OUP user menu

★ Editor's Choice ★

Dabigatran vs. placebo in patients with acute coronary syndromes on dual antiplatelet therapy: a randomized, double-blind, phase II trial

Jonas Oldgren, Andrzej Budaj, Christopher B. Granger, Yasser Khder, Juliet Roberts, Agneta Siegbahn, Jan G.P. Tijssen, Frans Van de Werf, Lars Wallentin
DOI: http://dx.doi.org/10.1093/eurheartj/ehr113 2781-2789 First published online: 7 May 2011


Aim After an acute coronary syndrome, patients remain at risk of recurrent ischaemic events, despite contemporary treatment, including aspirin and clopidogrel. We evaluated the safety and indicators of efficacy of the novel oral direct thrombin inhibitor dabigatran.

Methods and results In this double-blind, placebo-controlled, dose-escalation trial, 1861 patients (99.2% on dual antiplatelet treatment) in 161 centres were enrolled at mean 7.5 days (SD 3.8) after an ST-elevation (60%) or non-ST-elevation (40%) myocardial infarction and randomized to twice daily treatment with dabigatran 50 mg (n= 369), 75 mg (n= 368), 110 mg (n= 406), 150 mg (n= 347), or placebo (n= 371). Primary outcome was the composite of major or clinically relevant minor bleeding during the 6-month treatment period. There were 96 primary outcome events and, compared with placebo, a dose-dependent increase with dabigatran, hazard ratio (HR) 1.77 (95% confidence intervals 0.70, 4.50) for 50 mg; HR 2.17 (0.88, 5.31) for 75 mg; HR 3.92 (1.72, 8.95) for 110 mg; and HR 4.27 (1.86, 9.81) for 150 mg. Compared with placebo, D-dimer concentrations were reduced in all dabigatran dose groups by an average of 37 and 45% at weeks 1 and 4, respectively (P< 0.001). Fourteen (3.8%) patients died, had a myocardial infarction or stroke in the placebo group compared with 17 (4.6%) in 50 mg, 18 (4.9%) in 75 mg, 12 (3.0%) in 110 mg, and 12 (3.5%) in the 150 mg dabigatran groups.

Conclusions Dabigatran, in addition to dual antiplatelet therapy, was associated with a dose-dependent increase in bleeding events and significantly reduced coagulation activity in patients with a recent myocardial infarction.

  • Acute coronary syndromes
  • Anticoagulants
  • Direct thrombin inhibitor
  • Myocardial infarction

See page 2734 for the editorial comment on this article (doi:10.1093/eurheartj/ehr051)


During the first year after both ST-elevation and non-ST-elevation acute coronary syndromes (ACS), patients remain at a high risk of recurrent ischaemic events, despite contemporary evidence-based care, including revascularization and dual antiplatelet treatment with aspirin and clopidogrel.14 Oral vitamin K antagonists have been shown to prevent recurrent ischaemic events after ACS, both as monotherapy and in combination with aspirin, but are cumbersome to use because of multiple interactions with food and drugs, the need for frequent laboratory monitoring, and the bleeding risk.5,6 Evidence regarding the efficacy and safety of the combination of dual antiplatelet therapy and oral anticoagulants is limited. The first experience with an oral direct thrombin inhibitor ximelagatran, later withdrawn from the market because of liver toxicity, showed a reduction in the composite of death, myocardial infarction, and stroke when added to aspirin for 6 months after an ACS.7 In the recent phase II ACS trials, the factor Xa antagonists rivaroxaban and apixaban seemed to provide a reduction in ischaemic event rates when added to aspirin alone, but less so when added on top of dual antiplatelet treatment. However, this potential benefit was counterbalanced by an increased bleeding risk.8,9

Dabigatran etexilate is an oral pro-drug and is rapidly converted by a serum esterase to dabigatran, a potent, direct, competitive inhibitor of thrombin. It has an absolute bioavailability of 6.5%, serum half-life of 12–17 h, 80% of the given dose is excreted by the kidneys, and it does not require regular monitoring. In the pivotal RE-LY trial,10 dabigatran 110 mg b.d. was non-inferior and dabigatran 150 mg b.d. superior to warfarin for stroke prevention in atrial fibrillation patients. The present dose-guiding study evaluated the safety and indicators of efficacy of four dose regimens of dabigatran etexilate compared with placebo when given in addition to dual antiplatelet treatment in patients with a recent ST-elevation or non-ST-elevation myocardial infarction at high risk of new ischaemic cardiovascular events.


