|
|
Background In a previous study of the prevention of venous thromboembolism after total knee replacement, the efficacy of ximelagatran, an oral direct thrombin inhibitor that does not require monitoring of coagulation or dose adjustment, was found to be similar to that of warfarin at a dose of 24 mg of ximelagatran twice daily. The purpose of the present study was to determine whether a higher dose of ximelagatran is superior to warfarin.
Methods This randomized, double-blind trial compared a regimen of 7 to 12 days of oral ximelagatran, at a dose of 24 or 36 mg twice daily, starting the morning after surgery, with warfarin therapy started the evening of the day of surgery. The composite end point of venous thromboembolism and death from all causes and the incidence of bleeding were the primary outcome measures.
Results Among the 1851 patients in the efficacy analysis, oral ximelagatran at a dose of 36 mg twice daily was superior to warfarin with respect to the primary composite end point of venous thromboembolism and death from all causes (20.3 percent vs. 27.6 percent; P=0.003). There were no significant differences between these two groups with respect to major bleeding (incidence, 0.8 percent and 0.7 percent, respectively), perioperative indicators of bleeding, wound characteristics, or the composite secondary end point of proximal deep-vein thrombosis, pulmonary embolism, and death (2.7 percent vs. 4.1 percent; P=0.17).
Conclusions The efficacy of oral ximelagatran, administered starting the morning after total knee replacement, was superior to that of warfarin for prevention of venous thromboembolism. Rates of hemorrhagic complications with the two drugs were similar.
Ximelagatran is an orally administered direct thrombin inhibitor under development as an anticoagulant agent for prophylaxis against and treatment of thromboembolism. It is rapidly absorbed and quickly converted into its active form, melagatran, a reversible, active-site inhibitor of both free and clot-bound thrombin7 that has stable and reproducible pharmacokinetic properties. No clinically relevant interactions with food or cytochrome P-450–metabolized drugs have been reported with ximelagatran.8,9,10,11,12,13 Initial studies have shown ximelagatran to have good efficacy and safety in the prevention of venous thromboembolism after total knee or total hip replacement.14,15,16,17 Results of a phase 2 trial15 suggested a wide range of doses with good efficacy and safety, and an initial phase 3 trial showed that treatment with ximelagatran at a dose of 24 mg twice daily was at least as effective as warfarin.14 The purpose of the present study was to determine whether a higher dose would be safe and more effective than warfarin.
Methods
Study Design
We conducted a prospective, randomized, double-blind trial at 116 centers in the United States, Canada, Israel, Mexico, and Brazil. The institutional review board at each center approved the protocol, and written informed consent was obtained from all patients before any study procedures were conducted. The study was conducted in accordance with the ethical principles stated in the Declaration of Helsinki. Patients were screened 1 to 30 days preoperatively, and they were randomly assigned after postoperative rescreening to receive ximelagatran at a dose of 24 mg twice daily, ximelagatran at a dose of 36 mg twice daily, or warfarin. Randomization was performed with the use of a computer-generated system, with stratification at each center according to whether patients underwent unilateral or bilateral surgery.
Patients
Eligible patients were women without childbearing potential and men who weighed between 40 and 136 kg. Only patients undergoing primary total knee replacement were included. The criteria for exclusion were pneumatic leg compression; immobilization for 3 or more days; major surgery, stroke, myocardial infarction, or receipt of any investigational drug within 30 days before surgery; intracranial, retroperitoneal, or intraocular bleeding or any other disorder associated with an increased risk of bleeding within 90 days before surgery; gastrointestinal bleeding within 90 days before surgery, ulcer disease verified by endoscopic examination within 30 days before surgery, or both; uncontrolled hypertension; cancer that required cytostatic treatment or was itself the reason for total knee replacement; an alanine or aspartate aminotransferase level greater than two times the upper limit of the normal range; thrombocytopenia; drug or alcohol abuse within the previous 6 months; allergy to contrast medium or iodine; a contraindication to warfarin therapy; impaired renal function (defined by an estimated creatinine clearance of less than 30 ml per minute)18; and traumatic epidural or lumbar puncture.
