REVIEWS

Oral anticoagulants in the management

of venous thromboembolism
John N. Makaryus, Jonathan L. Halperin and Joe F. Lau
Abstract | Despite advances in diagnosis, prevention, and management, venous thromboembolism (VTE) remains
a common cause of morbidity and mortality. For decades, antithrombotic therapy for prevention and treatment of
VTE was limited to parenteral agents related to heparin and oral vitamin K antagonists (VKAs). Both classes
of anticoagulants are effective, but have limitations, including considerable variability in dose–response, narrow
therapeutic margins between the risks of thrombosis and bleeding, and the need to monitor anticoagulation
intensity. Over the past decade, the introduction of new oral anticoagulants that specifically inhibit coagulation
factors IIa (thrombin) or Xa has changed practice in a variety of clinical situations, including VTE prophylaxis
and treatment. In this Review, we outline the use of the novel oral anticoagulants apixaban, dabigatran,
edoxaban, and rivaroxaban in the prevention and treatment of VTE, and discuss practical considerations
for choosing the appropriate drug for each patient. Although the introduction of novel anticoagulant drugs is
promising, selecting the optimum strategy for an individual patient requires an understanding of the specific
circumstances associated with thrombus formation and the pharmacological properties of each agent.
Makaryus, J. N. et al. Nat. Rev. Cardiol. 10, 397–409 (2013); published online 21 May 2013; doi:10.1038/nrcardio.2013.73

Introduction
Venous thromboembolism (VTE) can be difficult to anticoagulants that offer the effectiveness of VKAs with
prevent and manage because of the diversity of patient- fewer logistical limitations.
specific and disease-specific risk factors. Patients at Other anticoagulant agents for the management
risk of VTE include those with atrial fibrillation (AF), of  VTE include unfractionated heparin, the low-
valvular heart disease, or malignancy, and those under­ molecula­r-weight heparins (LMWHs, for example
going major orthopaedic surgery of the hip or knee, dalteparin or enoxaparin), and indirect-acting
all of whom are managed differently. Nevertheless, for factor Xa inhibitors (such as fondaparinux). Although
over half a century, long-term anticoagulant therapy has these drugs can be viable alternatives to VKAs in
been limited to vitamin K antagonists (VKAs), such as most clinical situations, the routes of administration
warfarin, which is the most-commonly used VKA in (subcutaneous­ and parental) make these drugs time-
the world.1 Although the clinical effectiveness of dose- consuming to use and less convenient than oral medica-
adjusted VKA therapy for prevention and treatment tions. Furthermore, the pharmacokinetic properties of
of thromboembolic disease is not disputed, long-term these anticoagulants have limited their use. For example,
use of VKAs is difficult, and anticoagulation intensity, unfractionated heparin requires continuous intravenous
measured using the international normalized ratio, can infusion or daily dosing, and the LMWHs have unpre-
be outside the intended therapeutic range for much of dictable bioavailability in some patient populations, such Department of
the treatment time. VKAs have a number of limitations, as those with obesity or renal failure. Cardiology, Hofstra
including a narrow therapeutic window between the These issues have stimulated the development of North Shore-Long
Island Jewish School of
risks of thrombosis and haemorrhage, which necessi- altern­ative agents with practical advantages over VKAs Medicine, Long Island
tates frequent blood-coagulation monitoring; pharmaco­ and the intravenous anticoagulant drugs. 4 A large Jewish Medical Center,
270‑05 76th Avenue,
kinetics that are subject to variation owing to genetic number of clinical trials of novel oral anticoagulant Suite O‑4000, New
and physiological factors; and frequent interactions drugs have been completed for indications such as atrial Hyde Park, NY 11040,
with other drugs and foods.2,3 The ageing patient popu- fibrillation and acute coronary syndrome (ACS) as well USA (J. N. Makaryus,
J. F. Lau). Zena and
lation, among whom anticoagulation is required for a as VTE. In this Review, we evaluate clinical trial data Michael A. Wiener
variety of indications, has increased the need for oral for the use of novel oral anticoagulants in comparison Cardiovascular
Institute, Mount Sinai
with conventional anticoagulation strategies for VTE, School of Medicine,
Competing interests and discuss the use of these agents in the prevention and One Gustave L. Levy
Place, Box 1030, New
J. L. Halperin declares associations with the following treatment of this condition.5
York, NY 10029, USA
companies: AstraZeneca, Bayer AG HealthCare, Biotronic, (J. L. Halperin).
Boehringer Ingelheim, Daiichi Sankyo, Johnson & Johnson,
Ortho–McNeil–Janssen Pharmaceuticals, and Sanofi–Aventis.
Novel oral anticoagulants
Correspondence to:
See the article online for full details of the relationships. VKAs inhibit γ‑glutamyl carboxylation of coagulation J. F. Lau
J. N. Makaryus and J. F. Lau declare no competing interests. factors II, VII, IX, X, and the coagulation inhibitor [email protected]

NATURE REVIEWS | CARDIOLOGY VOLUME 10  |  JULY 2013  |  397

© 2013 Macmillan Publishers Limited. All rights reserved
REVIEWS

Key points interactions with commonly prescribed medications

such as atorvastatin and digoxin. Transport of a substan-
■■ Vitamin K antagonists (VKAs), such as warfarin, are the mainstay of the
tial proportion of the active drug to the intestinal lumen
management of venous thromboembolism (VTE)
■■ The logistical difficulties associated with the use of VKAs include a narrow
is, however, mediated by P‑glycoprotein, and inhibitors
therapeutic window, unpredictable pharmacodynamics, and the need for routine and activators of the P‑glycoprotein efflux mechanism
monitoring; novel oral anticoagulants address some of these concerns can alter the plasma level of dabigatran. Care should also
■■ Novel anticoagulants include direct thrombin inhibitors, such as dabigatran, be taken, and the dose adjusted, in patients with renal
and factor Xa inhibitors such as apixaban, edoxaban, and rivaroxaban insuff­iciency, because approximately 80% of dabigatran
■■ Novel anticoagulants are in various stages of approval in the USA and Europe is cleared by the kidneys.9 The therapeutic effect of dabi-
for several indications, including stroke prevention in atrial fibrillation, and the gatran does not require routine blood-test monitor­ing.
prophylaxis and treatment of VTE
Thrombin time correlates most closely with the blood level
■■ The inability to quickly reverse or properly monitor the effects of the novel
anticoagulants must be considered in the clinician’s decision-making algorithm and anticoagulant effect of dabigatran, but measurement
■■ The results of randomized trials will continue to clarify the role of these new requires a diluted assay methodology.10,11 Prolongation of
agents in the management and prevention of VTE events the activated partial thromboplastin time correlates more-
closely with blood levels and the anti­coagulant effect of
dabigatran than the prothromb­in time, and is used where
the assay for thrombin time is u­navailable. Activated
XII XIIa
partial thromboplastin times between 1.5 and 2.0 times
Vitamin K the control value are typically observed during therapeutic
antagonists
XI XIa dosing. Monitoring of dabigatran is not routine; there-
fore, clinicians are underexperience­d in monitoring its
Vitamin K
IX IXa VIIa VII effects, if indicated.
antagonists
On the basis of the VTE prophylaxis trials described
VIIIa TF below, dabigatran was approved in Europe and Canada
Factor Xa inhibitors: in 2008 for prevention of VTE in patients undergoing
■ Apixaban total hip or knee replacement surgery, but the FDA has
■ Betrixaban not yet granted approval for these indications (Table 2).
■ Edoxaban X Xa X
■ Rivaroxaban In countries where dabigatran has been approved, this
agent is initiated at a dose of 110 mg within 1–4 h of
Vitamin K Va the operation, and continued at a daily dose of 220 mg.
antagonists
A lower dose of 75 mg twice daily is recommended for
Direct thrombin patients with moderate or severe renal insufficiency
Prothrombin Thrombin inhibitor: (creatinine clearance 30–50 ml/min and 15–30 ml/min,
II IIa • Dabigatran
respectively), those aged >75 years, or patients taking
Cross-
Fibrinogen Fibrin linked medications such as amiodarone, quinidine, or vera-
I Ia fibrin pamil, which inhibit P‑glycoproteins (Table  2).
clot
Dabigatran should be prescribed for a minimum of
Figure 1 | Coagulation cascade and point of effect of the common oral 10 days after knee or hip arthroplasty, with recom-
anticoagulants. Vitamin K antagonists, such as warfarin, inhibit factors II, VII, mendations to extend therapy for least 28–35 days if
IX, and X. Dabigatran directly inhibits factor IIa (thrombin). Apixaban, betrixaban, possible in patients who have undergone hip arthro-
edoxaban, and rivaroxaban inhibit factor Xa. Abbreviation: TF, tissue factor. plasty. Dabigatran is not approved for use in the treat-
ment of acute VTE. Filing with the FDA to approve
proteins C and S. By contrast, novel oral anti­coagulants the use of dabigatran for this indication is likely to take
target specific proteins in the coagulation cascade an important step forward with the completion of the
(Figure 1), and have predictable pharmacokinetic pro- RE‑COVER II trial12 in 2012.
files that simplify dosing and render routine coagulation
monitoring unnecessary. However, the benefits of the VTE prophylaxis
newer agents come at a cost. Strategies to rapidly quan- The efficacy of dabigatran for the prevention of VTE has
tify or reverse the anticoagulant effects of these drugs been examined in four large-scale clinical trials. In the
are lacking.6,7 The current indications and data from RE‑MODEL trial,13 dabigatran (150 mg or 220 mg daily)
trials on VTE prophylaxis and treatment of the novel was compared with enoxaparin (40 mg subcutaneously
anticoagulant­s are discussed below for each drug. daily) in patients undergoing knee replacement surgery
(Table 3). Dabigatran was noninferior to enoxaparin
Dabigatran with respect to the primary end point—a composite of
Dabigatran, a direct thrombin inhibitor, is administered VTE events and all-cause mortality. However, the results
orally as dabigatran etexilate, a prodrug with an overall of the RE‑MOBILIZE trial,14 in which two daily doses of
bioavailability of about 3–7% (Table 1).8 The half-life is dabigatran were compared with enoxaparin (30 mg sub-
approximately 12–17 h. Dabigatran is not dependent on cutaneously twice daily) in knee replacement surgery,
the cytochrome P450 system for conversion to the active did not show the noninferiority of dabigatran in prevent-
drug or for its metabolism, which avoids drug–drug ing the same end point (34% with dabigatran 150 mg;

