Introduction of acute deep venous thrombosis


Deep venous thrombosis (DVT) most commonly involves the deep veins of the leg or arm, often resulting in potentially life-threatening emboli to the lungs or debilitating venous alular dysfunction and chronic leg swelling. Deep venous thrombosis (DVT) is also one of the most prevalent medical problems today, with an annual incidence of 117 cases per 100,000. Each year in the United States, more than 200,000 people develop venous thrombosis; of those, 50,000 cases are complicated by pulmonary embolism.1 Early recognition and appropriate treatment of deep venous thrombosis (DVT) and its complications can save many lives.




Over a century ago, Rudolf Virchow described 3 factors that are critically important in the development of venous thrombosis: (1) venous stasis, (2) activation of blood coagulation, and (3) vein damage. Over time, refinements have been made in their description and importance to the development of venous thrombosis. The origin of venous thrombosis is frequently multifactorial, with components of the triad of variable importance in individual patients.

Studies have shown that low flow sites, such as the soleal sinuses, behind venous valve pockets, and at venous confluences, are at most risk for the development of venous thrombi. However, stasis alone is not enough to facilitate the development of venous thrombosis. Experimental ligation of rabbit jugular veins for periods of up to 60 minutes have failed to consistently cause venous thrombosis. Although, patients that are immobilized for long periods of time seem to be at high risk for the development of venous thrombosis, an additional stimulus is required to develop deep venous thrombosis (DVTs).


Mechanical injury to the vein wall appears to provide an added stimulus for venous thrombosis. Hip arthroplasty patients with the associated femoral vein manipulation represent a high-risk group that cannot be explained by just immobilization, with 57% of thrombi originating in the affected femoral vein rather than the usual site of stasis in the calf.6 Endothelial injury can convert the normally antithrombogenic endothelium to become prothrombotic by stimulating the production of tissue factor, von Willebrand factor, and fibronectin.

 Genetic mutations within the blood’s coagulation cascade represent those at highest risk for the development of venous thrombosis (See Table 1).


Table 1. Relative Risk for Venous Thrombosis



 Primary deficiencies of coagulation inhibitors antithrombin, protein C, and protein S are associated with 5-10% of all thrombotic events. Resistance of procoagulant factors to an intact anticoagulation system has also recently been described with the recognition of factor V Leiden mutation, representing 10-65% of patients with deep venous thrombosis (DVT).8 In the setting of venous stasis, these factors are allowed to accumulate in thrombosis prone sites, where mechanical vessel injury has occurred, stimulating the endothelium to become prothrombotic.


Components of the Virchow triad are of variable importance in individual patients, but the end result is early thrombus interaction with the endothelium. This interaction stimulates local cytokine production and facilitates leukocyte adhesion to the endothelium, both of which promote venous thrombosis. Depending on the relative balance between activated coagulation and thrombolysis, thrombus propagation occurs.


Over time, thrombus organization begins with the infiltration of inflammatory cells into the clot. This results in a fibroelastic intimal thickening at the site of thrombus attachment in most patients and a fibrous synechiae in up to 11%. In many patients, this interaction between vessel wall and thrombus leads to alular dysfunction and overall vein wall fibrosis. Histological examination of vein wall remodeling after venous thrombosis has demonstrated an imbalance in connective tissue matrix regulation and a loss of regulatory venous contractility that contributes to the development of chronic venous insufficiency.


Risk factors


Many factors have been identified as known risk factors for the development of venous thrombosis. The single most powerful risk marker remains a prior history of DVT with up to 25% of acute venous thrombosis occurring in such patients. Pathologically, remnants of previous thrombi are often seen within the specimens of new acute thrombi. However, recurrent thrombosis may actually be the result of primary hypercoagulable states. Abnormalities within the coagulation cascade are the direct result of discrete genetic mutations within the coagulation cascade. Deficiencies of protein C, protein S, or antithrombin III account for approximately 5-10% of all cases of deep venous thrombosis (DVT).


Age has been well studied as an independent risk factor for venous thrombosis development. Although a 30-fold increase in incidence is noted from age 30 to age 80, the effect appears to be multifactorial, with more thrombogenic risk factors occurring in the elderly than in those younger than 40 years. Venous stasis, as seen in immobilized patients and paralyzed limbs, also contributes to the development of venous thrombosis. Autopsy studies parallel the duration of bed rest to the incidence of venous thrombosis, with 15% of patients in those studies dying within 7 days of bedrest to greater than 80% in those dying after 12 weeks.2 Within stroke patients, deep venous thrombosis (DVT) is found in 53% of paralyzed limbs, compared with only 7% on the nonaffected side.

Malignancy is noted in up to 30% of patients with venous thrombosis. The thrombogenic mechanisms involve abnormal coagulation, as evidenced by 90% of cancer patients having some abnormal coagulation factors.18 Chemotherapy may increase the risk of venous thrombosis by affecting the vascular endothelium, coagulation cascades, and tumor cell lysis. The incidence has been shown to increase in those patients undergoing longer courses of therapy for breast cancer, from 4.9% for 12 weeks of treatment to 8.8% for 36 weeks. Additionally, deep venous thrombosis (DVT) complicates 29% of surgical procedures done for malignancy.

Postoperative venous thrombosis varies depending on a multitude of patient factors, including the type of surgery undertaken. Without prophylaxis, general surgery operations typically have an incidence of deep venous thrombosis (DVT) around 20%, while orthopedic hip surgery can occur in up to 50% of patients. Based on radioactive labeled fibrinogen, about half of lower extremity thrombi develop intraoperatively. Perioperative immobilization, coagulation abnormalities, and venous injury all contribute to the development of surgical venous thrombosis.


Other clinical settings commonly reported as risk factors have also been identified and are shown in Table 2, 3



Table 2. Risk Factors for Venous Thromboemobolic Disease



Table 3.  Risk Factors for Venous Thromboembolism



Clinical and diagnostic evaluation


The clinical diagnosis of deep venous thrombosis (DVT) is difficult and fraught with uncertainty. The classic signs and symptoms of deep venous thrombosis (DVT) are those associated with obstruction to venous drainage and include pain, tenderness, and unilateral leg swelling. Other associated nonspecific findings are warmth, erythema, a palpable cord, and pain upon passive dorsiflexion of the foot (Homan sign). However, even with patients with classic symptoms, up to 46% have negative venograms. Furthermore, up to 50% of those with image-documented venous thrombosis lack any specific symptom. Deep venous thrombosis (DVT) simply cannot be diagnosed or excluded based on clinical findings; thus, diagnostic tests must be performed whenever the diagnosis of deep venous thrombosis (DVT) is being considered.


When a patient has deep venous thrombosis (DVT), symptoms may be present or absent, unilateral or bilateral, or mild or severe. Thrombus that does not cause a net venous outflow obstruction is often asymptomatic. Thrombus that involves the iliac bifurcation, the pelvic veins, or the vena cava produces leg edema that is usually bilateral rather than unilateral. High partial obstruction often produces mild bilateral edema that is mistaken for the dependent edema of right-sided heart failure, fluid overload, or hepatic or renal insufficiency.


Severe venous congestion produces a clinical appearance that can be indistinguishable from the appearance of cellulitis. Patients with a warm, swollen, tender leg should be evaluated for both cellulitis and deep venous thrombosis (DVT) because patients with primary deep venous thrombosis (DVT) often develop a secondary cellulitis, while patients with primary cellulitis often develop a secondary deep venous thrombosis (DVT). Superficial thrombophlebitis, likewise, is often associated with a clinically inapparent underlying DVT.


If a patient is thought to have pulmonary embolism (PE) or has documented PE, the absence of tenderness, erythema, edema, or a palpable cord upon examination of the lower extremities does not rule out thrombophlebitis, nor does it imply a source other than a leg vein. More than two thirds of patients with proven PE lack any clinically evident phlebitis. Nearly one third of patients with proven PE have no identifiable source of deep venous thrombosis (DVT), despite a thorough investigation. Autopsy studies suggest that even when the source is clinically inapparent, it lies undetected within the deep venous system of the lower extremity and pelvis in 90% of cases.


