thromboembolism. All values are numbers, or mean (range), or mean ± standard deviation. *Patients received epidural anesthesia without an indwelling catheter, as
                                                                            Enoxaparin 30 mg b.i.d.
                                                                                                                                                                                                                                                                                                                                                             to-treat population. The main difference was observed in the
                                                                                                                                                                                                                                                                                                                                                             BAY 59-7939 2.5 mg b.i.d. group, in which an incidence rate
                                                                                                                                                                                                                                                                                                                                                             of 36.1% was observed for the primary endpoint in the

                                                                                                                89.3 ± 16.7

                                                                                                                              90.5 ± 29.4
                                                                                                                31.8 ± 6.0

                                                                                                                                            7.2 ± 1.6
                                                                                                                                            7.4 ± 1.5
                                                                                                                                                                                                                                                                                                                                                             intention-to-treat population, compared with 31.7% in the per-

                                                                                                                66 (47–83)
                                                                            (n ¼ 104)

                                                                                                                              49 (47.1)
                                                                                                                              51 (49.0)
                                                                                                                 9 (8.7)

                                                                                                                               4 (3.8)
                                                                                                                57 (55)                                                                                                                                                                                                                                      protocol population.
                                                                                                                                                                                                                                                                                                                                                                Comparison of individual BAY 59-7939 doses with enox-
                                                                                                                                                                                                                                                                                                                                                             aparin showed that there was a lower incidence of the primary
                                                                                                                                                                                                                                                                                                                                                             efficacy endpoint in the BAY 59-7939 10 and 30 mg b.i.d. dose
                                                                                                                                                                                                                                                                                                                                                             groups (absolute reduction: 21.0%; 95% CI, 5.1–36.8%; and
                                                                                           30 mg (n ¼ 106)

                                                                                                                                                                                                                                                                                                                                                             18.9%; 95% CI, 2.8–35.0%, respectively). Major VTE events
                                                                                                                90.1 ± 18.4

                                                                                                                              91.1 ± 31.4
                                                                                                                                                                                                                                                                                                                                                             ranged from 0% to 6.7% among the BAY 59-7939 groups
                                                                                                                32.4 ± 6.2

                                                                                                                                            7.3 ± 1.8
                                                                                                                                            7.1 ± 1.8
                                                                                                                66 (39–86)

                                                                                                                12 (11.3)

                                                                                                                              63 (59.4)
                                                                                                                              41 (38.7)
                                                                                                                               2 (1.9)                                                                                                                                                                                                                       compared with 4.3% for the enoxaparin group.
                                                                                                                63 (59)

                                                                                                                                                                                                                                                                                                                                                                During the treatment period, six patients developed symp-
                                                                                                                                                                                                                                                                                                                                                             tomatic VTE. Two PE events occurred in the BAY 59-7939
                                                                                                                                                                                                                                                                                                                                                             5 mg b.i.d. group, one DVT in each of the 2.5 and 20 mg b.i.d.
                                                                                                                                                                                                                                                                                                                                                             groups, and two DVT in the enoxaparin group. There were no
                                                                                                                                                                                                                                                                                                                                                             deaths. During the follow-up period, a further four patients
                                                                                           20 mg (n ¼ 98)

                                                                                                                                                                                                                                                                                                                                                             developed symptomatic VTE (three PE events and one DVT).
                                                                                                                87.1 ± 20.4

                                                                                                                              89.2 ± 30.0
                                                                                                                31.2 ± 6.6

                                                                                                                                            7.0 ± 1.5
                                                                                                                                            7.2 ± 1.8
                                                                                                                68 (45–92)

                                                                                                                                                                                                                                                                                                                                                             Three patients died in the follow-up period, 5, 21 and 31 days
                                                                                                                10 (10.2)

                                                                                                                              48 (49.0)
                                                                                                                              48 (49.0)
                                                                                                                               2 (2.0)
                                                                                                                67 (68)

                                                                                                                                                                                                                                                                                                                                                             after treatment had ceased. Two of the deaths occurred in the
                                                                                                                                                                                                                                                                                                                                                             BAY 59-7939 2.5 mg b.i.d. group (PE and cardiorespiratory
                                                                                                                                                                                                                                                                                                                                                             failure) and one death in the 10 mg b.i.d. group (PE). None of
                                                                                                                                                                                                                                                                                                                                                             these patients received thromboprophylaxis beyond cessation
                                                               BAY 59-7939 b.i.d.

