Current strategies in prophylaxis and treatment of venous thromboembolism

Abstract

During the past two decades well designed randomized controlled trials have addressed every important aspect of prophylaxis against venous thromboembolism (VTE) and treatment of established VTE. In addition, systematic reviews and meta-analyses have been performed in most of these areas. The most important advances are the widely accepted and effective VTE prophylaxis in major orthopedic surgery and the shift from in-hospital to outpatient treatment of a majority of patients with established VTE. The major hurdles at this point are efficient selection of the appropriate medically ill patients for prophylaxis and the inconvenience of long-term secondary prophylaxis with vitamin K antagonists. Although new, orally available anticoagulants without need for monitoring or dose adjustments have been in clinical trials for a decade, none is approved yet, and will not be discussed in this review.

• Deep vein thrombosis
• heparin
• prophylaxis
• pulmonary embolism
• surgery
• treatment
• venous thromboembolism
• vitamin K antagonists

Incidence of venous thromboembolism

Epidemiological studies in the Western hemisphere have shown that the first event of symptomatic venous thromboembolism (VTE) occurs in approximately 100 patients (71–117 patients) per 100,000 inhabitants per year 1. The disease increases exponentially with age and is more common among Caucasians than Asians. One-fifth each of the events is associated with surgery or cancer, but unexplained (‘idiopathic’) VTE constitutes in many materials up to 50% of the cases. In studies with clinical materials two-thirds of the patients have deep vein thrombosis (DVT), and one-third have pulmonary embolism (PE), whereas in autopsy studies a majority of the cases have PE.

Without effective prophylaxis against VTE the risk of fatal PE after surgery is considerable. Autopsies of 1,234 patients who died within 30 days after surgery showed that PE was the cause of death in 29% 2. Implementation of protocols for thromboprophylaxis in hospitalized patients after surgery is easy and results in significant reduction of morbidity and mortality due to VTE. For some subsets of surgical patients, and particularly for medical patients, the increased risk of VTE continues beyond the period of hospitalization, and extended prophylaxis may be warranted but the optimal duration has been difficult to define.

Prophylaxis against venous thromboembolism

Surgery and trauma

The methods and agents used for prophylaxis against VTE are shown in Table I. In addition, choice of optimal surgical procedure, e.g. endoscopic instead of open, other methods to reduce surgical trauma, mode of anesthesia, e.g. neuraxial instead of general, avoidance of infectious complications, adequate fluid substitution, and early mobilization contribute to reduction of postoperative VTE. Some surgical procedures carry such a high risk of VTE that prophylaxis is universally required, for example major orthopedic surgery and major surgery for cancer. In other cases the individual risk factors as well as the duration of surgery should decide whether prophylaxis is needed. Thus, patients with previous VTE, previously diagnosed thrombophilic defects, and those with active cancer should be adequately protected for any surgery except very minor procedures. Other risk factors that justify thromboprophylaxis are obesity (body mass index >30), pregnancy, current medication with estrogens, and extensive, symptomatic varicose veins. Surgical risk factors that should prompt prophylaxis are reoperation within 30 days and surgical procedures exceeding 90 minutes. The incidence of VTE also increases with the age of the patient, so that with increasing age prophylaxis should be offered also for shorter procedures.

Key messages

• Appropriate prophylaxis against venous thromboembolism (VTE) relies on knowledge of evidence-based regimens and adjustment for additional risk factors for thrombosis and risk of bleeding.

• Patients at high risk of (recurrent) VTE should receive prophylaxis for most surgical procedures.

• In medically ill patients with high risk of VTE, prophylaxis will reduce the incidence of symptomatic events but not mortality, and user-friendly routines are crucial for compliance.

• The majority of patients with VTE should be treated as outpatients with low-molecular-weight heparin and receive individually tailored duration of secondary prophylaxis.

• The duration of secondary prophylaxis is controversial and influenced by priorities regarding reduction of recurrences, reduction of bleeding complications, and convenience.

