Dr. Antani is a fellow in interventional radiology at The
Institute for Vascular Health and Disease, Albany Medical
Center, Albany, NY.
Deep venous thrombosis (DVT) of the lower extremity is the most
common manifestation of venous thrombosis in the body. The
incidence of lower-extremity DVT in the United States has been
estimated at 300,000 cases annually.
Virchow's triad refers to the existence of several factors
predisposing to the occurrence of venous thrombosis, and includes
venous stasis, hypercoagulability, and vessel injury. One or more
of these predisposing factors are commonly identified in patients
with DVT. Clinical risk factors for DVT include malignancy,
prolonged immobilization, and prior venous thrombosis, among
In most patients, thrombus originates and remains in the calf
veins, and the clinical significance of isolated calf thrombosis is
controversial, as an associated high risk for pulmonary embolism
has not been identified.
Thrombus involving the popliteal and iliofemoral veins is of
considerably greater clinical concern due to the higher risk for
pulmonary embolism and is the focus of most treatment strategies.
Symptoms of acute DVT include leg swelling, tenderness, and pain.
Homans' sign, although classically associated with DVT, is observed
infrequently, and is described as the onset of calf pain with
forced dorsiflexion of the foot.
Pulmonary embolism is the major complication of DVT, and the lower
extremity veins are the source of pulmonary embolism in at least
80% of patients.
Furthermore, pulmonary emboli associated with proximal DVT are
larger and more frequent compared with calf vein thrombosis, and
are more likely to be fatal.
The natural history of acute DVT includes several possibilities
occurring over the course of approximately 6 months. The thrombus
may undergo complete lysis; organization with venous occlusion; or
recanalization resulting in narrowing of the vein lumen, thickening
of the vein walls, and destruction of the valves. Up to 50% of
veins demonstrate residual disease after acute DVT.
These chronic venous changes may lead to incompetency of the venous
valvular system, with secondary venous hypertension. This has been
termed the post-thrombotic syndrome.
The clinical symptoms of the post-thrombotic syndrome include limb
edema, pain, hyperpigmentation, and chronic venous stasis ulcers.
Phlegmasia cerulea dolens is a condition seen in a small percentage
of patients with chronic DVT in whom marked venous insufficiency
leads to severe pain and leg swelling, followed by signs of
arterial insufficiency. Complications include gangrene of the
affected extremity, pulmonary embolism, and even death.
In addition, the elevated pressure in the deep venous system may
result in reversal of flow in the superficial veins, leading to the
development of varicose veins.
Standard treatment for DVT is anticoagulation and supportive
care, including bed rest, leg elevation, and compression stockings.
Anticoagulation is typically started with approximately 5 days of
intravenous unfractionated heparin administration followed by the
conversion to long-term oral warfarin sodium. The recent
introduction of subcutaneously administered low molecular-weight
heparin compounds has permitted outpatient management of acute DVT
in select cases.
The benefits of anticoagulant therapy include the prevention of
thrombus propagation, reduction in the rate of new thrombus
formation, and reduction in the rate of pulmonary embolism.
However, anticoagulation does not eliminate the existing thrombus.
This relies upon the patient's own fibrinolytic system, which can
recanalize the lumen of the occluded vein, thereby permitting
re-establishment of venous patency.
Johnson et al
have shown that in veins >8 mm in diameter (such as the
iliofemoral veins), the endogenous fibrinolytic system may be
unable to completely recanalize the vessel lumen, ultimately
predisposing patients to the post-thrombotic syndrome.
Systemic administration of thrombolytic agents has been reported
in the literature and has been shown to be clinically superior to
heparin administration alone. However, systemic thrombolytic
delivery is usually not successful with extended areas of
Catheter-directed thrombolysis was initially reported in the
early 1990s as an alternative to anticoagulation and systemic
The primary goals of catheter-directed thrombolysis include rapid
clot lysis that produces rapid symptomatic relief, and improved
therapeutic efficiency with local delivery of the thrombolytic
agent. Urokinase had been the primary thrombolytic agent until late
1998, with technical success rates reported at 80% to 90% for
In late 1998, the U.S. Food and Drug Administration passed a ban on
further use of urokinase due to concerns about sterility.
Alteplase (rt-PA; Activase, Genentech, Inc., South San Francisco,
CA) has now become the most commonly used agent for
catheter-directed thrombolysis. Alteplase is a recombinantly
engineered analog of human tissue plasminogen activator. Current
indications for alteplase include bolus administration for acute
myocardial infarction, pulmonary embolism, and acute nonhemorrhagic
stroke, but it has been used with success for DVT thrombolysis.
