Deep Vein Thrombosis (DVT)

Deep vein thrombosis has two important sequelae:

• Pulmonary embolism
• Post-thrombotic syndrome (see section on prognosis)

Both are common disorders. A thrombus either arises spontaneously or is predisposed by such conditions as surgery, trauma or prolonged bed rest. They usually form in the deep veins of the lower limbs but may extend higher and into the pelvic veins. The intimate relationship between DVT and pulmonary embolism (PE) is such that the term venous thrombo-embolism (VTE) is often used to cover both conditions.

DVT can be very difficult to diagnose but early recognition and appropriate treatment can save many lives.

• Exact numbers are impossible to give as it is often subclinical but in developed countries the incidence may be as high as 1 in 1,000, the incidence rising with age.¹
• Male:female ratio = 1.2 to 1. When not pregnant or using oral contraceptives or HRT, women have a lower risk than men.
• Two thirds of patients with proven pulmonary embolism (PE) have no symptoms of DVT and in one third of cases it is impossible to find the original site of a DVT without an autopsy.
• Autopsy studies demonstrate that approximately 80% of all cases of DVT and PE remain undiagnosed, even when they are the immediate cause of death.

The risk factors for DVT (and hence VTE) arise from the 3 underlying components of the Virchow triad. They are:

• Venous stasis
• Hypercoagulability
• Injury to the intima of veins

There are many factors that increase the risk of DVT. Risk increases with age and with obesity but these are often associated with other risk factors. Smoking also increases risk.

High risk

• Previous history of VTE
  • Previous episode of DVT is the strongest risk factor for DVT with a 5-fold increase over baseline risk
  • Risk of new postoperative DVT rises from 26% to 68%
  • History of prior clinically apparent PE increases the risk of new postoperative DVT to nearly 100%
• Major surgery or cancer, especially if operation on abdomen or lower limb
• Immobilisation is a great risk factor
  • In patients on bed rest in a general medical ward the rate of DVT is 10% but in an intensive care unit it is 29%
  • Within 5 days of a stroke half will have DVT unless prophylaxis is given
  • 40% of postoperative neurosurgical patients develop DVT
  • In patients with a stroke, 60% develop DVT in the paralyzed legs compared with just 7% in the contralateral one
• The antiphospholipid syndrome occurs in a number of conditions, especially systemic lupus erythematosis
  • 9% of patients with SLE develop spontaneous DVT
  • The lupus anticoagulant responsible is also found in people with AIDS and with many autoimmune diseases
  • It can be produced in healthy patients by phenothiazines
• Increased haematocrit increases viscosity and risk of sludging and thrombosis. 40% of deaths in patients with polycythaemia rubra vera are due to thrombosis, but only a third are from venous thrombosis. Not just erythrocytes but white cells and platelets are raised too. Thrombocytosis may increase the risk of thrombosis or bleeding.
• Thrombophilia may be the underlying disease process
  • Deficiencies of protein C, protein S, antithrombin III and factor V Leiden (resistance to activated protein C) are well-recognized abnormalities
  • Impaired fibrinolysis can be inherited but is most often acquired in postoperative patients, those taking synthetic oestrogens, and women who are pregnant or postpartum

Medium risk

• Malignancy is an important risk and DVT may be the presenting feature of many occult cases
  • 16% of patients with angiographically proven PE are diagnosed with cancer within 2 years
• Some types of chemotherapy increase the risk for DVT either by reducing the level of certain anticoagulants or by increasing the levels of procoagulants
• Myocardial infarction and congestive heart failure increase the likelihood of DVT and PE, independent of bed rest or immobilisation
  • In patients with acute myocardial infarction who are not receiving anticoagulation there is a 26 to 38% rate of DVT, compared with a much lower rate in those in whom MI is suspected but then excluded.

Relatively low risk

• Pregnancy and the puerperium are a time of increased risk
• Combined oral contraceptives
  • Case-control and cohort studies suggest an increased risk of 3 to 12 times.^2,3
  • HRT uses natural oestrogens and the risk is lower but it is still higher than for controls
  • There appears to be no added risk in the use of progestogen only contraception whether as pills, injection or implant. (See "Internet and further reading" from RCOG).
• Tamoxifen also increases risk
• DVT is common in patients with local trauma and stasis, as with a leg cast.
  • The role of cramped conditions on long haul flights is discussed below.
• Hyperlipidaemia increases the speed and extent of thrombus formation in vascular injury
• Patients with ulcerative colitis or Crohn's disease have increased fibrinogen, factor VIII, and platelet activity and depressed levels of antithrombin III and alpha 2-macroglobulin. They are also at higher risk.