398  |  JULY 2013  |  VOLUME 10  www.nature.com/nrcardio

© 2013 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Table 1 | Novel oral anticoagulants in VTE prophylaxis and treatment

Characteristic Apixaban Dabigatran Edoxaban Rivaroxaban
Mechanism of Direct factor Xa inhibitor Direct thrombin (IIa) inhibitor Direct factor Xa Direct factor Xa inhibitor
action inhibitor
Oral bioavailability 50% 3–7% Approximately 60% 10 mg: 80–100%; 20 mg: 66%
Half-life 8–15 h 12–17 h 8–10 h 5–9 h
Metabolism and 27% renal, 73% faecal 80% renal, 20% faecal 33% renal 66% renal, 33% faecal
excretion or biliary
Protein binding 87% 35% 40–59% 92–95%
Time to peak effect 1–3 h 1–3 h 1–2 h 1–3 h
Dosing for VTE *Initiate at 2.5 mg twice *Normal renal function: initiate ‡
5, 15, 30 and 60 mg §
Initiate at 10 mg (6–10 h after surgery);
prophylaxis daily (12‑–24 h after at 110 mg (1–4 h after surgery); daily have been continue at 10 mg daily for 12 days
surgery); continue at 2.5 mg continue at 220 mg for 28–35 days evaluated, optimal (knee surgery) or 25 days (hip surgery)
twice daily for 10–14 days Abnormal renal function (creatinine dose is unknown
(knee surgery) or clearance <50 ml/min): initiate at 75 mg;
32–38 days (hip surgery) continue at 150 mg daily for 28–35 days
Dosing for VTE Not currently approved Not currently approved Not currently Acute VTE: initiate at 15 mg twice daily
treatment approved for 3 weeks, then 20 mg daily for
treatment period
Extended therapy: after an initial 6 months
of anticoagulation, initiate at 20 mg daily
Antidote None None None None
Coagulation No No No No
monitoring
Drug interactions CYP3A4, P‑glycoprotein P-glycoprotein CYP3A4, P‑glycoprotein CYP3A4, P‑glycoprotein
*Canada and Europe. Japan. Canada, Europe, and USA. Abbreviations: CYP3A4, cytochrome P450 3A4; VTE, venous thromboembolism.
‡ §

31% with dabigatran 220 mg; 25% with enoxaparin). designed, phase III RE‑COVER II study, 12 in which
In the RE‑NOVATE15 and RE‑NOVATE II16 trials, dabi- 6 months of treatment with dabigatran or war­farin for the
gatran (220 mg daily) was noninferior to enoxaparin treatment of acute VTE was compared in 2,589 patients,
(40 mg subcutaneously daily) in reducing the risk of has just been completed and the results are forthcom-
the composite end point of VTE or all-cause mortality ing. 18 The main difference betweenRE‑COVER and
in patients undergoing hip replacement surgery. Safety RE‑COVER II is that participants in RE‑COVER II
profiles for dabigatran and enoxaparin were similar in were deemed to be at high risk of recurrent VTE on the
all four studies (Table 3). On the basis of the results of basis of the site investigator’s evaluation. Otherwise,
the RE‑MODEL,13 RE‑MOBILIZE,14 RE‑NOVATE,15 the f­ollow-up study was designed to replicate the initial
and RE‑NOVATE II16 trials, dabigatran seems to be a study with a larger patient cohort.12
safe and effective agent for VTE prophylaxis in patients Two parallel trials, RE‑MEDY19 and RE‑SONATE,19
undergoing orthopaedic surgery, with a similar safety were motivated by the uncertainty over the optimal
profile to traditional therapy with enoxaparin. d­uration of anticoagulation treatment in patients with
idiopathic or unprovoked VTE. In these two studies,
VTE treatment the efficacy and safety of dabigatran were examined
Dabigatran has also been studied for the treat- in the secondary prevention of VTE (‘extended’ VTE
ment of VTE. The RE‑COVER trial17 was a phase III, therapy). Patients enrolled in these two studies of
­d ouble-blind, randomized, noninferiority study, in extended VTE therapy were deemed to have at least a
which 2,539 patients with acute pulmonary embolism moderate risk of VTE recurrence. In the RE‑MEDY
(PE) or proximal deep vein thrombosis (DVT) were study,19 almost 3,000 patients were randomly assigned to
enrolled. In this study, long-term (6 months) therapy receive either dabigatran (150 mg twice daily) or warfarin
with war­farin was compared with dabigatran (150 mg for an additional period of 6–36 months after an initial
twice daily), after initial therapy with LMWH for a 3‑month treatment course with anticoagulation (Table 4).
median of 9 days (Table 4). The primary outcome was The primary outcome measures were fatal and non­fatal
the incidence of recurrent VTE or VTE-related deaths PE and DVT, and VTE-related deaths. The hazard ratio
at 6 months (2.4% in the dabigatran group, 2.1% in the for the composite primary end point for dabigatran
warfarin group; HR 1.10, 95% CI 0.65–1.84, P <0.001 compared with warfarin was 1.44 (95% CI 0.78–2.64,
for non­inferiority).17 Although the incidence of major P = 0.01 for noninferiority) over the 6–36 month follow-
bleeding was similar in both groups, the overall rate up period.19 Of note, significantly fewer major or clini-
of bleeding was higher in the warfarin cohort than in cally relevant bleeding events occurred in the dabigatran
the dabigatran cohort (21.9% and 16.1%, respectively; arm than in the warfarin arm (HR 0.54, 95% CI 0.41–
HR 0.71, 95% CI 0.59–0.85, P <0.001).17 The similarly 0.71, P <0.001).19 In the RE‑SONATE study,19 a similarly

NATURE REVIEWS | CARDIOLOGY VOLUME 10  |  JULY 2013  |  399

© 2013 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 2 | Approved clinical use, contraindications, and warnings for novel oral anticoagulants
Drug Approved indications Approved indications Contraindications Warnings/precautions
in Canada and Europe in the USA
Apixaban Nonvalvular AF Nonvalvular AF Bleeding Concomitant use of P‑glycoprotein inducers
VTE prophylaxis after Hypersensitivity or inhibitors
orthopaedic surgery Concomitant use of strong CYP3A4 inhibitors
Renal insufficiency
Hepatic impairment
Prosthetic heart valves
Pregnancy (FDA category B*)
Dabigatran Nonvalvular AF Nonvalvular AF Bleeding Concomitant use of P‑glycoprotein inducers
VTE prophylaxis after Hypersensitivity or inhibitors
orthopaedic surgery Renal insufficiency (creatinine clearance
15–30 ml/min: use 75 mg twice daily,
avoid if creatinine clearance <15 ml/min
or in end-stage renal disease)
Prosthetic heart valves
Pregnancy (FDA category C‡)
Rivaroxaban Nonvalvular AF Nonvalvular AF Bleeding Concomitant use of P‑glycoprotein inducers
VTE prophylaxis after VTE prophylaxis after Hypersensitivity or inhibitors
orthopaedic surgery orthopaedic surgery Concomitant use of strong CYP3A4 inhibitors
Acute VTE treatment Acute VTE treatment Renal insufficiency
Extended VTE therapy Extended VTE therapy Hepatic impairment
Prosthetic heart valves
Pregnancy (FDA category C‡)
*FDA Pregnancy category B: either animal reproduction studies have not shown a risk to the foetus and no adequate and well-controlled studies in pregnant
women exist; or animal studies have shown an adverse effect, but adequate and well-controlled studies in pregnant women have not shown a risk to the foetus
in any trimester. ‡FDA Pregnancy category C: animal reproduction studies have shown an adverse effect on the foetus and no adequate and well-controlled
studies in humans exist, but potential benefits might warrant the use of the drug in pregnant women despite potential risks. Abbreviations: AF, atrial fibrillation;
CYP3A4, cytochrome P450 3A4; VTE, venous thromboembolism.

