The burden of venous thromboembolic disease continues to increase despite increased awareness of risk factors and prevention options. The incidence of a first episode of venous thromboembolism (VTE) is approximately 1 to 2 per 1000 person/years.
VTE most commonly manifests as lower extremity deep venous thrombosis (DVT) or pulmonary embolism (PE). Many nosocomial VTEs are preventable, although thromboprophylaxis continues to be underused. D-dimer testing and imaging for VTE diagnosis should be used within the context of appropriate clinical algorithms.
Opinions differ regarding the relevance of thrombophilia testing, and results usually do not affect the length of anticoagulation. The landscape of treatment options continues to evolve.
Alterations in three primary mechanisms of thrombosis predispose persons to VTE. Described by the German pathologist Rudolph Virchow more than 150 years ago, reduced or otherwise turbulent blood flow, alterations or injury to the vessel wall, and changes in blood components that are prothrombotic or that inhibit fibrinolysis (or both) compose the Virchow triad. VTE usually develops as a result of the synergistic effect of multiple risk factors, which may be inherited, acquired, or a combination of both.
One aspect of the Virchow triad is blood hypercoagulability, or thrombophilia. Thrombophilia can be inherited or acquired. Thrombophilia testing should not be routinely pursued in all patients presenting with DVT or PE. Although guidelines differ, most experts agree that thrombophilia evaluation should be considered only in certain populations, including patients with thromboses at unusual sites or recurrent idiopathic thrombosis, patients younger than 45 years with unprovoked thrombosis, patients with a clear family history of thrombosis in one or more first-degree relatives, and patients with warfarin-induced skin necrosis.
Additionally, many variables can affect outcomes of thrombophilia testing, including acute thrombosis and anticoagulant use, which may lead to false-positive test results. For this reason, and because a known thrombophilia will not change immediate management, testing should not be pursued in the acute setting. Asymptomatic patients with a family history of thrombosis should not undergo thrombophilia testing.
Inherited thrombophilias typically affect components of the coagulation cascade (see Figure 18 in Bleeding Disorders) that keep the hemostatic system in balance, either causing the prothrombotic system to continue unsuppressed or inhibiting clot lysis. All known inherited thrombophilias are autosomal dominant, meaning that most affected patients are heterozygous for the disorder. The two most common inherited causes are factor V Leiden and prothrombin G20210A gene mutation. Less common mutations involve antithrombin deficiency and protein C and S deficiency, although these latter disorders seem to be more significant risk factors for VTE.
Failure to identify a thrombophilia does not mean a thrombophilia does not exist. Studies have shown that even when an inherited disorder is not identified, a family history of thrombosis remains an independent risk factor for VTE. It is also possible for two thrombophilic defects to coexist, such as protein S deficiency and factor V Leiden.
Although identification of the inherited thrombophilias has advanced our understanding of the pathophysiology of VTE, it has had less influence on clinical management. The acute management of patients with VTE does not differ based on the presence of an inherited thrombophilia. Management duration is typically determined by whether the VTE event was provoked by a reversible or self-limited insult and is not often influenced by the presence of an underlying inherited thrombophilia, especially the more common disorders. Even if a patient with VTE is found to have an inherited thrombophilia, no evidence indicates asymptomatic family members should be screened to determine whether they also have the mutation.
Factor V Leiden is the most common inherited thrombophilia. When factor V is activated, it combines with factor X to produce thrombin, which leads to clot formation. This process is regulated by activated protein C, which inactivates factor V to stop the process of ongoing clot formation. Factor V Leiden is resistant to cleavage by activated protein C, leading to predisposition of thrombus formation. Although persons who are heterozygous are at a fourfold to eightfold increased risk for developing a first VTE, most remain asymptomatic. Heterozygous factor V Leiden is found in about 5% of Whites, whereas the homozygous form is found in less than 1%. Factor V Leiden is rare in Asian, African, African American, and Native American populations. It does not appear to be associated with arterial thrombosis. Factor V Leiden genetic testing or activated protein C resistance testing can be used to diagnose this condition.
The prothrombin G20210A gene mutation occurs in approximately 2% of Whites and 0.5% of Blacks and causes increased production of prothrombin (factor II) through a mutation at nucleotide 20210 from guanine to adenine. Persons with this mutation are at a twofold to fourfold increased risk for developing a first VTE, although, as with factor V Leiden, most patients with this mutation do not experience VTE events. Data are unclear regarding risks with the homozygous state, which is rare.
Antithrombin III (ATIII) and proteins C and S serve as natural anticoagulants in the body. Mutations that lead to loss of function of these components contribute to a tendency to develop VTE.
ATIII deficiency, although rare, with a prevalence of 1 in 3000 to 5000 persons, is a more significant thrombophilic risk factor than factor V Leiden or the prothrombin G20210A gene mutation. The main role of ATIII is to inhibit thrombin and activated factors IX and X (IXa and Xa). VTE-related pregnancy loss and pregnancy morbidity is common. Acquired ATIII deficiency is much more common than the congenital version (Table 27), and repeat testing is typically required to determine whether the deficiency is persistent.
