Dabigatran With or Without Concomitant Aspirin Compared
With Warfarin Alone in Patients With Nonvalvular Atrial
Fibrillation (PETRO Study)
Michael D. Ezekowitz, MD, PhDa,*, Paul A. Reilly, PhDb, Gerhard Nehmiz, PhDc,
Timothy A. Simmers, MDe, Rangadham Nagarakanti, MDa, Kambiz Parcham-Azad, MDa,
K. Erik Pedersen, MDf, Dominick A. Lionetti, MAb, Joachim Stangier, PhDd, and
Lars Wallentin, MD, PhDg
This is the ﬁrst evaluation of dabigatran, an oral direct thrombin inhibitor, in patients with
atrial ﬁbrillation (AF). Patients (n 502) were randomized to receive blinded doses of 50-,
150-, or 300-mg dabigatran twice daily alone or combined with 81- or 325-mg aspirin or
open-label warfarin administered to achieve an international normalized ratio of 2 to 3 for
12 weeks. Dabigatran plasma concentrations, activated partial thromboplastin time,
D-dimer, urinary 11-dehydrothromboxane B2 (DTB2), and liver function were measured at
baseline and at 1, 2, 4, 8, and 12 weeks. Clinical end points were assessed according to the
treatment received at the time of the event. Overall, 92% of patients completed the study.
Major hemorrhages were limited to the group treated with 300-mg dabigatran plus aspirin
(4 of 64), and the incidence was signiﬁcant versus 300-mg dabigatran alone (0 of 105,
p <0.02). Total bleeding events were more frequent in the 300-mg (39 of 169, 23%) and
150-mg (30 of 169, 18%) dabigatran groups compared with the 50-mg groups (7 of 107, 7%;
p 0.0002 and p 0.01, respectively). Thromboembolic events were limited to the 50-mg
dabigatran dose groups (2 of 107, 2%). The mean trough D-dimer measurements were
suppressed for the 2 highest doses of dabigatran and warfarin (international normalized
ratio of 2 to 3). Aminotransferase levels >3 times the upper limit of normal were observed
in 0.9% of the dabigatran recipients and in none of the warfarin recipients. Two dabigatran
recipients had aminotransferase levels >5 times the upper limit of normal as a result of
gallstones, which resolved. Trough activated partial thromboplastin time values were 1.2,
1.5, and 1.8 times the baseline level for the 50-, 150-, and 300-mg dabigatran groups,
respectively. DTB2 concentrations after 12 weeks of 50-, 150-, and 300-mg dabigatran
treatment were increased by 31%, 17%, and 23%, respectively, versus baseline (p 0.02,
p 0.03, and p 0.0004). In conclusion, major bleeding events were limited to patients
treated with dabigatran 300 mg plus aspirin and thromboembolic episodes were limited to
the 50-mg dabigatran groups. The 2 highest doses of dabigatran suppress D-dimer concen-
trations. Serious liver toxicity was not seen. The signiﬁcance of the increase of DTB2
concentrations in dabigatran-treated patients needs resolution. © 2007 Elsevier Inc. All
rights reserved. (Am J Cardiol 2007;100:1419 –1426)
Vitamin K antagonists are the only orally active antico- antagonists have a highly variable inter- and intraindi-
agulants currently available. They decrease the risk of vidual anticoagulant response, with multiple drug and
stroke and systemic thromboembolism in patients with food interactions, necessitating regular monitoring and
atrial ﬁbrillation (AF) by 68%.1 However, vitamin K adjustment of dose.2 These factors, plus fear of hemor-
rhage, have resulted in substantial underuse of warfarin,
a particularly in patients with AF.3 A ﬁxed-dose, orally
Lankenau Institute for Medical Research and The Heart Center, Wynne-
wood, Pennsylvania; bDepartment of Clinical Research, Boehringer Ingelheim
effective anticoagulant without the need for monitoring
Pharmaceuticals, Ridgeﬁeld, Connecticut; Departments of cMedical Data would constitute a signiﬁcant improvement. Dabigatran
Services and dDrug Metabolism and Pharmacokinetics, Boehringer Ingelheim etexilate is an orally available prodrug that is converted
Pharmaceuticals, Biberach, Germany; eThorax Centre, Amphia Hospital, to dabigatran, the active moiety. It is a potent, competi-
Breda, The Netherlands; fOdense University Hospital, Odense, Denmark; and tive, and reversible direct inhibitor of thrombin. Peak
Uppsala Clinical Research Centre, Uppsala, Sweden. Manuscript received dabigatran plasma concentrations occur 0.5 to 2 hours
March 14, 2007; revised manuscript received and accepted June 10, 2007.