                                                                                                                                                                                                                                                                                                                                                             of study treatment.
                                                                                           10 mg (n ¼ 103)

                                                                                                                86.4 ± 19.0

                                                                                                                              88.0 ± 34.1
                                                                                                                31.8 ± 6.3

                                                                                                                                            7.5 ± 2.5
                                                                                                                                            7.3 ± 1.5
                                                                                                                67 (49–84)

                                                                                                                              53 (51.5)
                                                                                                                              47 (45.6)

                                                                                                                                                                                                                                                                                                                                                             Safety endpoints
                                                                                                                 8 (7.8)

                                                                                                                               3 (2.9)
                                                                                                                66 (64)

                                                                                                                                                        planned epidural anesthesia with an indwelling catheter was an exclusion criterion of the study.

                                                                                                                                                                                                                                                                                                                                                             The incidence of major, postoperative bleeding events were
                                                                                                                                                                                                                                                                                                                                                             1.0%, 0%, 1.9%, 3.1%, and 7.5% for the BAY 59-7939 doses
                                                                                                                                                                                                                                                                                                                                                             of 2.5, 5, 10, 20, and 30 mg b.i.d., respectively (Table 3). A
                                                                                                                                                                                                                                                                                                                                                             statistically significant dose trend was observed with increasing
                                                                                           5 mg (n ¼ 102)

                                                                                                                                                                                                                                                                                                                                                             doses of BAY 59-7939 (P ¼ 0.0007) (Fig. 2). Major, postop-
                                                                                                                86.8 ± 17.2

                                                                                                                              83.2 ± 25.7
                                                                                                                31.7 ± 6.1

                                                                                                                                            7.0 ± 1.1
                                                                                                                                            7.2 ± 1.5

                                                                                                                                                                                                                                                                                                                                                             erative bleeding occurred in 1.9% of patients in the enoxaparin
                                                                                                                66 (45–86)

                                                                                                                              62 (60.8)
                                                                                                                              34 (33.3)
                                                                                                                 6 (5.9)

                                                                                                                               6 (5.9)

                                                                                                                                                                                                                                                                                                                                                             group. There was no significant difference in the incidence of
                                                                                                                65 (64)

                                                                                                                                                                                                                                                                                                                                                             major, postoperative bleeding between any of the BAY 59-7939
                                                                                                                                                                                                                                                                                                                                                             dosage groups and enoxaparin.
                                                                                                                                                                                                                                                                                                                                                                Most major, postoperative bleeding events occurred
                                                                                                                                                                                                                                                                                                                                                             2–4 days after surgery [13 of 16 events (81%)]; three events
                                                                                           2.5 mg (n ¼ 100)

                                                                                                                                                                                                                                                                                                                                                             occurred on the day following surgery. No cases of fatal or
                                                                                                                                                                                                                                                                                                                                                             critical organ bleeding were observed. One patient (10 mg b.i.d.
                                                                                                                90.9 ± 18.6

                                                                                                                              90.1 ± 31.5
                                                                                                                32.1 ± 6.6

                                                                                                                                            7.2 ± 2.0
                                                                                                                                            7.3 ± 1.6
                                                                                                                66 (48–84)
    Table 1 Baseline characteristics of all treated patients

                                                                                                                                                                                                                                                                                                                                                             BAY 59-7939 group) required re-operation due to bleeding.
                                                                                                                              61 (61.0)
                                                                                                                              36 (36.0)
                                                                                                                 6 (6.0)

                                                                                                                               3 (3.0)
                                                                                                                59 (59)