Table I.  Methods of prophylaxis against venous thromboembolism.

Mechanical:
Graduated compression stockings
Intermittent pneumatic compression devices
Venous foot pumps
Electrical calf muscle stimulation
Pharmacological:
Unfractionated heparin
Low-molecular-weight heparin
Pentasaccharide (fondaparinux)
Heparinoid (danaparoid)
Vitamin K antagonists
Dextran
Factor Xa inhibitors ^a
Direct thrombin inhibitors ^a
Acetylsalicylic acid

^aSeveral agents are in clinical trials: some have completed phase III but have not yet been approved at the time of writing.

Some surgical procedures have in large studies been associated with a low risk of VTE and do not require prophylaxis (Table II) except in patients with a high risk due to, for example, previous VTE.

Table II.  Surgical procedures that do not require routine prophylaxis against venous thromboembolism.

Laparoscopy ^a

Appendectomy ^a

Inguinal hernia repair ^a

Mastectomy ^a

Thyroidectomy

Transurethral resections

Vascular surgery

Knee arthroscopy

Fixation of fractures of the upper limb or below the knee ^a

^a Patients with previous thromboembolism and the elderly (>70 years) should be considered for prophylaxis.

In orthopedic surgery of the lower limb there is an increasing risk of DVT the more proximal the anatomic site is. The risk increases even further with spinal cord injury, particularly if there is paraparesis.

For fractures in the leg below the knee venographic studies have shown DVT in 10%–45% of the patients 3, but the thrombi are generally small, confined to the calf veins, asymptomatic, and of disputable clinical significance. This is true both for fractures treated conservatively and with open reduction and internal fixation. Prophylaxis against VTE is not routinely required except in the presence of additional risk factors.

In patients with femoral shaft fractures and screening with ultrasound the incidence of DVT was 30% 4, and a few percent of the patients have symptomatic VTE. Thromboprophylaxis should be given routinely, preferably with low-molecular-weight heparin (LMWH) until discharge from hospital and at least for 1 week.

Hip fractures carry a high risk of DVT, and even more worrisome is the risk of fatal pulmonary embolism of 2.5%–7.5% in the absence of prophylaxis 3. In an autopsy study 14% of 581 deaths after hip fracture were caused by pulmonary embolism 5. Patients need prophylaxis already before surgery, which may be delayed by several days, either with LMWH or with unfractionated heparin (UFH), 5,000 U every 8 h. After surgery, fondaparinux 2.5 mg daily is more effective than the LMWH enoxaparin at a dose of 40 mg daily 3, but perhaps equal to enoxaparin 30 mg twice daily.

The incidence of DVT on venography is higher after total knee replacement (TKR) than after total hip replacement (THR), 41%–85% versus 42%–57% 3, but for both surgeries an increasing number of hospitals now provide prophylaxis routinely. In the US the agent of choice has traditionally been warfarin but is progressively shifting to LMWH, which is dominating in the rest of the world. Another alternative is fondaparinux, although there seems to be more surgical site bleeding with this agent. The anti-Xa inhibitor rivaroxaban, which has not yet been approved, appears to be more effective than LMWH after TKR without causing more bleeding 6. The direct thrombin inhibitor dabigatran provides similar thromboprophylactic effect and safety as enoxaparin 40 mg once daily after THR or TKR and has now been approved in Europe on these indications 7, 8. Both rivaroxaban and dabigatran are administered orally and do not require monitoring of coagulation parameters or dose adjustments. For elective surgery prophylaxis may be started after surgery, preferably within 6–8 h after end of surgery to optimize the efficacy. The duration of treatment should be at least 10 days after TKR, based on evidence from several trials 3. Patients with THR or hip fractures continue to have a high risk of symptomatic DVT for several weeks and require extended prophylaxis for about 1 month.