after oral administration, resulting in a rapid onset of
Boehringer Ingelheim Pharmaceuticals, Biberach, Germany, is the sponsor
of this study and has provided a research grant. The members and structure of
action. There is a bi-exponential distribution phase with
the PETRO study group are given in the Appendix at the end of this article. a terminal half-life of 12 to 17 hours. As much as 80% of
*Corresponding author: Tel: 610-645-8451; fax: 610-645-8460. the drug is excreted unchanged by the kidneys. The
E-mail address: email@example.com (M.D. Ezekowitz). average absolute bioavailability of dabigatran is 6.5%.4,5
0002-9149/07/$ – see front matter © 2007 Elsevier Inc. All rights reserved. www.AJConline.org
1420 The American Journal of Cardiology (www.AJConline.org)
Figure 1. Treatment distribution, down-titration, and discontinuations. If glomerular ﬁltration rate was 50 and aPTT was 2.5 times the normal value,
down-titration was performed. Broken lines, down-titration; solid lines, discontinuation.
This is the ﬁrst evaluation of dabigatran in patients tomatic heart failure or left ventricular dysfunction (ejec-
with AF. The main goal of the study was to identify a safe tion fraction 40%), previous stroke or transient isch-
dose of dabigatran in patients with AF as determined by emic attack, or age 75 years. After entry of
the occurrence of bleeding and clinical events, and to approximately half of the patients, the requirement for
identify anticoagulant activity through activated partial coronary artery disease was removed to facilitate recruit-
thromboplastin time (aPTT) changes and the inhibition of ment. Exclusion criteria were mitral stenosis, prosthetic
D-dimer generation.6 heart valves, planned cardioversion, recent ( 1 month)
myocardial infarction, recent stroke or transient ischemic
Methods attack, coronary stent placement within 6 months, any
contraindication to or another indication for anticoagu-
This study was conducted in 53 centers in Denmark, the lant therapy, major hemorrhage in the past 6 months,
Netherlands, Sweden, and the United States. The protocol severe renal impairment (glomerular ﬁltration rate 30
was developed by the Steering Committee of the Preven- ml/min), abnormal liver function, risk of pregnancy, in-
tion of Embolic and Thrombotic Events in Patients with vestigational drug use within 30 days, or any other con-
Persistent AF (PETRO) study group. An independent dition that would not allow participation in the study.
adjudication committee blinded to treatment evaluated all
The trial had a total of 10 treatment groups. Three
bleeding events. An independent data and safety moni-
doses of dabigatran etexilate (50, 150, and 300 mg twice
toring board monitored the study for safety. Boehringer
daily) were combined in a 3 3 factorial fashion with no
Ingelheim (Biberach, Germany) sponsored the study and
aspirin or 81- or 325-mg aspirin once daily. Patients
was responsible for the statistical analysis conducted
according to a prospectively designed plan approved by received warfarin alone (international normalized ratio
the steering committee. All authors had access to the [INR] of 2 to 3) in the comparator group. The trial was
data. double-blind with respect to dabigatran dose but open-
The protocol, protocol amendments, informed con- label for concomitant aspirin treatment and for random-
sent, and subject information forms were reviewed and ization between dabigatran and warfarin groups. Ran-
approved by a local institutional review board or inde- domization was stratiﬁed in the ratio 6:9:9:4 (50-, 150-,
pendent ethics committee before initiation. The study and 300-mg dabigatran, and warfarin, respectively) (Fig-
was conducted in accordance with the ethical principles ure 1). The screening visit was the baseline visit for all
of the Declaration of Helsinki, International Conference laboratory tests and measurements with the exception of
on Harmonisation, and the Tripartite Guideline of Good D-dimer, the baseline visit for which for pre-/post-treat-
Clinical Practice. Written informed consent was obtained ment comparisons was the randomization visit.
from each patient. Two identical matching capsules containing 50- or
The PETRO study was a randomized trial of patients 150-mg dabigatran or placebo were taken twice daily for
with AF at high risk for thromboembolic events. Inclu- 12 weeks. All patients had been treated with a vitamin K
sion criteria were documented AF with coronary artery antagonist for 8 weeks before inclusion and required an
disease plus 1 of the following: hypertension requiring INR between 2 and 3 at study entry. This ensured that
medical treatment, diabetes mellitus (type 1 or 2), symp- baseline D-dimer measurements were obtained under the
Arrhythmias and Conduction Disturbances/Dabigatran in Nonvalvular AF 1421
condition of full anticoagulation. At the screening visit, linked ﬁbrin degradation products (Dade Behring, Mar-
warfarin pretreatment was terminated. Patients random- burg, Germany). Soluble ﬁbrin was determined by an
ized to receive warfarin or dabigatran started study treat- automated latex photometric assay (LPIA-Iatro SF; Mit-
ment when the INR was 1.5 to avoid excessive antico- subishi Kagaku Iatron, Gauting, Germany). Urinary sam-
agulation. ples for 11-dehydrothromboxane B2 (DTB2) excretion
Participants were monitored as outpatients at 1, 2, 4, 8, were obtained at screening and randomization and at
and 12 weeks after randomization. Renal function was cessation of treatment after the12-week visit. The deter-
measured with the Levey equation at the initial visit.7 minations of DTB2 in urine were performed with an
After 4 to 7 days of treatment, all patients with glomer- DTB2 enzyme immunoassay kit (Correlate-EIA; Assay
ular ﬁltration rates 50 ml/min underwent down-titration Designs, Ann Arbor, Michigan).