If the use of an epidural or spinal catheter continued into the treatment period, the catheter was to be removed during trough levels of melagatran (the active metabolite of ximelagatran). Treatment with thrombolytic, anticoagulant, or antiplatelet agents, including heparins, warfarin, direct thrombin inhibitors, dipyridamole, sulfinpyrazone, ticlopidine, clopidogrel, acetylsalicylic acid at a dose greater than 500 mg per day, and dextran, was not allowed within seven days before surgery or during the period of administration of the study drug.
Treatment Regimens
Warfarin (Coumadin, Bristol-Myers Squibb) or a warfarin placebo was administered each evening, with the first dose given on the evening of the day of surgery and with the dose adjusted to achieve an international normalized ratio (INR) of 2.5 (range, 1.8 to 3.0). Ximelagatran at a dose of 24 mg or 36 mg in tablet form (Exanta, AstraZeneca) or a ximelagatran placebo was given in the morning and evening starting 12 hours or more after surgery, when adequate hemostasis had been achieved. The treatment was continued until venography was performed. To guide the adjustment of the warfarin dosage, INR values were measured in a blinded fashion on days 1 to 3 after surgery, on the day when venography was performed, and as needed. INR values were measured either at participating centers, with the use of encrypted point-of-care devices provided by AstraZeneca, or at local laboratories equipped with a system that prevented access to the INR values by local personnel. All INR values were reported to an anticoagulation management center, which relayed real or sham values to local study personnel. A warfarin-dosing nomogram was provided, but the dose of either warfarin or warfarin placebo was chosen at the investigator's discretion. For patients receiving ximelagatran, sham INR values were generated to mimic the usual values in patients receiving warfarin. Compliance with oral treatment was assessed by counting tablets used in the hospital, dispensed at discharge, and returned after the treatment had ended.
Assessments of Efficacy
The composite primary end point for efficacy comprised total (distal and proximal) deep- vein thrombosis, pulmonary embolism, and death from all causes during treatment, as determined by an independent central adjudication committee. The composite secondary end point for efficacy comprised proximal deep-vein thrombosis, pulmonary embolism, and death from all causes during treatment.
Deep-vein thrombosis was evaluated by means of bilateral ascending venography19,20 7 to 12 days after the study treatment was initiated. In addition to central adjudication, venograms were assessed locally by personnel who had no knowledge of the assigned study treatment. The criterion for the diagnosis of deep-vein thrombosis was a consistent intraluminal filling defect on at least two images. To be considered adequate for evaluation, venograms had to show all deep veins except the deep femoral vein, the muscular veins of the calf, and the anterior tibial veins, although these were included in the evaluation if they were visible or if a clot was detected. Venograms were classified as indeterminate if they showed a lack of filling of a region of the deep-vein system without a filling defect in the same region. Symptomatic proximal deep-vein thrombosis could be diagnosed by compression ultrasonography, but a diagnosis of symptomatic distal deep-vein thrombosis required venography.
A diagnosis of pulmonary embolism was made if a lung scan showed one or more segmental perfusion defects in at least two views with corresponding normal ventilation, if pulmonary angiography showed a persistent intraluminal defect or an abrupt cutoff of a vessel that was greater than 2.5 mm in diameter, if a spiral computed tomographic scan showed a distinct filling defect in a large vessel, if embolectomy was performed, or if a pulmonary embolism was identified on autopsy. Deaths were classified as due to pulmonary embolism or a bleeding event or as not associated with venous thromboembolism or bleeding. All cases of confirmed symptomatic deep-vein thrombosis, suspected pulmonary embolism, and death from any cause were adjudicated centrally. Treatment of confirmed thrombosis was at the discretion of the attending physician.