In patients with pelvic fractures the high risk of thrombosis has to be balanced against the bleeding, which may be severe after pelvic trauma. When the patient is hemostatically stable a high prophylactic dose of LMWH, e.g. enoxaparin 30 mg twice daily, is warranted.

In elective spine surgery the risk of VTE is generally low when the patient can be mobilized early, and prophylaxis is only required in case of additional risk factors 3. Mechanical methods or LMWH have been studied and are effective. With spinal cord trauma and surgery the risk of DVT increases so that 60%–100% of the patients have DVT on screening if no prophylaxis is given 3. Enoxaparin 30 mg twice daily is best studied, but the risk of bleeding complications is not negligible (approximately 2% major bleeding). It is necessary to delay prophylaxis until 36 h after surgery. Perispinal hematoma and uncontrolled coagulopathy are contraindications, and in those cases mechanical methods should be used. In case of permanent paraplegia the treatment should be extended to 3 months, either with once daily LMWH, e.g. enoxaparin 40 mg, or by switching to a vitamin K antagonist (VKA).

All of the above discussed trauma situations may be combined with head injury and intracranial bleeding, which is a contraindication for antithrombotic agents. When there are also leg injuries, precluding application of intermittent pneumatic compression or graduated compression stockings, screening with leg Doppler for DVT should be considered. In case a proximal DVT develops and at that time prophylaxis still is unfeasible, a retrievable inferior vena cava (IVC) filter should be inserted. However, prophylactic insertion of IVC filters before occurrence of DVT has not demonstrated any reduction of PE in trauma patients.

In general surgery low-dose UFH (5,000 U twice daily), LMWH (2,500–3,400 IU once daily) or fondaparinux may be used, according to the guidelines mentioned initially in this section. In patients with high risk of bleeding mechanical methods should be used instead. After extensive surgical procedures for malignancy the prophylactic dose should be increased (UFH 5,000 U three times daily; LMWH 3,500–5,000 IU daily) and extended duration to 1 month considered 3.

In neurosurgery the risk of any VTE is 15%–40%. Patients with surgery for intracranial tumors have a very high risk with an incidence of symptomatic DVT of 30% during the first month without prophylaxis 9. Graduated compression stockings reduce the risk of VTE by 50% 10, and all patients with neurosurgery should wear knee-high stockings. For cancer patients postoperative prophylaxis with UFH 5,000 U twice daily should be considered, starting 24 h after surgery if a computed tomography (CT) scan is negative for intracranial hematoma. The risk of bleeding appears to be lower with UFH than with LMWH in this setting, at least during the early postoperative phase.

Medically ill patients

Medical patients have a risk of symptomatic DVT during hospitalization of 0.5%–1.5% 11. Whereas 70%–80% of all fatal PE occur in medical patients, studies on anticoagulation prophylaxis have not demonstrated any reduction in mortality 12. There is, however, a statistically significant decrease in the incidence of symptomatic PE and asymptomatic DVT without a corresponding increase of major bleeding. Only about 40% of medically ill patients, who fulfill current criteria for prophylaxis during hospitalization, actually receive it according to the worldwide ENDORSE survey 13. The prevailing difficulty is to select the correct patients for prophylaxis, and audits have shown that even when a moderately high proportion of hospitalized patients receive prophylaxis it is often given to those who should not have been treated 14. Electronic alerts were shown to significantly reduce the incidence of symptomatic VTE at 90 days by 41% (hazard ratio 0.59; 95% confidence interval (CI) 0.43–0.81) 15 and were based on a relatively simple algorithm (Table III). The dose of LMWH should be in the high prophylactic range, i.e. 3,500–5,000 IU once daily, based on the MEDENOX trial, which demonstrated that 20 mg of enoxaparin (2,000 IU) was not more effective than placebo 16. If UFH is used, a dose of 5,000 U three times daily has a similar effect to the recommended dose of LMWH 17, 18, but there are fewer adverse events and improved convenience with LMWH. Fondaparinux at a dose of 2.5 mg daily is another alternative, which in the ARTEMIS trial resulted a reduction of symptomatic VTE versus placebo (0 and 5 cases, P=0.029) and a trend to lower mortality at 32 days (P=0.06) 19. Since many medically ill patients continue to suffer from, for example, congestive heart failure or respiratory failure after discharge from hospital and remain immobile, it is not surprising that the risk of VTE continues to be high for several weeks or perhaps longer. Extended prophylaxis from 10 days to 28 days with enoxaparin 40 mg once daily versus placebo resulted in a reduction of symptomatic VTE (P=0.003), proximal VTE (P=0.0019), and all VTE (P=0.0011) but unfortunately also in an increase of major bleeding (P=0.038) 20. Thus, extended prophylaxis for medical patients is not recommended until better selection criteria have been identified.