designed trial comparing dabigatran to placebo in Rivaroxaban

extended VTE therapy, the hazard ratio for the primary Rivaroxaban, the first commercially available oral
end point was 0.08 (95% CI 0.02–0.25, P <0.001), and the factor Xa inhibitor, has a half-life of 5–9 h (Table 1).
hazard ratio for major or clinically relevant bleeding was Approximately one-third of the drug is cleared by the
2.92 (95% CI 1.52–5.60, P = 0.001). kidneys and the remainder is eliminated by the hepato-
Importantly, in the RE‑MEDY trial, 19 a significant biliary route. Care should be taken when the drug is used
increase in the incidence of ACS events was noted with in patients with hepatic or renal impairment; the dose
dabigatran compared with warfarin (0.9% versus 0.2%; should be reduced if creatinine clearance is 30–50 ml/min,
P <0.02). A similarly significant increase in the rate of and rivaroxaban should be avoided if the creatinine clear-
ACS events was previously noted in the RE‑LY trial,20 ance is <30 ml/min.22 Similarly to dabigatran, riva­roxaban
in which the role of dabigatran in nonvalvular atrial has no documented clinically important drug–drug
fibrillation was evaluated. The increase in ACS events interactions with commonly used medications, such as
observed in the RE‑MEDY and RE‑LY trials might a­torvastatin, clopidogrel, or naproxen. However, rivar-
have been driven by the increased prevalence of comor­ oxaban is partly metabolized via cytochrome P450 3A4,
bidities, such as coronary artery disease, hypertension, and interactions have been observed with strong inducers
and type 2 diabetes mellitus, in the dabigatran arm. This (such as ketoconazole) or inhibitors (such as rifampicin)
difference in the rate of ACS events was also described of the cytochrome P450 system.22,23
in a meta-analysis, in which an increased risk of ACS, The recommended dose of rivaroxaban for the
but decreased mortality, was reported with dabigatran prophylaxis of DVT in patients undergoing hip or knee
compared with warfarin.21 However, the results from the replacement surgery is 10 mg daily, with the initial dose
meta-analysis were driven mostly by data from the RE‑LY given 6–10 h after surgery (once haemostasis has been
trial.20 Notably, despite the significant difference in the achieved). The recommended treatment duration is
rate of ACS events, the overall number of patients with 12 days for patients undergoing knee replacement surgery,
ACS remains low in all the dabigatran trials. Long-term and 25 days for those undergoing hip replacement
follow-up and further multivariate analyses are needed to surgery (Table 2). The FDA announced in November 2012
investigate the relationship between dabigatran and the that rivaroxaban had been approved for the treatment of
rate of ACS events. patients with acute DVT or PE at a dose of 15 mg twice
In summary, dabigatran has been found to be equally daily for 3 weeks, followed by 20 mg daily for the remain-
as effective as warfarin for the secondary prevention ing treatment period. The drug has also been approved to
of VTE. Dabigatran seems to have a similar, if not reduce the risk of recurrent DVT and PE at a daily dose of
s­uperior, safety profile for the treatment of patients with 20 mg, after an initial 6‑month treatment course for acute
established DVT or PE, as measured by the incidence VTE. In 2008, rivaroxaban was approved for use in the
of recurren­t VTE and VTE-related deaths. prevention and treatment of VTE in Europe and Canada.

400  |  JULY 2013  |  VOLUME 10  www.nature.com/nrcardio

© 2013 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Table 3 | Phase III clinical trials of dabigatran for VTE prophylaxis

Trial Patient population Dose and Enoxaparin dose DVT, PE, or death Major bleeding
and cohort size duration of and duration (dabigatran vs (dabigatran vs
dabigatran enoxaparin) enoxaparin)
RE‑MODEL13* Total knee replacement 150 mg or 40 mg daily 40.5% (150 mg), 36.4% 1.3% (150 mg), 1.5%
n = 2,101 220 mg daily 6–10 days (220 mg) vs 37.7% (220 mg) vs 1.3%
6–10 days P = 0.82 (150 mg), P = 1.00 (150 mg),
P = 0.38 (220 mg) P = 0.82 (220 mg)
Dabigatran did not
achieve noninferiority
RE‑MOBILIZE14* Total knee replacement 150 mg or 30 mg twice daily 33.7% (150 mg), 31.1% 0.6% (150 mg), 0.6%
n = 2,615 220 mg daily 12–15 days (220 mg) vs 25.3% (220 mg) vs 1.4%
12–15 days P = 0.0009 (150 mg),
P = 0.0234 (220 mg)
RE‑NOVATE15* Total hip replacement 150 mg or 40 mg daily 8.6% (150 mg), 6.0% 1.3% (150 mg), 2.0%
n = 3,493 220 mg daily 28‑­35 days (220 mg) vs 6.7% (220 mg) vs 1.6%
28–35 days P <0.0001 (150 mg), P = 0.60 (150 mg),
P <0.0001 (220 mg) P = 0.44 (220 mg)
RE‑NOVATE II16* Total hip replacement 220 mg daily 40 mg daily 7.7% vs 8.8% 1.4% vs 0.9%
n = 2,055 28–35 days 28‑­35 days P <0.0001 P = 0.40
*Randomized, double-blind, noninferiority trial. Abbreviations: DVT, deep vein thrombosis; PE, pulmonary embolism; VTE, venous thromboembolism.

VTE prophylaxis VTE therapy

On the basis of pooled results from the four RECORD Rivaroxaban has been compared with either a placebo
trials,24–27 rivaroxaban (10 mg once daily) compared favour- or a LMWH and a VKA for treatment of patients with
ably with enoxaparin regimens administered over various established VTE in two trials. In EINSTEIN‑DVT,31 a ran-
intervals after hip or knee surgery, with a relative risk of domized, noninferiority trial, rivaroxaban (15 mg twice
VTE of 0.38 when compared with subcutaneous enoxa- daily for 3 weeks followed by 20 mg once daily for 3, 6,
parin 40 mg daily (P <0.0001) and 0.77 when compared or 12 months) was compared with LMWH followed by
with subcutaneous enoxaparin 30 mg twice daily (P = 0.05; VKA in 3,449 patients who had DVT without sympto­
Table 5). Rivaroxaban therapy was continued for an average matic PE  (Table  4). 31 The incidence of recurrent,
of 12 days after knee surgery and 35 days after hip surgery. s­ymptomatic VTE was 3.0% in the conventional-therap­y
Consistently with safety profiles for dabigatran, the rate arm compared with 2.1% in the rivaroxaban arm (HR 0.68,
of major bleeding did not differ significantly between 95% CI 0.44–1.04, P <0.0001 for noninferiorit­y). The rate
r­ivaroxaban and enoxaparin in the RECORD trials.28 of major and nonmajor clinically relevant bleeding was
In the MAGELLAN trial,29 extended oral rivaroxaban similar in both groups.31 EINSTEIN‑PE32 was of a similar
therapy (10 mg daily for 35 ± 4 days) was compared with study design to EINSTEIN‑DVT, but in EINSTEIN‑PE,
short-term prophylaxis with enoxaparin (40 mg sub­ 4,832 patients with symptomatic, acute PE, with or
cutaneously daily for 10 ± 4 days) followed by placebo in without DVT were enrolled. Rivaroxaban was deemed
medically ill patients aged ≥40 years with a life expectancy to be noninferior to the standard therapy of LMWH
≥6 months, reduced mobility for ≤3 days before enrol- and VKA (noninferiority margin 2.0, P = 0.003) for the
ment, and who were likely to have reduced mobility for primary efficacy outcome of symptomatic recurrent VTE,
≥3 days after enrolment.30 Enrolled patients suffered from with 50 events (2.1%) recorded in the rivaroxaban arm and
a wide variety of medical illnesses, including infectious 44 events (1.8%) in the standard-therapy arm (HR 1.12,
diseases, heart failure, respiratory disease, and various 95% CI 0.75–1.68, P = 0.003). Of note, fewer major bleed-
malignancies. Rivaroxaban was found to be noninferior to ing events occurred in the rivaroxaban group than in
enoxaparin in preventing the primary end point—a com- the standard-therapy group (1.1% versus 2.2%; HR 0.49,
posite of asymptomatic proximal or symptomatic VTE— 95% CI 0.31–0.79; P = 0.003), suggesting that a fixed-dose
after both 10 days and 35 days of follow-up.29 After10 days, regimen of rivaroxaban might have a better safety profile
the relative risk with rivaroxaban treatment was 0.97 than standard therapy.32,33
(95% CI 0.71–1.31, P = 0.003) when compared with enoxa- In the EINSTEIN-Extension trial,31 rivaroxaban 20 mg
parin plus placebo. The relative risk was 0.77 (95% CI daily for 6–12 months was compared with placebo for
0.62–0.96, P = 0.02 for noninferiority) after 35 days. In secondary prevention of VTE in 1,197 patients who had
this study, however, rivaroxaban was associated with a sig- completed a prescribed 6–12 month course of therapy.
nificantly increased risk of bleeding at both day 10 (2.8% Rivaroxaban was found to be more effective than placebo,
with rivaroxaban versus 1.2% with enoxaparin; P <0.001) with 1.3% of patients in the rivaroxaban group and 7.1%
and day 35 (4.1% and 1.7%, respectively; P <0.001).29 The of patients in the placebo group developing symptomatic,
four RECORD trials24–27 and the MAGELLAN trial29 have recurrent VTE (82% relative risk reduction; P = 0.001),
shown rivaroxaban to be at least as effective as enoxaparin with a major bleeding incidence of 0.7% in the rivaroxa-
in VTE prophylaxis in patients undergoing orthopaedic ban group and 0.0% in the placebo group (Table 4).31,34 On
surgery, with a similar, if not superior, safety profile. the basis of these trial data, rivaroxaban seems to be a safe

NATURE REVIEWS | CARDIOLOGY VOLUME 10  |  JULY 2013  |  401

© 2013 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 4 | Phase III clinical trials of novel OACs for VTE therapy

Trial Study design Patient population and Recurrent VTE Major bleeding
cohort size (OAC vs control) (OAC vs control)
Apixaban
AMPLIFY-EXT40 Randomized, double-blind, Patients with VTE who 1.7% (2.5 mg) and 0.2% (2.5 mg) and 0.1%
superiority, intention to treat completed 6–12 months 1.7% (5.0 mg) vs 8.8% (5.0 mg) vs 0.5%
analysis* of anticoagulation therapy P <0.001 (superiority) RR (2.5 mg vs control)
n = 2,482 0.49, 95% CI 0.09–2.64
RR (5.0 mg vs control)
0.25, 95% CI 0.03–2.24
Dabigatran
RE-COVER17 Randomized, double-blind, Acute VTE treated with 2.4% vs 2.1% 1.6% vs 1.9%
noninferiority‡ parenteral anticoagulation P <0.001 HR with dabigatran,
(median 9 days) (noninferiority) 0.82; 95% CI 0.45–1.48)
n = 2,539
RE-MEDY19 Randomized, double-blind, VTE, already completed 1.8% vs 1.3% 0.9% vs 1.8%
active-controlled, 3 months of therapy P = 0.03 P = 0.06
noninferiority‡ n = 2,856
RE-SONATE19 Randomized, double-blind, VTE, already completed 0.4% vs 5.6% 0.3% vs 0.0%
placebo-controlled, 3 months of therapy
noninferiority‡ n = 1,343
Rivaroxaban
EINSTEIN‑DVT31 Randomized, open-label Acute, symptomatic DVT 2.1% vs 3.0% 8.0% vs 8.0%
noninferiority§ n = 3,449 P <0.001
(noninferiority)
EINSTEIN‑PE32 Randomized, double-blind Acute, symptomatic PE 2.1% vs 1.8% 10.3% vs 11.4%
noninferiority§ with or without DVT P = 0.003 P = 0.23
n = 4,832 (noninferiority)
EINSTEIN‑EXT31,34 Randomized, double-blind Patients with VTE who 1.3% vs 7.1% 0.7% vs 0.0%
superiority§ completed 6–12 months P <0.001 (superiority) P = 0.11
of anticoagulation therapy
n = 1,196
*Apixaban (2.5 mg or 5.0 mg twice daily) vs placebo. ‡Dabigatran (150 mg twice daily) vs warfarin. §Rivaroxaban alone (15 mg twice daily for 3 weeks, followed by
20 mg daily) vs enoxaparin (1 mg/kg), followed by a vitamin K antagonist for 3, 6, or 12 months. Abbreviations: DVT, deep vein thrombosis; HR, hazard ratio;
OAC, oral anticoagulant; PE, pulmonary embolism; RR, relative risk; VTE, venous thromboembolism.