Safety Assessments
The primary variables used to assess safety were major bleeding and major or minor bleeding occurring up to 48 hours after the last dose of the study drug had been administered. All bleeding events were reviewed by the independent central adjudication committee and classified as major if they were clinically overt, with one or more of the following findings: involvement of a critical site (intracranial, retroperitoneal, intraocular, intraspinal, or pericardial), a bleeding index of 2.0 or more (calculated as the hemoglobin level [in grams per liter] at the base-line visit minus the postbleeding hemoglobin level, plus the number of units of packed red cells or whole blood transfused), a need for medical or surgical intervention at the operative site, or fatal bleeding. Other episodes of clinically overt bleeding were classified as minor. The investigators also assessed the appearance and characteristics of the surgical wound (swelling, drainage, erythema, and bleeding) according to the following categories: as expected, better than expected, or worse than expected. Laboratory results were assessed at the time of screening for the study, on the last day of administration of the study drug, and at the follow-up examination four to six weeks after surgery.
Statistical Analysis
We estimated the size of the sample by assuming that the rate of venous thromboembolism with warfarin would be 30 percent and predicting a risk reduction of 33.3 percent with ximelagatran at a dose of 36 mg and of 25 percent with ximelagatran at a dose of 24 mg. To demonstrate these differences at a 5 percent level of significance would require the enrollment of 500 patients per treatment group (for 90 percent power) or 565 patients (for 80 percent power) for the high and low doses of ximelagatran, respectively. We planned to enroll approximately 2250 patients in order to detect a 25 percent risk reduction with the lower dose, given the assumption that 25 percent of venograms would be inadequate for evaluation.
The primary analysis was designed to test the superior efficacy of ximelagatran at a dose of 36 mg twice daily as compared with warfarin. If the difference was significant (as indicated by a two-tailed P value of <0.05), then ximelagatran at a dose of 24 mg twice daily was to be compared with warfarin. The efficacy analyses included all patients who had received at least one dose of a study medication and who had a venogram that was adequate for evaluation or had symptomatic, objectively confirmed deep- vein thrombosis or pulmonary embolism or died during treatment. Bilateral venography was required, but patients undergoing unilateral surgery in whom a clot was detected were included in the efficacy analysis even if a venogram that was adequate for evaluation could be obtained only in the leg that had undergone surgery or only in the contralateral side. For patients who underwent bilateral surgery, venograms that were adequate for evaluation were required in both legs, unless deep-vein thrombosis was found in one leg. Symptomatic venous thromboembolism and deaths that occurred within two days after venography, or up to day 12 if no venography was performed, were included in the analyses.
The frequency of bleeding events was estimated with the use of the observed proportions (with 95 percent confidence intervals) for each treatment group. Differences between each of the two groups receiving ximelagatran and the warfarin group were analyzed with the use of the Cochran–Mantel–Haenszel chi- square test. Differences in blood loss and transfusion requirements were tested with the use of analysis of variance. Differences in ratings of the appearance of the wound and bleeding complications at the surgical site were tested with the use of the Cochran–Mantel–Haenszel chi-square test.
The study was designed by the members of the executive committee of the EXULT A (Exanta Used to Lessen Thrombosis A) Study Group, who also analyzed and interpreted the data. The data were collected by Omnicare Clinical Research at the direction of the executive committee.
Results
Study Groups
Of 2656 patients who were enrolled, a total of 2301 patients were randomly assigned to a study group and 1851 were included in the analysis of efficacy (Table 1). Of the 450 patients not included, 16 underwent randomization but did not receive treatment; in the case of the remaining 434 patients, the presence or absence of an end point could not be determined because venography either was not performed or was indeterminate or because there was no confirmed symptomatic venous thromboembolic event or death. Base-line characteristics, details of the surgery, and the hospital course were similar in the three groups (Table 2 and Table 3). Characteristics of the 2285 patients included in the safety analysis and those of the 1851 patients included in the efficacy analysis did not differ significantly. Venography was adequate for evaluation in 80.7 percent of the patients. In the efficacy analysis, the mean time to the first dose of medication on day 1 was 20.4 hours for all three treatment groups combined; 89 percent of the patients received the assigned study medication for a period of 7 to 12 days. Approximately one third of the patients (701) received additional anticoagulant therapy after discontinuing the study medication at the discretion of the attending physician; in two thirds of these patients (478 patients) the purpose was extended prophylaxis, and in one third (223 patients) it was acute treatment of venous thromboembolism. Only two patients were lost to follow-up; 92 percent of the patients who underwent randomization completed the study, with a follow-up visit at weeks 4 to 6.