Vascular Lab and Radiologic Evaluation


Duplex Ultrasound


DUS is now the most commonly performed test for the detection of infrainguinal DVT, both above and below the knee, and has a sensitivity and specificity of >95% in symptomatic patients. DUS combines real-time B-mode ultrasound with pulsed Doppler capability. Color flow imaging is useful in more technically difficult examinations, such as in the evaluation of possible calf vein DVT. This combination offers the ability to noninvasively visualize the venous anatomy, detect occluded and partially occluded venous segments, and demonstrate physiologic flow characteristics using a mobile self-contained device.

In the supine patient, normal lower extremity venous flow is phasic (Fig. 1), decreasing with inspiration in response to increased intra-abdominal pressure with the descent of the diaphragm and then increasing with expiration. When the patient is upright, the decrease in intra-abdominal pressure with expiration cannot overcome the hydrostatic column of pressure existing between the right atrium and the calf. Muscular contractions of the calf, along with the one-way venous valves, are then required to promote venous return to the heart. Flow also can be increased by leg elevation or compression and decreased by sudden elevation of intra-abdominal pressure (Valsalva's maneuver). In a venous DUS examination performed with the patient supine, spontaneous flow, variation of flow with respiration, and response of flow to Valsalva's maneuver are all assessed. However, the primary method of detecting DVT with ultrasound is demonstration of the lack of compressibility of the vein with probe pressure on B-mode imaging. Normally, in transverse section, the vein walls should coapt with pressure. Lack of coaptation indicates thrombus.



Fig. 1. Duplex ultrasound scan of a normal femoral vein with phasic flow signals.


The examination begins at the ankle and continues proximally to the groin. Each vein is visualized, and the flow signal is assessed with distal and proximal compression. Lower extremity DVT can be diagnosed by any of the following DUS findings: lack of spontaneous flow (Fig. 2), inability to compress the vein (Fig. 3), absence of color filling of the lumen by color flow DUS, loss of respiratory flow variation, and venous distention. Again, lack of venous compression on B-mode imaging is the primary diagnostic variable. Several studies comparing B-mode ultrasound to venography for the detection of femoropopliteal DVT in patients clinically suspected to have DVT report sensitivities of >91% and specificities of >97%. The ability of DUS to assess isolated calf vein DVT varies greatly, with sensitivities ranging from 50 to 93% and specificities approaching 100%.



Fig. 2. Duplex ultrasound of a femoral vein containing thrombus demonstrating no flow within the femoral vein



Fig. 3. B-mode ultrasound of the femoral vein in cross-section. The femoral vein does not collapse with external compression



Impedance Plethysmography


Impedance plethysmography (IPG) was the primary noninvasive method of diagnosing DVT before the widespread use of DUS but is infrequently used today. IPG is based on the principle that resistance to the flow of electricity between two electrodes, or electrical impedance, occurs as the volume of the extremity changes in response to blood flow. Two pairs of electrodes containing aluminum strips are placed circumferentially around the leg approximately 10 cm apart and a low-level current is delivered to the two outer electrodes. A pneumatic cuff is inflated over the thigh for venous outflow obstruction and then rapidly deflated. Changes in electrical resistance resulting from lower extremity blood volume changes are quantified. IPG is less accurate than DUS for the detection of proximal DVT, with an 83% sensitivity in symptomatic patients. It is a poor detector of calf vein DVT.


Iodine 125 Fibrinogen Uptake


Iodine 125 fibrinogen uptake (FUT) is a seldom used technique that involves IV administration of radioactive fibrinogen and monitoring for increased uptake in fibrin clots. An increase of 20% or more in one area of a limb indicates an area of thrombus.23 FUT can detect DVT in the calf, but high background radiation from the pelvis and the urinary tract limits its ability to detect proximal DVT. It also cannot be used in an extremity that has recently undergone surgery or has active inflammation. In a prospective study, FUT had a sensitivity of 73% and specificity of 71% for identification of DVT in a group of symptomatic and asymptomatic patients. Currently, FUT is primarily a research tool of historic interest.




Venography is the most definitive test for the diagnosis of DVT in both symptomatic and asymptomatic patients. It is the gold standard to which other modalities are compared. This procedure involves placement of a small catheter in the dorsum of the foot and injection of a radiopaque contrast agent. Radiographs are obtained in at least two projections. A positive study result is failure to fill the deep system with passage of the contrast medium into the superficial system or demonstration of discrete filling defects (Fig. 4). A normal study result virtually excludes the presence of DVT. In a study of 160 patients with a normal venogram followed for 3 months, only two patients (1.3%) subsequently developed DVT and no patients experienced symptoms of PE.


Fig. 4. Venogram showing a filling defect in the popliteal vein (arrows).


Venography is not routinely used for the evaluation of lower extremity DVT because of the associated complications discussed previously. Currently, venography is reserved for imaging before operative venous reconstruction and catheter-based therapy. It does, however, remain the procedure of choice in research studies evaluating methods of prophylaxis for DVT.


Laboratory analysis has also been used in aiding the diagnosis of venous thrombosis. D-dimers are degradation products of cross-linked fibrin by plasmin that are detected by diagnostic assays. Although highly sensitive, up to 97%, elevated levels are not specific with rates as low as 35%.27 Many other clinical situations can result in elevated D-dimer levels, including infection, trauma, postoperative states, and malignancy.28 Additional blood work should include coagulation studies to evaluate for a hypercoagulable state, if clinically indicated. A prolonged prothrombin time or activated partial thromboplastin time does not imply a lower risk of new thrombosis. Progression of deep venous thrombosis (DVT) and PE can occur despite full therapeutic anticoagulation in 13% of patients.




Once the diagnosis of VTE has been made, antithrombotic therapy should be initiated promptly. If clinical suspicion for VTE is high, it may be prudent to start treatment while the diagnosis is being objectively confirmed. The theoretic goals of VTE treatment are the prevention of mortality and morbidity associated with PE and the prevention of the postphlebitic syndrome. However, the only proven benefit of anticoagulant treatment for DVT is the prevention of death from PE. Treatment regimens may include antithrombotic therapy, vena caval interruption, catheter-directed or systemic thrombolytic therapy, and operative thrombectomy.


Antithrombotic Therapy


Antithrombotic therapy may be initiated with IV or SC unfractionated heparin, SC low molecular weight heparin, or SC fondaparinux (a synthetic pentasaccharide). This initial therapy usually is continued for at least 5 days, while oral vitamin K antagonists are being simultaneously administered. The initial therapy typically is discontinued when the international normalized ratio (INR) is ≥2.0 for 24 hours.

Unfractionated heparin (UFH) binds to antithrombin via a specific 18-saccharide sequence, which increases its activity over 1000-fold. This antithrombin-heparin complex primarily inhibits factor IIa (thrombin) and factor Xa and, to a lesser degree, factors IXa, XIa, and XIIa. In addition, UFH also binds to tissue factor pathway inhibitor, which inhibits the conversion of factor X to Xa, and factor IX to IXa. Finally, UFH catalyzes the inhibition of thrombin by heparin cofactor II via a mechanism that is independent of antithrombin.

UFH therapy is most commonly administered with an initial IV bolus of 80 units/kg or 5000 units. Weight-based UFH dosages have been shown to be more effective than standard fixed boluses in rapidly achieving therapeutic levels. The initial bolus is followed by a continuous IV drip, initially at 18 units/kg per hour or 1300 units per hour. The half-life of IV UFH ranges from 45 to 90 minutes and is dose dependent. The level of antithrombotic therapy should be monitored every 6 hours using the activated partial thromboplastin time (aPTT), with the goal range of 1.5 to 2.5 times control values. This should correspond with plasma heparin anti-Xa activity levels of 0.3 to 0.7 IU/mL.

Initial anticoagulation with UFH may be administered SC, although this route is less commonly used. Adjusted-dose therapeutic SC UFH is initiated with 17,500 units, followed by 250 units/kg twice daily, and dosing is adjusted to an aPTT goal range similar to that for IV UFH. Fixed-dose unmonitored SC UFH is started with a bolus of 333 units/kg, followed by 250 units/kg twice daily.

Hemorrhage is the primary complication of UFH therapy. The rate of major hemorrhage (fatal, intracranial, retroperitoneal, or requiring transfusion of >2 units of packed red blood cells) is approximately 5% in hospitalized patients undergoing UFH therapy (1% in medical patients and 8% in surgical patients).27 For patients with UFH-related bleeding complications, cessation of UFH is required, and anticoagulation may be reversed with protamine sulfate. Protamine sulfate binds to UFH and forms an inactive salt compound. Each milligram of protamine neutralizes 90 to 115 units of heparin, and the dosage should not exceed 50 mg IV over any 10-minute period. Side effects of protamine sulfate include hypotension, pulmonary edema, and anaphylaxis. Patients with prior exposure to protamine-containing insulin (NPH) and patients with allergy to fish may have an increased risk of hypersensitivity, although no direct relationship has been established. The protamine infusion should be terminated if any side effects occur.