For patients in whom heparin is initiated and titration to a therapeutic range is difficult, ATIII deficiency should be considered because heparin requires ATIII to be effective. ATIII concentrate can be used to treat this condition.
Protein C is a vitamin K–dependent protein that degrades activated factors V and VIII. Heterozygous protein C deficiency is uncommon, with a prevalence of 2 to 5 per 1000 persons. Many persons with this deficiency will experience a thrombotic event or pregnancy morbidity before 50 years of age, with a strong family history of thrombosis. Patients can also develop warfarin-induced skin necrosis because of further rapid depletion of protein C, which proceeds more rapidly than depletion of the coagulation factors. Homozygous deficiency is rare and causes neonatal purpura fulminans. If protein C deficiency is found, acquired causes should be ruled out (see Table 27). Repeat testing is often necessary to confirm a hereditary deficiency. Patients should not be tested during acute VTE events or while receiving warfarin. Protein C functional testing can be ordered to evaluate for evidence of deficiency.
Protein S is a cofactor for protein C to degrade activated factors V and VIII. Deficiency is uncommon and bears many similarities to protein C deficiency. Patients who are heterozygous for protein S deficiency typically experience VTE at a younger age (<50 years). Protein S is a vitamin K–dependent factor synthesized by the liver; it circulates in a free form and bound to a complement-binding protein. Although rare case reports show patients with a functional protein S deficiency, immunoassay of the free form of protein S is usually sufficient to make the diagnosis. Protein S deficiency is likely the most difficult hereditary thrombophilia to confirm because multiple laboratory assays for protein S are available, with cutoffs between normal and deficient that may be imprecise.
Methylene tetrahydrofolate reductase (MTHFR) gene polymorphisms cause mild elevations in homocysteine levels, which are associated with a mildly increased risk of cardiovascular and thrombotic disease. The heterozygous mutation is found in 20% of Whites and 2% of Blacks. Vitamin B6 and B12 supplementation can lower homocysteine levels without lowering thrombotic risk, which suggests the mutation may be a marker of thrombotic risk rather than a cause of thrombosis. Testing for the MTHFR mutation and measuring homocysteine levels should not be done in the evaluation of thrombophilia.
Factor VIII levels and plasminogen activator inhibitor activity should not be part of the standard thrombophilia evaluation because clinical trials regarding their importance have been inconclusive and results do not influence management.
VTE is more likely to occur in the setting of an acquired rather than an inherited thrombophilia. Many conditions predispose patients to the development of thrombosis.
Surgery, trauma, hospitalization, and immobilization are some of the most significant risk factors for VTE. VTE occurs frequently in medical and surgical patients. Approximately half of all new VTEs are diagnosed during or within 3 months of a hospital stay or surgical procedure. If prophylaxis is not used, the risk of DVT in the general surgical patient is 15% to 30%. In the orthopedic patient, the risk of DVT is approximately 60% after hip fracture surgery. Patients with cancer who undergo surgery and those undergoing orthopedic procedures, including knee arthroplasty, hip fracture repair, or hip replacement, are at particularly high risk. Nosocomial VTE risk is also increased for nonsurgical hospitalized patients, more so for immobilized patients, patients with acute neurologic illness, and patients in the medical ICU.
Certain medical conditions, including inflammatory conditions, nephrotic syndrome, and inflammatory bowel disease, have also been associated with increased thrombotic risk. In nephrotic syndrome, this risk is attributed to loss of antithrombin and proteins C and S in the urine. Obesity is also associated with increased thrombotic risk.
Thrombosis remains a leading cause of death in patients with cancer and is a significant source of morbidity. Increased thrombotic risk has been associated with numerous malignancies, including prostate and breast cancer.
Cancer is diagnosed in 10% of patients within 1 year of an unprovoked VTE occurrence. Cancer of the ovary, pancreas, and liver are most often found. The only randomized controlled clinical trial that compared routine age- and gender-indicated screening with extensive malignancy screening using CT of the thorax, abdomen, and pelvis showed that extensive malignancy screening provided no survival benefit. Extensive cancer screening should not be performed beyond recommendations for gender and age, independent of the VTE event.
In addition, other factors, such as hormonal therapy, can further increase risk.
Hormones used in oral contraceptives and in the treatment of menopause increase the risk of VTE. The risk in women using oral contraceptives is increased approximately threefold, but the absolute number of patients affected in this young healthy population remains small. VTE risk correlates with the specific progestin agent and is somewhat higher in oral contraceptives containing desogestrel and gestodene and somewhat lower with levonorgestrel. Injectable progestin agents do not increase the risk. Regardless of the type of contraceptive, VTE risk tends to be greater in women who have obesity and those who are older than 39 years. Women with a previous VTE event and a known inherited thrombophilia should not take oral contraceptives because the thrombotic risk is further increased. However, experts do not recommend routine thrombophilia screening before beginning contraceptive therapy because many women would need to be screened to prevent one adverse event from pulmonary embolism (PE). VTE risk is also increased by approximately twofold in menopausal women taking conjugated estrogen-medroxyprogesterone hormone replacement therapy, but the absolute risk remains small. The VTE risk in menopausal women seems lower in those taking estrogen only and in those using transdermal hormone replacement.