RE-DEEM was a multi-centre, prospective, randomized, double-blind, placebo-controlled, dose-escalation trial. Patients were enrolled at 161 sites in 24 countries in Asia, Europe, and North America. Patients had follow-up visits at Weeks 1, 4, 8, 12, 16 (telephone), 20, 26, and 28 with blood-sampling and assessment for bleeding and ischaemic outcome events. The trial was conducted between 14 March 2008 and 2 October 2009.

Eligible patients were males and females aged 18 years or older, hospitalized with non-ST or ST-segment elevation myocardial infarction within the last 14 days, and receiving treatment with dual antiplatelet therapy (aspirin and clopidogrel or another thienopyridine). The index event had to be documented by elevated values of cardiac biomarkers (preferably troponin T or I) above the 99th percentile of the upper reference limit together with ischaemic symptoms or ECG changes (ST-T changes, new left bundle branch block, or new Q-waves). Additionally, participants were required to have at least one risk factor for subsequent cardiovascular complications: age 65 years or above, diabetes mellitus on treatment, previous myocardial infarction, left bundle branch block, congestive heart failure requiring treatment or left ventricular ejection fraction <40%, peripheral arterial disease, moderate renal insufficiency [creatinine clearance (CrCl) ≥30–60 mL/min], or no revascularization for the index event.

Major exclusion criteria included: ongoing or planned treatment with vitamin K antagonists, severe disabling stroke within the previous 6 months or any stroke within the previous 14 days, conditions associated with an increased risk of bleeding such as major surgery (including bypass surgery) in the previous month, history of severe bleeding, gastrointestinal haemorrhage within the past year, gastroduodenal ulcer in the previous 30 days, fibrinolytic agents within 48 h of study entry, uncontrolled hypertension, haemoglobin <10 g/dL or platelet count <100 × 109/L, normal coronary arteries at angiogram for index event, congestive heart failure New York Heart Association Class IV, and severe renal impairment (CrCl <30 mL/min).

Ethics approval for the protocol was obtained from the local Ethics Committees, and the study was performed in accordance with the Declaration of Helsinki. All patients gave written informed consent to participate in the study.

Randomization and study drug

This trial used a dose-escalation, adaptive randomization plan. Patients were randomized via a centralized interactive voice response system (IVRS) and initially only to placebo or one of the two lower doses (50 and 75 mg b.d.) in a 1:1:1 ratio (Figure 1). An independent data safety monitoring board reviewed unblinded data and advised the release of the 110 mg dose in a second stage and the 150 mg dose group in a third stage. In the last (fourth) stage of the trial, the IVRS was re-programmed to achieve balance between the five groups.

Figure 1

Study design (adaptive dose-escalation) and patient flow.

Patients with moderate renal impairment (CrCl 30–50 mL/min) have higher plasma concentrations of dabigatran compared with patients with normal renal function.11 Therefore, patients who entered the study with CrCl <50 mL/min and who were randomized to 75, 110, or 150 mg dose groups were assigned to treatment with the next lower dabigatran dose in order to achieve similar drug exposures to those with normal renal function. Accordingly, these patients were analysed according to the dose level to which they were randomized. Patients stayed on the same medication dose level throughout the trial.

In case of a major bleeding, the study drug should be stopped. In case of a minor bleeding or stroke, study drug was continued or temporarily or permanently stopped at the discretion of the investigator. Study drug should be temporarily discontinued at least 24 h before planned invasive procedures and re-commenced after cessation of parenteral anticoagulants administered at percutaneous coronary interventions (PCI) and within 1 month after major surgery.

Concomitant medication

The daily aspirin dose was strongly advised to be ≤100 mg, but higher doses were permitted according to local practice or at the time of interventional procedures. The recommended daily dose of clopidogrel was 75 mg although loading doses of 300–600 mg were allowed for PCI procedures. The duration of clopidogrel treatment was at the discretion of the local investigator and in accordance with local practice/guidelines. Permitted alternatives to clopidogrel were ticlopidine or prasugrel, although the latter was not used for any study patient.


