49. Jones WT 3rd, Hagino RT, Chiou AC, et al. Graft patency is not the only clinical predictor of

success after exclusion and bypass of popliteal artery aneurysms. J Vasc Surg 2003;37:392–398.

50. Dorigo W, Pulli R, Turini F, et al. Acute leg ischaemia from thrombosed popliteal artery aneurysms:

role of preoperative thrombolysis. Eur J Vasc Endovasc Surg 2002;23:251–254.

2790

Chapter 98

Venous Disease

Thomas W. Wakefield and Michael C. Dalsing

Key Points

1 The diagnosis and treatment of acute and chronic venous disease have made significant advances in

the last decade.

2 New drug therapies exist or are on the verge of impacting the treatment of acute deep venous

thrombosis (DVT).

3 Chronic venous disease has experienced a resurgence of interest in the United States with new

endoluminal and surgical interventions being investigated and utilized clinically.

4 Characterization of the disease is being scrutinized to allow for a better understanding of how

treatment truly impacts patient outcomes.

5 The care of venous disease is an active and maturing discipline within the realm of the surgical

disciplines.

INCIDENCE, RISK FACTORS, AND CATEGORIES

Venous thromboembolism (VTE), including deep venous thrombosis (DVT) and pulmonary embolism

(PE), is a national healthcare concern. DVT affects greater than 250,000 patients annually, whereas PE

affects greater than 200,000 patients per year. The incidence has remained constant since 1980 and

increases with age.1 The cost of treatment for both DVT and PE (termed VTE) is in the billions of dollars

per year. Although Virchow triad of stasis, vessel wall injury, and hypercoagulability have defined the

events that predispose to DVT formation for the past 150 years, today the understanding of events that

occur at the level of the vein wall including the influence of the inflammatory response on

thrombogenesis is increasingly being recognized.

Acquired risk factors for VTE include increasing age, malignancy, immobilization, surgery and

trauma, oral contraceptive use, hormone replacement therapy, pregnancy and the puerperium,

neurologic disease, cardiac disease, obesity, and antiphospholipid antibodies.2 Genetic risk factors

include deficiencies of antithrombin, protein C and protein S, factor V Leiden, prothrombin 20210A,

blood group non-O, dysfibrinogenemia, dysplasminogenemia, hyperhomocystinemia, reduced heparin

cofactor II activity, elevated levels of clotting factors (e.g., factors XI, IX, VII, VIII, X, and II), and

elevations in plasminogen activator inhibitor-1 (PAI-1).3 When a patient presents with an idiopathic

VTE, there is family history of VTE, there is recurrent thrombosis, or there is thrombosis in unusual

locations, workup for a hypercoagulable state may be indicated (hypercoagulable testing; Table 98-1).2,3

Hematologic diseases associated with VTE include heparin-induced thrombocytopenia and thrombosis

syndrome, disseminated intravascular coagulation, antiphospholipid antibody syndrome, thrombotic

thrombocytopenic purpura, hemolytic uremic syndrome, and myeloproliferative disorders.

Most DVTs affect the iliac, femoral, or popliteal lower limb veins. Presenting symptoms include

unilateral leg pain and swelling, but some DVTs are silent, with the first manifestation a PE. A recent

study of 5,451 patients with ultrasound confirmed DVT revealed the most common comorbidities to be

hypertension, surgery within 3 months, immobility within 30 days, cancer, and obesity.4

Venous Thromboembolism Diagnosis

1 Venous color duplex imaging is now standard, although D-dimer may be considered in low probability

settings. The positive and negative predictive values for gray-scale imaging are inferior to color

imaging. A meta-analysis of 13 studies comparing CT with ultrasound for proximal DVT found a

sensitivity of 71% to 100% and specificity between 93% and 100%.5 In symptomatic patients, sensitivity

is approximately 93%, specificity 98%, accuracy 97%, with approximately 30% of studies indeterminate

2791

in the calf.6 Data from the Prospective Investigation of Pulmonary Embolism Diagnosis II, a multicenter

study of 711 patients, were used to compare the clinical value of CT Venography (CTV) after

multidetector CT angiography (CTA) with venous compression ultrasonography for the diagnosis of

VTE. There was 95.5% concordance between CTV and ultrasonography for the diagnosis or exclusion of

DVT (κ = 0.809). Attesting to the accuracy of ultrasound, the sensitivity and specificity of combined