The antiestrogen, tamoxifen, also increases VTE risk in women with estrogen receptor-positive breast cancer, and the risk increases further, approximately three times baseline, in women receiving tamoxifen with systemic chemotherapy. The risk for VTE with aromatase inhibitors, such as anastrozole, is lower than that seen with tamoxifen.
Patients with multiple myeloma receiving thalidomide and its analogs as part of combination chemotherapy have a significant risk of VTE that warrants prophylaxis. The vascular endothelial growth factor inhibitor bevacizumab and newer multitargeted tyrosine kinase inhibitors, such as sunitinib and sorafenib, also increase VTE risk.
Glucocorticoid therapy has also been identified to increase the risk for VTE.
The antiphospholipid antibody syndrome is an autoimmune disorder in which thrombosis and fetal demise (in pregnancy) may occur. Patients with antiphospholipid antibody syndrome are at risk for arterial and venous thrombosis.
Antiphospholipid antibodies are the anticardiolipin antibodies and the lupus anticoagulant. The diagnosis of antiphospholipid antibody syndrome is based on the clinical criteria of thromboembolism or pregnancy morbidity and laboratory findings of medium or high titer antiphospholipid antibodies present on two or more occasions at least 12 weeks apart (Table 28). A clue to the presence of the lupus anticoagulant is activated partial thromboplastin time prolongation.
Typically, patients who are diagnosed with antiphospholipid antibody syndrome require long-term anticoagulation owing to the risk of recurrent thrombosis.
The myeloproliferative neoplasms have been found to carry a particularly increased risk of thrombosis; although these thromboses include PE and DVT, portal vein thrombosis and Budd-Chiari syndrome (hepatic venous outflow obstruction) (see MKSAP 18 Gastroenterology and Hepatology) are often found. Evidence of a myeloproliferative neoplasm is found in approximately 50% of patients with Budd-Chiari syndrome, even when the complete blood count is normal. In the setting of splanchnic vein thrombosis (which includes Budd-Chiari syndrome and portal vein thrombosis), evaluation for evidence of a myeloproliferative neoplasm should be considered, including evaluation for the JAK2 tyrosine kinase mutation.
Paroxysmal nocturnal hemoglobinuria is another acquired stem cell disorder associated with hemolytic anemia, bone marrow failure, and thrombosis (see Erythrocyte Disorders).
A previous VTE event is one of the most powerful predictors of a subsequent VTE, regardless of whether an additional inherited or acquired thrombophilic risk factor is identified.
All hospitalized patients should be assessed for the risk of developing a VTE and treated with appropriate prophylaxis (see Section mk18_b_gm_s17_4_1 MKSAP 18 General Internal Medicine) because VTE is a major preventable cause of hospital morbidity and mortality. Generally, unless a clear contraindication to prophylaxis exists, pharmacologic treatment is indicated as opposed to mechanical prophylaxis. In acutely ill patients with risk for thrombosis, low-molecular-weight heparin (LMWH) and fondaparinux are preferable to unfractionated heparin for prophylaxis. Patients with cancer or stroke and those in the ICU have a particularly high risk for VTE. Despite the well-recognized risks of VTE, the rate of appropriate prophylaxis remains low in general hospitalized patients. Most patients do not require continued pharmacologic VTE prevention after discharge. However, patients with cancer who are undergoing major surgical procedures should continue VTE prophylaxis with LMWH or unfractionated heparin for 7 to 10 days following the procedure. Patients undergoing major abdominal or pelvic cancer resection surgery may require continued VTE prophylaxis with LMWH for up to 4 weeks. Patients undergoing knee arthroplasty and those with hip fracture repair or hip replacement require VTE prophylaxis for as long as 4 weeks after discharge.
VTE prophylaxis with apixaban, rivaroxaban, or LMWH may also be indicated for outpatients who are beginning new cancer chemotherapy and who have a heightened VTE risk (Khorana score of 2 or higher); the Khorana score provides decision support in identifying these patients (Table 49). Patients with multiple myeloma who are receiving thalidomide- or lenalidomide-based regimens should receive VTE prophylaxis with either aspirin or LMWH.
DVT and PE cause significant morbidity and require efficient evaluation and diagnosis. Previous VTE, immobilization, and other thrombophilia risk factors, especially cancer, should be assessed. History pertinent to other potential causes of leg or respiratory symptoms should be elicited. The typical clinical presentation of DVT involves unilateral swelling, pain, warmth, and erythema of the extremity. Patients with PE may present with chest pain, dyspnea, and tachypnea. Less commonly, symptoms may include cough, fever, cyanosis, syncope, or shock.