The primary outcome of the trial was the incidence of major or clinically relevant minor bleeding. Major bleeding events were assessed by the International Society of Thrombosis and Haemostasis (ISTH) definition,12 including bleeding that was fatal, occurred in a critical location (intra-cranial, intra-spinal, intra-ocular, retroperitoneal, intra-articular, intra-muscular with compartment syndrome, or pericardial), or was associated with a fall in haemoglobin of 2 g/dL or more, or a transfusion of two units or more of whole blood or packed red blood cells. Clinically relevant minor bleeding was defined as a clinically overt bleed that did not meet the criteria for major bleed but prompted a clinical response, i.e. hospital admission for bleeding, medical or surgical treatment, or a change in antithrombotic therapy including interruption or discontinuation of study drug. Components of the Thrombolysis in Myocardial Infarction (TIMI)13 and Global Utilization of Streptokinase and Tissue plasminogen activator for Occluded coronary arteries (GUSTO)14 bleeding definitions were also prospectively collected.

Secondary outcomes were indicators of efficacy such as reduction in D-dimer levels and incidences of cardiovascular ischaemic events. Plasma D-dimer concentrations were analysed with a commercial immunoassay, Biopool TintElize D-dimer (Trinity Biotech, Bray, Ireland) at the UCR laboratory, Uppsala, Sweden, in blood samples obtained at randomization, after 1, 4, and 26 weeks, and 2 weeks after cessation of study medication.

Cardiovascular ischaemic endpoints included the composite of cardiovascular death, non-fatal myocardial infarction, and non-haemorrhagic stroke; and the individual occurrence of death (cardiovascular and all-cause), non-fatal myocardial infarction, severe recurrent ischaemia, and non-haemorrhagic stroke. Myocardial infarction was defined in accordance with the international consensus document.15 Severe recurrent ischaemia was defined as worsening angina symptoms lasting at least 10 min with at least one of the following: dynamic ≥1 mm ST depression or elevation, hospitalization, or an unplanned or urgent revascularization procedure. Stroke was defined as an acute onset of a focal neurological deficit of presumed vascular origin lasting for 24 h or more or resulting in death. The stroke was categorized as haemorrhagic or non-haemorrhagic (based on computer tomography, magnetic resonance imaging, or autopsy) or type unknown.

All bleeds, deaths, and suspected cases of myocardial infarction, severe recurrent ischaemia, and stroke were evaluated independently by two experienced physicians blinded to study drug assignment.


The sample size was driven by the primary endpoint, i.e. to detect a statistically significant dose response (at the 5% significance level) for bleeding rates across the five treatment groups. With a sample size of 286 patients per-treatment group, a two-sided χ2 test of trend in proportions based on the logistic model would have 90% power to detect a difference in proportions predicted to be 5, 5, 7, 9, and 11% (placebo, 50, 75, 110, and 150 mg, respectively). Subsequently, at least 340 patients were included per-treatment group which allowed for potential drop outs or non-evaluable patients of ∼15%.

All analyses are, unless stated otherwise, presented for randomized patients who had received at least one dose of study drug, i.e. a modified intention to treat population. The proportions of patients experiencing the primary outcome, major or clinically relevant minor bleeding, per-treatment group were analysed by the Cochran–Armitage linear trend test. Kaplan–Meier plots were generated for the primary outcome. Hazard ratios (HR) and 95% confidence intervals (CI) were calculated for outcome events based on Cox proportional hazards regression models. Bleeding events occurring on or after the date of the first ingestion of study medication up to 3 days after last ingestion of study drug were also assessed in a pre-defined on-treatment analysis. Subgroup analyses were performed using linear logistic regression analysis with interaction tests. Age, gender, body mass index, renal function, ethnicity, haemoglobin at baseline, and diabetes mellitus were pre-defined subgroups, and revascularization for index event was added post hoc. Ischaemic cardiovascular endpoints were reported descriptively. Each patient was only counted once, i.e. first event, in composite endpoints.

The change in log10-transformed D-dimer concentration at Weeks 1 and 4 compared with baseline were analysed using analysis of covariance adjusted for log10 D-dimer concentration at baseline. Pairwise comparisons were performed between each individual dabigatran dose level and placebo. The proportion of patients with any reduction in D-dimer concentration was summarized per-treatment group and analysed using the Cochran–Armitage linear trend test.