The clinical diagnosis of DVT can be very difficult.

• Many DVTs progress to PE without DVT being clinically apparent.
• In those with classic clinical signs, only about 50% have DVT.
• History, examination and the presence of risk factors can be unreliable in making the diagnosis and ambulatory patients (as in primary care) may be rather different from hospital inpatients who are the group most studied.⁴

The last factor may make the GP ask,"Is the evidence base available applicable to the patient in front of me?"

Classical features are from obstruction to venous drainage:

• Limb pain and tenderness along the line of the deep veins
• Swelling of calf or thigh (usually unilateral). Involvement of the iliac bifurcation, the pelvic veins, or the vena cava produces leg oedema that is usually bilateral
• Distension of superficial veins
• Increase in skin temperature
• Skin discolouration (erythema or occasionally purple or cyanosed)
• A palpable cord (hard thickened palpable vein)
• Low grade fever. (Said to be rare but perhaps not sought often enough).
• There may be pain upon dorsiflexion of the foot and spontaneous maintenance of the relaxed foot in abnormal plantar flexion (Homan's sign is not helpful in diagnosis as it is very nonspecific).⁴

Cellulitis adds to the problem:

• Severe signs of DVT can resemble cellulitis
• Secondary cellulitis may develop with primary DVT
• Primary cellulitis may be followed by a secondary DVT
• Superficial thrombophlebitis may hide an underlying DVT

Because of the unreliability of clinical features several diagnostic scoring system have been validated whereby patients are classified as having a high, intermediate, or low probability of developing deep vein thrombosis, based on history and clinical examination. This is just one:⁵

Investigations may be classified as screening tests or definitive diagnosis.

Screening tests

Screening tests should have with a high negative predictive value (sensitivity) to exclude patients from the need for imaging.

• Initially arrange D-dimer tests or plethysmography according to local protocol. If patient presents "out of hours" consider subcutaneous low-molecular weight heparin now, and investigations tomorrow.
• D-dimers are specific cross linked products of fibrin degradation and are raised in patients with venous thromboembolism. Sensitivity is high but specificity poor.
  • High concentrations occur in other disorders such as malignancy and pregnancy and in other conditions where clots form, as after surgery. The investigation of VTE in pregnancy is fraught with many difficulties.⁷
  • Nevertheless, a negative test can exclude the need for further investigation.^8,9 Several D-dimer assays are available - eg ELISA tests or SimpliRED whole blood agglutination test suitable for near patient testing. Some evidence suggests that D-dimer should be measured after the end of the anticoagulation period, to assess the risk of re-thrombosis.¹⁰
• Plethysmography involves recording of changes in the size of the limb due to tissue fluid or pooled blood in the veins. Techniques include photoplethysmography (absorption of light by haemoglobin), strain gauge and electrical impedance (change in calf size or electrical impedance of skin during and after occluding venous outflow).

Definitive investigations

Imaging can be by venography, by 2-dimensional ultrasound, or by magnetic resonance imaging.

• All these tests are most sensitive when DVT is symptomatic, when thrombus causes complete venous outflow obstruction, and when the clot extends into the upper thigh.
• False negatives are highest when the thrombus is solely below the knee or above the groin, when obstruction is incomplete and when the patient is asymptomatic.

The choice of definitive test may depend upon local protocols as none is perfect.

• Duplex ultrasound is the initial investigation of choice in nearly all patients with suspected DVT.
  • Its reliability is dependent upon the skill of the user.
  • It may miss a non-occlusive thrombus in up to 40% of cases involving the iliac or pelvic veins.
  • Major axial veins of the lower limb are well displayed.
• Magnetic resonance venography is the most sensitive and specific test for deep and superficial venous disease in the lower legs and in the pelvis.
  • A great value of this test is that unsuspected non-vascular causes for leg pain and oedema often may be seen on the scan.
  • It may become the initial imaging of choice because it appears to be as sensitive and specific as either venography or ultrasound and offers a diagnosis in 60% of cases when a patient does not have DVT.
• Contrast venography has long been "the gold standard" of diagnosis for DVT.
  • An intravenous catheter is placed in a dorsal vein of the foot and contrast medium is infused into the vein.
  • A tourniquet around the leg occludes the superficial veins and forces contrast into the deep veins.
  • A positive result tends to be conclusive but a negative result is less reassuring. It is time-consuming and requires much technical skill to obtain good results.
  • It is highly invasive and has substantial morbidity and mortality unlike the other diagnostic tests for DVT.
  • Up to 10% of patients have new thrombosis shortly after a negative venogram. This may be because it missed the original DVT or because intravenous contrast can trigger thrombosis by causing endothelial injury.
  • Extravasation of contrast material into the dorsum of the foot may cause sloughing of tissue.
  • Anaphylactoid reactions to contrast material occur in 3% of patients and can cause death.