CT angiography has significantly improved the accuracy of evaluating PE, generally replacing ventilation-perfusion scanning, which lacks specificity, and avoiding the need for more invasive pulmonary arteriography. However, the overuse of CT angiography and D-dimer measurement in outpatients at low risk for PE has needlessly exposed patients to the additional radiation and expense of these procedures. For patients who present with symptoms suspicious for an acute VTE, validated prediction rules have been developed that use D-dimer testing to help effectively evaluate this condition. The Wells criteria for diagnosis of DVT (Table 29) and PE (Table 30) and the Geneva Score (Table 31) for diagnosis of PE are well-studied tools in this setting. Based on these criteria, patients with low pretest probability and low D-dimer levels do not require imaging because a VTE diagnosis is unlikely. Recent studies suggest that a subset of patients at very low risk can be identified using the Pulmonary Embolism Rule-Out Criteria (PERC) (Table 32); these patients do not require D-dimer testing to eliminate the need for additional imaging. American College of Physicians guidelines published in 2015 recommend using the PERC as the initial step in evaluating patients at low risk. If the PERC score is zero, no D-dimer testing is needed, and no CT angiography should be performed. In a recent meta-analysis of 12 studies, it was found that if the PERC were applied, only 0.3% of PEs would have been missed, and 22% of D-dimer testing would have been safely avoided. PERC will help eliminate unnecessary D-dimer testing in patients at low risk.
In patients at low risk but who have a PERC score greater than zero, D-dimer testing should be pursued. If the result is negative, no imaging is warranted. If the result is positive, further evaluation is merited. If a patient has a moderate or high pretest probability, imaging studies are indicated. D-dimer testing may be considered in patients with moderate pretest probability of pulmonary embolism (~20%) but should not be pursued in patients with high pretest probability because results would not change the need for imaging.
Duplex ultrasonography is the imaging modality of choice for suspected DVT. Lower extremity DVT is considered proximal if the popliteal veins are involved and is considered distal if only the calf veins are involved. CT angiography is the study of choice for suspected PE. The American Society of Hematology (ASH) recommends a strategy of D-dimer measurement to exclude DVT in a population with low probability of DVT (≤10%) followed by proximal lower extremity ultrasonography or whole-leg ultrasonography for patients requiring additional testing. For patients with moderate probability of DVT (~25%), the guideline recommends beginning the evaluation with whole-leg ultrasonography without D-dimer testing. No additional testing is recommended if the whole-leg ultrasound is negative. If the initial negative ultrasound was limited to a proximal study, it should be followed by a second proximal leg ultrasonography 1 week later if no alternative diagnosis for the leg symptoms is identified.
In patients with kidney disease or in whom intravenous contrast is contraindicated, a ventilation-perfusion lung scan can be pursued. A normal ventilation-perfusion scan result effectively rules out PE, and a high probability study result in a patient with a high likelihood of disease has a strong positive predictive value. The sensitivity and specificity of low probability or intermediate probability study results may not be accurate enough to establish or rule out the diagnosis. MRI can visualize intraluminal filling defects in the pulmonary vasculature, but not as well as CT, and avoids the ionizing radiation of CT. New MRI techniques are being evaluated that may enhance its role in diagnosis. CT is still considered standard of care for evaluation of PE.
Patients with study results that establish the diagnosis of PE do not require routine duplex imaging of the lower extremities, and patients with acute DVT in the absence of respiratory symptoms do not require CT angiography. Patients with established PE should undergo cardiac ultrasonography to evaluate acute pulmonary artery hypertension and right ventricular strain that may signify a more massive PE. Serum troponin and B-type natriuretic peptide measurements also help stratify risk in patients with PE. An elevated serum troponin level is associated with increased mortality.
Most patients with DVT can be efficiently and safely diagnosed and treated without hospitalization. More recent literature has shown a subset of patients with PE with an excellent prognosis who can also avoid inpatient care. Clinical predication models have been developed to help determine the outcome of patients with acute PE, such as the Pulmonary Embolism Severity Index (PESI), which predicts clinical severity and outcome of patients with PE using 11 clinical criteria. In a multicenter, prospective, open-label, randomized trial of patients with low-risk PE as determined by the PESI score, no difference was found between outpatient and inpatient management in recurrent VTE, major bleeding, or 90-day mortality. A simplified version of the PESI defines patients who are younger than 80 years, who are without significant comorbidity, and who have a pulse rate less than 110/min, systolic blood pressure greater than 100 mm Hg, and oxygen saturation greater than 90% breathing ambient air as low risk for adverse outcomes.
For most patients, anticoagulation is the primary treatment for VTE. Anticoagulant options for acute VTE include unfractionated heparin, which usually requires hospitalization, low-molecular-weight heparin (LMWH), fondaparinux, or one of the non–vitamin K oral anticoagulants, all of which are safe and effective for immediate outpatient management, although patients must learn injection technique for LMWH and fondaparinux. Traditional vitamin K antagonists are not effective without at least 5 days of concomitant parenteral heparin therapy, and dabigatran and edoxaban have not been evaluated in acute VTE without previous parenteral heparin therapy. Apixaban and rivaroxaban are safe and effective as monotherapy. For patients receiving anticoagulation therapy for VTE who survive an episode of major bleeding, the ASH suggests restarting oral anticoagulation therapy within 90 days. Patients who require hospitalization should avoid initial treatment with unfractionated heparin because of its unpredictable bioavailability compared with LMWH; however, unfractionated heparin, with a short half-life of residual anticoagulation after the infusion is stopped, may be preferred in patients who are not stable and who may need emergent surgery or transition to thrombolytic therapy.