to once-daily dabigatran if the aPTT ratio at trough was No formal statistical hypothesis was tested. However,
2.5 times the local laboratory reference value. These the goal of the study was to determine whether there was
patients were analyzed according to the group to which a dose-related incidence of bleeding and to collect phar-
they were initially assigned. macodynamic data to assist in the selection of a dose(s)
Major hemorrhages were observed in the group that for the phase 3 study. The primary end point was inci-
received dabigatran 300 mg twice daily plus aspirin: 3 at dence of bleeding. A secondary objective was to measure
the 325-mg aspirin dose and 1 at the 81-mg dose. The suppression of D-dimer to evaluate the activity of dabig-
data and safety monitoring board recommended, and the atran; week 12 values were compared with baseline val-
steering committee agreed, that patients receiving ues, deﬁned as the value at the randomization visit for
300-mg dabigatran twice daily plus aspirin should be each treatment group. The mean trough measurements for
switched to 300-mg dabigatran twice daily without aspi- each group were calculated and a 1-sample Wilcoxon test
rin. This decision occurred after patient recruitment had was used to determine statistical signiﬁcance of changes from
been completed. baseline. Generally, data were reported as means SD or as
Stroke was deﬁned as an acute onset of a focal neu- median interquartile range. Differences between groups
rologic deﬁcit of vascular origin lasting for 24 hours. were analyzed using the chi-square test or the Fisher exact
Systemic thromboembolism was deﬁned as an acute non- test, and for quantitative data with 1-way analysis of vari-
intracerebral or noncoronary vascular event. ance or the Kruskal-Wallis test as appropriate. All analyses
Bleeding events were classiﬁed as major or minor and were performed with SAS software (version 8.2; SAS In-
were assigned to the treatment the patient was receiving
stitute, Cary, North Carolina).
at the time of onset. Major bleeding was deﬁned as fatal
or life-threatening retroperitoneal, intracranial, intraocu-
lar, or intraspinal bleeding; or bleeding requiring surgery Results
or transfusion of 2 U or associated with a decrease in
hemoglobin of 2.0 g/L. Minor bleeding was further Of 502 patients randomized, 411 (81.9%) were men, with
subdivided into clinically relevant or nuisance bleeding 192 (38.2%) having permanent AF, 195 (38.8%) persis-
episodes. Clinically relevant bleeding was deﬁned as skin tent AF, and 115 (22.9%) paroxysmal AF. The median
hematoma 25 cm2, spontaneous nose bleed of 5 min- duration of AF was 4 years (range 0.05 to 30). Mean ages
utes duration, macroscopic hematuria, spontaneous rectal were 70.9 7.9 years in patients with CAD (n 306)
bleeding, gingival bleeding for 5 minutes, any bleeding and 68.0 8.8 years in those without CAD. The baseline
leading to hospitalization, any bleeding leading to trans- characteristics were balanced among the 10 randomized
fusion 2 U, or any other bleeding considered relevant treatment groups (Table 1). Patients had a median of 3
by the investigator. risk factors for stroke.
Safety laboratory assessments were performed at The mean rate of compliance with dabigatran treat-
screening and 1, 4, 8, and 12 weeks after randomization ment (excluding patients who withdrew or were discon-
and, if necessary, at a further follow-up after the 12-week tinued) was 99.5%. Twelve patients underwent a dose
visit. Treatment was withdrawn if aspartate aminotrans- down-titration (Figure 1). Thirteen patients stopped as-
ferase, alanine aminotransferase, bilirubin, or alkaline pirin treatment during the trial. In the warfarin group, the
phosphatase concentrations were 5 times the upper INR was within the target range (i.e., 2 to 3) during
limit of normal at any time or 3 times the upper limit of 57.2% of the treatment time. The percent time in target
normal for 4 weeks, or if the patient developed clinical range increased from 48.2% in the ﬁrst 4 weeks to 62.1%
signs or symptoms of hepatic insufﬁciency. in weeks 5 through 8 and to 64.5% in weeks 9 through 12.