|
|
|
The composite primary end point of total venous thromboembolism (deep-vein thrombosis and pulmonary embolism) and death from all causes occurred in 20.3 percent of patients in the group receiving 36 mg of ximelagatran (128 of 629) and in 27.6 percent of patients in the warfarin group (168 of 608) (relative risk reduction, 26.4 percent; P=0.003) (Table 4). The incidence of the composite primary end point in the group receiving 24 mg of ximelagatran was 24.9 percent (153 of 614 patients; P=0.28 for the comparison with warfarin). Because only unilateral venograms were available for some patients, a sensitivity analysis that included only the patients with bilateral venograms (76.1 percent of the total) was performed, with a similar result. In the 4 percent of patients who had undergone bilateral surgery, the primary end point of total venous thromboembolism or death occurred in 22.2 percent of patients receiving the 36-mg dose of ximelagatran, 19.0 percent of patients receiving the 24-mg dose of ximelagatran, and 34.8 percent of patients receiving warfarin.
|
The prespecified composite secondary end point of proximal deep-vein thrombosis, pulmonary embolism, and death from all causes occurred in 17 patients (2.7 percent) in the higher-dose ximelagatran group, 15 patients (2.5 percent) in the lower-dose ximelagatran group, and 25 patients (4.1 percent) in the warfarin group (P=0.17 for higher-dose ximelagatran vs. warfarin; P=0.10 for lower-dose ximelagatran vs. warfarin). Slightly more than 1 percent of patients in each group had symptomatic deep-vein thrombosis or died (Table 4).
The INR value was in the target range (1.8 to 3.0) or higher in 65 percent of patients in the warfarin group by postoperative day 3 (mean INR, 2.3) and in 76 percent of patients by the day of venography (mean INR, 2.4). There were no appreciable differences in mean INR values between patients with and those without venous thromboembolism on either day.
Safety
Major bleeding occurred during treatment in six patients in each of the two ximelagatran groups and in five patients in the warfarin group (Table 5). Additional major bleeding events occurred during follow-up in four patients in the lower-dose ximelagatran group and in one patient in the warfarin group. One bleeding complication was fatal; gastric-ulcer bleeding developed in a patient who had received two 36-mg doses of ximelagatran. The bleeding led to multiorgan-system failure and death on day 46. This patient had also received perioperative enoxaparin (as part of the anesthesia protocol) and diclofenac.
|
|
|
Our results show that ximelagatran given orally at a dose of 36 mg twice daily starting postoperatively (a mean of 20.4 hours after surgery) was significantly more effective than warfarin in preventing venous thromboembolism after total knee replacement, with an absolute risk reduction of 7.3 percent and a relative risk reduction of 26.4 percent. The number needed to treat was 14. This benefit was due to a reduction in the rate of asymptomatic deep-vein thrombosis, whereas the rates of proximal deep-vein thrombosis and of symptomatic venous thromboembolism were low in all three treatment groups and did not differ significantly between the group receiving 36 mg of ximelagatran and the warfarin group. Studies of the natural history of venous thromboembolism suggest that asymptomatic deep-vein thrombosis identified by postoperative venography is a predictor of the development of symptomatic venous thromboembolism.1,21
As compared with the rates of total and proximal venous thrombosis in seven other multicenter trials of warfarin for prophylaxis after total knee replacement, all conducted in the past 10 years, the rates in our large study (27.4 percent and 3.8 percent, respectively) were among the lowest reported for warfarin. The other studies reported rates of 38 to 55 percent for deep-vein thrombosis and 7 to 12 percent for proximal deep-vein thrombosis.22,23,24,25,26,27,28 The low rates in our study may be related to improvements in general surgical care and to the strict, predefined guidelines used for central adjudication. Also, between two thirds and three quarters of the patients in our study had therapeutic INR values on day 3 and on the day of venography, which is a somewhat higher proportion than usual. Local interpretation of venograms, which was a secondary end point for efficacy, confirmed the superiority of ximelagatran at a dose of 36 mg twice daily, even though the rates of deep-vein thrombosis were higher with local interpretation, most likely because of less stringent application of diagnostic criteria.