Duration of therapy varies based on the clinical scenario surrounding the event (Table 33). In provoked thrombosis with reversible risk factors, 3 to 6 months of anticoagulation is adequate. Extended therapy should be considered in patients at low bleeding risk with unprovoked VTE or with irreversible risk factors for recurrent VTE, such as underlying heart failure or stroke with long-term ambulatory dysfunction. If extended therapy is chosen, the risks, benefits, and choice of anticoagulant should be re-evaluated yearly.
In patients with unprovoked VTE in whom anticoagulation is discontinued, initiating aspirin is associated with approximately a 30% to 40% risk reduction in recurrent VTE. The American College of Chest Physicians (ACCP) guidelines published in 2016 recommend aspirin if a patient with unprovoked VTE does not continue long-term anticoagulation. The American Society of Hematology (ASH) guideline published in 2020 emphasizes that continued anticoagulation is more effective as secondary VTE prevention than aspirin but with the tradeoff of increased bleeding risk.
In patients with malignancy, LMWH remains preferable to warfarin; the CLOT trial, in which patients were randomly assigned to LMWH or warfarin, found that 15% of patients treated with warfarin developed recurrent VTE compared with 7.9% of patients treated with LMWH. Anticoagulation should be continued as long as the cancer is active.
For patients with active cancer and VTE, the 2020 American Society of Clinical Oncology guidelines recommend initial anticoagulation with LMWH, unfractionated heparin, fondaparinux, or rivaroxaban. LMWH is preferred over unfractionated heparin for the initial 5 to 10 days of anticoagulation for patients who do not have severe kidney injury (creatinine clearance <30 mL/min). For long-term anticoagulation, LMWH, edoxaban, apixaban, or rivaroxaban for at least 6 months is preferred over vitamin K antagonists. The major bleeding risk is increased with non–vitamin K antagonist oral anticoagulants (NOACs), particularly in gastrointestinal, and potentially genitourinary, malignancies. Caution is also warranted in other settings with high risk for mucosal bleeding. Anticoagulation with LMWH or NOACs beyond the initial 6 months should be offered to select patients with active cancer, such as those with metastatic disease or those receiving chemotherapy if the benefit is determined to exceed the risk. Patients with primary or metastatic central nervous system malignancies and acute VTE should be treated with anticoagulants analogous to those used for other sites of cancer, except when active intracranial or intraspinal bleeding is present.
Thrombolytic therapy is necessary to treat patients with massive PE and shock from low cardiac output. The ASH guideline suggests anticoagulation alone over the routine use of thrombolysis in addition to anticoagulation for patients who are clinically stable but have poor prognostic features on cardiac ultrasonography along with elevated serum troponin and B-type natriuretic peptide levels. For acute DVT, thrombolysis is indicated for massive thrombus leading to impaired venous drainage, severe edema, and acute limb ischemia. The main function of an inferior vena cava (IVC) filter is to prevent death from PE. In 2012, the ACCP recommended IVC filters for those with a contraindication to anticoagulation who either have acute PE or acute proximal (above the knee) DVT. The ASH further notes that no evidence supports the use of IVC filters with anticoagulation in managing acute VTE in patients with advanced lung disease or hemodynamic instability. If an IVC filter is placed, a temporary filter should be used.
Distal DVT does not usually require anticoagulation. Isolated distal DVT can be monitored with serial Doppler ultrasonography performed 5 to 7 days after the initial event in otherwise healthy, asymptomatic patients. According to ACCP guidelines, anticoagulation similar to that for proximal DVT is suggested in patients with certain risk factors for extension, including a positive D-dimer test result, extensive thrombosis or proximity to proximal veins, no reversible provoking factor for DVT, active cancer, history of VTE, and inpatient status.
Patients with DVT or PE can develop long-term complications affecting function and quality of life. Approximately 25% to 40% of patients with symptomatic DVT can develop aspects of postthrombotic syndrome (PTS) and chronic venous insufficiency, which often develop within 2 years of diagnosis. Symptoms of postthrombotic syndrome include pain in the affected limb, heaviness, swelling, stasis dermatitis, and ulceration. It often leads to poor quality of life and contributes to work disability. Treatment includes leg exercises, avoiding dependent positions for lengthy periods, and using compression stockings. Skin moisturizers and a low-moderate potency topical glucocorticoid may be used for stasis dermatitis (see MKSAP 18 Dermatology).
Patients with PE can also develop chronic thromboembolic pulmonary hypertension (see MKSAP 18 Pulmonary and Critical Care Medicine), cardiopulmonary dysfunction, or decreased exercise tolerance.
Superficial thrombophlebitis describes thrombus in a vein located near the skin's surface; it is a common inflammatory-thrombotic disorder that does not usually cause significant morbidity or progress to PE. It typically is treated with supportive care, analgesia, warm compresses, and NSAIDs. Imaging is indicated if symptoms progress or swelling occurs. Cannulated veins of the hands and arms often thrombose after infusions and intravenous catheter placement; this condition does not require anticoagulant therapy.