For the pharmacokinetic and pharmacodynamic anal- Overall, 38 patients discontinued trial treatment, 29
yses, samples were obtained at screening and randomiza- because of adverse events (Table 2); 3 withdrew consent;
tion and then at 1, 2, 4, 8, and 12 weeks after random- and 1 patient was not compliant with study medication.
ization. Dabigatran plasma concentrations were assayed Another patient returned for the ﬁnal visit, but it was not
using a liquid chromatograph–tandem mass spectrometry possible to determine whether the patient was compliant
method at AAI Deutschland, Neu-Ulm, Germany. The with study medication. The patient was classiﬁed as “dis-
aPTT was measured centrally by a coagulation analyzer continued.” The 4 remaining patients withdrew, 1 each as
(Roche Diagnostics, Nutley, New Jersey) at Boehringer a result of percutaneous coronary intervention for coro-
Ingelheim. Analyses of D-dimer were performed by the nary artery disease requiring clopidogrel, angiography
PLUS Assay for the turbidimetric determination of cross- planned before trial entry, difﬁculty with blood draws,
1422 The American Journal of Cardiology (www.AJConline.org)
Variable All Patients Dabigatran 50 mg Dabigatran 150 mg Dabigatran 300 mg Warfarin INR 2–3 p Value for
(n 502) Twice Daily Twice Daily Twice Daily (n 70) Equality of
(n 105) (n 166) (n 161) Groups
Age (mean SD, yrs) 70 (8.3) 70 (8.8) 70 (8.1) 69.5 (8.4) 69 (8.3) 0.7
Duration of AF median 4 (6.5) 3.6 (6.9) 3.9 (6.6) 6.4 (4.3) 3.4 (5.0) 0.4
(interquartile range, yrs)
Women 91 (18%) 21 (20%) 31 (18.7%) 28 (17.4%) 11 (15.7%) 0.9
Age 75 yrs 158 (31.5%) 36 (34.3%) 51 (30.7%) 52 (32.3%) 19 (27%) 0.8
TIA or stroke 87 (17.3%) 19 (18%) 29 (17.5%) 26 (16%) 13 (18.6%) 1.0
Hypertension 356 (71%) 70 (66.7%) 118 (71%) 119 (74%) 49 (70%) 0.6
Diabetes mellitus 126 (25%) 27 (25.7%) 45 (27%) 39 (24%) 15 (21.4%) 0.8
Heart failure 147 (29.3%) 35 (33.3%) 52 (31.3%) 36 (22.4%) 24 (34.3%) 0.1
Coronary artery disease 306 (61%) 64 (61%) 104 (63%) 96 (59.6%) 42 (60%) 1.0
Current or former smoker 365 (72.7%) 76 (72.4%) 120 (72.3%) 116 (72%) 53 (75.7%) 1.0
blockers 349 (69.5%) 69 (65.7%) 121 (73%) 110 (68.3%) 49 (70%) 0.6
ACE inhibitor/angiotensin 355 (70.7%) 70 (66.7%) 116 (69.8%) 112 (69.5%) 57 (81.4%) 0.2
Verapamil/diltiazem 95 (19%) 16 (15%) 31 (18.7%) 34 (21%) 14 (20%) 0.7
Other calcium inhibitors 115 (22.9%) 26 (24.7%) 37 (22.3%) 38 (23.6%) 14 (20%) 0.9
Amiodarone 37 (7.4%) 9 (8.5%) 9 (5.4%) 13 (8%) 6 (8.5%) 0.7
Digoxin 219 (43.6%) 46 (43.8%) 75 (45%) 66 (41%) 32 (45.7%) 0.9
Diuretic 289 (57.5%) 59 (56%) 89 (53.6%) 97 (60%) 44 (63%) 0.5
Statin 290 (57.7%) 58 (55%) 100 (60%) 95 (59%) 37 (53%) 0.7
ACE angiotensin-converting enzyme; TIA transient ischemic attack.
Adverse events leading to discontinuation
Event Dabigatran Dose (twice daily) Warfarin to INR of 2–3
50 mg (n 107) 150 mg (n 169) 300 mg (n 169)
Patients discontinuing with adverse events* 5 9 15 0
Cardiovascular and peripheral embolic events† 4 2 1 0
Major/clinically relevant hemorrhage 0 4 7 0
Gastrointestinal symptoms‡ 2 2 8 0
ALT/AST increased 0 0 1 0
Other symptoms§ 2 1 4 0
* Some patients had 1 event.
Cardiac failure, acute coronary syndrome, cerebrovascular accident, and peripheral emboli, including renal infarction.
Abdominal pain, dyspepsia, and nausea.
Fatigue, dyspnea, visual disturbance, worsening dizziness, renal pain, and chest pain.