In addition to hemorrhage, heparin also has unique complications. Heparin-induced thrombocytopenia (HIT) results from heparin-associated antiplatelet antibodies (HAAbs) directed against platelet factor 4 complexed with heparin. HIT occurs in 1 to 5% of patients being treated with heparin. In patients with repeat heparin exposure (such as vascular surgery patients), the incidence of HAAb may be as high as 21%.HIT occurs most frequently in the second week of therapy and may lead to disastrous venous or arterial thrombotic complications. Therefore, platelet counts should be monitored periodically in patients receiving continuous heparin therapy. All forms of heparin should be stopped if there is a high clinical suspicion or confirmation of HIT [usually accompanied by an unexplained thrombocytopenia (<100,000/L) or platelet count decrease of 30 to 50%]. Fortunately, direct thrombin inhibitors (recombinant hirudin, argatroban, bivalirudin) now are available as alternative antithrombotic agents (see later). Another complication of prolonged high-dose heparin therapy is osteopenia, which results from impairment of bone formation and enhancement of bone resorption by heparin.

Low molecular weight heparins (LMWHs) are derived from the depolymerization of porcine UFH. Like UFH, LMWHs bind to antithrombin via a specific pentasaccharide sequence to expose an active site for the neutralization of factor Xa. However, LMWHs lack the sufficient number of additional saccharide units (18 or more), which results in less inactivation of thrombin (factor IIa). In comparison to UFH, LMWHs have increased bioavailability (>90% after SC injection), longer half-lives (approximately 4 to 6 hours), and more predictable elimination rates. Weight-based once- or twice-daily SC LMWH injections, for which no monitoring is needed, provide a distinct advantage over continuous IV infusions of UFH for treatment of VTE.

Most patients who receive therapeutic LMWH do not require monitoring. Patients who do require monitoring include those with significant renal insufficiency or failure, pediatric patients, obese patients of >120 kg, and patients who are pregnant. Monitoring may be performed using anti-Xa activity assays. However, the therapeutic anti-Xa goal range will depend on the type of LMWH and the frequency of dosing. Numerous LMWHs are commercially available. The various preparations differ in their anti-Xa and anti-IIa activities, and the treatment dosing for one LMWH cannot be extrapolated for use with another. The action of LMWHs may be partially reversed (approximately 60%) with protamine sulfate.

Numerous well-designed trials comparing SC LMWH with IV and SC UFH for the treatment of DVT have been critically evaluated in several meta-analyses. The more recent studies demonstrate a decrease in thrombotic complications, bleeding, and mortality with LMWHs. LMWHs also are associated with a decreased rate of HAAb formation and HIT (<2%) compared with UFH (at least in prophylactic doses). However, patients with established HIT should not subsequently receive LMWHs due to significant rates of cross reactivity. A major benefit of LMWHs is the ability to treat patients with VTE as outpatients. In a randomized study comparing IV UFH and the LMWH nadroparin calcium, there was no significant difference in recurrent thromboembolism (8.6% for UFH vs. 6.9% for LMWH) or major bleeding complications (2.0% for UFH vs. 0.5% for LMWH). There was a 67% reduction in mean days in the hospital for the LMWH group.

A patient with VTE should meet several criteria before receiving outpatient LMWH therapy. First, the patient should not require hospitalization for any associated conditions. The patient should not require monitoring of the LMWH therapy (which is necessary in patients with severe renal insufficiency, pediatric patients, obese patients, and pregnant patients). The patient should be hemodynamically stable with a low suspicion of PE and have a low bleeding risk. An established outpatient system to administer LMWH and warfarin, as well as to monitor for recurrent VTE and bleeding complications, should be present. In addition, the patient's symptoms of pain and edema should be controllable at home.

Fondaparinux currently is the only synthetic pentasaccharide that has been approved by the U.S. Food and Drug Administration (FDA) for the initial treatment of DVT and PE. Its five-polysaccharide sequence binds and activates antithrombin, causing specific inhibition of factor Xa. In two large noninferiority trials, fondaparinux was compared with the LMWH enoxaparin for the initial treatment of DVT and with IV UFH for the initial treatment of PE.The rates of recurrent VTE ranged from 3.8 to 5%, with rates of major bleeding of 2 to 2.6%, for all treatment arms. The drug is administered SC once daily with a weight-based dosing protocol: 5 mg, 7.5 mg, or 10 mg for patients weighing <50 kg, 50 to 100 kg, or >100 kg, respectively. The half-life of fondaparinux is approximately 17 hours in patients with normal renal function. There are rare case reports of fondaparinux-induced thrombocytopenia.

Direct-thrombin inhibitors (DTIs) include recombinant hirudin, argatroban, and bivalirudin. These antithrombotic agents bind to thrombin, inhibiting the conversion of fibrinogen to fibrin as well as thrombin-induced platelet activation. These actions are independent of antithrombin. The direct thrombin inhibitors should be reserved for (a) patients in whom there is a high clinical suspicion or confirmation of HIT, and (b) patients who have a history of HIT or test positive for heparin-associated antibodies. In patients with established HIT, DTIs should be administered for at least 7 days, or until the platelet count normalizes. Warfarin may then be introduced slowly, overlapping therapy with a DTI for at least 5 days.41 Because bivalirudin is approved primarily for patients with or without HIT who undergo percutaneous coronary intervention, it is not discussed here in further detail.

Commercially available hirudin is manufactured using recombinant DNA technology. It is indicated for the prophylaxis and treatment of patients with HIT. In patients with normal renal function, recombinant hirudin is administered in an IV bolus dose of 0.4 mg/kg, followed by a continuous IV infusion of 0.15 mg/kg per hour. The half-life ranges from 30 to 60 minutes. The aPTT is monitored, starting approximately 4 hours after initiation of therapy, and dosage is adjusted to maintain an aPTT of 1.5 to 2.5 times the laboratory normal value. The less commonly used ecarin clotting time is an alternative method of monitoring. Because recombinant hirudin is eliminated via renal excretion, significant dosage adjustments are required in patients with renal insufficiency.

Argatroban is indicated for the prophylaxis and treatment of thrombosis in HIT. It also is approved for patients with, or at risk for, HIT who undergo percutaneous coronary intervention. Antithrombotic prophylaxis and therapy are initiated with a continuous IV infusion of 2 g/kg per minute, without the need for a bolus. The half-life ranges from 39 to 51 minutes, and the dosage is adjusted to maintain an aPTT of 1.5 to 3 times normal. Large initial boluses and higher rates of continuous infusion are reserved for patients with coronary artery thrombosis and myocardial infarction. In these patients, therapy is monitored using the activated clotting time. Argatroban is metabolized by the liver, and the majority is excreted via the biliary tract. Significant dosage adjustments are needed in patients with hepatic impairment. There is no reversal agent for argatroban.

Vitamin K antagonists, which include warfarin and other coumarin derivatives, are the mainstay of long-term antithrombotic therapy in patients with VTE. Warfarin inhibits the -carboxylation of vitamin K–dependent procoagulants (factors II, VII, IX, X) and anticoagulants (proteins C and S), which results in the formation of less functional proteins. Warfarin usually requires several days to achieve its full effect, because normal circulating coagulation proteins must first undergo their normal degradation. Factors X and II have the longest half-lives, in the range of 36 and 72 hours, respectively. In addition, the steady-state concentration of warfarin is usually not reached for 4 to 5 days.

Warfarin therapy usually is monitored by measuring the INR, calculated using the following equation:



where ISI is the international sensitivity index. The ISI describes the strength of the thromboplastin that is added to activate the extrinsic coagulation pathway. The therapeutic target INR range is usually 2.0 to 3.0, but the response to warfarin is variable and depends on liver function, diet, age, and concomitant medications. In patients receiving anticoagulation therapy without concomitant thrombolysis or venous thrombectomy, the vitamin K antagonist may be started on the same day as the initial parenteral anticoagulant, usually at doses ranging from 5 to 10 mg. Smaller initial doses may be needed in older and malnourished patients, in those with liver disease or congestive heart failure, and in those who have recently undergone major surgery.