                                                                                                                                                                                                                                                                                                                                                             Bleeding warranting treatment discontinuation was observed in
                                                                                                                                                                                                                                                                                                                                                             the three higher BAY 59-7939 dose groups (one patient each in
                                                                                                                                                                                                                                                                                                                                                             the 10 and 20 mg b.i.d. groups, and two patients in the 30 mg
                                                                                                                                                                                                                                                                                                                                                             b.i.d. group).
                                                                                                                                                                                                                                                                                                                                                                The incidence of clinically relevant, non-major bleeding and
                                                                                                                  Duration of treatment – days
                                                                                                                  Epidural anesthesia* – n (%)

                                                                                                                                                                                                                                                                                                                                                             minor bleeding were of a similar pattern to the primary
                                                                                                                  General anesthesia – n (%)

                                                                                                                  Time to first oral dose – h
                                                                                                                  Spinal anesthesia – n (%)

                                                                                                                  Duration of surgery, min
                                                                                                                Body mass index – kg m)2

                                                                                                                Study drug administration

                                                                                                                                                                                                                                                                                                                                                             endpoint with a lower incidence in the BAY 59-7939 2.5–10 mg
                                                                                                                VTE risk factor – n (%)

                                                                                                                                                                                                                                                                                                                                                             b.i.d. groups. Similarly, the composite of major, postoperative
                                                                                                                Age – years (range)

                                                                                                                                                                                                                                                                                                                                                             bleeding or clinically relevant, non-major bleeding was lower in
                                                                                                                Female – n (%)

                                                                                                                                                                                                                                                                                                                                                             the BAY 59-7939 2.5–10 mg b.i.d. groups. No significant
                                                                                                                Surgery details

                                                                                                                Weight – kg

                                                                                                                                                                                                                                                                                                                                                             difference was observed compared with enoxaparin. Mean
                                                                                                                                                                                                                                                                                                                                                             transfusion volumes were lowest in patients receiving
                                                                                                                                                                                                                                                                                                                                                             BAY 59-7939 2.5 mg b.i.d. (149 mL) and highest in the
                                                                                                                                                                                                                                                                                                                                                             30 mg b.i.d. group (230 mL) (Table 3).

                                                                                                                                                                                                                                                                                                                                                                   Ó 2005 International Society on Thrombosis and Haemostasis
                                                                                                       BAY 59-7939 for prevention of VTE 2483

Table 2 Incidence of efficacy endpoints and individual components*
                                                                     BAY 59-7939 b.i.d.
                                2.5 mg (n ¼ 63)     5 mg (n ¼ 57)      10 mg (n ¼ 60)      20 mg (n ¼ 57)      30 mg (n ¼ 59)      30 mg b.i.d. (n ¼ 70)
Characteristic – n (%)

Primary efficacy endpoint         20 (31.7)           23 (40.4)          14 (23.3)           20 (35.1)           15 (25.4)           31 (44.3)
  95% CI (%)                    20.6–44.7           27.6–54.2          13.4–36.0           22.9–48.9           15.0–38.4           32.4–56.7
  Death (all-cause)              0                   0                  0                   0                   0                   0
  PE                             0                   2 (3.5)            0                   0                   0                   0
  DVT                           20 (31.7)           21 (36.8)          14 (23.3)           20 (35.1)           15 (25.4)           31 (44.3)
    Proximalà                    2 (3.2)             1 (1.8)            4 (6.7)             2 (3.5)             0                   3 (4.3)
    Distal only                 18 (28.6)           20 (35.1)          10 (16.7)           18 (31.6)           15 (25.4)           28 (40.0)
    Symptomatic                  1 (1.5)             0                  0                   1 (1.8)             0                   2 (2.9)
Major VTE                        2 (3.2)             3 (5.3)            4 (6.7)             2 (3.5)             0                   3 (4.3)

b.i.d., twice daily; CI, confidence interval. *Per-protocol population.  Includes any DVT, symptomatic confirmed non-fatal PE, and all-cause
mortality. àPatients with both proximal and distal DVT are included once in the proximal category.