Adequate compliance was defined as within 80–120% of capsules that should have been taken, excluding days when study medication was temporarily or permanently discontinued.

All analyses were performed with SAS statistical software version 8.2 (SAS Inc., Cary, NC, USA).


A total of 1861 patients were randomized and received at least one dose of study drug (Figure 1). Mean exposure to study drug ranged from 158.9 to 164.6 days for the dabigatran groups and 164.7 days for the placebo group. Average compliance was 95.9% in the placebo group and 91.2, 94.8, 90.3, and 92.1% in the respective 50, 75, 110, and 150 mg dabigatran groups.

Baseline characteristics

The baseline characteristics are presented in Table 1. Sixty per cent of the patients were enrolled after an ST-elevation myocardial infarction and 40.0% after non-ST-elevation myocardial infarction and the 54.5% underwent PCI for the index event. Risk factors for subsequent cardiovascular events included age >65 years in 44.1%, diabetes mellitus in 31.3%, previous myocardial infarction in 29.0%, and congestive heart failure in 11.7% of the patients. The mean time from index event to the first administration of trial drug was 7.5 (SD 3.8) days. In the 110 and 150 mg dabigatran groups, a higher proportion of patients (56.7 and 62.5%, respectively) were recruited from Central Europe and consequently a lower proportion of patients from Western Europe and Asia. In addition, a higher proportion of patients in the 110 and 150 mg dabigatran groups were females and of higher age (Table 1). A total of 47 (12.8%), 51 (12.6%), and 30 (8.6%) of the patients assigned to the 75, 110, and 150 mg groups, respectively, received the next lower dabigatran dose due to moderately impaired renal function.

View this table:
Table 1

Baseline characteristics, concomitant treatment at baseline, and index events

50 mg b.d.75 mg b.d.110 mg b.d.150 mg b.d.
Treated patients, n (total, N = 1861)371369368406347
 Age, mean (SD), years61.5 (11.3)61.9 (12.3)60.9 (11.5)62.3 (11.2)62.3 (10.8)
 Age ≥65 years, %42.444.740.545.347.8
 Body mass index, mean (SD), kg/m227.0 (4.4)27.0 (4.2)27.1 (4.7)27.3 (4.5)27.8 (4.4)
 Males, %78.477.279.971.273.2
Medical history
 Diabetes mellitus, %29.430.132.632.032.3
 Creatinine clearance ≥30 and <50 mL/min, %a7.
 Hypertension, %65.566.462.268.070.9
 Congestive heart failure, %13.510.011.413.310.1
 Peripheral artery disease, %
 Previous myocardial infarction, %28.629.034.226.626.5
 Left bundle branch block, %
 North America, %
 Central Europe, %44.544.742.156.762.5
 Western Europe, %22.426.325.018.720.5
 Asia, %27.822.226.419.513.8
Index event
 ST-elevation myocardial infarction, %61.756.961.162.157.9
 Non-ST-elevation myocardial infarction, %38.343.138.937.942.1
 Percutaneous coronary interventionb, %54.252.658.452.255.3
Days from index event to randomization, mean (SD)7.5 (3.7)7.7 (3.8)7.6 (4.1)7.4 (3.6)7.4 (3.9)
Baseline medication
 Beta blocking agents, %85.284.883.786.986.7
 Angiotensin converting enzyme inhibitors, %
 Angiotensin II antagonists11.09.512.711.48.3
 Selective calcium channel blockers, %25.323.324.023.124.5
  • aMeasured by central laboratory.

  • bIncluding primary percutaneous coronary intervention for ST-elevation myocardial infarction.

Concomitant medications

The use of concomitant cardiovascular medications was generally balanced across the treatment groups (Table 1). At randomization, 99.2% of patients received dual antiplatelet therapy. At the end of study treatment, the proportion of patients receiving dual antiplatelet therapy was 83.8%, while 14.2% received aspirin only (Table 2).