Table III.  Criteria for selection of medically ill patients for prophylaxis against venous thromboembolism used in conjunction with electronic alerts.

1 point each:

Age ≥70

Obesity (body mass index >29)

Bed rest

Hormone replacement therapy or combined oral contraceptives

2 points:

Major surgery

3 points each:

Cancer

Prior venous thromboembolism

Hypercoagulability

Increased risk if at least 4 points at any time point after admission.

Patients with central venous catheters have an increased risk of DVT in the vein where the catheter is placed due to reduced blood flow and endothelial damage by the high concentration of infused medication or nutrients as well as by the catheter. Recent studies failed, however, to show any benefit of anticoagulant thromboprophylaxis due to an overall low event rate. This may be due to improved catheter materials or other routines, and routine prophylaxis is therefore not recommended 3.

Initial treatment of venous thromboembolism

Thrombolytic therapy

The objectives of the initial treatment of VTE are to reduce the risk of extension of the disease, which ultimately may be fatal. For patients with symptomatic DVT alone the goal is to avoid involvement of additional venous segments in order to limit the damage to the venous valves and also to abolish the risk of embolization to the lungs. For patients with symptomatic PE the main objective is to prevent additional emboli with ensuing circulatory failure. In addition, for patients with massive pulmonary embolism a rapid reduction or preferably elimination of the embolic burden is needed to reduce mortality. This can be achieved with thrombolytic therapy or by mechanical removal of the emboli.

In DVT thrombolytic therapy is less in demand since mortality will not be reduced. A systematic review of 12 studies showed that the DVT was resolved more often with thrombolysis than with heparin, relative risk (RR) 0.24 (95% CI 0.07–0.82), and the clinically most relevant end point postthrombotic syndrome was also reduced (RR 0.66, 95% CI 0.47–0.94) 21. The disadvantage was more bleeding complications (RR 1.73, 95% CI 1.04–2.88), and there were two strokes in the thrombolysis group 21. In clinical practice thrombolysis for DVT is used for a very small subset of relatively young patients with extensive DVT, particularly when the leg is very painful and cyanotic (phlegmasia coerulea dolens) or so edematous that it becomes pale from compromised arterial circulation (phlegmasia alba dolens). There are a variety of treatment regimens including 1) systemic infusion, usually with streptokinase but sometimes with tissue plasminogen activator (tPA), 2) loco-regional infusion with streptokinase, tPA, or urokinase, and 3) catheter-directed with tPA. The dose regimens have also varied widely, and it may be preferable to follow local experience and protocols.

In massive PE when the patient is hemodynamically unstable thrombolytic therapy appears to reduce mortality 22, 23, but in a systematic review of all randomized trials there was no such benefit 24. In a trial with 256 patients suffering from submassive PE with pulmonary hypertension or right ventricle dysfunction, thrombolytic treatment with alteplase and heparin was better than heparin alone on a composite efficacy end point, but there was no difference in mortality 25. Thus, it is not clear which patients with submassive PE should be selected for thrombolysis, but the risk of non-major bleeding is higher with thrombolysis than with heparin, as demonstrated in a meta-analysis of 11 trials (odds ratio (OR) 2.63, 95% CI 1.53–4.54) 23. The best-documented regimen is with alteplase, 100 mg infused systemically over 2 hours with 10% of the dose given as an initial bolus.