The recommended duration of warfarin antithrombotic therapy is increasingly being stratified based on whether the DVT was provoked or unprovoked, whether it was the first or a recurrent episode, where the DVT is located, and whether malignancy is present. Current American College of Chest Physicians (ACCP) recommendations for duration of warfarin therapy are summarized in Table 24-3. In patients with proximal DVT, several randomized clinical trials have demonstrated that shorter-term antithrombotic therapy (4 to 6 weeks) is associated with a higher rate of recurrence than 3 to 6 months of anticoagulation. In these trials, most of the patients with transient risk factors had a low rate of recurrent VTE; most of the recurrences were in patients with continuing risk factors. These studies support the ACCP recommendation that 3 months of anticoagulation is sufficient to prevent recurrent VTE in patients whose DVT occurred around the time of a transient risk factor (e.g., hospitalization, orthopedic or major general surgery).


Table 4.  Summary of American College of Chest Physicians Recommendations Regarding Duration of Long-Term Antithrombotic Therapy for Deep Vein Thrombosis (DVT)


LMWH = low molecular weight heparin; VKA = vitamin K antagonist.

Source: Adapted with permission from Kearon C, Kahn SR, Agnelli G, et al: Antithrombotic therapy for venous thromboembolic disease: American College of Chest Physicians Evidence-Based Clinical Practice Guidelines (8th edition). Chest 133:454S, 2008.


In contrast to patients with thrombosis related to transient risk factors, patients with idiopathic VTE are much more likely to develop recurrence (rates as high as 40% at 10 years). In this latter group of patients, numerous clinical trials have compared 3 to 6 months of anticoagulation therapy with extended-duration warfarin therapy, both at low intensity (INR of 1.5 to 2.0) and at conventional intensity (INR of 2.0 to 3.0). In patients with idiopathic DVT, extended-duration antithrombotic therapy is associated with a relative reduction in the rate of recurrent VTE by 75% to >90%. In addition, conventional-intensity warfarin reduces the risk even further compared with low-intensity warfarin (0.7 events per 100 person-years vs. 1.9 events per 100 person-years), but the rate of bleeding complications is no different.

In patients with VTE in association with a hypercoagulable condition, the optimal duration of anticoagulation therapy is influenced more by the clinical circumstances at the time of the VTE (idiopathic vs. secondary) than by the actual presence or absence of the more common thrombophilic conditions. In patients with VTE related to malignancy, increasing evidence suggests that longer-term therapy with LMWH (up to 6 months) is associated with a lower VTE recurrence rate than treatment using conventional vitamin K antagonists.

The primary complication of warfarin therapy is hemorrhage, and the risk is related to the magnitude of INR prolongation. Depending on the INR and the presence of bleeding, warfarin anticoagulation may be reversed by (a) omitting or decreasing subsequent dosages, (b) administering oral or parenteral vitamin K, or (c) administering fresh-frozen plasma, prothrombin complex concentrate, or recombinant factor VIIa. Warfarin therapy rarely may be associated with the development of skin necrosis and limb gangrene. These conditions occur more commonly in women (4:1), and the most commonly affected areas are the breast, buttocks, and thighs. This complication, which usually occurs in the first days of therapy, is occasionally, but not exclusively, associated with protein C or S deficiency and malignancy. Patients who require continued anticoagulation may restart low-dose warfarin (2 mg) while receiving concomitant therapeutic heparin. The warfarin dosage is then gradually increased over a 1- to 2-week period.


Systemic and Catheter-Directed Thrombolysis


Patients with extensive proximal DVT may benefit from systemic thrombolysis or catheter-directed thrombolysis, which can potentially reduce acute symptoms more rapidly than anticoagulation alone. These techniques also may decrease the development of postthrombotic syndrome. Several thrombolysis preparations are available, including streptokinase, urokinase, alteplase (recombinant tissue plasminogen activator), reteplase, and tenecteplase. All these agents share the ability to convert plasminogen to plasmin, which leads to the degradation of fibrin. They differ with regard to their half-lives, their potential for inducing fibrinogenolysis (generalized lytic state), their potential for antigenicity, and their FDA-approved indications for use.

Streptokinase is purified from beta-hemolytic Streptococcus and is approved for the treatment of acute myocardial infarction, PE, DVT, arterial thromboembolism, and occluded central lines and arteriovenous shunts. It is not specific for fibrin-bound plasminogen, however, and its use is limited by its significant rates of antigenicity. Fevers and shivering occur in 1 to 4% of patients. Urokinase is derived from human neonatal kidney cells, grown in tissue culture. Currently, it is only approved for lysis of massive PE or PE associated with unstable hemodynamics. Alteplase, reteplase, and tenecteplase all are recombinant variants of tissue plasminogen activator. Alteplase is indicated for the treatment of acute myocardial infarction, acute ischemic stroke, and acute massive PE. However, it often is used for catheter-directed thrombolysis of DVT. Reteplase and tenecteplase are indicated only for the treatment of acute myocardial infarction.


Systemic thrombolysis was evaluated in numerous older prospective and randomized clinical trials, and its efficacy was summarized in a recent Cochrane Review. In 12 studies involving over 700 patients, systemic thrombolysis was associated with significantly more clot lysis [relative risk (RR) 0.24 to 0.37] and significantly less postthrombotic syndrome (RR 0.66). However, venous function was not significantly improved. In addition, more bleeding complications did occur (RR 1.73), but the incidence appears to have decreased in later studies, probably due to improved patient selection.

In an effort to minimize bleeding complications and increase efficacy, catheter-directed thrombolytic techniques have been developed for the treatment of symptomatic DVT. With catheter-directed therapy, venous access may be achieved through percutaneous catheterization of the ipsilateral popliteal vein, retrograde catheterization through the contralateral femoral vein, or retrograde cannulation from the internal jugular vein. Multi–side-hole infusion catheters, with or without infusion wires, are used to deliver the lytic agent directly into the thrombus.

The efficacy of catheter-directed urokinase for the treatment of symptomatic lower extremity DVT has been reported in a large multicenter registry. Two hundred twenty-one patients with iliofemoral DVT and 79 patients with femoropopliteal DVT were treated with catheter-directed urokinase for a mean of 53 hours. Complete lysis was seen in 31% of the limbs, 50 to 99% lysis in 52% of the limbs, and <50% lysis in 17%. Overall, 1-year primary patency was 60%. Patency was higher in patients with iliofemoral DVT than in patients with femoropopliteal DVT (64% vs. 47%, P <.01). In addition, patients with acute symptoms (≤10 days) had a greater likelihood of complete lysis (34%) than patients with chronic symptoms (>10 days; 19%). Major bleeding occurred in 11%, but neurologic involvement and mortality were rare (both 0.4%). Adjunctive stent placement to treat residual stenosis and/or short segment occlusion was required in 103 limbs.


One small randomized trial and numerous other retrospective studies have demonstrated similar rates of thrombolysis, with some also showing improved valve preservation and quality of life. Combining thrombolysis with percutaneous thrombus fragmentation and extraction has the added benefit of decreasing the infusion time, the hospital stay, and the overall cost of treatment. These studies, as well as the current ACCP guidelines, suggest that catheter-directed thrombolysis (with adjunctive angioplasty, venous stenting, and pharmacomechanical fragmentation and extraction) may be useful in selected patients with extensive iliofemoral DVT. Patients should have a recent onset of symptoms (<14 days), good functional status, decent life expectancy, and low bleeding risk.


Inferior Vena Caval Filters


Since the introduction of the Kimray-Greenfield filter in the United States in 1973, numerous vena caval filters have been developed. Although the designs are variable, they all prevent pulmonary emboli, while allowing continuation of venous blood flow through the IVC. Early filters were placed surgically through the femoral vein. Currently, less invasive techniques allow percutaneous filter placement through a femoral vein, internal jugular vein, or small peripheral vein under fluoroscopic or ultrasound guidance. Complications associated with IVC filter placement include insertion site thrombosis, filter migration, erosion of the filter into the IVC wall, and IVC thrombosis. The rate of fatal complications is <0.12%.