CTA and CTV were equivalent to combined CTA and ultrasonography.7

The usefulness and absolute need for duplex imaging depend, to some degree, on the pretest

probability for DVT. In a patient with a low pretest probability for DVT (such as a Wells score), if a Ddimer biomarker test is negative, no further testing is necessary.8 D-Dimer requires that a highsensitivity D-dimer assay (such as the advanced turbidimetric method or ELISA) is used. If the D-dimer is

positive, then duplex must be performed even if the pretest probability is high, as a positive D-dimer

and a positive risk assessment is associated with DVT in only approximately 70% of cases.9 A highly

soluble P-selectin, along with a positive Wells score, may allow for the diagnosis of DVT in greater than

90% of cases.10 The high negative predictive value (NPV) of color duplex Doppler ultrasound both

above and below the knee supports withholding anticoagulation on the basis of a good-quality negative

study alone, and clinical data also support that strategy.11 A single good-quality color duplex Doppler

study is sufficient (NPV >99.5%) to exclude proximal DVT; repeat scanning is seldom indicated unless

the initial study was technically suboptimal.12 A truly negative study means that all the segments of the

leg, including the external iliac, common femoral, femoral, popliteal, and calf veins, are negative by

ultrasound imaging and clear by Doppler flow. In a patient with a moderate pretest probability for DVT,

if duplex ultrasound is negative, no further testing is necessary. On the contrary, in a patient with a

high pretest probability for DVT, the D-dimer is less useful and even a negative result would not

supersede the need for imaging. The positive predictive value is high for symptomatic patients with a

positive duplex ultrasound and such patients should undergo treatment. If duplex ultrasound of proximal

veins is negative, either perform D-dimer (if negative, withhold anticoagulation and if positive, repeat in

1 week) or just repeat in 3 to 7 days (3 days if very suspicious, 7 days if less suspicious). If whole leg

duplex is negative, no further testing is necessary. During pregnancy the aforementioned

recommendations remain, but it is very important to attempt to assess for iliac vein thrombosis.

Table 98-1 Alternative Anticoagulants

For the diagnosis of superficial vein thrombophlebitis (SVT), duplex ultrasound is also important.13 In

a prospective epidemiologic study, including 844 patients with symptomatic SVT, 210 (24.9%) had

concomitant DVT at the time of diagnosis of SVT. Of the 600 patients without DVT, over the next 3

months, 58 (10.2%) developed thromboembolic complications, including 3 (0.5%) with PE, 15 (2.8%)

with DVT, 18 (3.3%) with extension, and 10 (1.9%) with recurrence of SVT. These events occurred

despite the use of anticoagulant treatment. Thus, it is important to perform good duplex imaging in

patients diagnosed with SVT.

Finally, in the CALISTO study, fondaparinux (Arixtra) was found effective in limiting a series of

endpoints including death, PE, DVT, and extension or recurrence at days 47 and 77 after SVT

diagnosis.14 This was in patients with axial vein SVT of at least 5-cm length with the thrombus at least 3

cm from the saphenofemoral junction (SFJ). Less extensive SVT requires only symptom control with

oral or topical nonsteroidal anti-inflammatory drugs (NSAIDs). Thus, it is important to document the

length of the SVT, the location, and the extent of the SVT to the SFJ, when imaging.

Other conditions may be confused with DVT and include lymphedema, muscle strains, and muscle

contusion. Iliac vein obstruction may lead to unilateral leg edema (May–Thurner syndrome), while the

presence of a cyst behind the knee may produce unilateral lower leg pain and edema. Other causes of

2792

leg swelling include systematic problems such as cardiac, renal, or hepatic abnormalities. These

systemic problems usually lead to bilateral edema.

Pulmonary Embolism

Testing for PE begins with prior probability estimation for all patients. Two of the most widely known

and validated diagnostic scoring systems are the Wells criteria (or modified/dichotomous Wells criteria)

and Geneva score.15–17 They use a combination of physical examination, history, and vital signs to

predict the likelihood of PE. D-Dimer testing is used to exclude PE for patients with low prior

probability for DVT. D-Dimer testing has no role in diagnosis if prior probability is not low. Use of Ddimer testing in PE diagnosis requires that a high-sensitivity D-dimer assay (such as the advanced

turbidimetric method or ELISA) validated at the local institution be used. Patients with both low

probability and negative D-dimer require no further investigation as the NPV is 99%.18,19

The diagnosis of PE has involved ventilation-perfusion (V/Q) scanning and pulmonary angiography.

