Deep vein thrombosis (DVT) management considers diagnosis and therapy of DVT as well as prevention and risks of recurrence on stopping therapy. DVTs are treated in order to prevent pulmonary embolism (PE) fatality, morbidity and long-term pulmonary hypertension. DVT may also cause changes to the leg—post-thrombotic syndrome.
Recently published clinical guidance from NICE on therapy for varicose veins discounts treatment with compression hosiery,2 largely because about 50% of patients are non-compliant and many more are only partially compliant with them.
Unless the patient is unable or unwilling to undertake interventional therapy then compression hosiery for varicose veins is consigned to past practice. It is used for a few days post- varicose vein therapy but retains an important place in therapy post-DVT. Most patients with complications or symptoms from varicose veins should be referred to the local vascular service for interventional therapy, of which endothermal ablation is first line and has better outcomes than compression hosiery.
Incidences of DVT and PE as complications of varicose veins therapy in trials, comparing any of endothermal ablation, foam sclerotherapy, laser therapy or stripping, are very low. Anticoagulants are not recommended.
DVTs are subdivided into provoked and unprovoked: 25–50% are unprovoked, and the rest are provoked by a predisposing temporary risk factor. NICE1 defines provoking factors in the older patient as temporary significant immobility, surgery, trauma, and hormone replacement therapy (HRT).
Some 10% of DVT patients will develop severe post-thrombotic syndrome within five years.3 Some may develop a PE and some will present only with PE. Therapy is targeted at treatment but also at prevention of DVTs in those vulnerable, especially knee and hip replacement patients and lastly in prevention of recurrence in those with predisposing risk factors like cancer. This has led to variations in length of therapy compared to previous practice.
As medical students we learnt Virchow’s triad of causes of DVT: disorders of the blood, injury to the vein or activation of blood coagulation. This is still pertinent but an expansion of which the pathologist, Virchow, would have approved includes risk factors like hyperviscosity syndromes eg. leukaemia, dehydration, hereditary thrombophilias, cancer, some chemotherapy agents, increased age, previous DVT, damage to veins from catheters, thrombophlebitis, venous stasis, hospitalisation, immobility from travel or illness or paralysis or surgery, obesity, congestive cardiac failure, nephritic syndrome, HRT, and intravenous drug abusers.
A recent study also identified among older patients recent histories of infection, therapy with erythropoetin and blood transfusion as less well known risk factors.4
It is difficult to know how many patients develop a DVT as the rates are much increased if imaging is used to make the diagnosis compared to clinical findings. Given that proximal DVT is the clinical concern then a 2001 review states that in surgical patients without thromboprophylaxis the risk of proximal DVT is 0.4–10%, depending on the patient’s risk factors. Thromboprophylaxis reduced the incidence of DVT after general surgery (detected by fibrinogen leg scanning) from 25% in untreated patients to 6% in those taking low molecular weight heparin (LMWH).5
Most clinical DVTs affect legs and have proximal venous extension from the calf. A study of 542 clinical DVTs showed that only 35% (189) were in fact DVTs and 88% of those had proximal thrombosis and 99% of those also had calf vein thrombosis.6 DVTs isolated to the calf are at low risk of PE.
Not all DVTs are symptomatic or have clinical signs and some will present with PE. Some may be pelvic. One month mortality from DVT is 1.4%.7
Post-thrombotic syndrome refers to the sequel of skin changes in the affected leg: chronic oedema, discomfort, pain, venous ulceration and lipodermatosclerosis.
Patients present to the GP with leg pain, discomfort, oedema, swelling (compared to the other calf when measured at the same level), redness, warmth but findings can be variable or absent. It is more difficult if the patient has pre-existing leg oedema especially after orthopaedic surgery but it is well recognised that DVTs occur post-operation and there is an incidence on anticoagulation.
A DVT may resolve or progress proximally from the calf vein. It may embolise and ascend the inferior vena cava, entering the right side of the heart and is then ejected into the pulmonary arteries, becoming lodged as an arterial PE and causing infarction of lung tissue.
The GP should also assess the patient for alternative diagnoses and provoking factors. The GP or A&E department assesses the clinical probability of DVT using the two level Wells score which indicates if a DVT is likely (two points or more) or unlikely (score of one point or less) and is mainly a measure of a number of provoking factors for DVT. GPs should have it readily accessible on their computers and it is available to download as a free app for smart phones, like most medical score charts.
The body’s response to DVT is to activate plasminogen to plasmin in order to lyse the thrombus so fibrin is broken down to soluble fibrin degradation products, one of which is d-dimer which circulates at increased levels (see part one of this series).