ALT alanine aminotransferase; AST aspartate aminotransferase.
and personal reasons. Patients were followed for 2 weeks When the doses of dabigatran were compared with
after discontinuation of study medication. each other, irrespective of aspirin assignment, there were
The primary outcome was the frequency of bleeding differences in total bleeding episodes in the 300-mg twice
events (Table 3). Major bleeding events were limited to daily and 150-mg twice daily groups versus the 50-mg
the group treated with 300-mg dabigatran twice daily twice daily group (37 of 169 and 30 of 169 vs 7 of 107,
plus aspirin (4 of 64). The rate was statistically different p 0.0002 and p 0.01, respectively).
compared with the group treated with dabigatran 300 mg There were 2 patients with systemic thromboembolic
twice daily without aspirin (0 of 105, p 0.02). events (Table 3), both of whom received 50-mg dabigatran
There was a signiﬁcant difference in major plus clin- twice daily (1.96%). One patient had a peripheral embolism
ically relevant bleeding episodes (11 of 64 vs 6 of 105, p to the toe and the other patient had a stroke and a renal
0.03) and total bleeding episodes (25 of 64 vs 14 of infarction.
105, p 0.0003) between 300-mg dabigatran twice daily Seven patients reported angina, of which 2 patients
plus aspirin and 300-mg dabigatran twice daily without were classiﬁed as having acute coronary syndrome: 1
aspirin. The frequency of bleeding in the group treated treated with 50-mg dabigatran twice daily plus 81-mg
with 50-mg dabigatran twice daily was signiﬁcantly aspirin and the other treated with 300-mg dabigatran
lower than that in the warfarin group: 7 of 107 vs 12 of twice daily plus 81-mg aspirin. Both patients were dis-
70 (p 0.044). continued from the trial (Table 2). Four patients devel-
Arrhythmias and Conduction Disturbances/Dabigatran in Nonvalvular AF 1423
Major or clinically relevant bleeding episodes and thromboembolic events
Dabigatran Dose Aspirin Dose (mg) No. of Patients* Bleeding Events Thromboembolic Events
(mg twice daily)
Major Clinically Relevant Plus Major Total
50 0 59 0 0 2 (3.4%) 1 (1.7%)
50 81 21 0 1 (4.8%) 2 (9.5%) 1 (4.8%)
50 325 27 0 1 (3.7%) 3 (11.1%) 0
150 0 100 0 9 (9%) 15 (15%) 0
150 81 36 0 2 (5.6%) 8 (22.2) 0
150 325 33 0 2 (6.1%) 7 (21.2%) 0
300 0 105 0 6 (5.7%) 14 (13.3%) 0
300 81 34 1 (2.9%) 5 (14.7%) 11 (32.4%) 0
300 325 30 3 (10%) 6 (20%) 14 (46.7%) 0
Warfarin once daily 0 70 0 4 (5.7%) 12 (17.1%) 0
* Thirteen patients assigned to dabigatran plus aspirin stopped receiving aspirin during the trial (2 receiving 50-mg dabigatran, 3 receiving 150-mg
dabigatran, and 8 receiving 300-mg dabigatran twice daily) and then were analyzed in dabigatran-alone groups.
Figure 2. Box and whisker plots of median trough plasma concentration of
dabigatran (in nanograms per milliliter) on maintenance treatment for each
Figure 3. Relation between dabigatran plasma concentration and aPTT.
dabigatran treatment group.
oped congestive heart failure, with 1 patient receiving trough levels with 50-, 150-, and 300-mg twice daily
150-mg dabigatran twice daily discontinuing the trial. dabigatran doses were 1.2, 1.5, and 1.8, respectively
None of these events resulted in statistically signiﬁcant (Figure 4).
differences between treatment groups. Plasma D-dimer measurements at an INR 1.5 were
Dabigatran (plasma half-life of 12 to 17 hours) at- higher than measurements at an INR 2, with geometric
tained steady state by the time of the ﬁrst follow-up at 4 means of 76.5 to 83.6 ng/ml (n 432, p 0.001),
to 7 days. The mean trough plasma concentration in- reﬂecting a washout of the warfarin effect. A comparison
creased linearly with increasing dose (Figure 2). Reduced of the change in median D-dimer measurements of each
creatinine clearance resulted in higher dabigatran plasma group at baseline versus the last on-treatment measure-
concentrations (data not shown). The mean creatinine ment (Table 4), generally at 12 weeks, showed a 13%
clearance was 81 ml/min; it was 69 ml/min in women and relative increase of D-dimer concentrations in the group
83 ml/min in men. Trough plasma concentrations in treated with 50-mg dabigatran twice daily (p 0.0008)
women were 12% to 20% higher than those in men. Body and a 3% relative increase in the group treated with
mass index and weight did not inﬂuence dabigatran con- 150-mg dabigatran twice daily (p 0.027), but no sig-
centrations (data not shown). niﬁcant changes with 300-mg dabigatran twice daily
There was a general correlation between aPTT and ( 0%, p 0.413) or warfarin ( 1%, p 0.267). There
dabigatran plasma concentrations with a ﬂattening re- was poor correlation between plasma measurements of
sponse at higher plasma concentrations (Figure 3). soluble ﬁbrin for dabigatran and warfarin (data not
Trough aPTT values for each dose group had low inter- shown). Aspirin treatment had no effect on any of these
individual variability, with coefﬁcients of variation be- analyses.