At the 2001 Society of Cardiovascular and Interventional
Radiology (SCVIR) Annual Meeting, Sugimoto et al
reported on the treatment of a total of 93 affected limbs (54 DVT,
39 acute peripheral arterial obstruction) with either alteplase
with subtherapeutic heparin, or urokinase with full heparin. The
investigators found that compared with urokinase, alteplase (at
<2 mg/hr) with subtherapeutic heparin was equally efficacious
with a similar complication rate, yet significantly (
<0.05) faster and less costly. Recently, the SCVIR Advisory
Panel on Catheter-Directed Thrombolytic Therapy published
preliminary guidelines for alteplase dosing to minimize the
associated risk of bleeding (Table 1).
Further recommendations for alteplase for catheter-directed
thrombolysis are being developed.
Reteplase (r-PA; Retevase, Centocor/Johnson & Johnson,
Malvern, PA) is a more recently investigated recombinant analog of
alteplase, and is presently approved for acute myocardial ischemia.
Reteplase demonstrates an increased circulating half-life (15
versus 5 minutes) and a decreased fibrin affinity when compared
with alteplase. The reduced affinity for fibrin may allow improved
absorption by the thrombus compared with alteplase, although its
true clinical significance has not been demonstrated. Recently,
investigators have reported the use of reteplase for peripheral
DVT. Castaneda et al,
at the 2001 SCVIR Annual Meeting, described preliminary results in
25 patients with acute DVT treated with reteplase. More than 90% of
patients were treated with infusion rates of 0.5 to 1.0 U/hr.
Success was obtained in 89% of patients, and there was no
significant difference in success with the lower 0.5 U/hr infusion
rate. Further research with reteplase comparing its efficacy with
other thrombolytic agents is being conducted. In addition, recent
research has suggested that reteplase in combination with
abciximab, a glycoprotein IIb/IIIa receptor antibody, may improve
the time to achieve thrombolysis and reduce the dose of the
thrombolytic compared with reteplase alone.
Patients with acute iliofemoral DVT, with <10-day duration of
symptoms, documented by venography or color Doppler ultrasound, and
without contraindications for thrombolysis and chronic
anticoagulation therapy, may be considered suitable candidates for
catheter-directed thrombolytic therapy.
Contraindications to thrombolysis include prior cerebrovascular
accident, active internal bleeding, a neurosurgical procedure
(intracranial or spinal) within the last 12 months, pregnancy, or
Typically, a multiple side-hole infusion catheter is placed across
the thrombus via the ipsilateral popliteal vein, which is punctured
under ultrasound guidance to avoid inadvertent popliteal artery
puncture. Although the femoral vein can also be used as a venous
access site, retrograde placement of the infusion catheter may
result in injury to the valve leaflets. Venography is performed at
selected intervals at the discretion of the physician to monitor
the progress of thrombolysis. Thrombolysis is discontinued when
venography shows resolution of thrombus, or when no additional
improvement is noted after further infusion over 12 hours. Balloon
angioplasty may be performed to macerate residual thrombus, and
stent placement may be used to treat venous wall elastic recoil,
venous rupture, or residual stenosis. Often, successful DVT
thrombolysis may uncover the primary cause of the patient's DVT,
such as iliac vein compression (May-Thurner) syndrome, which can be
treated with additional procedures such as stent placement
(figure 1). Inferior vena cava filters are indicated infrequently
because the occurrence of significant pulmonary embolism during
catheter-directed thrombolysis is uncommon.
Filter placement may be considered if there is a history of
recurrent pulmonary emboli despite adequate anticoagulation, or if
a "free-floating" thrombus in the iliac veins or inferior vena cava
has been identified.
The most significant potential complication of thrombolysis is
bleeding, either major (venous entry site, intracranial,
retroperitoneal, gastrointestinal, genitourinary, or
musculoskeletal) or minor (venous entry site). In the series of 473
patients reported by Mewissen et al,
11% had major bleeding complications and 16% had minor bleeding
complications. Pulmonary embolism is another major complication
reported in the literature, and in the same series by Mewissen et
al, a total of 6 patients experienced pulmonary embolism, one of
which was fatal.
Various investigators have attempted to devise methods to
"stage" DVT to predict the potential success of thrombolysis. At
the 2001 SCVIR Meeting, Roh et al
reported results assessing the value of selected CT findings from
24 patients with DVT in predicting the success of catheter-directed
thrombolysis. Patients who responded favorably to thrombolysis had
significantly higher thrombus attenuation coefficients (measured as
Hounsfield units) compared with those of patients who responded
poorly (66.1 ± 8.7 versus 45.9 ± 9.6,
<0.0001). The attenuation coefficient was found to be the most
significant of several CT variables by multivariate regression
Mechanical thrombectomy for lower extremity DVT, either
exclusively or in combination with thrombolysis, remains
controversial and the focus of ongoing investigation (figure 2).