However, newer techniques include spiral computed tomographic scanning and magnetic resonance

imaging. The sensitivity of V/Q scanning was defined in PIOPED I at 98%, but specificity was low at

10%.20 By combining clinical factors with V/Q scan, levels of sensitivity and specificity greater than

95% were achieved. With a high-probability V/Q scan, two risk factors positive for PE, the sensitivity

was 97%; with one risk factor, 84%; and with no risk factors, 82%. Similarly, with a normal V/Q scan,

the chance of PE was essentially 0, irrespective of the risk factor status.21 Thus, a normal V/Q scan or a

high-probability scan provided good diagnostic information that could be used to base treatment on.

However, only a small portion of V/Q scans are in one of these two categories, leaving many patients

needing further testing. Because of its invasive nature, pulmonary angiography is used less often today.

Pulmonary arteriography is included with acute massive PE, inferior vena cava (IVC) interruption, and

when planning interventional therapy, such as thrombolysis or pulmonary embolectomy.

Multidetector helical CTA is the primary imaging modality, although V/Q scanning remains viable

(particularly for patients with otherwise normal lungs in whom interpretability of the test is optimal),

and a positive lower-extremity color duplex Doppler study in a high-probability patient can establish the

diagnosis of VTE without lung imaging. However, the problem with using only lower extremity duplex

imaging is that a baseline study in the chest is not obtained for later comparisons. Multidetector

scanners have significantly improved the sensitivity and specificity, as well as the positive and negative

predictive value of CTA. Recent outcome studies have found the sensitivity and specificity of CTA to be

greater than 95%, and a negative CTA carries a 3-month risk of VTE of 1% to 2%. CTA establishes the

diagnosis of PE if it is positive in an intermediate- or high-probability patient, or in a low-probability

patient with findings of a main-stem or lobar embolus. It excludes the diagnosis of PE in low-probability

patients with negative scans. Discordance between prior probability and CTA findings requires further

investigation, as well as technically inadequate studies.7 Other diagnostic tests may include V/Q

imaging or pulmonary angiography.

Outcome studies have found comparable results between pulmonary angiography and CTA; a negative

result with either study confers approximately a 1% VTE rate within 6 months. However, because

angiography is invasive, it carries a greater risk of complications and mortality. The mortality from

angiography has been estimated at 0.5%, while 1% may experience major complications, including

arrhythmias, hypotension, bleeding, and nephrotoxicity.22 Without a higher standard to appeal to, one

cannot discuss specificity and sensitivity of pulmonary angiography using commonly accepted

definitions of these terms. Instead, the accuracy of pulmonary angiography is discussed in terms of

interobserver variability in the reading of pulmonary angiograms obtained in the context of large

multicenter trials. Studies demonstrate that the larger the embolus, the better the interobserver

agreement. For segmental and larger emboli, agreement exceeds 95%. For subsegmental emboli,

agreement is considerably less. We do not recommend pulmonary angiography, except in certain

circumstances, such as with inadequacy of V/Q imaging or when catheter-directed thrombolysis is

recommended.

Recently, the efficacy of magnetic resonance angiography (MRA), magnetic resonance venography

(MRV), or the combination of the two in the diagnosis of acute PE has been studied.23 The gold

standard used for comparison was a composite end-point of CTA, CTA-CTV, V/Q scan, lower extremity

ultrasonography, D-dimer assay, and clinical assessment. Overall, MRA and MRA-MRV were found to be

poor tests for the diagnosis of PE. Approximately 25% of the 371 patients enrolled in a large

multicenter study had a technically inadequate MRA, and 48% had either an inadequate MRV or MRA.

Considering all patients enrolled, MRA alone identified only 57% of PEs and had a sensitivity of 78%

2793

when only patients with adequate studies were considered. The combined MRA-MRV studies had a

sensitivity of 92%, but only about half of the patients had technically adequate studies. These poor

results are generally felt related to the technical difficulties of MRA in identifying abrupt vessel

termination and capturing adequate images of the chest vessels secondary to motion artifact. At this

time, MRA-MRV is not generally recommended and only considered in centers with a great deal of

experience.