tween 13% and 21%. Mean trough aPTT measurements DTB2 values after 12 weeks of treatment with dabig-
obtained at steady state showed minimal variation be- atran (Table 4) showed a non– dose-related increase be-
tween visits. The ratios of aPTT values at baseline and tween 17% and 31%. Values during warfarin treatment
1424 The American Journal of Cardiology (www.AJConline.org)
clinical observations of bleeding at higher doses and
thromboembolic events in the lowest dose group (50 mg
twice daily). The median aPTT measurements in patients
receiving dabigatran 150 mg twice daily are comparable
to measurements reported with a therapeutic dose of
ximelagatran (36 mg twice daily).10 Mean D-dimer con-
centrations were almost completely suppressed except at
the lowest dose of dabigatran. The highest dose of dabig-
atran achieved D-dimer suppression similar to that
achieved with warfarin. Based on bleeding rates and
aPTT and D-dimer measurements, dabigatran 150 mg
twice daily is well tolerated and shows anticoagulant
As expected, there was a pronounced depression of
DTB2 excretion for all doses of aspirin treatment. Unex-
pectedly, urinary excretion of DTB2 was approximately
20% higher after 12 weeks of treatment with all dabig-
Figure 4. Box and whisker plots of median trough aPTT measurements atran doses compared with warfarin in patients who did
during treatment with different dabigatran dosages. not receive aspirin. There was no dose effect for dabig-
atran. There have been no previous reports of effects of
were unchanged from baseline ( 4%, p 0.32). In the direct thrombin inhibitors on thromboxane excretion. An
presence of aspirin, DTB2 concentrations decreased by increase of thromboxane excretion may suggest a plate-
40% to 57% compared with baseline measurements. let-activating effect of treatment in the absence of con-
There was a 0.9% incidence of increased aminotransfer- comitant aspirin treatment. This might theoretically cause
ase levels 3 times the upper limit of normal in dabigatran- a paradoxic increase in thrombotic risk, but this would
treated patients (4 of 432; Table 5). need to be conﬁrmed in a clinical outcomes trial. In
Adverse events were more frequent in the dabigatran contrast, ximelagatran, an oral thrombin inhibitor, de-
groups than in the warfarin-treated patients. The most creased cardiac events in patients with recent myocardial
commonly reported adverse events were gastrointestinal infarction when administered in addition to aspirin.11
disorders such as diarrhea, nausea, or vomiting (26%); Despite initial regulatory concerns by the United States
followed by general system disorders such as fatigue or Food and Drug Administration, trials of ximelagatran in
edema (12%), dizziness and headache (12%), and infec- patients with AF showed no excess myocardial infarc-
tions. Most of these were mild and required no change in tions in patients receiving ximelagatran with or without
treatment. All adverse events leading to treatment dis- aspirin compared with warfarin.12 A previous meta-anal-
continuation are listed in Table 2. ysis in acute coronary syndromes has reported a possible
increased risk of cardiac events with univalent oral
Discussion thrombin inhibitors compared with heparin, but this anal-
ysis included mostly dose-ﬁnding studies.13
This phase 2 trial of several ﬁxed doses of a direct Because of liver function abnormalities associated
thrombin inhibitor with and without aspirin compared with ximelagatran, the present study had an extensive
with warfarin alone in AF established a dose response for surveillance of liver function. Only 0.9% of patients had
bleeding and an upper limit of tolerability (300 mg twice increased levels of alanine aminotransferase or aspartate
daily plus aspirin) based on the frequency of major and clini- aminotransferase 3 times the upper limit of normal. No
cally signiﬁcant bleeding events. As anticipated, the frequency patient had a drug-related increase in bilirubin 2 times
of thromboembolic events was too low to reach conclusions, the upper limit of normal within 30 days after an amino-
but the only 2 strokes in this 12-week trial occurred in patients transferase level increase 3 times the upper limit of
receiving the lowest dose of dabigatran (50 mg twice daily). normal.14 However, the duration of the study was only 12
Because many patients with AF have concomitant coronary weeks. The long-term effects of dabigatran on liver func-
disease and may require aspirin, the statistically signiﬁcant tion are currently being evaluated in the extension of the
effect of aspirin on bleeding rates is important in designing PETRO study and the phase 3 Randomized Evaluation of
further trials. Efﬁcacy and safety of dabigatran in prevention of Long term anticoagulation therapy (RE-LY) study.