and effective option for the secondary prevention of VTE, VTE prophylaxis
with a similar adverse event rate to standard therapy. In the ADVANCE trials,35–37 the efficacy and safety of
apixaban was evaluated for the prevention of VTE in
Apixaban approximately 6,000 patients undergoing knee replace-
Apixaban, a potent, oral, direct factor Xa inhibitor, ment surgery. The primary end point was a composite
is eliminated primarily via the faecal route (approxi- of nonfatal PE, symptomatic and asymptomatic DVT,
mately 50%) with smaller portions metabolized via the and all-cause mortality (Table 6). In the ADVANCE‑1
cytochrome P450 3A4-dependent pathway in the liver trial,35 apixaban 2.5 mg twice daily did not meet the pre-
(approximately 25%) and via the urine (approximately specified criteria for noninferiority when compared with
25%). Apixaban might, therefore, be the preferred agent enoxaparin 30 mg twice daily, with event rates of 9.0%
for patients with renal insufficiency. This possibility and 8.9%, respectively (relative risk 1.02, 95% CI 0.78–
remains unsubstantiated by evidence as there have been 1.32). However, the noninferiority criteria were met in
no direct comparative studies to date. The peak plasma the ADVANCE‑2 trial,36 in which apixaban 2.5 mg twice
level is achieved 1–3 h after oral intake, and apixaban has daily (started the day after surgery) compared favour-
a h­alf‑life of 8–15 h (Table 1).34 ably with enoxaparin 40 mg once daily in a randomized,
Apixaban was approved in Canada and Europe in 2011 double-blind, phase III study involving 3,057 patients. In
for the primary prevention of VTE in patients under­ both studies, the rate of major bleeding was similar in the
going knee or hip replacement surgery (Table 2). The two treatment arms. The ADVANCE‑3 trial37 involved
recommended dosage is 2.5 mg twice daily, to be initiated 3,866 patients undergoing hip surgery treated with
12–24 h after orthopaedic surgery. The recommended apixaban or enoxaparin (40 mg daily). The primary end
treatment duration is 10–14 days for patients who have point occurred in 1.4% of patients in the apixaban group
had knee replacement surgery, and 32–38 days for compared with 3.9% in the enoxaparin group (relative
patients who have had hip replacement surgery. The drug risk 0.36, 95% CI 0.22–0.54, P <0.001 for s­uperiority;
is not approved for the prophylaxis or treatment of VTE absolute risk reduction 2.5%, 95% CI 1.5–3.5).37 Again,
in the USA (Table 2). The FDA is expected to review the safety data were similar in each group.37 Although apixa-
use of apixaban for this indication in early 2013. ban has not been approved by the FDA, and despite

402  |  JULY 2013  |  VOLUME 10  www.nature.com/nrcardio

© 2013 Macmillan Publishers Limited. All rights reserved
 REVIEWS

Table 5 | Phase III clinical trials of rivaroxaban for VTE prophylaxis

Trial Patient population Dose and Enoxaparin DVT, PE, or death Major bleeding
and cohort size duration of dose and (rivaroxaban vs (rivaroxaban vs
rivaroxaban duration enoxaparin) enoxaparin)
RECORD 124* Total hip replacement 10 mg daily 40 mg daily 1.1% vs 3.7% 0.3% vs 0.1%
n = 4,501 35 days 35 days P <0.001 P = 0.18
RECORD 225* Total hip replacement 10 mg daily 40 mg daily 2.0% vs 9.3% <0.1% vs <0.1%
n = 2,509 31–39 days 14 days P <0.001
RECORD 326* Total knee replacement 10 mg daily 40 mg daily 9.6% vs 18.9% 0.6% vs 0.5%
n = 2,615 10‑­14 days 10–14 days P <0.001 P = 0.77
RECORD 427* Total knee replacement 10 mg daily 30 mg twice daily 6.9% vs 10.1% 0.7% vs 0.3%
n = 2,055 10‑­14 days 10–14 days P = 0.0118 P = 0.11
MAGELLAN29‡ Acute medical illness 10 mg daily 40 mg daily 2.7% vs 2.7% (day 10) 2.8% vs 1.2% (day 10)
n = 8,101 35 ± 4 days 10 ± 4 days 4.4% vs 5.7% (day 35) 4.1% vs 1.7% (day 35)
(Oral placebo: P = 0.003 (noninferiority), P = 0.001 for both
35 ± 4 days) P = 0.02 (superiority) time points
*Randomized, double-blind, noninferiority trial. ‡Randomized, double-blind noninferiority (day 10) and superiority (day 35) trial. Abbreviations: DVT, deep vein
thrombosis; PE, pulmonary embolism; VTE, venous thromboembolism.

failure to meet noninferiority criteria in the ADVANCE‑1 patients who had completed treatment for DVT or PE.
trial, 35 apixaban seems to be at least as effective (in In this trial, patients were randomly assigned to receive
the ADVANCE‑2 trial36) and potentially superior (in the 2.5 mg apixaban twice daily (n = 840), 5 mg apixaban
ADVANCE‑3 trial37) to enoxaparin in preventing VTE (n = 813), or placebo (n = 829) for 12 months (Table 4).40
in patients undergoing orthopaedic surgery. All 2,486 patients enrolled in the AMPLIFY‑EXT trial
In the ADOPT trial, 38 the safety and efficacy of had pre­viously finished 6–12 months of anticoagulation
extended prophylaxis with apixaban (2.5 mg twice daily therapy. The primary efficacy outcome was the composite
for 30 days) was compared with short-term prophy- of symptomatic recurrent VTE or death from any cause.
laxis with enoxaparin (40 mg once daily for 6–14 days) The primary safety outcome was major bleeding, and the
in an initial cohort of 6,528 acutely ill medical patients secondary safety outcome was the composite of signifi-
(Table 6). The most-common medical indications for cant, nonmajor bleeding. The AMPLIFY‑EXT data were
admission were heart failure, acute respiratory failure, analysed on the basis of intention-to-treat. The investi­
infection, and acute rheumatic disorders. No significant gators found that each dose of apixaban reduced the risk
difference in the incidence of VTE events was observed of recurrent VTE compared with placebo, without an
between the two groups. However, at 30 days, a signifi- increased risk of bleeding. Approximately 8.8% of the
cant increase in the risk of major bleeding occurred in placebo group had a symptomatic recurrent VTE or died
the apixaban group compared with the enoxaparin group of a VTE-related cause (suggesting that these patients
(0.47% versus 0.19%, respectively; relative risk 2.58, had a high baseline risk of recurrent VTE), compared
95% CI 1.02–7.24, P = 0.04). Notably, the bleeding rates with 1.7% of patients taking 2.5 mg of apixaban and
between the cohorts seemed to diverge primarily when 1.7% of patients taking 5 mg of apixaban.40 The rate of
patients in the enoxaparin arm were taking placebo (the major bleeding was unexpectedly low in the treatment
short-term prophylaxis with enoxaparin was only 6–14 groups, but this difference might be attributable to chance
days) while the patients in the apixaban arm were still because of the small number of overall events. The rate
taking the study drug.38 On the basis of the ADOPT trial of nonmajor bleeding was similar in all three arms. In an
results, not enough evidence exists to support the use of appropriate population with high risk of VTE recurrence,
apixaban for extended VTE prophylaxis (after hospital apixaban can, therefore, be a safe and effective treat-
discharge) in patients admitted with an acute medical ment option over a long-term period (1 year). However,
illness, particularly because of the increased bleeding rate r­egulatory approval for this indication is pending.
in the apixaban cohort. Further studies are needed to
determine the potential utility of extended prophylaxis Edoxaban
with apixaban. Edoxaban is a reversible direct factor Xa inhibitor with
good oral bioavailability (about 50%), and a half-life
VTE treatment of approximately 8–10 h. The peak plasma concen­
Two trials have focused specifically on the use of tration is reached 1–2 h after oral intake. Approximately
a­pixaban for the treatment of VTE. The AMPLIFY one-third of edoxaban is renally cleared. Accordingly,
trial39 is an ongoing safety and efficacy study, in which patients with renal insufficiency (creatinine clearance
a­pixaban (plus placebo) is being compared with enoxa- 30–50 ml/min) should take a reduced dose. Similarly
parin and warfarin (plus placebo) in patients with symp- to the other factor Xa inhibitors, edoxaban does not
tomatic PE or DVT. In the double-blind companion require regular monitoring and no antidote is readily
trial, AMPLIFY‑EXT (Extension),40 apixaban was eval- available. Edoxaban is a substrate for P‑glycoprotein and,
uated for the prevention of death or recurrent VTE in as a result, the concurrent use of strong P‑glycoprotein