Initial anticoagulant therapy

In the majority of patients the initial treatment consists of LMWH or UFH. The advantages of LMWH are shown in Table IV. In a meta-analysis of 22 trials initial LMWH resulted in a lower risk of recurrent VTE (OR 0.68, 95% CI 0.55–0.84), less major hemorrhage (OR 0.57, 95% CI 0.39–0.83), and lower mortality (OR 0.76, 95% CI 0.62–0.92) 26. Once daily subcutaneous injection with LMWH allows for outpatient treatment, usually managed by the patient. There is no advantage of twice daily injections with LMWH 27. Treatment with fondaparinux is equally effective and safe, as shown in the MATISSE studies for DVT 28 and PE 29. The dose is 7.5 mg once daily subcutaneously, except 5 mg for patients with <50 kg body weight and 10 mg for those with >100 kg body weight. If cost of LMWH or fondaparinux is an issue, subcutaneous injection with UFH twice daily is effective and safe 30, and this can probably also be done without monitoring, as recently demonstrated 31.

Table IV.  Advantages of low-molecular-weight heparin (LMWH) compared to unfractionated heparin (UFH).

Characteristic	Low-molecular-weight heparin	Clinical implication
Half-life	Longer (3–6 h versus 30–60 min)	Once daily injection possible
Bioavailability, subcutaneous injection	Better (≈100%)	Subcutaneous equal to intravenous route
Protein binding	Less	Predictable response, no monitoring
Interaction with platelets	Less	Lower risk of heparin-induced thrombocytopenia
Activation of osteoclasts	Less	Minimal risk of osteoporosis

Intravenous infusion of UFH remains the treatment of choice for very ill patients, in situations where quick reversal of the anticoagulant effect may be necessary, and in patients with massive thromboembolism. The initial dose should be according to body weight 32 and thereafter adjusted according to activated partial thromboplastin time.

In patients with renal failure the elimination of LMWH and fondaparinux is impaired. The risk of hemorrhage increases generally when the creatinine clearance is reduced below 30 mL/min (OR 2.25, 95% CI 1.19–4.27) 33. Tinzaparin has the highest molecular weight among the LMWHs and can be used with creatinine clearance down to 20 mL/min 34. Beyond that limit UFH is required.

Patients with morbid obesity should have the dose of LMWH increased linearly according to body weight 35 rather than capping the dose. Monitoring with anti-factor Xa levels is prudent in such cases.

The initial treatment with LMWH/UFH continues normally until a vitamin K antagonist (VKA), started in parallel, achieves the therapeutic range of international normalized ratio (INR) 2.0–3.0 for two consecutive days and also for at least 5 days. Patients with VTE and cancer should, however, preferably remain on treatment with LMWH for 3–6 months, which in the CLOT study with 672 patients as well as in three smaller studies provided better efficacy than VKAs 36.

Secondary prophylaxis with LMWH instead of VKA may also be considered in a few other situations (Table V).

Table V.  Clinical conditions suitable for secondary prophylaxis with low-molecular-weight heparin (LMWH)

Condition with venous thromboembolism	Rationale
Cancer	Improved efficacy (and survival)
Lupus anticoagulant with elevated baseline INR	False therapeutic INR with VKA; LMWH does not require monitoring
Pregnancy	Avoid teratogenicity in first trimester; reduce bleeding complications in premature delivery
Planned very short treatment duration	Avoid initial phase of frequent blood tests with VKA
Intestinal disease with poor absorption	VKA therapy may be impossible to stabilize

INR = international normalized ratio; VKA = vitamin K antagonist.