Fig. 2. Dose–response relationship for the primary efficacy and safety endpoints. Dose–response relationship between BAY 59-7939 and the primary
efficacy endpoint (DVT, non-fatal PE, all-cause mortality) and the primary safety endpoint (major, postoperative bleeding). The solid lines are the dose–
response curves for BAY 59-7939, estimated by logistic regression, including total daily dose as a covariate. The dashed lines are the lower and upper 95%
CIs, and the dots represent the observed frequencies.

   A post hoc analysis of a composite of major VTE and major,                  observed to have elevations of lipase and AST of >3 · upper
postoperative bleeding was performed. The lowest values were                   limit of normal (ULN), respectively. The corresponding rates
observed in the BAY 59-7939 2.5 and 5 mg b.i.d. dose groups                    in patients receiving enoxaparin were 5.1% and 3.0%,
(2.8% and 4.8% respectively), compared with 6.7% in the                        respectively. No significant increase in AST was observed
enoxaparin group. A higher incidence was observed in the                       during the treatment period and lipase levels fell. Elevated
BAY 59-7939 10, 20, and 30 mg b.i.d. dose groups (8.8%,                        levels of ALT or AST (>3 · ULN) and of bilirubin
7.7% and 11.8%, respectively).                                                 (2 · ULN) were noted in four patients receiving
                                                                               BAY 59-7939 (one patient in 10 mg b.i.d. group, one in
                                                                               20 mg group, and two in 30 mg group). Both patients with
Adverse events
                                                                               elevated ALT and bilirubin remained on study medication
A total of 23 patients experienced treatment-related serious                   without any clinical symptoms, with liver function tests
adverse events with a dose-dependent increase noted across the                 returning to normal by follow-up. There were no cases of
BAY 59-7939 groups. The incidence rates in the lower doses                     clinically relevant thrombocytopenia. The incidence of nausea
(2.5–10 mg b.i.d.) ranged between 0% and 4%, with 3%                           and vomiting with early postoperative administration of
observed with enoxaparin. Treatment-related serious adverse                    BAY 59-7939 was low in all groups. BAY 59-7939 did not
events were more frequent in BAY 59-7939 20 and 30 mg b.i.d.                   have untoward effects on ECG parameters; in particular, there
groups (7% and 8%, respectively), predominantly due to                         was no treatment-related QTc-prolongation in patients with
major, postoperative bleeding events. Mild increases in the liver              normal QTc interval at baseline.
enzymes aspartate transaminase (AST) and alanine transami-
nase (ALT) were observed in all treatment groups, with no
dose–response relationship observed in the BAY 59-7939
groups. Following surgery and prior to treatment administra-                   In this study, oral administration of a wide, 12-fold dose range
tion, 3.5% and 5.2% of patients receiving BAY 59-7939 were                     (2.5–30 mg b.i.d.) of BAY 59-7939, commenced early in the

Ó 2005 International Society on Thrombosis and Haemostasis
2484 A. G. G. Turpie et al

Table 3 Incidence of postoperative bleeding events
                                                                                 BAY 59-7939 b.i.d.

                                                           2.5 mg      5 mg          10 mg        20 mg         30 mg         Enoxaparin 30 mg
Characteristic                                             (n ¼ 100)   (n ¼ 102)     (n ¼ 103)    (n ¼ 98)      (n ¼ 106)     b.i.d. (n ¼ 104)