D-dimer concentrations are very sensitive for thrombus production in the body, NICE examined all evidence and concluded that d-dimers had sensitivities of about 90% so nine out 10 DVTs would be picked up by d-dimer testing, but specificity was only 40–60% so many patients had raised d-dimer without having a DVT. This is because d-dimer can be raised in MI, any thrombus formation, DIC, age, liver disease, inflammation, malignancy, pregnancy, trauma and recent surgery.1 A raised d-dimer is
not reliable for diagnosing DVT but a patient with low Wells score (low clinical probability of DVT) and a low d-dimer helps confirm that thrombus formation has not occurred.
Patients with Wells score of two or more and access to ultrasound (USS) within four hours
The only patients who do not get d-dimer testing are those with Wells scores of two or more and who have USS within four hours and in whom it is positive. In these patients there is no point wasting resources on a d-dimer. Everyone else though gets a d-dimer.
The patients above who have an unexpectedly negative USS get a d-dimer and if it is too low then DVT is excluded. The patients above who have a negative USS but a positive d-dimer get a check USS 6–8 days later to ensure there is no DVT formation and then definitively treated as DVT or not depending on the outcome.
Patients with Wells scores of two or more having to wait 4–24 hours for an USS
They all have a d-dimer test and are anti-coagulated whilst waiting. If the USS is positive they are treated as a DVT. If the USS is negative and the d-dimer negative they do not have a DVT and anticoagulation is stopped. If USS is negative and the d-dimer is raised they have a second check USS at 6–8 days and then definitively treated as DVT or not depending on the outcome.
Patients with Wells scores low at one or 0
If the d-dimer is also low the patient does not have a DVT and does not need USS or anticoagulation. If the d-dimer is raised and the patient can have an USS within four hours then they can be diagnosed as a DVT or not. If the d-dimer is raised and the USS is arranged for between 4–24 hours later then the patient should receive anticoagulation whilst waiting and then treated definitively as a DVT or not depending on the outcome.
Patients told that they have not got a DVT should be safety netted by education about symptoms and signs which would prompt re-attendance to the GP.
There is discussion currently about improving the specificity of d-dimer testing by using higher cut off values in the elderly patient with possible DVT.8 In some areas local clinical pathways by CCGs have been produced to allow diagnosis and management of DVT in primary care through Local Enhanced Service specifications. This requires rapid access to d-dimer testing and USS.
NICE 2012 VTE guidance suggests starting patients usually on LMWH, like enoxaparin and using unfractionated heparin (UH) if they are in renal failure. Heparin activates antithrombin and amplifies its action 100-fold. Antithrombin inactivates F10 and thrombin to prevent thrombus formation. An alternative is fondaparinux, given subcutaneously, a F10 inhibitor which prevents the cleavage of prothrombin to active thrombin and there appears to be no differences between rates of DVT recurrence and major bleeding between fondaparinux and LMWH. These drugs are continued for about five days until the oral warfarin has reached therapeutic INR levels.
A patient with uncomplicated unprovoked DVT would have warfarin therapy for three months. The risk of PE reduces over time but the risk of bleeding remains the same and so there is a point at which the risk of recurrence of DVT or of PE is out-weighed by the risk of bleeding and anticoagulation should be stopped. NICE guidance suggests estimating each patient’s risk of recurrence of VTE and risk of bleeding in deciding whether to continue their anticoagulation beyond three months in unprovoked DVT along with their risk of falls, mental state and desire to adhere to medication.
Warfarin is quoted as a risk of haemorrhage of 3–4% per annum,5 has a narrow therapeutic window and needs monitoring. It requires a lot of input to have patients in their therapeutic INR range for as little as 70% of the time.9 All patients taking warfarin should have an anticoagulant booklet and a helpline in order to get advice if they have altered their doses, changed medications or diet. They should carry an anticoagulation alert card at all times. The link to the anticoagulant booklet provided in many languages by the National Patient Safety Agency (NPSA) is: http://www.nrls.npsa.nhs.uk/resources; Actions that can make oral anticoagulant therapy safer: Information for patients and carer and covers topics like dental care, alcohol, missed doses. As there is no evidence of benefit with patient self-monitoring over usual anticoagulant care with warfarin it is not currently routinely recommended.
In patients with cancer then LMWH therapy is recommended for a minimum of six months and again consideration of the patient’s best interests in deciding whether to stop therapy or not with them. Therefore cancer patients are more likely than previously to be on long-term LMWH therapy. As discussed in part two of this series side-effects of heparin are heparin induced thrombocytopenia (HIT), which usually occurs in the first 14 days of use at a rate of about 0.2% and can cause thrombosis. It is less common in LMWH than in UH and is due to antibody formation against heparin. Rarely there may be hyperkalaemia and neutropenia. Like all anticoagulants there is a risk of bleeding, especially if the patient has severe liver disease or severe hypertension. LMWH should not be used if the eGFR is <30mls/min. LMWH is less likely than UH to cause osteoporosis but there is still an increased risk long-term.