Superficial vein thrombosis (SVT) often affects the lower extremities and is thought to account for 10% of lower extremity thromboses. When affecting the great saphenous vein (also referred to as the greater or long saphenous vein), SVT may progress into the deep venous system. In a randomized trial of fondaparinux versus placebo for lower extremity SVT, it was found that fondaparinux was safe and effective in preventing PE. In patients with lower extremity SVT of at least 5 cm in length or close to the deep venous system, fondaparinux or an alternate anticoagulant is recommended. Anticoagulation may also be indicated for patients with SVT and other thrombophilic risk factors, including cancer or previous DVT. Patients with lower extremity SVT who are not treated initially with anticoagulants should undergo follow-up evaluation in 1 week to assess signs of thrombus progression. Imaging is necessary for persistent or worsening symptoms.
Thrombophilias do not play a significant role in arterial thrombosis. The primary causes of arterial thrombosis are arteriosclerosis and atrial fibrillation with systemic arterial embolism. Patients with arterial thrombosis due to arteriosclerosis are typically treated with antiplatelet therapy. It is unknown whether patients with arterial clots in whom a strong thrombophilia is found are more effectively treated with antiplatelet therapy or anticoagulants.
Upper extremity DVT accounts for 10% all DVT occurrences. Secondary DVT of the upper extremity is much more common than primary (two thirds versus one third). Secondary upper extremity DVT usually occurs with central venous catheter use or malignancy; treatment consists of 3 months of anticoagulation. However, if the catheter will not be removed in the setting of a proximal DVT, anticoagulation should continue as long as the catheter remains in place.
Primary upper extremity DVT is uncommon and usually caused by anatomic abnormalities of the thoracic outlet system leading to axillosubclavian compression and thrombosis (venous thoracic outlet syndrome). Patients are usually young, and thrombus occurs with strenuous upper extremity activity. Expert recommendations vary regarding the use of thrombolysis or thoracic outlet decompression surgery in addition to anticoagulation. ACCP guidelines recommend that treatment of primary and secondary upper extremity DVT follow similar guidelines as lower extremity DVT. Provoked upper extremity DVT should be treated for 3 months.
Unfractionated heparin works by binding to antithrombin, which leads to activation and potentiation of its action, resulting in inactivation of thrombin and factor Xa.
The activated partial thromboplastin time (aPTT) is used in monitoring patients receiving heparin therapy. In the setting of lupus anticoagulant (which prolongs the aPTT), heparin resistance, or markedly elevated factor VIII, antifactor Xa monitoring can be used. Although ideal dosing has been controversial, a weight-based nomogram is usually used, and most hospitals follow a specific heparin dosing algorithm. Typically, an initial bolus dose of 80 to 100 U/kg is given.
Heparin is available in intravenous and subcutaneous preparations for the treatment of VTE, although the intravenous form is typically used. A parenteral agent should be overlapped with warfarin for 5 days and until the INR is 2 or greater for at least 24 hours.
The rate of heparin-associated major bleeding is approximately 3%. Failure to follow a dosage adjustment algorithm is associated with increased bleeding risk. When major bleeding occurs, protamine sulfate can be administered to reverse anticoagulation. A dose of 1 mg of protamine per 100 units of heparin is recommended. Protamine has its own significant adverse effects, which include allergic reactions, hypotension, bradycardia, and respiratory toxicity.
Although weight-based nomograms for instituting heparin therapy and specific algorithms for adjusting dose based on aPTT results have enhanced the safety and efficacy of unfractionated heparin, variations in bioavailability and potential delay in arriving at a therapeutic dose are still more likely than in patients treated with LMWH. Unfractionated heparin should generally be reserved for patients for whom LMWH is contraindicated or in those who require anticoagulation that can be stopped quickly, generally in anticipation of an invasive procedure or surgery.
Heparin-induced thrombocytopenia is a paradoxical adverse effect of heparin that can result in life-threatening thrombosis (see Platelet Disorders).
LMWH is derived from unfractionated heparin through a chemical depolymerization producing smaller fragments that are one third the size of heparin.
LMWH does not affect the aPTT because the smaller fragment size does not bind as readily to thrombin but retains the ability to inactivate factor Xa. Dosing is more predictable and laboratory testing is generally unnecessary. LMWH is cleared through the kidney, and the biological half-life is increased in patients with kidney disease. For patients with kidney dysfunction (creatinine clearance <30 mL/min) receiving LMWH therapy, the ASH suggests against using anti–factor Xa concentration monitoring to guide LMWH dose adjustment but rather consider using doses adjusted for kidney function as recommended in product labeling or switching to an alternative anticoagulant with lower kidney clearance, such as unfractionated heparin. In patients with obesity, twice daily dosing is suggested based on actual body weight.
LMWH is preferred to unfractionated heparin. In a meta-analysis of DVT treatment comparing LMWH with unfractionated heparin, LMWH was associated with less major bleeding, decreased mortality, and decreased thrombotic recurrence.
Protamine does not fully reverse the anti-Xa effect of LMWH but provides some benefit in restoring hemostasis; it should be given at a dose of 0.5 to 1 mg of protamine per 1 mg of enoxaparin.