Major, postoperative bleeding – n (%)                      1 (1.0)     0             2 (1.9)      3 (3.1)       8 (7.5)       2 (1.9)
    95% CI (%)                                             0.02–5.4    0–3.6         0.2–6.8      0.6–8.7       3.3–14.3      0.2–6.8
   Fatal/critical bleeding – n (%)                         0           0             0            0             0             0
   Clinically overt bleeding associated with fall in Hbà   1 (1.0)     0             1 (1.0)      1 (1.0)       8 (7.5)       2 (1.9)
   Clinically overt bleeding leading to transfusion§       1 (1.0)     0             2 (1.9)      3 (3.1)       7 (6.6)       2 (1.9)
   Bleeding leading to re-operation                        0           0             1 (1.0)      0             0             0
   Clinically overt bleeding and treatment cessation       0           0             1 (1.0)      1 (1.0)       2 (1.9)       0
Site of major, postoperative bleeding events – n (%)
   Surgical site bleeding                                  1   (1.0)   0             2   (1.9)    3 (3.1)       6 (5.7)       2   (1.9)
   Extrasurgical site bleeding                             0           0             0            0             2 (1.9)       0
Non-major, clinically relevant bleeding – n (%)            2   (2.0)   3 (2.9)       1   (1.0)    5 (5.1)       7 (6.6)       3   (2.9)
Minor bleeding – n (%)                                     6   (6.0)   6 (5.9)       6   (5.8)    10 (10.2)     12 (11.3)     3   (2.9)
Composite major, postoperative or clinically relevant,     3   (3.0)   3 (2.9)       2   (1.9)    8 (8.2)       14 (13.2)     5   (4.8)
 non-major bleeding – n (%)
Patients receiving blood transfusions – n (%)              24 (24.0)   33 (32.4)     29 (28.2)    33 (33.7)     35 (33.0)     20 (19.2)
   Volume (mL; mean ± SD)                                  149 ± 364   162 ± 262     179 ± 335    178 ± 291     230 ± 400     117 ± 263
Postoperative drainage volume (mL; mean ± SD)              403 ± 430   318 ± 320     427 ± 412    429 ± 424     478 ± 450     411 ± 425

Hb, hemoglobin; SD, standard deviation. *Safety population.  Bleeding events starting more than 2 days after last study medication intake were
not considered. àAssociated with a fall in Hb of ‡2 g dL)1 within 24 h from first postoperative day. §Leading to transfusion of ‡2 units of blood.

postoperative period to patients undergoing elective total knee            b.i.d.) had rates comparable with enoxaparin (0–1.9% vs.
replacement, was equally efficacious for the prevention of VTE,             1.9%, respectively), and these findings are similar to those
with no dose–response apparent. Furthermore, all dose groups               reported for patients treated with placebo (2%) in a previous
of BAY 59-7939 had comparable levels of efficacy with                       total knee replacement study [13]. The observed incidence of
subcutaneous twice-daily enoxaparin, initiated the day after               major, postoperative bleeding was higher in the 20 and 30 mg
surgery.                                                                   b.i.d. doses, although none of the doses was significantly higher
   Of the 613 patients treated, 60% were evaluable for the                 than enoxaparin and none was stopped because of excessive
primary efficacy analysis, which was lower than the anticipated             bleeding. This contrasts with a recent study investigating the
figure of 78% in the sample size calculation. This reflects fewer            oral, direct FXa inhibitor razaxaban, in which the highest three
than expected evaluable venograms, either because bilateral                doses were stopped due to excessive major bleeding [14].
venography was not performed, or because it was adjudicated                   Individual bleeding components showed consistent out-
to be indeterminate by the central reading committee. This                 comes with regard to all parameters measured, in particular
resulted in a reduction in the power of the study to demonstrate           the incidence of clinically relevant, non-major bleeding,
an efficacy dose response.                                                  minor bleeding, and blood transfusion volumes. The pro-
   Using the same rigorous, specific venogram adjudication                  portion of patients requiring transfusion was similar to that
techniques, other dose-ranging studies have reported similar               reported in a similar dose-finding study conducted in North
VTE rates with enoxaparin. Studies involving ximelagatran                  America comparing ximelagatran with enoxaparin [11]. As is
(METHRO III and EXPRESS) [8,9] and dabigatran etexilate                    common in these studies, treatment-related adverse events
(BISTRO II) [10] showed almost identical VTE rates for patients            were mainly related to bleeding; other adverse events were
undergoing total knee replacement treated with enoxaparin,                 infrequent and there were no differences compared with
ranging from 44.1% to 46% compared with 44.3% in the                       enoxaparin. Nausea and vomiting were uncommon following
present study. The event rates in the present study are higher             administration of the first oral dose a mean of 7 h following
than those reported in two North American elective total knee              completion of surgery. Two baseline liver enzyme screens,
replacement trials [11,12], which reported VTE rates of                    which were conducted at study entry and just prior to the
23% to 27.8% in patients receiving enoxaparin 30 mg b.i.d.,                commencement of study drug postsurgery, indicated that
initiated 12–24 h after surgery. This is probably due to the use of        mild elevations were present before study drug was admin-
different criteria for venogram assessment i.e. the present study          istered. This suggests an influence of surgery and associated
employed a more detailed, bilateral assessment for DVT that                medications, including anesthesia, on liver enzyme activity.
required optimal visualization of muscular calf and anterior               Indeed, similar increases in liver enzymes have been reported
tibial veins.                                                              in clinical studies using placebo and with anticoagulants such
   Across the BAY 59-7939 treatment groups, a significant dose              as dextran and LMWH [15–17]. This safety profile, consid-
trend was apparent with respect to major, postoperative                    ered with the efficacy findings described, suggests that
bleeding events. Lower doses of BAY 59-7939 (2.5–10 mg                     BAY 59-7939 has a wide therapeutic window.