NATURE REVIEWS | CARDIOLOGY VOLUME 10  |  JULY 2013  |  403

© 2013 Macmillan Publishers Limited. All rights reserved
REVIEWS

Table 6 | Phase III clinical trials of apixaban for VTE prophylaxis

Trial Patient population Dose and Enoxaparin DVT, PE, or death Major bleeding
and cohort size duration dose and (apixaban vs (apixaban vs
of apixaban duration enoxaparin) enoxaparin)
ADVANCE‑135* Total knee replacement 2.5 mg twice daily 30 mg twice daily 9.0% vs 8.8% 0.7% vs 1.4%
n = 3,195 10–14 days 12 days P = 0.06 P = 0.05
Apixaban did not achieve
noninferiority
ADVANCE‑236* Total knee replacement 2.5 mg twice daily 40 mg daily 15.1% vs 24.4% 0.6% vs 0.9%
n = 3,057 10–14 days 10–14 days P <0.0001 (superiority) P = 0.30
ADVANCE‑337‡ Total hip replacement 2.5 mg twice daily 40 mg daily 1.4% vs 3.9% 0.8% vs 0.7%
n = 5,407 35 days 35 days P <0.001 (superiority) P = 0.54
ADOPT38‡ Medical illness 2.5 mg twice daily 40 mg twice daily 2.71% vs 3.06% 0.46% vs 0.19%
n = 4,495 30 days 6–14 days P = 0.44 P = 0.04
Apixaban did not achieve
superiority
*Randomized, double-blind, noninferiority trial. ‡Randomized, double-blind, superiority trial. Abbreviations: DVT, deep vein thrombosis; PE, pulmonary embolism;
VTE, venous thromboembolism.

inhibitors, such as amiodarone, dronederone, quinidine, noted across all doses of edoxaban (29.5%, 26.1%, 12.5%,
and verapamil, could increase the half-life and serum and 9.1% in the groups treated with 5, 15, 30, or 60 mg,
drug level of edoxaban. respectively, compared with 48.3% in the placebo group).
Edoxaban is approved in Japan for VTE prevention Of note, the incidence of major and clinically-relevant
after lower limb orthopaedic surgery. This agent has not nonmajor bleeding was not significantly different across
received approval for any indication outside of Japan, all doses of edoxaban, or between edoxaban and placebo.
although several international trials are currently under-
way, such as ENGAGE AF TIMI 48,41 in which the use VTE treatment
of edoxaban for stroke prevention in patients with atrial Edoxaban is also currently under investigation for the
fibrillation will be evaluated, and HOKUSAI VTE42 in prevention of recurrent VTE in patients with previously
which the use of edoxaban in the treatment of VTE or diagnosed DVT, PE, or both, in the large-scale, multi-
PE will be evaluate. Nevertheless, partly as a result of the national, double-blind, randomized HOKUSAI VTE
low prophylactic dose of enoxaparin (2,000 inter­national phase III noninferiority trial. 42 In this trial, several
units or 20 mg) used in the Japanese trials in which anticoagulant protocols are being investigated. Patients
edoxaban was compared with standard therapy, the data will initially be treated with heparin (either unfrac-
outlined below do not seem to warrant w­idespread use tionated heparin or LMWH), and will then trans­ition
of this agent.43 to either warfarin or edoxaban (60 mg once daily or
30 mg in patients with body mass <60 kg, creatinine
VTE prophylaxis clearance 30–50 ml/min, or concomitant use of potent
In an analysis of data pooled from two phase III trials, P‑glycoprotein inhibitors) for varying treatment periods
STARS E‑344 and STARS J‑V,44 in which a total of 1,326 of 3, 6, or 12 months. The treatment duration is at
Japanese patients undergoing orthopaedic surgery were the discretion of the clinician, with consideration of the
evaluated, edoxaban 30 mg once daily significantly patients’ preferences and their risk of recurrent VTE
reduced the risk of VTE compared with enoxaparin and bleeding. The variable length of treatment is meant
(5.1% and 10.7%, respectively; P <0.001). No signifi- to provide flexibility to clinicians and simulate actual
cant difference in bleeding risk between the groups was clinical practice, as duration of anticoagulant therapy
reported.45 Edoxaban was also evaluated in a random­ in patients with unprovoked VTE remains unclear, but
ized, double-blind, dose-response study in the USA. treatment beyond 3 months is likely to provide a benefit.
Edoxaban (15 mg, 30 mg, 60 mg, or 90 mg once daily) was The primary efficacy outcome is symptomatic, recur-
compared with dalteparin (5,000 units subcutaneously rent VTE during the 12‑month study. The enrolment
daily) for 7–10 days after hip surgery. The incidence of phase of this study, involving >8,250 patients, was com-
VTE in each edoxaban dose cohort was significantly less pleted in late 2012, and the study should be the largest
than in the dalteparin group (28.2%, 21.2%, 15.2%, and clinical trial to date for the treatment and secondary
10.6% in the edoxaban groups, respectively, and 43.8% in preventio­n of VTE.
the dalteparin group; P <0.005).46 In another trial, various
doses of edoxaban (5, 15, 30, and 60 mg daily) were eval- Combination with antiplatelet agents
uated for the prevention of VTE in patients undergoing Patients with atherosclerosis and associated risk factors
total knee arthroplasty.47 A total of 523 Japanese patients for coronary artery disease, such as type 2 diabetes, have
were randomly assigned to receive edoxaban or placebo an increased risk of VTE.48–52 In the appropriate clinical
once daily for 11–14 days after total knee arthroplasty. context, using antiplatelet agents together with VKAs
A dose-dependent decrease in the incidence of VTE was might, therefore, become necessary. For example, the

404  |  JULY 2013  |  VOLUME 10  www.nature.com/nrcardio

© 2013 Macmillan Publishers Limited. All rights reserved
 REVIEWS

2013 ACC Foundation/AHA guidelines for the manage- One of the major, frequently cited potential draw-
ment of ST‑segment elevation myocardial infarction53 backs of the novel oral anticoagulant agents is the lack
outline class 1A recommendations for the combined use of reversibility of their anticoagulant effect. Bleeding
of warfarin and aspirin, clopidogrel, or both for patients is of particular concern for the large percentage of the
with an indication for anticoagulant therapy, such as the patients receiving these agents who tend to be >65
presence of paroxysmal or chronic atrial fibrillation or years of age. Media reports and litigation have raised
atrial flutter, or left ventricular thrombus. This recom- questions as to the safety of these agents as a result of
mendation, however, is specific to patients with ACS and anecdotal cases of irreversible bleeding, with one report
does not apply to patients with stable coronary artery citing that 542 deaths attributable to dabigatran (the
disease or VTE. Furthermore, current guidelines indi- most-commonl­y prescribed novel anticoagulant) were
cate that the duration of triple antithrombotic therapy reported to the FDA in 2011 alone.59
with aspirin, a P2Y12 receptor inhibitor, such as clopi- Outside of clinical trials, the lack of options when
dogrel or ticagrelor, and a VKA must be limited because treating a patient who is bleeding is a concern for
of the increased risk of associated bleeding.53 Additional patients and physicians. Proponents of the novel oral
data for the optimal duration of therapy, as well as target anti­c oagulant agents argue that the shorter half-life
values of the international normalized ratio, are needed and predictable pharmacokinetics of these drugs
to clarify therapeutic regimens for these patients. The use renders the need for reversal less likely than for VKAs.8
of the novel oral anticoagulants has not been evaluated in Nevertheless, the lack of readily available reversal agents
the context of concurrent ACS and VTE, and no specific has resulted in some physician trepidation to use novel
recommendations regarding their use in such a clinical oral a­nticoagulants, particularly for elderly patients.
scenario have been published. Prothrombin complex concentrate (PCC) is a poten-
The extent of atherothrombotic protection afforded tial, but unproven, means of reversing the anticoagulant
by the novel oral anticoagulants needs further investi­ effects of apixaban, dabigatran, and rivaroxaban.60–62
gation. First, in some clinical trials, an increased inci- PCCs are concentrated, pooled plasma products con-
dence of ACS among patients taking direct thrombin taining inactivated clotting factors and usually a small
inhibitors, such as dabigatran (for example, the RE‑LY20 amount of protein C and protein S. In April 2013,
and RE‑MEDY19 trials) or ximelagatran (for example, the the FDA approved the first nonactivated, four-factor
THRIVE,54 SPORTIF III,55 and SPORTIF V56 trials) alone PCC (which includes Factors II, VII, IX, and X) to
has been observed, but not in those taking anti-factor Xa treat patients who are taking warfarin or other VKAs.
medications. This finding could suggest that the range However, there are minimal data from clinical trials
of atheroembolic protection using these agents does not supporting the use of PCCs to reverse bleeding asso-
approach the degree provided by traditional antiplatelet ciated with the novel oral anti­coagulants. In a small,
agents, such as aspirin, clopidogrel, or warfarin, although randomized, double-blind, placebo-controlled study,
this hypothesis is debated.21,57 Second, the risk:benefit 12 healthy male volunteers received rivaroxaban 20 mg
ratio for the combined use of novel oral anticoagulants twice daily (n = 6) or dabigatran 150 mg twice daily
and antiplatelet medications remains to be determined. (n = 6) for 2.5 days, followed by either a bolus of saline
In a meta-analysis, the use of novel anticoagulants in or a bolus of PCC of 50 international units per kg. The
patients receiving antiplatelet therapy after ACS did not PCC successfully reversed the laboratory effect of rivar-
improve overall clinical outcome, but increased the risk oxaban (reduced partial thromboplastin time), but not
of major bleeding.58 Antiplatelet agents should, there- the effect of dabigatran.60 In an animal study, an acti-
fore, continue to be used as directed in current guide- vated PCC did not correct the prolonged activated partial
lines, along with oral anticoagulation if appropriate, in thromboplastin time caused by dabigatran treatment, but
patients with concomitant risk of atheroembolic events substantially reduced bleeding time.63 Ex vivo analysis
(for example, patients with coronary or peripheral artery has provided evidence for the potential utility of PCCs in
disease) and VTE. reversing nearly all aspects of the laboratory effects of the
novel oral anticoagulants,62 but their potential use in clin-
Practical concerns ical settings remains understudied. Furthermore, PCCs
Bleeding and lack of antidotes might pose a high risk for thrombosis, and co­nsultation
In the large, randomized safety and efficacy trials, the with a haematologist is recommended before use.64
major bleeding rate associated with novel anticoagulants Recombinant factor VIIa could also potentially be used
has been similar to, or even lower than, those associ- to reverse the effects of apixaban.60–62 In animal models of
ated with LMWH or VKAs (Tables 3, 4, 5, and 6). For rivaroxaban overdose, recombinant factor VIIa improved
patients undergoing surgery who have normal renal the coagulopathic profile (including bleeding time, clot-
function and a nonelevated risk of bleeding, dabigatran ting time, and thrombin generation), but was ineffective
(half-life 12–17 h) or rivaroxaban (half-life 5–9 h) should at reversing rivaroxaban-associated bleeding.65 Minimal
be discontinued at least 24 h before surgery. An addi- data on the use of recombinant factor VIIa in humans
tional 1–3 days could be necessary to allow for sufficient exist, and thus its use is primarily on the basis of anec­
clearance in patients of advanced age (≥75 years), with dotal evidence. Regular use is, therefore, not recom-
moderate or severe renal insufficiency, or at high risk mended and any use in the management of bleeding
of bleeding. related to the use of oral anticoagulants is off-label.65