LMWH is of pig origin. It had less recurrence of VTE than warfarin in trials but similar rates of major bleeding. This is surprising as in practice a number of patients with cancer on warfarin might be expected to have nausea, vomiting and dietary problems affecting coagulation which LMWH would circumvent and NICE recognises this.
Rivaroxaban is an oral F10 antagonist and NICE guidance10 explores its use in VTE. It offers an alternative anticoagulant for patients who are needle phobic for LMWH and have trouble with warfarin for any reason, especially monitoring and adherence issues. Rates of bleeding in long-term use were comparable to usual LMWH and warfarin, as were rates of VTE recurrence and in addition there were reduced costs, lack of need for INR monitoring and increased patient convenience. Unlike warfarin it does not have a specific antidote.
Rivaroxaban is affected by renal impairment and may interact with liver metabolised drugs. A number of CCGs have produced guidance for GPs using rivaroxaban.11 At diagnosis the d-dimer result must be returned urgently as those with low risk Wells score and negative d-dimer can be excluded from DVT diagnosis as above and all others are issued with a prescription for rivaroxaban to take if their USS is positive, or whilst waiting if the scan is between 4–24 hours. All patients are issued with a patient alert card and information about therapy. Usual dosing is 15mg bd with food for three weeks and 20mg od for another three months in provoked DVT with consideration of longer therapy if required and for unprovoked DVT.
Other oral factor 10 inhibitors are edoxaban tosylate and apixaban but these are not licensed for DVT treatment in the UK at present.
NICE guidance also recommends that patients wear a below knee graduated compression stocking on the affected leg for two years to help prevent post-thrombotic syndrome if there are no contraindications and that patients with unprovoked DVTs should be investigated for cancer. Usual contraindications to stockings for which GPs need to be alert are: peripheral arterial disease, peripheral neuropathy, leg oedema, local skin and soft tissue conditions like dermatitis and infection, deformity of the leg and ankle:brachial pressure index <0.8.
NICE1 states that 11% of patients with an unprovoked DVT will have cancer detected in the next one to two years and suggests investigating by history, examination, CXR, FBC, LFT, calcium and urinalysis. In addition the clinician should consider mammography and abdo-pelvic CT scan for those patients over 40 years old.
Indications for thrombolysis via catheter to the DVT site (urokinase, rt-plasminogen activator) for iliofemoral DVTs are: if it is less than 14 days old, the patient has low bleeding risks, is expected to live over a year and has good functional status.
Compared to using heparin, thrombolysis reduced the incidence of post-thrombotic leg changes and reduced hospital stay without a significant difference in recurrent VTE or bleeding rates.1 Other treatments are thrombectomy and insertion of an IVC filter.
There is little about pulmonary vein thrombosis in the literature but presumably patients, particularly with lung diseases, may incur DVTs of the pulmonary circulation. Present therapy for patients with VTE involves the thrombin inhibitors, F10 inhibitors and warfarin.
The fifth article in this series reviews prevention of PE and DVT.
Conflict of interest: none declared
1. Venous thromboembolic diseases: the management of venous thromboembolic diseases and the role of thrombophilia testing. June 2012. www.nice.org.uk
2. Varicose veins in the legs: full guideline. cg 168 July 2013. www. nice.org.uk
3. Kearon C. Four Topics in Venous Thromboembolism. Circulation 2003; 107: 1-18
4. Rogers MAM, Blumberg N, Flanders SA, et al. Triggers of Hospitalization for Venous Thromboembolism Circulation 2012; 125: 2092–99
5. Geerts WH, Heit JA, Clagett GP, et al. Prevention of venous thromboembolism. Chest 2001; 119(1 Suppl): 132S–5S
6. Cogo A, Lensing AW, Prandoni P, Hirsh J. Distribution of thrombosis in patients with symptomatic deep vein thrombosis. Implications for simplifying the diagnostic process with compression ultrasound. Arch Intern Med 1993; 153(24): 2777–80
7. Huerta C, Johansson S, Wallander MA et al. Risk factors and short-term mortality of venous thromboembolism diagnosed in the primary care setting in the United Kingdom. Archives of internal medicine 2007; 167(9): 935–43
8. Schouten HJ, GeersingG J, Koek HL, et al. Diagnostic accuracy of conventional or age adjusted D-dimer cut-off values in older patients with suspected venous thromboembolism: systematic review and meta-analysis. BMJ 2013; 346: 2492
9. Lader E, Martin N, Cohen G, et al. Warfarin therapeutic monitoring: is 70% time in the therapeutic range the best we can do? J Clin Pharm Ther. 2012; 37(4): 375–77
10. Rivaroxaban for the treatment of deep vein thrombosis and prevention of recurrent deep vein thrombosis and pulmonary embolism. July 2012 TA 261. www.nice.org.uk