                                                                                   Ó 2005 International Society on Thrombosis and Haemostasis
                                                                                              BAY 59-7939 for prevention of VTE 2485

   BAY 59-7939 has undergone separate, phase II assessment              3 Kubitza D, Becka M, Wensing G, Voith B, Zuehlsdorf M. Multiple
in dose-ranging studies of total hip and total knee replacement,          dose escalation study investigating the pharmacodynamics, safety, and
                                                                          pharmacokinetics of BAY 59-7939 an oral, direct Factor Xa inhibitor
in contrast to studies undertaken with other anticoagulants
                                                                          in healthy male subjects. Blood 2003; 102: 811A.
that have included mixed groups of patients undergoing                  4 Rabinov K, Paulin S. Roentgen diagnosis of venous thrombosis in the
elective arthroplasty [9,10]. In a dose-ranging study with a              leg. Arch Surg 1972; 104: 134–44.
similar design as the current study, but in patients undergoing         5 Kalebo P, Ekman S, Lindbratt S, Eriksson BI, Pauli U, Zachrisson
total hip replacement [18], twice-daily administration of a range         BE, Close P. Percentage of inadequate phlebograms and observer
                                                                          agreement in thromboprophylactic multicenter trials using standard-
of BAY 59-7939 doses was also demonstrated to be potentially
                                                                          ized methodology and central assessment. Thromb Haemost 1996; 76:
effective for the prevention of VTE, but without a significant             893–6.
dose trend. However, in both this and the current study, a              6 Kalebo P, Anthmyr BA, Eriksson BI, Zachrisson BE. Optimization
significant dose trend for major, postoperative bleeding was               of ascending phlebography of the leg for screening of deep vein
found. In view of the lack of efficacy dose response, the post hoc         thrombosis in thromboprophylactic trials. Acta Radiologica 1997; 38:
analyses described offer important insight for the choice of the
                                                                        7 Biello DR, Mattar AG, McKnight RC, Siegel BA. Ventilation-per-
optimal dose for definitive, phase III studies. When major,                fusion studies in suspected pulmonary embolism. AJR Am J Roent-
postoperative bleeding and clinically relevant, non-major                 genol 1979; 133: 1033–7.
bleeding were considered together, the incidence was lower in           8 Eriksson BI, Agnelli G, Cohen AT, Dahl OE, Mouret P,
the BAY 59-7939 2.5–10 mg b.i.d. groups. An analysis of the               Rosencher N, Eskilson C, Nylander I, Frison L, Ogren M. Direct
                                                                          thrombin inhibitor melagatran followed by oral ximelagatran in
composite of major VTE, and major, postoperative bleeding,
                                                                          comparison with enoxaparin for prevention of venous thrombo-
showed the lowest values were observed in the BAY 59-7939                 embolism after total hip or knee replacement. Thromb Haemost 2003;
2.5 and 5 mg b.i.d. dose groups.                                          89: 288–96.
   In summary, this study demonstrates the potential efficacy of         9 Eriksson BI, Agnelli G, Cohen AT, Dahl OE, Lassen MR, Mouret P,
postoperative administration of BAY 59-7939 for prevention                Rosencher N, Kalebo P, Panfilov S, Eskilson C, Andersson M. The
                                                                          direct thrombin inhibitor melagatran followed by oral ximelagatran
of VTE after total knee replacement. The incidence of major,
                                                                          compared with enoxaparin for the prevention of venous thromboem-
postoperative bleeding of <2% in the three lower doses was                bolism after total hip or knee replacement: the EXPRESS study.
comparable with that for enoxaparin. Although the test for                J Thromb Haemost 2003; 1: 2490–6.
trend in dose efficacy did not reach statistical significance,           10 Eriksson BI, Dahl OE, Buller HR, Hettiarachchi R, Rosencher N,
based on the available data, the optimal dose range of BAY                Bravo ML, Ahnfelt L, Piovella F, Stangier J, Kalebo P, Reilly P. A