• Treat the suspicion of DVT with subcutaneous low molecular weight heparins (LMWH) if no contraindications.¹¹
• Patients with a DVT should wear compression stockings to reduce the rate of post-thrombotic syndrome. In one study of 194 patients with a first episode of proximal DVT, the rate of post thrombotic syndrome was reduced by 50% when graded compression stockings were used.¹²
• Low molecular weight heparins have made outpatient treatment of DVT possible. Trials have suggested that it is both feasible and safe.¹³
• Filters in the inferior vena cava are occasionally inserted to reduce pulmonary embolism. The indications for their use include pulmonary embolism with contraindication to anticoagulation and recurrent pulmonary embolism despite adequate anticoagulation. In the UK most vena cava filters are removed 3 weeks after the period of greatest risk and are not permanent. There is no consensus about whether patients with long term filters should receive long term treatment anticoagulation.

Target ranges

According to the BNF¹⁴:

• Give a loading first dose of 10mg of warfarin (less if prolonged prothrombin time, LFT's abnormal, cardiac failure, parenteral feeding, patient underweight or on drugs known to potentiate warfarin).
• First episode of a proximal vein thrombosis should receive anticoagulants for 6 months, with a target INR of 2.5 (± 0.5 is acceptable). The duration of anticoagulation is still under debate. Use higher target of 3.5 in recurrent DVT/PE in patients who had recurrence whilst on lower dose of warfarin or mechanical heart valves.

The British Thoracic Society concluded that if venous thromboembolism arises after surgery, 4 weeks of anticoagulation should be adequate. In other settings, patients with new DVT, PE, or both, who do not have a persisting underlying cause or risk factor should receive anticoagulants for 3 months.¹⁵ In patients with no other risk factors, there is no additional benefit of anticoagulation for 6 months rather than 3.¹⁶

There is limited evidence that thrombolytics such as streptokinase may resolve symptoms more rapidly and preserve venous valve integrity and hence decrease the incidence of the post-phlebitic syndrome,¹⁷ although there is an increased risk of major haemorrhage.

Fondaparinux sodium is a synthetic pentasaccharide that inhibits activated factor X. This drug has recently been licensed in the United Kingdom for prophylaxis of VTE in patients having major orthopaedic surgery of the legs. The potential of this drug for treating DVT has not been fully evaluated.

• There may well be recurrence of DVT. Recurrence after the first event can be as high as 60% but halved by compression stockings.¹²
• Death occurs in approximately 6% of DVT cases and 12% of PE cases within 1 month of diagnosis.¹ Long term outcome tends to be good if it is not associated with malignancy.¹⁸

Post thrombotic syndrome

• Isolated calf vein thrombophlebitis produces post-phlebitic syndrome in 20 to 40% of cases.
  • Thrombosis destroys the valves of the deep veins, leading to chronically elevated venous pressure within the legs.
  • Valve incompetence need not be extensive to produce symptoms. Isolated incompetence of the valves in the popliteal segments leads to elevated venous pressures, and over 60% of those with isolated popliteal valve failure develop severe chronic venous insufficiency.
  • The syndrome produces chronic pain, oedema, hyperpigmentation and ulceration.

What is effective to prevent DVT in one situation (eg during gynaecological or orthopaedic surgery) may not be effective in another situation (eg ambulatory patient with previous DVT). It is important to look for relevant evidence. Groups of patients may include:

• Surgical patients having orthopaedic,gynaecological or other surgery
• Medical patients after myocardial infarction or stroke
• Patients with malignancy
• Ambulatory people with previous VTE
• Ambulatory people with predisposing conditions such as antiphospholipid syndrome or polycythaemia rubra vera

Heparin should not be seen as the same as unfractionated heparin (UH) or low molecular weight heparin (UMWH).