In a clinical trial, fondaparinux, dose adjusted based on patients' weights, was noninferior to enoxaparin with respect to the primary endpoint of recurrent VTE at 3 months (3.9% vs. 4.1%). Fondaparinux is also cleared through the kidney and should be avoided in patients with creatinine clearance less than 30 mL/min. As with LMWH or unfractionated heparin, treatment with fondaparinux and warfarin should overlap for 5 days.
Fondaparinux has no reversal agent. Caution should be used in patients at risk for bleeding because the half-life is 17 hours. Prothrombin complex concentrates (PCCs) and fresh frozen plasma (FFP) have been administered with positive outcomes in patients experiencing bleeding.
Warfarin is a vitamin K antagonist. It inhibits vitamin K epoxide reductase, which leads to inhibition of γ carboxylation of precursor coagulation factors II, VII, IX, and X and proteins C and S. Laboratory monitoring involves the prothrombin time and INR.
Because warfarin lowers protein C levels before inducing its anticoagulant effect, it can initially cause a prothrombotic state. For this reason, for the treatment of acute VTE, it must be administered initially with a parenteral anticoagulant. Warfarin should be initiated as soon as possible after diagnosis of VTE. Typically, unfractionated heparin or LMWH is used with warfarin. As noted previously, the parenteral agent should be given for at least 5 days, and the INR can be measured on day 3; heparin is continued until the INR is 2 or greater for at least 24 hours.
Although non–vitamin K antagonist oral anticoagulants have changed the landscape of treatment for patients with VTE, warfarin remains a reasonable anticoagulant for some patients. This may include patients with known kidney disease and obesity or patients with a mechanical heart valve, for whom alternate oral anticoagulants have not been approved.
Patients must have access to continued outpatient INR monitoring. Studies evaluating the use of cytochrome 2C9 and vitamin K epoxide reductase (VKORC1) pharmacogenetics to guide warfarin therapy have not shown benefit. Common reasons for fluctuations in INR include changes in vitamin K intake, medications, and nonadherence. Studies attempting to decrease INR variability with low-dose daily vitamin K supplementation were unsuccessful, so this supplementation is not indicated.
Bleeding is the most significant complication in patients treated with warfarin, occurring in 1% to 3% of patients per year. The risk is higher when warfarin therapy is initiated and during episodes of concurrent acute illness. Bleeding risk increases further in patients with an INR greater than 5. Independent of INR, bleeding risk is increased in patients older than 75 years or in those with previous stroke, gastrointestinal bleeding, or most other chronic comorbidities. Concomitant aspirin, clopidogrel, and NSAID use increases the bleeding risk. The indication for antiplatelet agents for patients taking warfarin should be carefully reviewed, and dual antiplatelet therapy avoided if possible. Acetaminophen should be used instead of NSAIDs when feasible.
Concern is often expressed when older adults begin oral anticoagulation, often for atrial fibrillation, because age is an important risk factor for stroke and bleeding complications associated with warfarin. Oral anticoagulation may be prematurely excluded as a therapeutic option in these patients because of concerns regarding a “falls risk.” The true risk of serious bleeding related to a fall while taking an anticoagulant is unclear, although small studies have not shown an increased risk of major bleeding in patients taking oral anticoagulants who were considered at high risk for falls. Risk factors for falls should be thoroughly evaluated and appropriate steps employed for prevention (see MKSAP 18 General Internal Medicine). Recommendations suggest that neither age nor a risk of falls is reason to withhold warfarin anticoagulation from a patient who has clinical criteria warranting such therapy for VTE or stroke prevention.
Patients with asymptomatic INR elevation between 4.5 and 10 can often be managed by simply withholding warfarin. For INRs greater than 10 in patients without bleeding, oral vitamin K, 2.5 mg, should be given. In patients experiencing bleeding, in addition to vitamin K, four-factor PCCs should be given rather than FFP. Although FFP contains the appropriate clotting factors, it requires thawing and large volumes to correct the INR. Three- and four-factor PCCs contain proteins C and S and factors II, IX, and X; four-factor PCC also contains factor VII. In a clinical trial of vitamin K antagonist–related bleeding, four-factor PCC was found to be noninferior to FFP for hemostatic efficacy. Four-factor PCC is preferred because of its rapid reversal of INR, rapid infusion and administration, and lack of volume overload. Recombinant factor VIIa is not recommended for warfarin reversal.
Bridging therapy, which uses heparin or LMWH for patients in whom warfarin has been stopped for an invasive procedure and will be resumed, is not necessary for most patients and is associated with more bleeding complications without additional anticoagulant benefit. The exception to this may be patients with recent VTE (within the past 4 weeks), history of VTE during anticoagulant interruption for surgery, or a procedure with very high VTE risk, such as orthopedic surgery. Bridging is also indicated in patients with atrial fibrillation who have had a stroke or transient ischemic attack in the preceding year, in patients who have multiple risk factors for stroke (CHADS2 score of 5-6), and in most patients with a mechanical heart valve.