Placement of an IVC filter is indicated for patients who develop recurrent DVT (significant propagation of the original thrombus or proximal DVT at a new site) or PE despite adequate anticoagulation therapy and for patients with pulmonary hypertension who experience recurrent PE. In patients who receive IVC filters for these indications, therapeutic anticoagulation should be continued. The duration of anticoagulation is determined by the underlying VTE and not by the presence of the IVC filter itself. Practically speaking, however, many patients who require an IVC filter for recurrent VTE are the same ones who would benefit most from indefinite anticoagulation. The other major indication for placement of an IVC filter is a contraindication to, or complication of, anticoagulation therapy in the presence of an acute proximal DVT. In patients who are not able to receive anticoagulants due to recent surgery or trauma, the clinician should continually reassess if antithrombotic agents may be started safely at a later date. Even some patients who develop anticoagulation-associated bleeding complications may be able to restart therapy at a lower intensity of anticoagulation later in the hospital course. As before, the clinical circumstances surrounding the VTE should determine the duration of anticoagulation.


Placement of permanent IVC filters has been evaluated as an adjunct to routine anticoagulation in patients with proximal DVT. In this study, routine IVC filter placement did not prolong early or late survival in patients with proximal DVT but did decrease the rate of PE (hazard ratio, 0.22; 95% confidence interval, 0.05 to 0.90). An increased rate of recurrent DVT was seen in patients with IVC filters (hazard ratio, 1.87; 95% confidence interval, 1.10 to 3.20). More controversial indications for IVC filter placement include prophylaxis against PE in patients receiving catheter-directed thrombolysis and in high-risk patients without established DVT or PE.


Hypercoagulability States


Certain patients seem to have a tendency to clot spontaneously. So-called hypercoagulability states were long thought to exist, but they were difficult to document except on clinical grounds. Currently, however, these clotting tendencies are better understood, thanks in large part to recognition of the role of antithrombins. If an antithrombin deficiency exists and clotting goes unchecked, activation of a clotting cascade could theoretically progress to clotting throughout the entire vasculature. Another important development was the recognition that deficiencies of certain natural clot-removing substances in the blood may lead to a clinical thrombotic tendency. Both types of deficiency can be either acquired or congenital.




When the etiology of a clotting episode is unclear, the family history should be reviewed for evidence of a congenital disorder. Even if the history is negative, the patient should be screened for both acquired and congenital disorders (table 6).



Acquired Clotting Conditions


Screening for acquired clotting conditions [see Table 7] is based on the history, physical examination, and laboratory assessment. The history should include medications, diseases, and surgical procedures or other injuries.Examination may disclose causes of hypercoagulability. Soft tissue injury, for example, is a potent activator of the coagulation system. If the injury is severe enough, it may be capable of causing a severe acquired coagulopathy. The problem is usually obvious, but on occasion, detailed study may be necessary to identify tissue damage or ischemic injury to bowel or extremities. Hypovolemia”especially hypovolemic shock—markedly reduces clotting time: blood from a patient in profound shock may clot instantaneously in the syringe as it is being drawn. The breakdown of red cells in a hemolytic transfusion reaction can cause clotting. Severe infection, especially from gram-negative organisms, is a potent activator of coagulation.


Table 7. Etiology of Acquired Hypercoagulability



Of the acquired hypercoagulability syndromes, Trousseau syndrome is a particularly important condition for surgeons to recognize because it occurs in the surgical population (cancer patients) and must be treated with heparin (it is unresponsive to warfarin). It occurs when an adenocarcinoma secretes a protein recognized by the body as tissue factor, resulting in multiple episodes of venous thromboembolism over time (migratory thrombophlebitis). Simple depletion of vitamin K-dependent factors is ineffective. Patients should receive therapeutic-dose heparin indefinitely or until the cancer is brought into remission.

Laboratory screening may facilitate diagnosis. A complete blood count may document the presence of polycythemia or leukemia. Thrombocythemia may be a manifestation of a hypercoagulable disorder, and thrombocytopenia after the administration of heparin raises the possibility of intravascular platelet aggregation. A prolonged aPTT is suggestive of lupuslike anticoagulant. Increased levels of D-dimers, fibrin degradation products (FDPs), or fibrin monomers in the plasma may reflect low-grade intravascular coagulation.


Congenital Clotting Conditions


Congenital clotting tendencies can result from deficiencies in inhibitors of thrombosis (antithrombin, proteins C and S, and possibly heparin cofactor II), dysfibrinogenemias, or dysfibrinolysis [see Table 8]. Most congenital clotting defects are transmitted as an autosomal dominant trait. A negative family history does not preclude inherited thrombophilia, because the defects have a low penetrance, and fresh mutations may have occurred.

Table 8. Congenital Clotting Disorders





Initial evaluation of a patient with an unexplained thrombotic episode should be directed at the most common causes of hypercoagulability. Acquired causes of clotting are more commonly seen by surgeons than congenital causes and therefore must be excluded first. If a clotting disorder is determined to be congenital, a laboratory workup should be undertaken. Several of the relevant assays ”specifically, the functional assaysâ”should be performed after the acute phase of the disorder has passed. If they are performed during the acute phase, levels of several antithrombotics (e.g., antithrombin and proteins C and S) will be misleadingly low not because deficiencies of these substances caused the underlying thrombotic process but because they were consumed in that process.


Specific Causes Of Thrombotic Tendency


The most common congenital causes of accelerated clotting are mutations of prothrombin (prothrombin G20210A mutation) and factor V (Leiden mutation, or activated protein C resistance).The prevalence of each of these ranges from 1% to 5% in the general population and may be much higher in specific ethnic subpopulations.1 Each mutation may be identified conclusively by means of polymerase chain reaction (PCR) techniques. Detection of these mutations, unlike assays for antithrombin and proteins C and S, is not dependent on the patient's current inflammatory state. It must be remembered that the presence of one of these mutations, especially in the heterozygous form, does not imply that it is the sole cause of thrombosis. In many patients, a second precipitating factor must be present for the pathologic genetic thrombotic potential to be manifested.


Prothrombin G20210A Mutation


The prothrombin G20210A mutation is known to involve a single amino acid substitution in the prothrombin gene, but precisely how this increases the risk of venous thromboembolism is unclear. The one apparent manifestation of the mutation is a 15% to 40% increase in circulating prothrombin. Regardless of the mechanism at work, patients who are at least heterozygous for the trait are at two- to sixfold greater risk for venous thromboembolism than those without the mutation.


Resistance to Activated Protein C (Factor V Leiden)


Resistance of human clotting factors to inactivation by activated protein C is believed to be the most common inherited procoagulant disorder.114 Normally, activated factor V is degraded by activated protein C in the presence of membrane surface as part of normal regulation of thrombosis. Activated protein C resistance is caused by a single substitution mutation in the factor V gene, which is passed in an autosomal dominant fashion. The mutant factor V that results, termed factor V Leiden, is resistant to inactivation by activated protein C and thus has a greater ability to activate thrombin and accelerate clotting.

Two techniques are commonly used to diagnose this disorder. The first is a functional assay that compares a standard aPTT to one performed in the presence of exogenous activated protein C. If the latter aPTT does not exhibit significant prolongation, the patient is probably resistant to activated protein C. The results of this assay must be interpreted with caution if the patient is still in the acute phase of the illness. The second technique, which is more reliable, involves direct detection of the mutation via PCR analysis of DNA.


Antithrombin Deficiency


Antithrombin (once termed antithrombin III) is a 65 kd protein that decelerates the coagulation system by inactivating activated factors—primarily factor Xa and thrombin but also factors XII, XI, and IX. Antithrombin therefore acts as a scavenger of activated clotting factors. Its activity is enhanced 100-fold by the presence of heparans on the endothelial surface and 1,000-fold by administration of exogenous heparin.

Congenital antithrombin deficiency occurs in approximately 0.01% to 0.05% of the general population and 2% to 4% of patients with venous thrombosis. The trait is passed on as an autosomal dominant trait, with the heterozygous genotype being incompatible with life. Antithrombin-deficient patients are at increased risk for thromboembolism when their antithrombin activity falls below 70% of normal.

Patients with congenital antithrombin deficiency frequently present after a stressful event. They usually have DVT but sometimes have PE. If anticoagulation is not contraindicated, the treatment of choice is heparin at a dosage sufficient to raise the aPTT to the desired level, followed by warfarin. If anticoagulation is contraindicated (as it is during the peripartum period), antithrombin concentrate should be given to raise the antithrombin activity to 80% to 120% of normal during the period when anticoagulants cannot be given.