venous thrombosis has already been demonstrated.8
The pharmacokinetic and pharmacodynamic measure- Appendix
ments added further insights for dose selection. Dabigat-
ran plasma concentrations increased consistently with Steering Committee: L. Wallentin (chairman), M.
dose, as did aPTT and D-dimer suppression. The aPTT is Ezekowitz (co-chairman United States national coordina-
a readily accessible test that allows a real-time measure- tor), T. Simmers (Netherlands national coordinator), and K.
ment of the effect of dabigatran on the coagulation cas- Erik Pedersen (Denmark national coordinator). The follow-
cade. D-dimer suppression may be a valuable marker of ing representatives of Boehringer Ingelheim also partici-
long-term effects on the level of anticoagulation.9 The pated in the Steering Committee as nonvoting members:
changes in these parameters were consistent with the L.E. Lins, G. Nehmiz, P. Reilly, J. Stangier, H. Paalum
Arrhythmias and Conduction Disturbances/Dabigatran in Nonvalvular AF 1425
Median plasma concentrations of D-dimer and median urinary 11-dehydrothromboxane B2 excretion: baseline versus ﬁnal on-treatment value
Measurement Dabigatran (mg twice daily) Warfarin to INR of 2–3
50 150 300
D-dimer n 95 n 137 n 131
Baseline (ng/ml) (IQR) 75 (98.5) 70 (92) 85 (112) 84 (106)
Change at 12 wks (IQR) 10 (43) 2 (34) 0 (34) 1 (32)
p Value* 0.0008 0.027 0.413 0.267
DTB2† n 44 n 73 n 73 n 51
Baseline (pg/mg/creatinine) (IQR) 3,382 (2,176) 3,602 (2,585) 3,178 (1,810) 3,409 (2,983)
Change at 12 wks (pg/mg/creatinine) (IQR) 1,040 (2,681) 610 (2,740) 720 (2,470) 140 (1,970)
p Value* 0.019 0.028 0.0004 0.322
* p value determined by a 2-sided test of no difference from baseline: the differences among treatments were p 0.003 and p 0.01 for D-dimer and
DTB2 for patients without aspirin.
IQR interquartile range.
Occurrence of increased alanine aminotransferase measurements
ALT Measurement Dabigatran Total Dabigatran (mg twice daily) Warfarin
(n 432) (n 70)
50 (n 105) 150 (n 166) 300 (n 161)
1 times ULN 24 (5.6%) 6 (5.7%) 11 (6.6%) 7 (4.3%) 7 (10%)
2 times ULN 6 (1.4%) 0 3 (1.8%) 3 (1.9%) 0
3 times ULN 4 (0.9%)* 0 2 (1.2%)* 2 (1.2%)* 0
5 times ULN 2 (0.5%)† 0 2 (1.2%)† 0 0
* No simultaneous elevation of bilirubin 2 times upper limits of normal (ULN). Concomitant increase of aspartate aminotransferase was observed in 1
case but did not exceed the ALT concentration.
One patient had obstructive jaundice and pancreatitis as a result of gallstones in the common bile duct and recovered completely after cholecystectomy.
The other had gallstones and spontaneously recovered.
Abbreviations as in Table 2.
(Trial Clinical Monitor until June 2004), A. Meijer (Trial 1. Atrial Fibrillation Investigators. Risk factors for stroke and efﬁcacy of
Clinical Monitor from June 1, 2004, to December 1, 2004). antithrombotic therapy in atrial ﬁbrillation: analysis of pooled data
from ﬁve randomized controlled trials. Arch Intern Med 1994;154:
The Data and Safety Monitoring Board of the PETRO Study 1449 –1457.
group are: J. Godtfredsen (chairman), Kobenhavn, Den- 2. Hirsh J, Dalen JE, Deykin F, Poller L, Bussey H. Oral anticoagulants.
mark; J.G.P. Tijssen, Academic Medical Center, University Mechanism of action, clinical effectiveness, and optimal therapeutic
of Amsterdam, The Netherlands; and H.R. Büller, Aca- range. Chest 1995;108(suppl):231S–246S.
3. Go AS, Hylek EM, Borowsky LH, Phillips KA, Selby JV, Singer DE.
demic Medical Center, University of Amsterdam, The Neth- Warfarin use among ambulatory patients with nonvalvular atrial ﬁbril-
erlands. The Bleeding Adjudication Committee of the lation: the Anticoagulation and Risk Factors in Atrial Fibrillation
PETRO Study group are: M.H. Prins, University Hospital (ATRIA) study. Ann Intern Med 1999;131:927–934.