Table 2 reviews six presently available mechanical thrombectomy
devices that are approved for use in polytetrafluoroethylene
hemodialysis grafts. These devices can be classified into two
categories, based upon their mechanism of action. Rotational
devices employ a mechanical device such as a wire basket
(Arrow-Trerotola Percutaneous Thrombolytic Device, Arrow
International, Reading, PA) or helical propeller (Amplatz
Thrombectomy Device, Microvena, White Bear Lake, MN) to macerate
the clot. Hydrolysing devices create a hydrodynamic vortex to
homogenize the clot into a slurry, which may be aspirated,
depending upon the specific device. A major disadvantage of
hydrolysing devices is that they are limited to relatively acute
thrombus (within 10 days old) for maximum efficacy.
Vedantham et al
reported the outcomes of 13 patients with 17 symptomatic limbs over
3 years with iliofemoral, caval, or femoropopliteal DVT treated
with both thrombolysis and various mechanical thrombectomy devices,
including the Amplatz Thrombectomy Device (ATD; n=9), AngioJet
(n=6), Oasis (n=1), and the Arrow-Trerotola Percutaneous
Thrombolytic Device (PTD; n=1). Nine cases were acute or subacute,
and 4 were chronic. Six patients were treated with urokinase (with
full heparin), 4 with alteplase, and 3 with retavase (both with
subtherapeutic heparin). Clinical success was obtained in 14 of 17
limbs; in 9 patients, patency was fully restored, and in 6
patients, iliac vein stents were required for residual stenosis (1
patient experienced pectineus muscle bleeding that required
transfusion). Of those limbs not treated for freshly formed
intraprocedure thrombus, mechanical thrombectomy cleared at least
one venous segment of thrombus in 4 of 7 when used before
thrombolysis, and in 8 of 8 cases when used after thrombolysis.
However, in only one of these cases was mechanical thrombectomy
able to clear thrombus over the entire occlusion; the remainder
demonstrated only partial clearing. Mean thrombolysis doses and
times were less than those reported by other investigators who used
different thrombolytic agents but who did not use mechanical
In a recent paper by Delomez et al,
mechanical thormbectomy with the Amplatz Thrombectomy Device was
performed on a total of 18 patients with iliofemoral or IVC DVT,
with technical success in 83%. Of 11 living patients followed-up on
average at 30 months, 10 had no or minimal sequelae.
A significant concern about mechanical thrombectomy that has
limited its widespread use is the risk for pulmonary embolism
during the procedure. The use of temporary IVC filtration to
prevent intraprocedure pulmonary embolism was reported by Trerotola
In this investigation, mechanical thrombectomy with the
Arrow-Trerotola Percutaneous Thrombolytic Device was performed in
12 canine models with subacute iliocaval thrombus after placement
of a nitinol expandable sheath into the suprarenal IVC for
temporary caval filtration. Results of pulmonary arteriography
after thrombectomy were compared with another group who underwent
the same procedure without caval filtration. A significant (
< 0.002) reduction in pulmonary emboli was noted after
filtration, although pulmonary emboli were not entirely eliminated,
as indicated by segmental and subsegmental emboli that were
observed after the procedure in the filtered group. This report
suggests that temporary caval filtration, although not a perfect
solution, may represent a significant advance that improves the
risk profile of mechanical thrombectomy for acute lower-extremity
Catheter-directed thrombolysis is a safe and effective treatment
for selected patients with symptomatic acute DVT, providing rapid
relief of symptoms and a reduction in the incidence of
complications associated with chronic DVT. The long-term benefits
of catheter-directed thrombolysis have yet to be established, but
should be accomplished with large randomized controlled trials
comparing this therapy with other therapies for acute DVT, such as
anticoagulation. Successful thrombolysis often may uncover an
underlying primary etiology for the patient's DVT, and adjunctive
procedures, such as angioplasty and stent placement, may increase
the primary patency rate compared with thrombolysis alone.
Mechanical thrombectomy, when used in conjunction with
catheter-directed thrombolysis, appears to be a promising procedure
that may significantly reduce the thrombolytic dose and time of
infusion. If, in fact, the thrombolytic dose can be reduced with
mechanical thrombectomy, perhaps the incidence of the feared
bleeding complications may be reduced even further. Temporary caval
filtration is an interesting technique that may reduce the
incidence of pulmonary emboli during mechanical thrombectomy, and
it deserves further investigation. *