View this table:
Table 2

Aspirin and clopidogrel use at randomization and at different time-points during treatment, n (%)

nAspirin onlyAspirin and clopidogrela
Randomization1861 (100.0)8 (0.4)1846 (99.2)
1 week1792 (100.0)12 (0.7)1765 (98.5)
4 weeks1724 (100.0)24 (1.4)1683 (97.6)
12 weeks1640 (100.0)71 (4.3)1547 (94.3)
26 weeks—end of trial1861 (100.0)264 (14.2)1559 (83.8)
28 weeks—follow-up1858 (100.0)338 (18.2)1479 (79.6)
  • aIncluding 10 patients receiving ticlopidine instead of clopidogrel.


The 6-month incidence of the primary endpoint, the composite of major or clinically relevant minor bleeding events, was 3.5, 4.3, 7.9, and 7.8% in the respective 50, 75, 110, and 150 mg dabigatran groups, compared with 2.2% in the placebo group, P< 0.001 for linear trend (Figure 2). Eighty-nine of the 96 primary outcome events occurred on-treatment, i.e. within 3 days of last intake of study drug (Table 3). Compared with placebo, the HR for the primary outcome were 1.77 (95% CI 0.70, 4.50) for 50 mg, HR 2.17 (95% CI 0.88, 5.31) for 75 mg, 3.92 (95% CI 1.72, 8.95) for 110 mg, and 4.27 (95% CI 1.86, 9.81) for 150 mg, respectively. In a sensitivity analysis, patients were analysed at the actually received dose level including those patients dose adjusted due to renal impairment. In this additional analysis, the HR for the primary outcome were 1.82 (95% CI 0.77, 4.29) for 50 mg, 2.44 (95% CI 1.05, 5.65) for 75 mg, 3.56 (95% CI 1.60, 7.91) for 110 mg, and 3.88 (95% CI 1.73, 8.74) for 150 mg, respectively, compared with placebo.

View this table:
Table 3

Bleeding events

50 mg b.d.75 mg b.d.110 mg b.d.150 mg b.d.
Number of patients, n (%)371 (100.0)369 (100.0)368 (100.0)406 (100.0)347 (100.0)
Patients intention to treat
 Primary outcomea8 (2.2)13 (3.5)16 (4.3)32 (7.9)27 (7.8)
  Major (ISTHb definition)2 (0.5)3 (0.8)1 (0.3)8 (2.0)4 (1.2)
  Clinically relevant minor bleeding6 (1.6)10 (2.7)16 (4.3)24 (5.9)23 (6.6)
  Non-clinically relevant minor bleeding17 (4.6)32 (8.7)29 (7.9)35 (8.6)20 (5.8)
TIMIc major1 (0.3)1 (0.3)0 (0.0)5 (1.2)1 (0.3)
GUSTOd severe1 (0.3)1 (0.3)0 (0.0)2 (0.5)0 (0.0)
Patients on-treatment
 Major (ISTHb definition)1 (0.3)2 (0.5)1 (0.3)6 (1.5)4 (1.2)
 Clinically relevant minor bleeding6 (1.6)9 (2.4)15 (4.1)23 (5.7)23 (6.6)
 Non-clinically relevant minor bleeding17 (4.6)31 (8.4)29 (7.9)34 (8.4)19 (5.5)
  • aPrimary outcome is defined as the composite of major (ISTH) or clinically relevant minor bleeding events.

  • bISTH, International Society of Thrombosis and Haemostasis.

  • cTIMI, Thrombolysis In Myocardial Infarction.

  • dGUSTO, Global Utilization of Streptokinase and Tissue plasminogen activator for Occluded coronary arteries.

Figure 2

Kaplan–Meier curve depicting the primary endpoint, i.e. the composite of major and clinically relevant minor bleeding. Compared with placebo, primary outcomes differed significantly (P< 0.001) for the 110 and 150 mg dabigatran doses using Cox proportional hazards regression models. Number of subjects at risk are given below the x-axis.

The incidence of major clinical bleeding events were low in the placebo, 50 and 75 mg dabigatran dose groups, both in the intention-to-treat as well as on-treatment analyses (Table 3). Although based on very few patients, compared with placebo there was ∼1% higher incidence of major bleedings in the 110 and 150 mg treatment groups combined, with the highest bleeding rate in the 110 mg dose group (Table 3). Major or severe bleeding rates according to the TIMI13 and GUSTO14 definitions were lower, with <1% absolute increase in any of the dabigatran groups compared with placebo (Table 3).