                                                                          new oral direct thrombin inhibitor, dabigatran etexilate, compared
59-7939 was determined to be 2.5–10 mg b.i.d., and these data
                                                                          with enoxaparin for prevention of thromboembolic events following
will determine the choice of dose for further investigation in            total hip or knee replacement: the BISTRO II randomized trial.
phase III studies in major orthopaedic surgery.                           J Thromb Haemost 2005; 3: 103–11.
                                                                       11 Heit JA, Colwell CW, Francis CW, Ginsberg JS, Berkowitz SD,
                                                                          Whipple J, Peters G. Comparison of the oral direct thrombin inhibitor
Acknowledgements                                                          ximelagatran with enoxaparin as prophylaxis against venous throm-
                                                                          boembolism after total knee replacement: a phase 2 dose-finding study.
We thank the investigators and study coordinators who                     Arch Intern Med 2001; 161: 2215–21.
participated in this study. We thank Andrea Nadel and Badri            12 Bauer KA, Eriksson BI, Lassen MR, Turpie AGG. Fondaparinux
Moridani from Bayer for their contribution to the conduct and             compared with enoxaparin for the prevention of venous thromboem-
analysis of the study.                                                    bolism after elective major knee surgery. New Engl J Med 2001; 345:
                                                                       13 Leclerc JR, Geerts WH, Desjardins L, Jobin F, Laroche F, Delorme F,
Conflict of interest disclosure                                            Haviernick S, Atkinson S, Bourgouin J. Prevention of deep vein
                                                                          thrombosis after major knee surgery – a randomized, double-blind trial
This study was supported by Bayer HealthCare AG, Germany.                 comparing a low molecular weight heparin fragment (enoxaparin) to
M. Irwin and F. Misselwitz are employees of Bayer. W. Fisher              placebo. Thromb Haemost 1992; 67: 417–23.
                                                                       14 Lassen MR, Davidson BL, Gallus A, Pineo G, Ansell J, Deitchman D.
and L. Kwong were investigators during the study and received
                                                                          A phase II randomized, double-blind, five-arm, parallel-group, dose–
grants for enrolling patients. P. Kalebo received honoraria for
                                    ¨                                     response study of a new oral directly-acting factor Xa inhibitor, Raz-
adjudication of the venograms during the study. All steering              axaban, for the prevention of deep vein thrombosis in knee replace-
committee members received honoraria for their participation.             ment surgery – on behalf of the Razaxaban investigators. Blood 2003;
                                                                          102: 15A.
                                                                       15 Christiansen HM, Lassen MR, Borris LC, Sorensen JV, Rahr HB,
References                                                                Jorgensen LN, Jorgensen PW, Hauch O. Biologic tolerance of two
                                                                          different low molecular weight heparins. Semin Thromb Hemost 1991;
 1 Geerts WH, Pineo GF, Heit JA, Bergqvist D, Lassen MR, Colwell          17: 450–4.
   CW, Ray JG. Prevention of venous thromboembolism: the seventh       16 Christiansen HM, Rud-Lassen M, Borris LC, Sorensen JV, Schmidt K,
   ACCP conference on antithrombotic and thrombolytic therapy. Chest      Boll KL, Eiskjaer S, Nielsen BW, Lucht U. Biological tolerance of
   2004; 126: 338S–400S.                                                  Logiparin, a low molecular weight heparin used in patients undergoing
 2 Perzborn E, Strassburger J, Wilmen A, Pohlmann J, Roehrig S,           total hip replacement. Semin Thromb Hemost 1991; 17(Suppl. 2): 224–7.
   Schlemmer KH, Straub A. In vitro and in vivo studies of the novel   17 Carlson MK, Gleason PP, Sen S. Elevation of hepatic transaminases
   antithrombotic agent BAY 59-7939-an oral, direct Factor Xa             after enoxaparin use: case report and review of unfractionated and
   inhibitor. J Thromb Haemost 2005; 3: 514–21.