The non–vitamin K antagonist oral anticoagulants (NOACs) have emerged as a safe and effective treatment for certain patients with VTE. In the 2016 CHEST guidelines for DVT and PE treatment, NOACs are suggested as the treatment of choice for anticoagulation in patients without cancer; however, 2020 American Society of Clinical Oncology guidelines recommend rivaroxaban for initial anticoagulation and add apixaban and edoxaban as options for long-term anticoagulation in patients with malignancies. The NOACs available for use in the United States are dabigatran, rivaroxaban, apixaban, edoxaban, and betrixaban. In clinical trials of patients with VTE, patients were initially treated with a parenteral agent and transitioned to dabigatran or edoxaban. Rivaroxaban and apixaban were studied without concomitant parenteral therapy and were approved as monotherapy for DVT and PE. Dabigatran functions as a direct thrombin inhibitor, whereas the other agents are factor Xa inhibitors. Betrixaban is only approved for DVT prophylaxis (Table 34).
Routine coagulation studies do not reliably measure the degree of coagulation activity. However, the thrombin time is quite sensitive to the presence of dabigatran and, if normal, indicates that the anticoagulant effect of dabigatran is no longer significant.
Advantages of the NOACs include no need for routine monitoring, rapid onset of action and short half-life, fixed dosing, and fewer drug-drug interactions. These drugs are as effective as warfarin in the prevention of VTE; although the overall bleeding risk was comparable, less central nervous system bleeding, fatal bleeding, and use of blood product support among patients taking NOACs was seen than with warfarin. The bleeding risk is higher in patients taking aspirin or clopidogrel with a NOAC and is further increased in patients taking dual antiplatelet drugs plus NOACs. These qualities must be considered when choosing the appropriate patient for these therapies. No head-to-head trials have been performed comparing the various NOACs. It must be noted that certain patient groups were excluded from the major trials of the NOACs, including patients with severe obesity (BMI >40), pregnant patients, and those with mechanical heart valves. Nonadherent patients should not be treated with NOACs, and the additional cost of these drugs compared with warfarin may be a barrier for some patients. In patients with antiphospholipid antibody syndrome, the role of NOACs remains unclear, although clinical trials are ongoing. Treatment failures with the use of NOACs in patients with antiphospholipid antibody syndrome have been reported. Dyspepsia and gastrointestinal bleeding were seen more frequently with dabigatran compared with warfarin in clinical trials. In patients with concern for gastrointestinal bleeding, dabigatran may not be the preferred option.
All of the NOACs are at least partially eliminated through the kidney (see Table 34), and the dose must be reduced in patients with advanced chronic kidney disease. Apixaban has the lowest renal elimination, so it is approved for patients undergoing dialysis; however, many physicians still prefer warfarin in patients with advanced kidney disease, and caution should be used. Data from a 2019 systematic review and meta-analysis suggest that NOACs may be preferred in patients with early stage chronic kidney disease (estimated glomerular filtration rate <60 mL/min/1.73 m2) and atrial fibrillation. Additionally, the risk of bleeding with NOACs appears to be no worse than with VKAs.
NOACs should either be used with caution or avoided in patients with Child-Pugh class B or C liver disease.
Bridging therapy is typically unnecessary in patients taking NOACs. Discontinuation of the NOAC depends on the half-life of the drug, the type of procedure, and the patient's kidney function. NOACs should be stopped 24 to 48 hours before surgery with moderate bleeding risk and 72 hours before surgery with higher bleeding risk. In patients with impaired kidney function, NOACs should be stopped earlier. For procedures with low bleeding risk, NOACs can be resumed promptly when effective hemostasis is secured. For procedures with higher rates of bleeding, reinstitution is usually delayed 2 to 3 days.
The standard approach to patients experiencing bleeding involves hemodynamic monitoring and resuscitation with fluid and blood products. Activated charcoal can be considered if the NOAC was ingested recently (<6 hours). Hemodialysis can be considered with dabigatran therapy if new kidney disease is found. For patients with life-threatening bleeding who are taking an oral direct Xa inhibitor, the ASH suggests stopping the oral direct Xa inhibitor and considering treatment with either four-factor PCC or andexanet alfa (off-label use for edoxaban and betrixaban). Because of the lack of comparative data on four-factor PCC and andexanet alfa, a preferred strategy cannot be identified. In the context of limited clinical evidence, high cost, and significant complications, experts recommend caution in using andexanet alfa and restricting its use to patients receiving anti-Xa anticoagulants who are experiencing life-threatening bleeding. Idarucizumab, a monoclonal antibody fragment, binds free and thrombin-bound dabigatran and neutralizes its activity. In a phase 3, multicenter, prospective, cohort trial, idarucizumab was found to be safe and effective in reversing the anticoagulant effects of dabigatran in patients who either experienced serious, overt, life-threatening bleeding determined to require a reversal agent or who required an urgent invasive procedure that could not be delayed. Idarucizumab was FDA approved for this indication in October 2015 and recommended in the ASH 2018 guidelines on VTE treatment. If idarucizumab is not available, activated prothrombin complex concentrate is an alternative for patients with dabigatran-associated major bleeding. Specific antidotes for this class of agents continue to be developed.