Because of its low sensitivity and specificity, transthoracic or transesophageal echocardiography has

limited diagnostic value for PE.24 For critically ill patients too unstable for transport, echocardiography

can suggest the diagnosis of PE by showing dilatation of the RV or hypokinesis. Commonly, acute

changes in the RV pressure, size, and function are seen, indicating increased RV strain and pulmonary

arterial pressures. These changes suggest PE in the absence of alternative diagnoses. Although of limited

value in the diagnosis of PE, echocardiography is of great prognostic use in stratifying risk for patients

with acute PE. Right ventricular dysfunction or dilatation in acute PE is associated with worse

outcomes, including increased mortality rate.

Biomarkers include B-type natriuretic peptide (BNP), released by ventricular myocardial cells in

response to wall stretch and volume overload. BNP is a prognostic (not diagnostic) biomarker for PE.

BNP levels generally indicate RV strain due to elevated pulmonary vascular resistance in the lungs. If

measured early (within 4 hours of admission for PE), elevated BNP levels (>90 pg/mL) demonstrate a

sensitivity of 85% and a specificity of 75% in predicting PE-related outcomes. Conversely, normal BNP

values in the setting of acute PE carry a 97% to 100% NPV for in-hospital death. Another biomarker is

troponin released from damaged myocardial cells. When elevated in acute PE, troponins represent

myocyte ischemia and microinfarction due to acute cardiac strain of the right ventricle. Approximately

30% to 50% of patients with large PE will have elevations in troponins I and T that are mild and shortlived. They correlate with worse RV function and a high incidence of complications. Normal troponin T

levels have a 97% to 100% NPV for in-hospital death. These two biomarkers are not part of routine

algorithms.19

Axillary/Subclavian Vein Thrombosis

Thrombosis of the axillary/subclavian vein accounts for less than 5% of all cases of acute DVT.

However, it may associated with PE in up to 10% to 15% of cases and additionally can also be the

source of significant disability.25 Primary axillary/subclavian vein thrombosis results from obstruction

of the axillary vein in the thoracic outlet, the so-called Paget–Schroetter syndrome, noted especially in

healthy muscular athletic individuals. Such thrombosis may also occur in patients with hypercoagulable

states. Secondary axillary/subclavian vein thrombosis results from mediastinal tumors, congestive heart

failure, and nephrotic syndrome. Patients with axillary–subclavian venous thrombosis often present with

arm pain, edema, and cyanosis. Superficial venous distension may be apparent over the arm, forearm,

shoulder, and even the anterior chest wall.

Upper extremity venous duplex ultrasound is used to diagnose suspected axillary–subclavian vein

thrombosis. Thrombolysis and phlebography are often considered next. If phlebography is performed, it

is important that the patient undergo positional phlebography with arm abduction to 120 degrees to

confirm extrinsic subclavian vein compression at the thoracic outlet. Venous compromise is further

evidenced by prominent collateral veins. Since a cervical rib may be the cause of such obstruction, chest

x-ray film should be obtained to exclude its presence.

STANDARD THERAPY FOR VTE

Treatment

The traditional treatment of VTE is systemic anticoagulation, which reduces the risk of PE, extension of

thrombosis, and thrombus recurrence. Immediate anticoagulation should be undertaken as the

recurrence rate for VTE is higher if anticoagulation is not therapeutic in the first 24 hours.26 For DVT,

since duplex imaging is rapidly obtained, usually testing precedes anticoagulation while for PE,

anticoagulation is recommended either simultaneous with or before testing. Recurrent DVT can still

occur in up to one third of patients over an 8-year period, even with appropriate anticoagulant

therapy.27

Unfractionated heparin or low–molecular-weight heparin (LMWH) is given along with oral

anticoagulation with vitamin K antagonists (usually warfarin [Coumadin]). It is recommended that the

2794

international normalized ratio (INR) be therapeutic for 2 consecutive days before stopping heparin or

LMWH.28 LMWH has become the standard for treatment because it is administered subcutaneously,

requires no monitoring (except in certain circumstances, such as renal insufficiency or morbid obesity),

and is associated with a decrease in bleeding potential.29 Compared to standard unfractionated heparin,

LMWH has significantly improved bioavailability, less endothelial cell binding and protein binding, and

improved pharmacokinetics.30 The half-life of LMWH is dose independent and it is administered in a

weight-based fashion. LMWHs may decrease indices of chronic venous insufficiency (CVI) compared

with standard therapy when used over an extended period. Based on all of the available evidence,

LMWH is now preferred over standard unfractionated heparin for the initial treatment of VTE with a

level of evidence given 2B (according to the 2012 Chest consensus guidelines).31 Warfarin should be

started after heparinization is therapeutic to prevent warfarin-induced skin necrosis, as warfarin causes

inhibition of protein C and S before factors II, IX, and X, leading to paradoxical hypercoagulability at

the initiation of therapy. For standard unfractionated heparin, this requires a therapeutic activated

partial thromboplastin time; for LMWH, warfarin is administered after an appropriate weight-based

dose of LMWH is administered and allowed to circulate.