Ó 2005 International Society on Thrombosis and Haemostasis
2486 A. G. G. Turpie et al

   low-molecular-weight heparin-induced hepatotoxicity. Pharmacother-   Adjudication Committee – M. Prins (Academic Hospital of
   apy 2001; 21: 108–13.                                                the University of Maastricht, Maastricht, the Netherlands),
18 Eriksson BI, Borris LC, Dahl OE, Haas SK, Huisman MV, Kakkar                                                                 ¨
                                                                        U. Angeras (Sahlgrenska University Hospital/Ostra,
   AK, Kalebo P, Misselwitz F, Muhlhofer EM. Prevention of venous
          ¨                        ¨
   thromboembolism with an oral, direct Factor Xa inhibitor –           Goteborg, Sweden), A. Falk (Carlanderska Hospital,
   BAY 59-7939 – in elective hip replacement: a dose-ranging study.     Goteborg, Sweden); Venography Adjudication Committee –
   J Thromb Haemost 2005; 3(Suppl. 1): [Abstract].                      P. Kalebo and B. Zachrisson (Sahlgrenska University
                                                                        Hospital/Ostra, Goteborg, Sweden); VTE Adjudication Com-
Appendix                                                                mittee – H. Eriksson, G. Sandgren and J. Wallin (Sahlgrenska
                                                                        University Hospital/Ostra, Goteborg, Sweden); Investigators –
The members of the ODIXa-KNEE (Oral DIrect FactorXa                     Canada: P. Ansari, L. Bredo, G. Crawford, A. Profitt,
inhibitor BAY 59-7939 in the prevention of VTE in patients              C. Watson, J. Cisa, V. Dekorompay, C. Demers, G. Dervin,
undergoing total KNEE replacement) Study Group were as                  E. Dessouki, C. Dobson, W. Fisher, L. Flores, R. Josefchak,
follows: Steering Committee – A.G.G. Turpie (Study Chair),              M. Mant, W. Pisesky, D. Puskas, L. Vickars, D. Stevens,
K.A. Bauer, W.D. Fisher, M. Gent, L.M. Kwong, M. Irwin                  R. Zarnett, A. Berkshire, K. Mckenzie; USA: R. Murray,
(non-voting member); Data and Safety Monitoring Board –                 J. Gimbel, D. Bramlet, L. Levy, G. Gill, M. Jove, G. Kantor,
R. Roberts (McMaster University, Hamilton, Ontario,                     L. Kwong, S. Slagis, M. Snyder, D. Whitaker, J. Papilion,
Canada), J. Weitz (McMaster University and Henderson                    P. Peters Jr., J. Schwappach.
Research Centre, Hamilton, Canada) and J. Ansell (Boston
University Medical Center, Boston, USA); Bleeding Event

                                                                              Ó 2005 International Society on Thrombosis and Haemostasis

To top