NATURE REVIEWS | CARDIOLOGY VOLUME 10  |  JULY 2013  |  405

© 2013 Macmillan Publishers Limited. All rights reserved
REVIEWS

Plasma-derived and recombinant factor Xa are also in patients with renal dysfunction might be lower for
being investigated as potential reversal agents for the apixaban and edoxaban, because these drugs are excreted
factor Xa inhibitors.66 Urgent haemodialysis could also be by the kidney to a lesser degree than dabigatran and rivar-
considered for patients with bleeding associated with dabi­ oxaban.69 Using unfractionated heparin and VKAs for
gatran or for patients with underlying renal insuf­ficiency. patients with severe renal dysfunction is probably still
In patients with end-stage renal disease, haemodialysis the most-prudent option.
has been shown to remove 62–68% of circulating, active
dabigatran.9 Insufficient data exist to determine whether Mechanical heart valves
haemodialysis is useful in patients taking ­apixaban or Currently, VKAs are the only agents approved for
rivaroxaban, although the high percentage of the drug that thrombo­prophylaxis in patients with mechanical heart
is bound to proteins in the blood renders effective dialy- valves.70 None of the novel oral anticoagulants is likely
sis unlikely (~66% of rivaroxaban is cleared via the renal to receive approval for use in these patients. In fact, the
route compared with ~80% of dabigatran).67 Despite some FDA has listed dabigatran as an absolute contra­indication
promising data, the results of trials in which ‘antidotes’ in these instances, after a case report of blood-clot for-
to these agents have been assessed have been mixed, and mation on mechanical prosthetic heart valves. 71 The
the lack of availability of reversal agents continues to be phase II RE‑ALIGN72 trial was designed to test the use of
a c­hallenge in the use of new oral anticoagulants. dabigatran in this patient population, but was abruptly
terminated in December 2012 when excessive blood-
Lack of monitoring clot formation was again found on prosthetic valves.
No standardized monitoring tests are currently available Furthermore, patients taking dabigatran experienced
for the novel oral anticoagulants. Although these agents more bleeding after surgery than patients taking warfarin,
do affect various coagulation parameters, the correlation and had an increased propensity for t­hromboembolism
between these effects and the associated risks of bleeding leading to stroke and myocardial infarction.72
and thrombosis have not been measured and standard-
ized. Low doses of the novel oral anticoagulants are avail- Pregnancy
able to reduce the risk of bleeding in certain subgroups of VTE during pregnancy and in the peripartum period is
patients. For most patients, routine monitoring to gauge an important cause of maternal morbidity and mor­tality.
the precise anticoagulant effect of these agents is prob- VTE is estimated to occur in approximately 0.5–2.2 per
ably unnecessary, as shown by randomized efficacy and 1,000 pregnancies.73,74 Negatively charged heparin particles
safety trials. Nevertheless, some clinical scenarios, such cannot cross the placenta, so this class of anticoagulants
as the management of a patient during the preoperative is associated with a very low risk to the foetus.75,76 The
period, or with active bleeding, might require knowledge LMWHs are, therefore, currently the preferred means
of the precise degree of anticoagulation.68 of prophylaxis and therapy for VTE during pregnancy.77
The lack of monitoring that is required for these novel No substantial published data currently exist for the use
agents has other potentially deleterious effects. The need of the novel oral anticoagulants in patients who are preg-
to monitor the international normalized ratio in patients nant. These agents are expected to cross the placental
taking VKAs is also a compliance tool, enabling health- barrier and could lead to undesired foetal anticoagulation.
care providers to assess patient progress and adherence Furthermore, the protease inhib­itory capacity of these
to prescribed regimens. Furthermore, the need for daily drugs might actually hinder foetal development.78 As
dosing of the novel oral anticoagulants and their relatively a result, the novel oral anticoagulants are currently contra­
short half-lives (compared with VKAs) could be prob- indicated for use in patients who are pregnant. Heparin,
lematic for some patients—missing a few doses of these LMWH, and warfarin have not been found in high enough
drugs might result in an abrupt decline in anti­coagulant concentrations to be clinically relevant in breast milk and,
effect, whereas a reduction of VKA dose usually leads to a therefore, are likely to pose minimal risk to newborns.79
more-gradual decrease. Thus, although the lack of moni- Further studies, however, are necessary to determine the
toring required for the novel oral anticoagulants might be concentrations of the novel oral anticoagulants secreted
a benefit for most patients, the potentially negative impact by lactating women and the potential risks posed to new-
of reduced monitoring must be kept in mind. borns. As a result, the use of the novel anticoagulants
in breastfeeding mothers in not recommended.
Contraindications
The use of novel oral anticoagulants is relatively or abso- Patients with cancer
lutely contraindicated in a number of clinical scenarios Malignancy is associated with a high rate of VTE events.
(Table 2). Patients at risk of bleeding or with a hyper- Patients with cancer are estimated to be at a fourfold to
sensitivity to the drug should not take any of the novel sevenfold increased risk of VTE, and 15–20% of patients
oral anticoagulants. with cancer experience at least one VTE episode during
the course of their disease.80–82 The annual incidence of
Renal failure VTE in patients with cancer is estimated to be in the range
Agents such as dabigatran and rivaroxaban should be of 0.5–20.0%, depending on clinical parameters, such as
avoided in patients with severe renal failure (c­reatinine the type of cancer and baseline risk factors.83 VTE is the
clearance <30 ml/min).9 The risk of bleeding complications second most-common cause of death in patients with

406  |  JULY 2013  |  VOLUME 10  www.nature.com/nrcardio

© 2013 Macmillan Publishers Limited. All rights reserved
 REVIEWS

cancer.84 For hospitalized patients with cancer, current the novel oral anticoagulants, these drugs should be used
guidelines generally recommend thrombo­prophylaxis cautiously in patients with substantial renal or hepatic
with LMWH (enoxaparin 40 mg daily or dalteparin impairment, elderly patients, or those prone to bleeding,
5,000 units daily), fondaparinux (75 international units because of the relative paucity of information on the man-
per kg, subcutaneous daily), or low-dose unfractionated agement of bleeding complications that could arise during
heparin (5,000 units every 8–12 h).85 Patients under­ treatment. In addition, given the rising cost of health care,
going surgery for cancer are at particularly high risk the widespread use of such agents might or might not be
of the development of VTE. Therapy must be tailored cost-effective in the long-term.
s­pecifically to each clinical scenario and risk-factor profile. The novel oral anticoagulants are an important leap
Unfractionated heparin, LMWH, and the VKAs remain forward in the clinical approach to the prophylaxis and
the mainstays of anticoagulant treatment in patients with management of venous thromboembolic disease. The
cancer, with data from the CLOT trial86 favouring the use success and comparative safety of the currently available
of LMWHs (specifically dalteparin) over warfarin in pre- agents, such as apixaban, dabigatran, and rivaroxaban, are
venting recurrent VTE.87 This finding is reflected in the likely to spur further research on these and other poten-
current guidelines.88 The benefits of the novel oral anti- tial new drugs. With the completion of the HOKUSAI
coagulants in patients without cancer are also likely to be trial,42 further information is expected on the efficacy and
applicable to patients with cancer,89 although this remains safety of edoxaban in the prevention of VTE recurrence.
to be tested in large-scale, randomized trials. The utility of other factor Xa inhibitors, such as betrixa-
ban, the subject of the EXPERT trial,90 will continue to be
Conclusions defined in ongoing studies. Otamixaban, another factor Xa
During the past decade, new oral anticoagulants for the inhibitor, is currently in late-stage clinical development for
prevention and treatment of VTE have become available. the management of ACS, and might also eventually be used
Apixaban, dabigatran, edoxaban, and rivaroxaban are for VTE prophylaxis or therapy. These and other agents are
at various stages of approval for VTE indications. These at various stages of development and await phase III cl­inical
agents are easy to administer without the need for routine trials to clarify their role in the management of VTE.
monitoring, and have several other advantages, including
predictable pharmacokinetics, infrequent drug and food
interactions, and rapid onset of action. One important Review criteria
limitation is the lack of proven reversal strategies for use in
The NCBI PubMed database was searched for full-
the event of bleeding or when urgent surgery is necessary. text, English language papers published since 1980.
Dabigatran and rivaroxaban have been approved for Search terms included “venous thromboembolism”,
VTE prophylaxis in Europe and Canada, but in the USA, “deep venous thrombosis”, “pulmonary embolism”,
only rivaroxaban is approved for prophylaxis against VTE “anticoagulation”, “novel oral anticoagulants”, “vitamin K
in patients undergoing major orthopaedic surgery of the antagonists”, “dabigatran”, “apixaban”, “rivaroxaban”,
hip or knee, or for treatment of patients with established and “edoxaban”. The search was performed most
recently in March 2013.
DVT or PE. Despite promising results in several trials of