The differences in bleeding events were consistent across the majority of the subgroups investigated, although the number of patients and/or events was limited in some subgroups (Table 4). Female gender and age >75 years were associated with higher bleeding incidences in the 110 and 150 mg dose groups, with significant treatment by subgroup interactions.

View this table:
Table 4

Subgroup analysis of the incidence [n, (%)] of major and clinically relevant minor bleeding events

Total, nPlaceboDabigatranP-valuea
50 mg b.d.75 mg b.d.110 mg b.d.150 mg b.d.
Female4482 (2.5)1 (1.2)2 (2.7)14 (12.0)15 (16.1)0.012
Male14136 (2.1)12 (4.2)14 (4.8)18 (6.2)12 (4.7)
Age (years)
 <6510406 (2.8)8 (3.9)8 (3.7)9 (4.1)8 (4.4)0.035
 ≥65 and <755750 (0.0)3 (2.8)7 (6.8)14 (10.9)13 (10.6)
 ≥752462 (4.6)2 (3.5)1 (2.2)9 (16.1)6 (14.0)
Body mass index (kg/m2)
 <255913 (2.5)4 (3.3)4 (3.2)11 (8.2)10 (11.0)0.616
 ≥25 and <308403 (1.8)5 (2.9)5 (3.2)13 (7.3)9 (5.6)
 ≥304282 (2.5)4 (5.4)7 (8.1)8 (8.5)8 (8.5)
White14424 (1.5)10 (3.5)12 (4.5)28 (8.7)23 (7.7)0.303
Asian4144 (3.9)3 (3.7)3 (3.0)3 (3.8)4 (8.3)
Baseline haemoglobin
 Within normal range14674 (1.4)11 (3.6)11 (3.8)27 (8.6)20 (7.1)0.509
 Below normal range3233 (4.3)2 (3.3)4 (6.3)4 (5.9)6 (10.0)
Creatinine clearance (mL/min)
 ≥30 to ≤501611 (3.0)1 (3.1)1 (3.0)5 (12.8)3 (12.5)0.529
 >5016997 (2.1)12 (3.6)15 (4.5)27 (7.4)24 (7.4)
Diabetes mellitus
 Yes5823 (2.8)4 (3.6)7 (5.8)9 (6.9)8 (7.1)0.433
 No12785 (1.9)9 (3.5)9 (3.6)23 (8.3)19 (8.1)
Revascularized for index event
 Yes10144 (2.0)8 (4.1)8 (3.7)16 (7.5)17 (8.9)0.641
 No8474 (2.4)5 (2.9)8 (5.2)16 (8.2)10 (6.5)
  • aLogistic regression interaction test

The most frequently reported bleeding events were gastrointestinal bleeds (placebo, 1.3%; 50 mg, 2.4%; 75 mg, 3.0%; 110 mg, 4.9%; 150 mg, 3.2%) and epistaxis (placebo, 0.8%; 50 mg, 3.3%; 75 mg, 2.2%; 110 mg, 4.2%; 150 mg, 3.7%). There were two fatal bleeding events, one in the placebo group and one in the 110 mg dose group. There were no intra-cranial or intra-spinal bleeding events in any of the treatment groups.


In total, 73 patients experienced at least one component of the composite of cardiovascular death, non-fatal myocardial infarction, or non-haemorrhagic stroke during the study period (Table 5). In patients receiving the two higher dabigatran doses, a numerically lower proportion reached this secondary composite outcome (110 mg, 3.0%; 150 mg, 3.5%) compared with the patients receiving the two lower dabigatran doses (50 mg, 4.6%; 75 mg, 4.9%).