The first large, randomized clinical trial of ximelagatran for the prevention of deep-vein thrombosis after total knee replacement in North America showed that a dose of 24 mg twice daily had an efficacy similar to that of warfarin; venous thromboembolism occurred in 19 percent of patients receiving this dose of ximelagatran and in 26 percent of those receiving warfarin (P=0.07).14 Because the rate of bleeding complications was low and ximelagatran was well tolerated over a range of doses, we chose a 36-mg dose, in addition to the 24-mg dose, for our study.
In six studies comparing low-molecular-weight heparin with warfarin after total knee replacement, the rates of venographically identified deep- vein thrombosis in the groups that received low-molecular-weight heparin ranged from 25 percent to 45 percent, with a mean of 31 percent, as compared with rates in the warfarin groups ranging from 38 percent to 55 percent, with a mean of 47 percent.1,22,23,24,25,26,27 In one study, which compared fondaparinux, a synthetic pentasaccharide, with enoxaparin in 724 patients, the fondaparinux group had a 12.5 percent rate of thromboembolism, whereas the rate with enoxaparin was 27.8 percent.29 The incidence of major hemorrhagic complications has been higher with low-molecular-weight heparins than with warfarin.23,24,25,27 A meta-analysis of four large clinical trials comparing fondaparinux with enoxaparin for prophylaxis after orthopedic surgery showed that major bleeding was significantly more frequent with fondaparinux.30
Warfarin is the single most commonly used pharmacologic form of prophylaxis prescribed after total knee replacement in North America.31,32 Among the many reasons for this preference are longer experience with the use of warfarin, oral administration, and a lower rate of early bleeding complications than with other agents. Nevertheless, warfarin has a delayed onset of action, requires monitoring of coagulation and dose adjustment, and has multiple interactions with food and drugs.33,34 The results of our study indicate that fixed-dose ximelagatran, administered without coagulation monitoring, was significantly more effective than warfarin and had similar safety. It could therefore be considered as an alternative to other available thromboprophylactic agents.
Supported by AstraZeneca.
Drs. Francis, Lieberman, Ginsberg, and Paiement report having served as paid consultants to AstraZeneca, Dr. Colwell reports having received lecture fees from the company, Drs. Colwell and Ginsberg report having received grant support from the company, and Ms. Roth reports owning equity in the company.
We are indebted to Joyce S. Weiner, Ph.D., an employee of AstraZeneca, for editorial support in the preparation of the manuscript.
Source Information
From the Department of Medicine, University of Rochester, Rochester, N.Y. (C.W.F.); the Department of Clinical Development, AstraZeneca, Wilmington, Del. (S.D.B., G.R.P., A.W.R., J.M.); the Department of Medicine, University of Oklahoma, Oklahoma City (P.C.C.); the Department of Orthopedic Surgery, University of California at Los Angeles, Los Angeles (J.R.L.); the Department of Medicine, McMaster University, Hamilton, Ont., Canada (J.S.G.); the Department of Orthopedic Surgery, University of California at Irvine, Irvine (G.P.); and the Department of Orthopaedic Surgery, Scripps Clinic, La Jolla, Calif. (C.W.C.).
Address reprint requests to Dr. Francis at the University of Rochester Medical Center, 610 Elmwood Ave., Box 610, Rochester, NY 14642, or at charles_francis{at}urmc.rochester.edu.
References
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||
Related Letters:
Ximelagatran versus Warfarin after Total Knee Replacement
Banarer S., Schillaci G., Pasqualini L., Mannarino E., Nishiguchi M., Takamura N., Aoyagi K., Francis C. W., Berkowitz S. D., Colwell C. W. Jr.
Extract |
Full Text |
PDF
N Engl J Med 2004;
350:616-617, Feb 5, 2004.
Correspondence
This article has been cited by other articles:
HOME | SUBSCRIBE | SEARCH | CURRENT ISSUE | PAST ISSUES | COLLECTIONS | PRIVACY | TERMS OF USE | HELP | beta.nejm.org Comments and questions? Please contact us. The New England Journal of Medicine is owned, published, and copyrighted © 2009 Massachusetts Medical Society. All rights reserved. |
Previous