Maastricht, The Netherlands; U. Angerås, Kirurgkliniken, 4. Stangier J, Rathgen K, Staehle H, Gansser D, Roth W. The pharma-
Göteborg, Sweden; A. Falk, Carlanderska Hospital, Göte- cokinetics, pharmacodynamics and tolerability of dabigatran etexi-
late, a new oral direct thrombin inhibitor, in healthy male subjects.
borg, Sweden; and G. Andersen, Arhus Kommunehospital, Br J Clin Pharmacol 2007;64:292–303.
Denmark. The investigators of the PETRO Study group are: 5. Di Nisio M, Middeldorp S, Büller HR. Direct thrombin inhibitors.
S. Lind Rasmussen, E. Agner, K. Skagen, S. Husted, K. N Engl J Med 2005;353:1028 –1040.
Egstrup, H. Kræmmer Nielsen, K. Skodebjerg Kristensen, 6. Lip GYH, Lip PL, Zariﬁs J, Watson RDS, Bareford D, Lowe GDO,
L. Hvilsted Rasmussen, K. Korsgaard Thomsen, M. Ege Beevers DG. Fibrin D-dimer and betathromboglobulin as markers of
thrombogenesis and platelet activation in atrial ﬁbrillation. Circulation
Olsen, P. Fruergaard, and K. Klarlund (Denmark); M. 1996;94:425– 431.
Rosenqvist, S. Bandh, A. Englund, J.E. Karlsson, F. 7. Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D, for the
Rönn, P.L. Ågren, O. Nilsson, O. Hansen, B. Holmberg, Modiﬁcation of Diet in Renal Disease Study Group. A more accurate
F. Maru, and K. Eggers (Sweden); F.R. den Hartog, W.L. method to estimate glomerular ﬁltration rate from serum creatinine: a
new prediction equation. Ann Intern Med 1999;130:461– 470.
ten Holt, G.J. de Weerd, J.B. Winter, M.J. de Leeuw, 8. Eriksson BI, Dahl OE, Büller HR, Hettiarachchi R, Rosencher N, Bravo
J.G.M. Tans, L. Cozijnsen, C.M. Leenders, H.J.M. Thi- ML, Ahnfelt L, Piovella F, Stangier J, Kälebo P, Reilly P, for the BISTRO
jssen, H.A. van de Klippe, T. Slagboom, and J.H. Fast II Study Group. A new oral direct thrombin inhibitor, dabigatran etexilate,
(The Netherlands); and F. Cardona, L.W. Sprinkle Jr, D. compared with enoxaparin for prevention of thromboembolic events fol-
Hotchkiss, F.A. McGrew, D. Philips, J. Cieszkowski, S. lowing total hip or knee replacement: the BISTRO II randomized trial. J
Thrombosis Haemostasis 2005;3:103–111.
Jerome, T.D. Bahnson, J. Cooper, P. Zwerner, J. Leppo, 9. Lip GYH, Lowe GDO, Rumley A, Dunn FG. Increased markers of
D.J. Slotwiner, M.B. Cohen, and S.O. Gottlieb (United thrombogenesis in chronic atrial ﬁbrillation: effects of warfarin treat-
States). ment. Br Heart J 1995;73:527–533.
1426 The American Journal of Cardiology (www.AJConline.org)
10. Carlsson SC, Mattsson C, Eriksson UG, Sarich TC, Wahlander K, ing aspirin with anticoagulant therapy in patients with atrial
Eliasson A, Karlson BW, Sheth SB, Held P. A review of the effects of ﬁbrillation: An exploratory analysis of stroke prevention using an oral
the oral direct thrombin inhibitor ximelagatran on coagulation assays. thrombin inhibitor in atrial ﬁbrillation (SPORTIF) trials. Am Heart J
Thromb Res 2005;115:9 –18. 2006;152:967–973.
11. Wallentin L, Wilcox R, Weaver WD, Emanuelsson H, Goodvin A, 13. Direct Thrombin Inhibitor Trialists Collaborative Group. Direct
Nyström P, Bylock A. Oral ximelagatran for secondary prophylaxis thrombin inhibitors in acute coronary syndromes: principal results of a
after myocardial infarction: the ESTEEM randomised controlled trial. meta-analysis based on individual patients’ data. Lancet 2002;359:
Lancet 2003;362:789 –797. 294 –302.
12. Flaker GC, Gruber M, Connolly SJ, Goldman S, Chaparro S, Vahanian 14. Willis M. Drug-induced hepatotoxicity. J Clin Gastroenterol 2005;
A, Halinen MD, Horrow J, Halperin JL. Risk and beneﬁts of combin- 39(suppl):S83–S89.