Acquired antithrombin deficiency is a well-recognized entity. In most patients undergoing severe systemic stress, antithrombin levels fall below normal. Patients with classic risk factors for venous thromboembolism tend to have the lowest levels.


Protein C and Protein S Deficiency


Protein C is a 62 kd glycoprotein with a half-life of 6 hours. Because it is vitamin K dependent, a deficiency will develop in the absence of vitamin K. Acquired protein C deficiency is seen in liver disease, malignancy, infection, the postoperative state, and disseminated intravascular coagulation. Protein C deficiency occurs in approximately 4% to 5% of patients younger than 40 to 45 years who present with unexplained venous thrombosis. It is transmitted as an autosomal dominant trait, and the family history is usually positive for a clotting tendency. Protein C levels range from 70% to 164% of normal in patients without a clotting tendency; levels below 70% of normal are associated with a thrombotic tendency. The most appropriate tests for screening are functional assays; there are cases of dysfunctional protein C deficiency in which protein C antigen levels are normal but protein C activity is low, and these would not be detected by the usual immunoassays.

Protein S is a vitamin K-dependent protein that acts as a cofactor for activated protein C by enhancing protein C-induced inactivation of activated factor V. The incidence of protein S deficiency is similar to that of protein C deficiency. It is transmitted as a dominant trait, and the family history is often positive for a thrombotic tendency.




Although hyperhomocysteinemia is more commonly associated with cardiac disease and arterial thrombosis, it may also be associated with an increased incidence of venous thromboembolism. This association is not as strong as those already discussed. Accordingly, anticoagulation of asymptomatic patients with elevated homocysteine levels is not currently recommended.




More than 100 qualitative abnormalities of fibrinogen (dysfibrinogenemias) have been reported. Dysfibrinogenemias are inherited in an autosomal dominant manner, with most patients being heterozygous. Most patients with dysfibrinogenemia have either no clinical symptoms or symptoms of a bleeding disorder; a minority (about 11%) have clinical features of a recurrent thromboembolic disorder. Congenital dysfibrinogenemias associated with thrombosis account for about 1% of cases of unexplained venous thrombosis occurring in young people. The most commonly observed functional defect in such dysfibrinogenemias is abnormal fibrin monomer polymerization combined with resistance to fibrinolysis. Decreased binding of plasminogen and increased resistance to lysis by plasmin have been noted.

In addition to a prolonged TT, patients who have dysfibrinogenemia associated with thromboembolism may have a prolonged INR. The diagnosis is confirmed if the reptilase time is also prolonged. Measured with clotting techniques, fibrinogen levels may be slightly or moderately low; measured immunologically, levels may be normal or even increased.




Fibrinolysis can be impaired by inherited deficiencies of plasminogen, defective release of t-PA from the vascular endothelium, and high plasma levels of regulatory proteins (e.g., t-PA inhibitors). In addition, factor XII (contact factor) deficiency may induce failure of fibrinolysis activation.

Inherited plasminogen deficiency is probably only rarely responsible for unexplained DVT in young patients. It is transmitted as an autosomal dominant trait. In heterozygous persons with a thrombotic tendency, plasminogen activity is about one half normal (3.9 to 8.4 µmol/ml). The euglobulin clot lysis time is prolonged. Functional assays should be carried out, and there should be full transformation of plasminogen into plasmin activators.

The important role of t-PA inhibitors I and II in the regulation of fibrinolysis is well defined.In normal plasma, t-PA inhibitor I is the primary inhibitor for both t-PA and urokinase. Release of t-PA inhibitor I by platelets results in locally increased concentrations where platelets accumulate. The ensuing local inhibition of fibrinolysis may help stabilize the hemostatic plug. t-PA inhibitor II is present in and secreted by monocytes and macrophages.

Factor XII deficiency is a rare cause of impaired fibrinolysis. Initial contact activation of factor XII not only results in activation of the clotting cascade and of the inflammatory response but also leads to plasmin generation. This intrinsic activation of fibrinolysis requires factor XII, prekallikrein, and high-molecular-weight kininogen. Patients with factor XII deficiencies can be identified by a prolonged aPTT in the absence of clinical bleeding.


Operative Venous Thrombectomy


In patients with acute iliofemoral DVT, surgical therapy is generally reserved for patients who worsen with anticoagulation therapy and those with phlegmasia cerulea dolens and impending venous gangrene. If the patient has phlegmasia cerulea dolens, a fasciotomy of the calf compartments is first performed. In iliofemoral DVT, a longitudinal venotomy is made in the common femoral vein and a venous balloon embolectomy catheter is passed through the thrombus into the IVC and pulled back several times until no further thrombus can be extracted. The distal thrombus in the leg is removed by manual pressure beginning in the foot. This is accomplished by application of a tight rubber elastic wrap beginning at the foot and extending to the thigh. If the thrombus in the femoral vein is old and cannot be extracted, the vein is ligated. For a thrombus that extends into the IVC, the IVC is exposed transperitoneally and the IVC is controlled below the renal veins. The IVC is opened and the thrombus is removed by gentle massage. An intraoperative completion venogram is obtained to determine if any residual thrombus or stenosis is present. If a residual iliac vein stenosis is present, intraoperative angioplasty and stenting can be performed. In most cases, an arteriovenous fistula is then created by anastomosing the great saphenous vein (GSV) end to side with the superficial femoral artery in an effort to maintain patency of the thrombectomized iliofemoral venous segment. Heparin is administered postoperatively for several days. Warfarin anticoagulation is maintained for at least 6 months after thrombectomy. Complications of iliofemoral thrombectomy include PE in up to 20% of patients and death in <1% of patients.

One study followed 77 limbs for a mean of 8.5 years after thrombectomy for acute iliofemoral DVT. In limbs with successful thrombectomies, valvular competence in the thrombectomized venous segment was 80% at 5 years and 56% at 10 years. More than 90% of patients had minimal or no symptoms of postthrombotic syndrome. There were 12 (16%) early thrombectomy failures. Patients were required to wear compression stockings for at least 1 year after thrombectomy.

Survival rates for surgical pulmonary embolectomy have improved over the past 20 years with the addition of cardiopulmonary bypass. Emergency pulmonary embolectomy for acute PE is rarely indicated. Patients with preterminal massive PE (Fig. 5) for whom thrombolysis has failed or who have contraindications to thrombolytics may be candidates for this procedure. Open pulmonary artery embolectomy is performed through a posterolateral thoracotomy with direct visualization of the pulmonary arteries. Mortality rates range between 20 and 40%.



Fig. 5. Autopsy specimen showing a massive pulmonary embolism.


Percutaneous catheter-based techniques for removal of a PE involve mechanical thrombus fragmentation or embolectomy using suction devices. Mechanical clot fragmentation is followed by catheter-directed thrombolysis. Results of catheter-based fragmentation are based on small case series. In a study in which a fragmentation device was used in 10 patients with acute massive PE, fragmentation was successful in 7 patients with a mortality rate of 20%.65 Transvenous catheter pulmonary suction embolectomy has also been performed for acute massive PE with a reported 76% successful extraction rate and a 30-day survival of 70%.66




Patients who undergo major general surgical, gynecologic, urologic, and neurosurgical procedures without thromboprophylaxis have a significant incidence of perioperative DVT (15 to 40%). The incidence is even higher with major trauma (40 to 80%), hip and knee replacement surgery (40 to 60%), and spinal cord injury (60 to 80%). The goal of prophylaxis is to reduce the mortality and morbidity associated with VTE. The first manifestation of VTE may be a life-threatening PE (Fig. 6), and as indicated earlier, clinical evaluation to detect DVT before PE is unreliable.


Fig. 6. Pulmonary angiogram showing a pulmonary embolism (arrow).


Effective methods of VTE prophylaxis involve the use of one or more pharmacologic or mechanical modalities. Currently available pharmacologic agents include low-dose UFH, LMWH, synthetic pentasaccharides, and vitamin K antagonists. Mechanical methods include intermittent pneumatic compression (IPC) and graduated compression stockings. Aspirin therapy alone is not adequate for DVT prophylaxis. These prophylaxis methods vary with regard to their efficacy, and the 2008 ACCP Clinical Practice Guidelines stratify their uses according to the patient's level of risk (Table 6).