Secondary prophylaxis

Treatment with VKA is started with the anticipated maintenance dose or a slightly higher dose, for example with 5 or 10 mg of warfarin. Monitoring of VKA therapy is challenging and should be handled by staff with experience and in charge of at least 30–40 patients on this medication. Dosing support with computer software or nomograms improves the quality of anticoagulant treatment 37–39. For highly selected, motivated patients, and provided that reimbursement is not an issue, self-testing and self-dosing (i.e. self-management) after appropriate training improves outcome. A meta-analysis of ten trials with 2,724 patients randomized to self-management or conventional monitoring demonstrated a reduction of mortality (RR 0.48, 95% CI 0.29–0.79) and other major complications (RR 0.58, 95% CI 0.42–0.81) 40. A substantial proportion of the variability in dose requirements of warfarin is explained by polymorphisms in the genes coding for the target enzyme vitamin K epoxide reductase (VKORC1) and the microsomal enzyme mainly responsible for metabolism, CYP2C9 41. It remains to be demonstrated that genetic screening for these polymorphisms confers any advantage regarding safety or health economy.

The duration of secondary prophylaxis is controversial and influenced by priorities regarding reduction of recurrences, reduction of bleeding complications, and convenience. The shortest duration that may be recommended is 6 weeks for patients with distal DVT and removable risk factor, e.g. postoperative calf vein thrombosis 42, 43. These patients have a risk of recurrence of about 1% per year over 6 years 44. At the other extreme there is essentially consensus that indefinite duration is applicable for patients with three episodes of any VTE, two or more unprovoked episodes of VTE, single life-threatening VTE in combination with any thrombophilic defect, single unprovoked VTE in combination with severe or combined thrombophilic defects, or any VTE in a patient with non-curable cancer. For the majority of patients, who do not fulfill these criteria, the decision is a balancing act. At the time of initiation of VKA therapy it can only be a preliminary decision. Reassessment has to be made after 6–12 months, reviewing patient compliance, bleeding events, and comorbidities. The influence of factors to be taken into account is illustrated in Figure 1. Patient preferences should weigh in heavily but only after the patient has been well informed of the objectives of treatment and the possible consequences of recurrent VTE without VKA versus bleeding with VKA. The majority of patients with DVT will in case of recurrence again have DVT—with an increasing risk of development of postthrombotic syndrome, which in about 5% of patients with VTE results in a venous leg ulcer 45. Likewise, the majority of patients with initial PE will in case of recurrence again have PE 45, with an annual risk of fatality in 0.5% 46 and a risk of pulmonary hypertension of 3%–4% 47. The risk of major bleeding is highest during the first months of treatment. For patients with more than 3 months of anticoagulation the rate of intracranial bleeding is 0.65 per 100 patient-years (95% CI 0.63–0.68 per 100 patient-years), and the case fatality rate of major bleeding is 13.4% (95% CI 9.4%–17.4%) 48.

Figure 1.  Influence of risk factors for bleeding and recurrent thromboembolism on decision regarding duration of secondary prophylaxis. VTE = venous thromboembolism; ASA = acetylsalicylic acid; INR = international normalized ratio; DVT = deep vein thrombosis.

Once the decision is made to discontinue VKA therapy, this can be done abruptly, since no study has shown a clinical advantage of tapering the VKA.

Supportive therapy

Supportive therapy consists of early mobilization and compression hosiery. Three studies with randomization to early mobilization versus bed rest during the initial treatment showed no difference in the risk of pulmonary embolism, extension of the DVT, or deaths 49–51. The endogenous fibrinolysis improves with physical activity, and outpatient treatment with LMWH goes well along this concept.

Two large studies with compression stockings (30–40 mmHg) versus control showed a reduction of any form of postthrombotic syndrome (OR 0.31, 95% CI 0.20–0.48) as well as of the severe syndrome (OR 0.39, 95% CI 0.20–0.76) after 2 years 52, 53. Conversely, a meta-analysis did not find that anti-inflammatory therapy during the initial phase contributed significantly to clinically relevant improvement 54.