1. Steffel, J., Luscher, T. F. & Tanner, F. C. Tissue 7. Brenner, B. & Hoffman, R. Emerging options [abstract 205]. Presented at the American
factor in cardiovascular diseases: molecular in the treatment of deep vein thrombosis and Society of Hematology 2011 Annual Meeting.
mechanisms and clinical implications. pulmonary embolism. Blood Rev. 25, 215–221 13. Eriksson, B. I. et al. Oral dabigatran etexilate
Circulation 113, 722–731 (2006). (2011). vs. subcutaneous enoxaparin for the prevention
2. Hart, R. G., Pearce, L. A. & Aguilar, M. I. Meta- 8. Blech, S., Ebner, T., Ludwig–Schwellinger, E., of venous thromboembolism after total knee
analysis: antithrombotic therapy to prevent Stangier, J. & Roth, W. The metabolism and replacement: the RE‑MODEL randomized trial.
stroke in patients who have nonvalvular atrial disposition of the oral direct thrombin inhibitor, J. Thromb. Haemost. 5, 2178–2185 (2007).
fibrillation. Ann. Intern. Med. 146, 857–867 dabigatran, in humans. Drug Metab. Dispos. 36, 14. RE-MOBILIZE Writing Committee et al. Oral
(2007). 386–399 (2008). thrombin inhibitor dabigatran etexilate vs North
3. Guyatt, G. H. et al. Executive summary: 9. Stangier, J., Rathgen, K., Stahle, H. & Mazur, D. American enoxaparin regimen for prevention
antithrombotic therapy and prevention of Influence of renal impairment on the of venous thromboembolism after knee
thrombosis, 9th ed: American College of Chest pharmacokinetics and pharmacodynamics of arthroplasty surgery. J. Arthroplasty 24, 1–9
Physicians evidence-based clinical practice oral dabigatran etexilate: an open-label, parallel- (2009).
guidelines. Chest 141 (Suppl. 2), 7S–47S group, single-centre study. Clin. Pharmacokinet. 15. Eriksson, B. I. et al. Dabigatran etexilate
(2012). 49, 259–268 (2010). versus enoxaparin for prevention of venous
4. Steffel, J. & Luscher, T. F. Vitamin K antagonists: 10. Stangier, J., Rathgen, K., Stahle, H., thromboembolism after total hip replacement:
ready to be replaced? Hamostaseologie 32, Gansser, D. & Roth, W. The pharmacokinetics, a randomised, double-blind, non-inferiority trial.
249–257 (2012). pharmacodynamics and tolerability of dabigatran Lancet 370, 949–956 (2007).
5. Giorgi, M. A., Cohen Arazi, H., Gonzalez, C. D. etexilate, a new oral direct thrombin inhibitor, in 16. Eriksson, B. I. et al. Oral dabigatran versus
& Di Girolamo, G. Changing anticoagulant healthy male subjects. Br. J. Clin. Pharmacol. 64, enoxaparin for thromboprophylaxis after
paradigms for atrial fibrillation: dabigatran, 292–303 (2007). primary total hip arthroplasty (RE-NOVATE II*):
apixaban and rivaroxaban. Expert Opin. 11. van Ryn, J. et al. Dabigatran etexilate—a novel, a randomised, double-blind, non-inferiority trial.
Pharmacother. 12, 567–577 (2011). reversible, oral direct thrombin inhibitor: Thromb. Haemost. 105, 721–729 (2011).
6. Harder, S., Parisius, J. & Picard-Willems, B. interpretation of coagulation assays and 17. Schulman, S. et al. Dabigatran versus
Monitoring direct FXa-inhibitors and fondaparinux reversal of anticoagulant activity. Thromb. warfarin in the treatment of acute venous
by prothrombinase-induced clotting time (PiCT): Haemost. 103, 1116–1127 (2010). thromboembolism. N. Engl. J. Med. 361,
relation to FXa-activity and influence of assay 12. Schulman, S. et al. A randomized trial of 2342–2352 (2009).
modifications. Thromb. Res. 123, 396–403 dabigatran versus warfarin in the treatment of 18. Prandoni, P. & Taher, A. Insights from the
(2008). acute venous thromboembolism (RE-COVER II) dabigatran versus warfarin trial in patients