The goal for warfarin dosing is an INR between 2.0 and 3.0. The duration of anticoagulation depends

on a number of factors, including the presence of continuing risk factors for thrombosis, the type of

thrombosis (idiopathic or provoked), the number of times thrombosis has occurred, the status of the

veins when stopping anticoagulation, and the level of D-dimer measured approximately 1 month after

stopping warfarin. One study demonstrated a statistically significant advantage to resuming warfarin if

the D-dimer assay is elevated over an average 1.4-year follow-up (odds ratio [OR], 4.26), and a metaanalysis has confirmed this relationship.32 The recommended duration of anticoagulation after a first

episode of VTE is 3 months.31 After a second episode of VTE, the usual recommendation is prolonged

warfarin unless the patient is very young at the time of presentation or there are other mitigating

factors. VTE recurrence is increased with homozygous factor V Leiden and prothrombin 20210A

mutation, protein C or protein S deficiency, antithrombin deficiency, antiphospholipid antibodies, and

cancer until resolved. In these situations, long-term warfarin is recommended. However, heterozygous

factor V Leiden and prothrombin 20210A do not carry the same risk as their homozygous counterparts,

and the length of oral anticoagulation is shortened for these, the most common of hypercoagulable

conditions. Regarding idiopathic DVT, most believe that this diagnosis requires longer than 6 months of

anticoagulation, but the actual length is unknown.33 Taken together, criteria for discontinuing

anticoagulation are given a level of evidence of 1B to 2B, depending on the clinical situation.31,33–37

Recent evidence suggests that the decision to continue anticoagulation indefinitely after a first

unprovoked proximal DVT is strengthened if the patient is male, the index event was a PE, and the Ddimer is positive 1 month after stopping anticoagulation.38 In addition, there is growing evidence that

in certain circumstances, such as active cancer, the use of LMWH is superior to LMWH converted to

warfarin for long-term treatment.

Complications

Bleeding is the most common complication of anticoagulation. With standard heparin, bleeding occurs

over the first 5 days in approximately 10% of patients.39 Heparin-induced thrombocytopenia (HIT)

occurs in 0.6% to 30% of patients. Although historically morbidity and mortality rates have been high,

it has been found that early diagnosis and appropriate treatment have decreased these rates.40 HIT

usually begins 3 to 14 days after heparin is begun, although it can occur earlier if there has been

exposure to heparin in the past. A heparin-dependent antibody binds to platelets, activates them with

the release of procoagulant microparticles leading to an increase in thrombocytopenia, resulting in

thrombosis.41 Both bovine and porcine unfractionated heparin and LMWH have been associated with

HIT, although the incidence and severity of the thrombosis are less with LMWH. Even small exposures

to heparin can cause the syndrome. The diagnosis should be suspected with a 50% or greater drop in

platelet count, when the platelet count falls less than 100,000/mL or when thrombosis occurs during

heparin or LMWH therapy.42 A highly sensitive but poorly specific ELISA test detects the antiheparin

antibody in the plasma. The serotonin release assay is another test that is more specific but less

sensitive.43 When the diagnosis is made, heparin must be stopped and warfarin should not be given until

an adequate alternative anticoagulant has been established and the platelet count has normalized.