Table 6. Thromboembolism Risk and Recommended Thromboprophylaxis in Surgical Patients




Complications after venous thrombosis can vary from life threatening to chronically debilitating. Pulmonary embolism develops as venous thrombi break off from their location of origin and travel through the right heart and into the pulmonary artery, causing a ventilation perfusion defect and cardiac strain. PE occurs in approximately 10% of patients with acute deep venous thrombosis and can cause up to 10% of in hospital deaths.60,61 However, most patients (up to 75%) are asymptomatic. Traditionally, proximal venous thrombosis are thought to be at highest risk for causing pulmonary emboli; however, the single largest autopsy series ever performed to specifically to look for the source of fatal PE was performed by Havig in 1977, who found that one third of the fatal emboli arose directly from the calf veins.6





More than 600,000 cases of venous thromboembolism are estimated to occur each year in the United States. Pulmonary embolism (PE) complicates approximately 50% of cases of untreated proximal deep venous thrombosis (DVT) and contributes to 10-15% of all hospital deaths. Less frequent manifestations of venous thrombosis include phlegmasia alba dolens, phlegmasia cerulea dolens (PCD), and venous gangrene. These form a clinical spectrum of the same disorder. All 3 manifestations result from acute massive venous thrombosis and obstruction of the venous drainage of an extremity


History of the Procedure


In the 16th century, Fabricius Hildanus first described the clinical syndrome of what is currently called PCD. In 1938, Gregoire made an outstanding description of the condition and used the term PCD to differentiate ischemia-associated massive venous thrombosis from phlegmasia alba dolens, which describes fulminant venous thrombosis without ischemia.1 The exact incidence of these disorders is not well reported.


In 1939, Leriche and Geissendorfer performed the first thrombectomy for cases of PCD.2 Historically, surgical thrombectomy has been the procedure of choice for PCD refractory to medical therapy and in patients with established or impeding gangrene.




More than 600,000 cases of venous thromboembolism are estimated to occur annually in the United States. Phlegmasia alba dolens, PCD, and venous gangrene occur at any age but are more common during the fifth and sixth decades of life. Incidence is higher in females than in males.




The main causative factor in phlegmasia is massive thrombosis and occlusion of major venous channels with significantly compromised venous outflow. Multiple triggering factors exist. Malignancy is the most common triggering factor and is present in approximately 20-40% of patients with PCD. Other associated risk factors include hypercoagulable syndrome, surgery, trauma, ulcerative colitis, gastroenteritis, heart failure, mitral valve stenosis, vena caval filter insertion, and May-Thurner syndrome (compression of the left iliac vein by the right iliac artery). Pregnancy has often been associated with phlegmasia alba dolens, especially during the third trimester when the uterus is large enough to compress the left common iliac vein against the pelvic rim (ie, milk leg syndrome). Finally, 10% of patients with phlegmasia have no apparent risk factors.




In phlegmasia alba dolens, the thrombosis involves only major deep venous channels of the extremity, therefore sparing collateral veins. The venous drainage is decreased but still present; the lack of venous congestion differentiates this entity from PCD.

In PCD, the thrombosis extends to collateral veins, resulting in venous congestions with massive fluid sequestration and more significant edema. Without established gangrene, these phases are reversible if proper measures are taken.


Of PCD cases, 40-60% also have capillary involvement, which results in irreversible venous gangrene that involves the skin, subcutaneous tissue, or muscle. Under these conditions, the hydrostatic pressure in arterial and venous capillaries exceeds the oncotic pressure, causing fluid sequestration in the interstitium. Venous pressure may increase rapidly, as much as 16- to 17-fold within 6 hours. Fluid sequestration may reach 6-10 L in the affected extremity within days. Circulatory shock, which is present in about one third of patients, and arterial insufficiency may ensue.


The exact mechanism for the compromised arterial circulation is debatable but may involve shock, increased venous outflow resistance, and collapse of arterioles due to increased interstitial pressure. Vasospasm of the resistance vessels has also been hypothesized but has never been observed experimentally or radiographically.




In the lower extremities, left-sided involvement is more common by a 3:1 or 4:1 ratio. Involvement of upper extremities occurs in less than 5% of patients with PCD. Manifestations may be gradual or fulminant. Of PCD cases, 50-60% are preceded by phlegmasia alba dolens, with symptoms of edema, pain, and blanching (alba) without cyanosis. The blanching, which previously was thought to be caused by arterial vasospasm, is caused by subcutaneous edema, without venous congestion.


Patients with PCD present with the clinical triad of edema, agonizing pain, and cyanosis. Massive fluid sequestration may lead to bleb and bullae formation. The pain is constant, usually starting at the femoral triangle and then progressing to the entire extremity. Cyanosis is the pathognomonic finding of PCD, progressing from distal to proximal areas.


When venous gangrene occurs, it has a similar distribution with the cyanosis. Arterial pulses may be present when the venous gangrene is superficial; however, gangrene that involves the muscular compartment may result in increased compartment pressures and a pulse deficit. In addition, the pulses may be difficult to appreciate because of the significant edema. Various degrees of shock may be present because of significant fluid loss.




Historically, surgical thrombectomy has been the procedure of choice for phlegmasia cerulea dolens (PCD) refractory to medical therapy and for patients with established or impeding gangrene. The standard treatment of phlegmasia and venous gangrene is evolving, but most clinicians attempt endovascular approaches to thrombolysis, if possible.


Relevant Anatomy


The main causative factor in phlegmasia is massive deep venous thrombosis (DVT) and occlusion of major venous channels with significantly compromised venous outflow. In phlegmasia alba dolens, the thrombosis involves only major deep venous channels of the extremity, therefore sparing collateral veins and preserving some venous outflow from the limb. In phlegmasia cerulea dolens (PCD), the thrombosis extends to collateral veins, with complete obstruction of venous outflow, resulting in massive venous congestion and fluid sequestration and more significant edema.




For phlegmasia alba dolens and mild nongangrenous forms of phlegmasia cerulea dolens (PCD), conservative medical treatment, rather than surgical thrombectomy, should be the initial course of therapy. Thrombolysis may be initiated if conservative management does not elicit a response and if the patient has no contraindication to lytic therapy.


Surgical thrombectomy cannot open the small venules that are affected in venous gangrene, and it does not prevent valvular incompetence or postphlebitic syndrome. For these reasons, thrombolysis seems to be an attractive alternative in the management of PCD and venous gangrene.


Imaging Studies


The diagnoses of phlegmasia alba dolens, phlegmasia cerulea dolens (PCD), and venous gangrene are established mainly via clinical criteria with the assistance of contrast venography and duplex ultrasonography.

Although venography is considered the criterion standard for diagnosis, technical difficulties may be encountered in as many as 20-25% of patients. Attempts to perform ascending venography when extensive deep system thrombosis is present may result in nonvisualization of the deep system and a nondiagnostic study result. In these cases, descending venography via the contralateral femoral vein or via the upper extremity veins may provide more information about the iliocaval system and proximal extent of the thrombus.

Recent improvements in ultrasonography have made this modality a more reliable and accurate way to assess for proximal deep venous thrombosis (DVT) with less morbidity. In addition, duplex imaging may be repeated as needed to monitor for thrombus propagation. Ultrasonography can also be performed at the bedside in patients who are critically ill or unstable. Ultrasonography is often used to guide the initial venipuncture for diagnostic venography and initiation of thrombolytic therapy.

Magnetic resonance venography (MRV) is an evolving modality of diagnostic imaging. Its principal advantage is its ability to easily reveal the proximal and distal extent of thrombus with a single study. Its principal disadvantage is the inability to image acutely ill patients with hemodynamic instability or motion artifacts due to pain.


Medical Therapy


The standard treatment of phlegmasia and venous gangrene is still evolving. The optimal therapeutic modality remains under debate. So far, the results of treatment have been moderately successful. For phlegmasia alba dolens and mild nongangrenous forms of phlegmasia cerulea dolens (PCD), conservative medical treatment, such as steep limb elevation, anticoagulation with intravenous administration of heparin, and fluid resuscitation, should be the initial course of therapy.


Initiate heparin administration with an intravenous bolus of 80-100 U/kg, followed by a continuous infusion of 15-18 U/kg/h. Frequently monitor the activated partial thromboplastin time (aPTT), with a goal range of 2-2.5 times the laboratory reference range. Frequently monitor platelet counts to allow the early detection of heparin-induced thrombocytopenia.