View this table:
Table 5

Cardiovascular ischaemic events

50 mg b.d.75 mg b.d.110 mg b.d.150 mg b.d.
Patients intention-to-treat, n (%)371 (100.0)369 (100.0)368 (100.0)406 (100.0)347 (100.0)
 Cardiovascular death, non-fatal myocardial infarction, or non-haemorrhagic stroke14 (3.8)17 (4.6)18 (4.9)12 (3.0)12 (3.5)
  Cardiovascular death9 (2.4)8 (2.2)9 (2.4)5 (1.2)4 (1.2)
  Non-fatal myocardial infarction4 (1.1)9 (2.4)8 (2.2)7 (1.7)8 (2.3)
  Non-haemorrhagic stroke3 (0.8)0 (0.0)1 (0.3)0 (0.0)0 (0.0)
  Severe recurrent ischaemia9 (2.4)9 (2.4)11 (3.0)9 (2.2)11 (3.2)
  All-cause death14 (3.8)8 (2.2)10 (2.7)7 (1.7)7 (2.0)


Seventy-four per cent of the patients in the placebo group had a reduction in D-dimer concentrations after 1 or 4 weeks. In the respective 50, 75, 110, and 150 mg dabigatran groups, the proportions of patients with a D-dimer reduction were 81, 82, 86, and 89%, P< 0.001 for linear trend. At Weeks 1 and 4, the median D-dimer levels were on average ∼37 and 45% lower, respectively, in patients receiving dabigatran treatment compared with patients in the placebo group (P< 0.001) (Figure 3). The D-dimer concentrations in the dabigatran groups returned to similar levels as the placebo group after study drug termination (Figure 3). A decreased D-dimer concentration at Week 1 was associated with a significant reduction in the composite endpoint of cardiovascular death, non-fatal myocardial infarction, and non-haemorrhagic stroke, HR 0.47 (95 % CI 0.26; 0.82) compared with patients without decreased D-dimer concentration.

Figure 3

Median D-dimer concentration over time. Compared with placebo, median D-dimer concentrations were significantly (P< 0.001) lower after 1 and 4 weeks in all four dabigatran dose groups using Cochran–Armitage linear trend test.


The RE-DEEM study was the first randomized evaluation of treatment with an oral anticoagulant in patients with a recent non-ST or ST-elevation myocardial infarction in which all patients were required to be on antiplatelet treatment with both aspirin and clopidogrel. The addition of dabigatran, a direct thrombin inhibitor, was associated with a dose-dependent increase in major or clinically relevant minor bleeding. These findings are consistent with previous experiences with the direct thrombin inhibitor ximelagatran in the ESTEEM study7 and the novel factor Xa-inhibitors rivaroxaban and apixaban in the ATLAS-TIMI46 and APPRAISE studies, respectively.8,9 In the ESTEEM study, patients received single antiplatelet treatment with aspirin only. In the recent factor Xa inhibitor studies, where treatment with either single or dual antiplatelets was allowed, 76–78% of the patients were on dual antiplatelet therapy at randomization. In the present study, 99% of the patients received guideline-recommended dual antiplatelet treatment at randomization, with the vast majority still on dual antiplatelets after 6 months, and the relevant comparisons between the recent trials with novel anticoagulants should therefore focus on patients receiving dual antiplatelet treatment.

The rate of major or clinically relevant minor bleeding was 7.8–7.9% with the two highest doses of dabigatran, an approximately four-fold increase compared with placebo. Similar bleeding definitions were used in the APPRAISE and ATLAS studies. On top of dual antiplatelet treatment, there was a 2.6 times relative increase and 7.9% absolute rate with a 10 mg daily dose of apixaban, and a 3.2 times relative increase and a 12.2 % absolute rate with 10 mg rivaroxaban daily (doses that have been carried forward in phase III trials).

In RE-DEEM, the major bleeding rate in the placebo group was only ≤0.5% depending on the definition used, i.e. ISTH, TIMI, or GUSTO,1214 despite dual antiplatelet treatment which previously has been shown to increase bleeding.16 The low major bleeding rate might, at least in part, be related to the low recommended aspirin dose, i.e. <100 mg daily.17 There was ≤1% absolute increase in major bleeding events in the highest dabigatran dose groups or the two highest doses combined, irrespective of bleeding definition. Similar major bleeding rates were observed with apixaban up to 10 mg daily and rivaroxaban up to 10 mg daily on top of dual antiplatelet treatment in the APPRAISE and ATLAS studies.8,9

D-dimer levels were consistently and significantly reduced with all four dabigatran doses without a clear dose–response relationship, which is similar to the findings with the oral direct thrombin inhibitor ximelagatran.18 The D-dimer reduction, reflecting a decrease in thrombin generation and fibrin turn-over, was associated with a decreased cardiovascular event rate, as has previously been shown.19