NATURE REVIEWS | CARDIOLOGY VOLUME 10  |  JULY 2013  |  407

© 2013 Macmillan Publishers Limited. All rights reserved
REVIEWS

with venous thromboembolism (the ill patients. N. Engl. J. Med. 365, 2167–2177 57. Artang, R., Rome, E. & Vidaillet, H. Dabigatran
RE‑COVER trial). Expert Opin. Pharmacother. (2011). and myocardial infarction, drug or class effect:
11, 1035–1037 (2010). 39. US National Library of Medicine. ClinicalTrials.gov meta-analysis of randomized trials with oral
19. Schulman, S. et al. Extended use of [online], http://www.clinicaltrials.gov/ct2/ direct thrombin inhibitors [abstract]. J. Am. Coll.
dabigatran, warfarin, or placebo in venous show/NCT00643201 (2012). Cardiol. 59, E571 (2012).
thromboembolism. N. Engl. J. Med. 368, 40. Agnelli, G. et al. Apixaban for extended treatment 58. Komocsi, A., Vorobcsuk, A., Kehl, D. & Aradi, D.
709–718 (2013). of venous thromboembolism. N. Engl. J. Med. Use of new-generation oral anticoagulant agents
20. Dans, A. L. et al. Concomitant use of 368, 699–708 (2013). in patients receiving antiplatelet therapy after an
antiplatelet therapy with dabigatran or warfarin 41. Ruff, C. T. et al. Evaluation of the novel acute coronary syndrome: systematic review and
in the Randomized Evaluation of Long-Term factor Xa inhibitor edoxaban compared with meta-analysis of randomized controlled trials.
Anticoagulation Therapy (RE-LY) trial. warfarin in patients with atrial fibrillation: design Arch. Intern. Med. 172, 1537–1545 (2012).
Circulation 127, 634–640 (2013). and rationale for the Effective aNticoaGulation 59. Thomas, K. A promising drug with a flaw.
21. Uchino, K. & Hernandez, A. V. Dabigatran with factor xA next GEneration in Atrial The New York Times (2 November 2012).
association with higher risk of acute coronary Fibrillation-Thrombolysis In Myocardial Infarction 60. Eerenberg, E. S. et al. Reversal of rivaroxaban
events: meta-analysis of noninferiority study 48 (ENGAGE AF‑TIMI 48). Am. Heart J. 160, and dabigatran by prothrombin complex
randomized controlled trials. Arch. Intern. Med. 635–641 (2010). concentrate: a randomized, placebo-controlled,
172, 397–402 (2012). 42. Raskob, G. et al. Edoxaban for the long-term crossover study in healthy subjects. Circulation
22. Perzborn, E. et al. Rivaroxaban: a new oral treatment of venous thromboembolism: 124, 1573–1579 (2011).
factor Xa inhibitor. Arterioscler. Thromb. Vasc. rationale and design of the Hokusai-VTE study. 61. Crowther, M. A. & Warkentin, T. E. Managing
Biol. 30, 376–381 (2010). J. Thromb. Haemost. http://dx.doi.org/ bleeding in anticoagulated patients with a focus
23. Baumann Kreuziger, L. M., Morton, C. T. & 10.1111/jth.12230. on novel therapeutic agents. J. Thromb. Haemost.
Dries, D. J. New anticoagulants: a concise review. 43. Yeh, C. H., Fredenburgh, J. C. & Weitz, J. I. 7 (Suppl. 1), 107–110 (2009).
J. Trauma Acute Care Surg. 73, 983–992 (2012). Oral direct factor Xa inhibitors. Circ. Res. 62. Marlu, R. et al. Effect of non-specific reversal
24. Eriksson, B. I. et al. Rivaraoxaban versus 111, 1069–1078 (2012). agents on anticoagulant activity of dabigatran
enoxaparin for thromboprophylaxis after hip 44. Fuji, T., Fujita, S., Tachibana, S. & Kawai, Y. and rivaroxaban: a randomised crossover ex vivo
arthroplasty. N. Engl. J. Med. 388, 2765–2775 Edoxaban versus enoxaparin for the prevention study in healthy volunteers. Thromb. Haemost.
(2008). of venous thromboembolism and bleeding from 108, 217–224 (2012).
25. Kakkar, A. K. et al. Extended duration STARS E‑3 and STARS J.‑V [abstract 208]. 63. van Ryn, J. et al. Dabigatran etexilate—a novel,
rivaroxaban versus short term enoxaparin for Presented at the American Society of reversible, oral direct thrombin inhibitor:
the prevention of VTE after total hip arthroplasty. Hematology 2011 Annual Meeting. interpretation of coagulation assays and reversal
Lancet 372, 31–39 (2008). 45. Ahrens, I. & Bode, C. Oral anticoagulation with of anticoagulant activity. Thromb. Haemost. 103,
26. Lassen, M. R. et al. Rivaroxaban versus edoxaban. Focus on current phase III clinical 1116–1127 (2010).
enoxaparin for thromboprophylaxis after development. Hamostaseologie 32, 212–215 64. Akwaa, F. & Spyropoulos, A. C. Treatment
total knee arthroplasty. N. Engl. J. Med. (2012). of bleeding complications when using oral
358, 2776–2786 (2008). 46. Raskob, G. et al. Oral direct factor Xa inhibition anticoagulants for prevention of strokes.
27. Turpie, A. G. et al. Rivaroxaban versus with edoxaban for thromboprophylaxis after Curr. Treat. Options Cardiovasc. Med. http://
enoxaparin for thromboprophylaxis after total elective total hip replacement: a randomised dx.doi.org/10.1007/s11936-013-0238-5.
knee arthroplasty (RECORD4): a randomized double-blind dose-response study. Thromb. 65. Godier, A. et al. Evaluation of prothrombin
trial. Lancet 373, 1673–1680 (2009). Haemost. 104, 642–649 (2010). complex concentrate and recombinant activated
28. Turun, S. et al. A systematic review of 47. Fuji, T. et al. A dose-ranging study evaluating factor VII to reverse rivaroxaban in a rabbit
rivaroxaban versus enoxaparin in the the oral factor Xa inhibitor edoxaban for model. Anesthesiology 116, 94–102 (2012).
prevention of venous thromboembolism after prevention of venous thromboembolism in 66. Hollenbach, S. J. et al. PRT064445 but not
hip or knee replacement. Thromb. Res. 127, patients undergoing total knee arthroplasty. recombinant FVIIa reverses rivaroxaban induced
525–534 (2011). J. Thromb. Haemost. 8, 2458–2468 (2010). anticoagulation as measured by reduction in
29. Cohen, A. T. et al. Rivaroxaban for 48. Piazza, G. et al. Venous thromboembolism blood loss in a rabbit liver laceration model
thromboprophylaxis in acutely ill medical in patients with symptomatic atherosclerosis. [abstract 3414]. Presented at the American
patients. N. Engl. J. Med. 368, 513–523 (2013). Thromb. Haemost. 106, 1095–1102 (2011). Society of Hematology 2012 Annual Meeting.
30. Raghavan, N. et al. Apixaban metabolism and 49. Piazza, G. et al. Venous thromboembolism in 67. Kaatz, S. et al. Guidance on the emergent
pharmacokinetics after oral administration to patients with diabetes mellitus. Am. J. Med. 125, reversal of oral thrombin and factor Xa inhibitors.
humans. Drug Metab. Dispos. 37, 74–81 (2009). 709–716 (2011). Am. J. Hematol. 87 (Suppl. 1), S141–S145
31. EINSTEIN Investigators. Oral rivaroxaban for 50. Prandoni, P. et al. An association between (2012).
symptomatic venous thromboembolism. N. Engl. atherosclerosis and venous thrombosis. N. Engl. 68. Samama, M. M. et al. Assessment of laboratory
J. Med. 363, 2499–2510 (2010). J. Med. 348, 1435–1441 (2003). assays to measure rivaroxaban—an oral, direct
32. EINSTEIN-PE Investigators. Oral rivaroxaban 51. Prandoni, P. Venous thromboembolism and factor Xa inhibitor. Thromb. Haemost. 103,
for the treatment of symptomatic pulmonary atherosclerosis: is there a link? J. Thromb. 815–825 (2010).
embolism. N. Engl. J. Med. 366, 1287–1297 Haemost. 5 (Suppl. 1), 270–275 (2007). 69. Poulsen, B. K., Grove, E. L. & Husted, S. E.
(2012). 52. Prandoni, P. Links between arterial and venous New oral anticoagulants: a review of the
33. Prandoni, P. Anticoagulant treatment of disease. J. Intern. Med. 262, 341–350 (2007). literature with particular emphasis on patients
pulmonary embolism: impact and implications 53. O’Gara, P. T. et al. 2013 ACCF/AHA guideline with impaired renal function. Drugs 72,
of the EINSTEIN PE study. Eur. J. Haematol. 89, for the management of ST-elevation myocardial 1739–1753 (2012).
281–287 (2012). infarction: a report of the American College 70. Baber, U., van der Zee, S. & Fuster, V.
34. Romualdi, E., Donadini, M. P. & Ageno, W. of Cardiology Foundation/American Heart Anticoagulation for mechanical heart valves
Oral rivaroxaban after symptomatic venous Association task force on practice guidelines. in patients with and without atrial fibrillation.
thromboembolism: the continued treatment Circulation 127, e362–e425 (2013). Curr. Cardiol. Rep. 12, 133–139 (2010).
study (EINSTEIN-extension study). Expert Rev. 54. Fiessinger, J. N. et al. Ximelagatran vs 71. Price, J., Hynes, M., Labinaz, M., Ruel, M.
Cardiovasc. Ther. 9, 841–844 (2011). low‑molecular-weight heparin and warfarin & Boodhwani, M. Mechanical valve thrombosis
35. Lassen, M. R. et al. Apixaban or enoxaparin for for the treatment of deep vein thrombosis: with dabigatran. J. Am. Coll. Cardiol. 60,
thromboprophylaxis after knee replacement. a randomized trial. JAMA 293, 681–689 (2005). 1710–1711 (2012).
N. Engl. J. Med. 361, 594–604 (2009). 55. Olsson, S. B. et al. Stroke prevention with 72. Van de Werf, F. et al. A comparison of dabigatran
36. Lassen, M. R. et al. Apixaban versus enoxaparin the oral direct thrombin inhibitor ximelagatran etexilate with warfarin in patients with
for thromboprophylaxis after knee replacement compared with warfarin in patients with mechanical heart valves: the randomized,
(ADVANCE‑2): a randomised double-blind trial. non‑valvular atrial fibrillation (SPORTIF III): phase II study to evaluate the safety and
Lancet 375, 807–815 (2010). randomised controlled trial. Lancet 362, pharmacokinetics of oral dabigatran etexilate
37. Lassen, M. R. et al. Apixaban versus enoxaparin 1691–1968 (2003). in patients after heart valve replacement
for thromboprophylaxis after hip replacement. 56. Albers, G. W. et al. Ximelagatran vs warfarin for (RE‑ALIGN). Am. Heart J. 163, 931–937, e1
N. Engl. J. Med. 363, 2487–2498 (2010). stroke prevention in patients with nonvalvular (2012).
38. Goldhaber, S. Z. et al. Apixaban versus atrial fibrillation: a randomized trial. JAMA 293, 73. Heit, J. A. et al. Trends in the incidence of
enoxaparin for thromboprophylaxis in medically 690–698 (2005). venous thromboembolism during pregnancy

408  |  JULY 2013  |  VOLUME 10  www.nature.com/nrcardio

© 2013 Macmillan Publishers Limited. All rights reserved
 REVIEWS

or postpartum: a 30-year population-based 79. Stone, S. E. & Morris, T. A. Pulmonary embolism 86. Lee, A. Y. et al. Low-molecular-weight
study. Ann. Intern. Med. 143, 697–706 (2005). during and after pregnancy. Crit. Care Med. heparin versus a coumarin for the prevention
74. James, A. H., Jamison, M. G., Brancazio, L. R. 33 (Suppl. 10), S294–S300 (2005). of recurrent venous thromboembolism in
& Myers, E. R. Venous thromboembolism 80. Sorensen, H. T., Mellemkjaer, L., Olsen, J. H. patients with cancer. N. Engl. J. Med. 349,
during pregnancy and the postpartum period: & Baron, J. A. Prognosis of cancers associated 146–153 (2003).
incidence, risk factors, and mortality. Am. J. with venous thromboembolism. N. Engl. J. Med. 87. Verso, M. & Agnelli, G. New and old
Obstet. Gynecol. 194, 1311–1315 (2006). 343, 1846–1850 (2000). anticoagulants in cancer. Thromb. Res.
75. Forestier, F., Daffos, F. & Capella-Pavlovsky, M. 81. Chew, H. K., Wun, T., Harvey, D., Zhou, H. 129 (Suppl. 1), S101–S105 (2012).
Low molecular weight heparin (PK 10169) & White, R. H. Incidence of venous 88. Farge, D. et al. International clinical practice
does not cross the placenta during the second thromboembolism and its effect on survival guidelines for the treatment and prophylaxis
trimester of pregnancy study by direct fetal blood among patients with common cancers. Arch. of venous thromboembolism in patients
sampling under ultrasound. Thromb. Res. 34, Intern. Med. 166, 458–464 (2006). with cancer. J. Thromb. Haemost. 11, 56–70
557–560 (1984). 82. Kuderer, N. M., Ortel, T. L. & Francis, C. W. (2013).
76. Ginsberg, J. S., Hirsh, J., Turner, D. C., Impact of venous thromboembolism and 89. Levine, M. N. New antithrombotic drugs:
Levine, M. N. & Burrows, R. Risks to the fetus anticoagulation on cancer and cancer survival. potential for use in oncology. J. Clin. Oncol.
of anticoagulant therapy during pregnancy. J. Clin. Oncol. 27, 4902–11 (2009). 27, 4912–4918 (2009).
Thromb. Haemost. 61, 197–203 (1989). 83. Horsted, F., West, J. & Grainge, M. J. Risk 90. Turpie, A. G. et al. A randomized evaluation
77. Romualdi, E. et al. Anticoagulant therapy for of venous thromboembolism in patients with of betrixaban, an oral factor Xa inhibitor, for
venous thromboembolism during pregnancy: cancer: a systematic review and meta-analysis. prevention of thromboembolic events after total
a systematic review and a meta-analysis of the PLoS Med. 9, e1001275 (2012). knee replacement (EXPERT). Thromb. Haemost.
literature. J. Thromb. Haemost. 11, 270–281 84. Young, A. et al. Thrombosis and cancer. 101, 68–76 (2009).
(2013). Nat. Rev. Clin. Oncol. 9, 437–449 (2012).
78. Jacobsen, A. F. & Sandset, P. M. Venous 85. Lyman, G. H. et al. American Society of Clinical Author contributions
thromboembolism associated with Oncology guideline: recommendations for All authors contributed substantially to researching
pregnancy and hormonal therapy. Best venous thromboembolism prophylaxis and data for the article, discussion of content, wrote the
Pract. Res. Clin. Haematol. 25, 319–332 treatment in patients with cancer. J. Clin. Oncol. article, and reviewed and edited the manuscript
(2012). 25, 5490–5505 (2007). before submission.

NATURE REVIEWS | CARDIOLOGY VOLUME 10  |  JULY 2013  |  409

© 2013 Macmillan Publishers Limited. All rights reserved