The purpose of rapid heparin anticoagulation is to decrease the risk of proximal clot propagation or thromboembolism. Heparin does not directly affect limb swelling. The best nonsurgical method to decrease edema is steep leg elevation.


Recent studies have demonstrated that low molecular weight heparins are safe and effective in the treatment of proximal deep venous thrombosis (DVT) and pulmonary embolism (PE); however, no good evidence supports the use of these newer agents in phlegmasia and venous gangrene.


If heparin-induced thrombocytopenia occurs, immediately discontinue the use of heparin and replace it with an alternative anticoagulant. Danaparoid and lepirudin are effective alternative agents; however, heparin-associated antibodies exhibit a 10-19% cross-reactivity with danaparoid. Thus, perform cross-reactivity testing before the initiation of danaparoid in patients with these antibodies. Lepirudin is a direct thrombin inhibitor that does not demonstrate any cross-reactivity. The recommended dosage of lepirudin in patients without renal failure is 0.4 mg/kg as an intravenous bolus followed by a continuous infusion of 0.15 mg/kg/h. Use aPTT to monitor therapy, with a goal range of 2-2.5 times the laboratory reference range.


Continue long-term anticoagulation with warfarin (or other coumarin derivatives) for at least 6 months. Life-long anticoagulation is recommended in patients with hypercoagulable states.


Patients should wear long-term prescription compression stockings with at least 30-40 mm Hg of graded pressure. Many physicians erroneously have the patient fitted for a prescription stocking while the limb is still severely edematous. Instead, the patient may use nonprescription stockings or an elastic bandage, in combination with elevation, to minimize edema prior to being fit for a prescription stocking.


Surgical Therapy


Surgical thrombectomy performed through a femoral venotomy allows instant decompression of the venous hypertension. An intraoperative Trendelenburg position may be used to decrease the risk of PE. Transabdominal cavotomy and thrombectomy is an alternative approach that permits better control of the cava above the thrombus and, thus, provides protection against PE. Procedures that have been performed in an effort to decrease the rethrombosis rate include  cross-pubic vein-to-vein reconstruction with polytetrafluoroethylene (PTFE) or the greater saphenous vein (GSV) or  the creation of an arteriovenous fistula between the femoral artery and the GSV. These adjuvant procedures may be especially beneficial in cases that involve proximal iliofemoral vein constriction, damage, or external compression.


Concomitant administration of heparin and long-term anticoagulation are mandatory. Regardless, thrombectomy in patients with PCD is associated with a high rate of rethrombosis. Surgical thrombectomy cannot open the small venules that are affected in venous gangrene, and it does not prevent valvular incompetence or postphlebitic syndrome. The incidence of postphlebitic syndrome may be as high as 94% among survivors.


For the above reasons, thrombolysis seems to be an attractive alternative in the management of PCD and venous gangrene. In 1970, Paquet was the first to use thrombolysis for the treatment of PCD.4 Some authors propose catheter-directed thrombolysis directly into the vein with high doses of urokinase or tissue plasminogen activator (t-PA). Other authors support the method of intra-arterial low-dose thrombolysis via the common femoral artery, reasoning that the arterial route delivers the thrombolytic agent to the arterial capillaries and, subsequently, to the venules. The intra-arterial approach seems to be more effective in cases with venous gangrene. Systemic thrombolysis has also been used. Many authors have strongly recommended the insertion of a vena caval filter prior to initiation of thrombolytic therapy. Combine thrombolysis with heparin administration and long-term oral anticoagulation.


Fasciotomy alone or in conjunction with thrombectomy or thrombolysis reduces compartmental pressures; however, it significantly increases morbidity because of the prolonged wound healing and the risk of infection.


Finally, if all efforts fail and amputation is required, delay the procedure as long as possible. Take all precautions to reduce edema, allow venous channels to recanalize, and allow necrotic tissue to demarcate.


Preoperative Details


Patients who require emergent venous thrombectomy should have heparin continued throughout the perioperative period. Banked red blood cells should be available. The proximal extent of the thrombosis must be defined using a combination of venous ultrasonography for infrainguinal veins and retrograde venography of the iliac veins and inferior vena cava using a jugular or contralateral femoral approach. If the thrombosis extends into the iliac veins and vena cava, preparations should be made to control the cava via a right retroperitoneal incision. A high-quality fluoroscopy unit should be available to aid in catheter manipulation and completion venography.


Intraoperative Details


Operative exposure depends on the proximal and distal extent of the thrombus. The involved veins should be controlled proximally and distally prior to venotomy.


Iliac venous thrombectomy should be performed with large-bore thrombectomy balloon catheters (as large as 10F). Extension of the thrombus into the inferior vena cava may require proximal control of the cava. The anesthesiologist should apply positive airway pressure while the thrombus is being extracted from the iliocaval system. Digital subtraction venography should be performed to ascertain completeness of the thrombectomy.

Infrainguinal extraction of the thrombus is aided by the intraoperative placement of an Esmarch bandage from foot to thigh. Thrombectomy balloon catheters (3F and 4F) are passed through the femoral veins and, possibly, the posterior tibial veins as well.

Thrombolytic agents may be administered intraoperatively through the posterior tibial veins. t-PA is the most commonly used agent and may be administered intraoperatively.

After the thrombectomy is performed, an arteriovenous fistula should be constructed, connecting the proximal greater saphenous vein or one of its larger tributaries to the superficial femoral artery in an end-to-side fashion.

Completion venography should be performed to exclude the presence of residual thrombus or proximal venous stenosis. If one is present, balloon angioplasty with or without stent placement may be necessary.

When percutaneous endovascular therapy is performed as a single treatment modality, and many centers are now reporting this as a first-line therapy, the popliteal veins are usually accessed with duplex ultrasonography as an aid. Prone positioning is rarely necessary. If extensive thrombus is present, access via the posterior tibial vein is usually successful. A 6-F sheath is usually adequate. An infusion wire is passed through the thrombus just to its proximal extent, often into the vena cava. Infusion is usually performed in the most proximal segment first, usually in the iliac veins.

A common protocol is to infuse tPA (1 mg/hr) through the infusion wire as well as through the sheath for 24 hours, then to change the sheath perfusion to lower dose heparin after 24 hours. The infusion is then performed in the superficial femoral and popliteal vein segments. Clinical improvement is often noted with clearing of the profunda venous segment. Performance of simultaneous percutaneous mechanical thrombectomy is controversial and may not give better results than postprocedure balloon dilation.


Postoperative Details


Intravenous heparin is administered throughout the postoperative period to prolong the aPTT (2-2.5 times the reference range for aPTT). This is continued until the patient is adequately anticoagulated with warfarin or one of the coumarin derivatives (international normalized ratio [INR] range of 2.0-3.0). The optimal duration of oral anticoagulation is not established.

A sequential compression device is also placed, or, at a minimum, an ace bandage is placed for control of edema. Once the edema is at its minimum, the patient may be fitted for a thigh-length compression stocking. Ambulation is encouraged, if the patient is able.





The incidence of postphlebitic syndrome may be as high as 94% among survivors. Pulmonary embolism is common, and prophylactic placement of an inferior vena cava filter is recommended in most cases. Thrombectomy in patients with phlegmasia cerulea dolens (PCD) is associated with a high rate of rethrombosis. Amputation and death are common.



Outcome and Prognosis


Despite all of the therapeutic modalities described above, phlegmasia cerulea dolens (PCD) and venous gangrene still remain life-threatening and limb-threatening conditions with overall mortality rates of 20-40%. Pulmonary embolism (PE) is responsible for 30% of the deaths reported from PCD. Overall, amputation rates of 12-50% have been reported among survivors. The postphlebitic sequelae are apparent in 60-94% of survivors. Strict adherence to the use of long-term compression stockings helps to control chronic edema.


Future and Controversies


Phlegmasia alba dolens, phlegmasia cerulea dolens (PCD), and venous gangrene still remain a challenge to the vascular surgeon. Treatment modalities continue to evolve. Endovascular management may offer hope of successful and more effective management, with less morbidity, than traditional surgery. The role of mechanical thrombectomy, compared with thrombolysis, is unclear. Small numbers of patients and lack of randomized trials preclude clear recommendations.