By and large, unfractionated heparin (UH) or low molecular weight heparin (UMWH) seem to be effective whilst heparin, warfarin and aspirin are of lesser or no value. The place of aspirin in preventing arterial thrombosis is strongly established but its place, if any, in preventing venous thrombosis is dubious.¹⁹

A pharmacological approach is just one of several considerations. Physical compression of veins, early mobilisation and prevention of dehydration all have a place.

• Assessing degree of risk in surgical patients depends upon the nature of the surgery as well as personal parameters. Possible interventions have been reviewed and these include early ambulation, elastic compression stockings, pneumatic compression devices, inferior vena cava filters, and a variety of drugs including unfractionated heparin, warfarin, aspirin, low molecular weight heparin, and pentasaccharides.²⁰
• Gynaecological surgery is high risk and subcutaneous heparins are usually given pre-operatively.²¹ The highest risk comes with major orthopaedic procedures to the lower limb. A Cochrane review concluded that low molecular weight and unfractionated heparin were both useful in preventing DVT but not necessarily PE. Foot and calf pumping devices appear to prevent DVT, may protect against pulmonary embolism, and reduce mortality, but compliance remains a problem. Good quality trials of mechanical methods as well as direct comparisons with heparin and low dose aspirin should be considered.²² A comprehensive approach with mechanical as well as pharmacological methods in orthopaedic patients is advised.²³
• In medical patients at risk, unfractionated heparin (UFH), low-molecular weight heparin (LMWH), and fondaparinux sodium have been shown to be effective in the prevention of VTE but UFH has a higher rate of bleeding complications than LMWH. There is no evidence supporting the use of aspirin, warfarin, or mechanical methods to prevent VTE in medical patients. LMWH or fondaparinux sodium are well tolerated and effective thromboprophylactic agents in medical patients.²⁴
• It has been recommended that for those with antiphospholipid syndrome that long term warfarin should be used with a target INR of 3.0 to 4.0.²⁵

Aeroplane travel

There has been much interest in recent years about DVT resulting from travel by aeroplane as stories abound about fit young people who undertake long haul flights and a few days later they suddenly drop dead. A study from South Africa investigated D-dimers in people after long flights, using this as a surrogate for possible DVT. They had positive results in over 10% but, contrary to expectations, there was no difference according to class flown.²⁶ It is usually assumed that cheaper fares in more cramped conditions are a greater risk. Passengers should be advised to keep well hydrated on flights. This also necessitates more frequent walks down the aisle to the toilet with use of the muscle pump.

Another study found that people who had aisle seats and so for whom getting up and walking around did not involve disturbing others, had a lower rate of DVT.²⁷ They found that DVTs were asymptomatic in 60% of subjects and they recommended a single dose of low molecular weight heparin for high risk people about to undertake long flights. The use of below-knee graduated compression stockings appears to be highly effective.²⁸ Mechanisms to exercise the calf muscle pump during flight have also been recommended. Aspirin has been recommended as a cheap and readily available prophylactic for all but the benefit is dubious and it is estimated that 17,000 people have to take aspirin to prevent 1 DVT.¹⁹ Hence risk probably exceeds benefit and aspirin should be reserved to prevent thrombosis in the arterial rather than the venous system.

For people at very high risk of DVT long term heparin would seem to be the most effective intervention but it is not without problems. There are concerns about long term use, including osteoporosis.²⁹

Clinical Knowlege Summaries advise everyone:

• Drink plenty of water.
• Avoid excessive alcohol (as this can cause dehydration).
• Avoid sleeping tablets (as these cause immobility).
• Regularly flex ankles to contract calf muscles.
• Take occasional short walks (if possible).
• Take advantage of refuelling stopovers where it may be possible to walk about.

They recommend that special precautions are required only for flights in excess of 6 hours. For recommendations about high risk people, see the guidance in "Internet and further reading".

• Schenk first described venous thrombosis in 1644, when he described an occlusion in the inferior vena cava.
• Virchow recognised the association between venous thrombosis in the legs and pulmonary embolism in 1846. Rudolf Virchow (the name of Slavic origin often mispronounced then as it is now, pronounced "Furko") was Professor of Pathology in Berlin in the Pathological Institute built specifically for him, and author of the most important pathology text in history, "Cellular-pathologie" in which his famous dictum "Omnis cellula e cellula" ("each cell stems from another cell") appeared for the first time. One of his many interesting original observations is that of leukaemia. He also found time to serve in the Reichstag for 13 years, as leader of the Radical Party, a persistent opponent of Bismark.