LMWHs cannot be used as substitutes as studies demonstrate high cross-reactivity with standard heparin

antibodies. Agents that have been approved by the Food and Drug Administration (FDA) as alternatives

include the direct thrombin inhibitor argatroban and bivalirudin (Table 98-1). Fondaparinux has also

2795

been found effective for treatment of HIT, but it is not FDA approved for this indication. The use of

these alternative agents is given 2C and 1C evidence.44

Low–Molecular-Weight Heparin Special Features

When considering once-a-day to twice-a-day LMWH dosing, a meta-analysis of greater than 1,500

patients with VTE demonstrated a nonsignificant difference in the incidence of recurrent

thromboembolism, thrombosis size, hemorrhagic events, and mortality.45 Twice-a-day dosing may still

be more appropriate than once-a-day dosing in patients with marked obesity and patients with cancer.46

LMWH has been suggested as a replacement for oral vitamin K antagonists. Rates of recanalization

have been reported to be higher in certain venous segments using LMWH and the use of LMWH has

been found to lead to improved outcomes in cancer patients compared to standard heparin or

LMWH/warfarin therapy when used for 6 months, without differences in rates of major bleeding.47

LMWH has also been found to provide better DVT prophylaxis than placebo when used for extended

prophylaxis over 4 weeks in patients undergoing abdominal and pelvic cancer surgery.48

ALTERNATIVE/FUTURE MEDICAL TREATMENTS FOR DVT

Fondaparinux targets factor Xa with a 17-hour half-life for fondaparinux (Table 98-1). It exhibits no

endothelial or protein binding and produces no thrombocytopenia. One disadvantage is the lack of a

readily available antidote. Fondaparinux has been tested in the prophylaxis of major orthopedic surgery.

In a meta-analysis involving greater than 7,000 patients, there was more than a 50% risk reduction

using fondaparinux begun 6 hours after surgery compared with LMWH begun 12 to 24 hours after

surgery.49 Critical bleeding was not different, although major bleeding was increased. Fondaparinux has

also been effective in prophylaxis of general medical patients, abdominal surgery patients, and for

extended prophylaxis after hip fracture.50–52 For DVT treatment, fondaparinux was found equal to

LMWH, while for PE, it was found equal to standard heparin.53,54 Dosage is based on body weight: 5 mg

per body weight <50 kg; 7.5 mg per body weight 50 to 100 kg; and 10 mg per body weight >100 kg.

Treatment at least for 5 days with concurrent administration of oral anticoagulation is recommended,

until the INR is therapeutic at a level of 2 to 3. Fondaparinux has been approved for the treatment of

DVT/PE, and for thrombosis prophylaxis in total hip, total knee, and hip-fractured patients, in the

extended prophylaxis of hip-fractured patients, and in abdominal surgery patients. A number of novel

oral anticoagulants are currently developed or in stages of development to either replace vitamin K

antagonists in concert with initial heparin or LMWH, or to replace both heparin/LMWH and vitamin K

antagonists totally as monotherapy. These agents hold the promise of not needing monitoring, being

safer in terms of bleeding risk than current agents, and being of equal or improved efficacy to

established anticoagulants.

2 Dabigatran targets activated factor II (factor IIa), while rivaroxaban, apixaban, and edoxaban target

activated factor X (factor Xa). Dabigatran etexilate is FDA approved for stroke and systemic

embolization prevention in patients with atrial fibrillation and for treating deep vein thrombosis (DVT)

and PE in patients who have been treated with a parenteral anticoagulant for 5 to 10 days. In trials for

VTE, out of six randomized controlled trials (RCTs), dabigatran was noninferior in three trials, superior

in two trials, and inferior in one trial. In the Recover trial, which compared dabigatran 150 mg to

therapeutic anticoagulation with vitamin K antagonists (INR, 2 to 3) in the treatment of DVT for 6

months, after both were given LMWH or unfractionated heparin for an average of 9 days, dabigatran

was noninferior in the 6-month rate of VTE recurrence. In addition, clinically significant bleeding was

not significantly different when compared with warfarin.55 In addition, in two other VTE trials,

dabigatran was found in extended duration treatment to have fewer recurrent VTEs compared with

placebo.56 Rivaroxaban is FDA approved for VTE prophylaxis in patients undergoing hip or knee

replacements, for stroke and systemic embolization prevention in patients with atrial fibrillation, and

for VTE treatment. Eight major trials studying VTE and rivaroxaban have been published, with

rivaroxaban noninferior in 6 and superior in 2 to standard therapy. The Einstein trial evaluated

rivaroxaban compared to standard anticoagulation in the treatment of acute DVT.57 As monotherapy,

rivaroxaban was found statistically noninferior to standard therapy, without increased bleeding risk. In

addition, the Einstein group added a continued treatment group compared to placebo for an additional 6

to 12 months. Extended rivaroxaban showed a significant decrease in recurrent VTE without an increase

in major bleeding. A similar finding with PE has been noted.58 In a trial for atrial fibrillation,

2796