The total number of patients experiencing ischaemic cardiovascular events during the study was low, with minor differences between the treatment groups. This finding is consistent with the results from APPRAISE and ATLAS studies, where patients receiving dual antiplatelet treatment also had low rates of ischaemic cardiovascular events, with small and statistically non-significant reductions by apixaban or rivaroxaban, respectively.8,9 Similar to these results with oral direct factor Xa inhibitors, short-term treatment in the acute phase with the intra-venous direct factor Xa inhibitor otamixaban, compared with unfractionated heparin and eptifibatide, showed a dose-dependent increase in bleedings by otamixaban and a tendency towards lower ischaemic events in ACS patients on dual antiplatelet therapy.20 Thus, the ideal time to start, the optimal duration of treatment, and the balance between efficacy and safety with all these new anticoagulants remain to elucidate, especially in combination with new and more potent antiplatelet agents such as prasugrel and ticagrelor.21,22 Notably, the phase III ACS trial with apixaban was prematurely terminated in November 2010 due to clear evidence of a clinically important increase in bleeding without meaningful reductions in ischaemic events among patients randomized to apixaban. These results emphasize the unclear role of triple antithrombotic therapy after an ACS.

In conclusion, 6 months treatment with dabigatran from 50 to 150 mg twice daily was associated with a dose-related two to four times increased risk of bleeding in post-myocardial infarction patients receiving dual antiplatelet treatment, although the absolute incidence of major bleeding events was low. Dabigatran significantly reduced coagulation activity and may have the potential to reduce cardiovascular events when added to dual antiplatelet treatment in the doses 110–150 mg twice daily. The net clinical benefit of dabigatran, balancing the reduction of thrombo-embolic events vs. the increase risk of bleeding, can only be appropriately evaluated in a large scale, adequately powered phase III study, for which there is currently no final decision.


The work was supported by Boehringer Ingelheim. An Operations Committee, with assistance of an international Steering Committee, was responsible for study design, conduct, and reporting. The lead authors of the manuscript had full availability to individual data and designed and led the performance of the statistical analyses. The decision to submit the manuscript was the responsibility of the Operations Committee.

Conflict of interest: J.O. reports receiving consulting fees and grant support from Boehringer Ingelheim; and lecture fees from Bayer, Merck, and Boehringer Ingelheim. A.B. reports receiving consulting fees from AstraZeneca, Eli Lilly, Novartis, and Sanofi-aventis; lecture fees from AstraZeneca, Boehringer Ingelheim, GlaxoSmithKline, and sanofi-aventis; grant support from AstraZeneca, Boehringer Ingelheim, Bristol-Myers Squibb, GlaxoSmithKline, and sanofi-aventis. C.B.G. reports consulting fees and grant support from Boehringer Ingelheim, sanofi-aventis, AstraZeneca, Bristol-Myers Squibb, The Medicines Company, and GlaxoSmithKline; consulting fees from Novartis; and grant support from Medtronic Foundation. A.S. reports consulting fees, lecture fees and grant support from Boehringer Ingelheim, lecture fees and grants from Eli Lilly and grant support from AstraZeneca. J.G.P.T. reports serving on various DSMBs and receiving lecture fees from Boehringer Ingelheim. F.V.W. reports receiving consulting and lecture fees from AstraZeneca, Eli Lilly, Boehringer Ingelheim, Schering-Plough, sanofi-aventis, and The Medicines Company; study grants from Boehringer Ingelheim, Schering-Plough and sanofi-aventis. L.W. reports receiving consulting fees, lecture fees, honoraria and grant support from Boehringer Ingelheim; consulting fees from Regado and Athera; lecture fees from AstraZeneca, Bristol-Myers Squibb, Eli Lilly, and GlaxoSmithKline; grant support from AstraZeneca, Bristol-Myers Squibb, GlaxoSmithKline, and Schering Plough. Y.K. and J.R. are Boehringer Ingelheim full-time employees.


We would like to acknowledge Ruth Harper and Jon Blatchford from Boehringer-Ingelheim and Susanne Heller from Uppsala Clinical Research Center (UCR) for their contribution to the trial, and to thank Ebba Bergman (UCR) for assistance with the report.


View Abstract