Pharmacology

LMWHs are obtained by chemical or enzymatic depolymerisation of UFH, a heterogeneous mixture of polysaccharides with an average molecular weight of 15,000 daltons. The LMWHs used clinically in the UK have mean molecular weights in the range 4,000-6,000 daltons.2 LMWHs are weaker inhibitors of thrombin (factor IIa) than UFH but inhibit the coagulation enzyme Xa to a similar degree. The ratios of anti-Xa to anti-IIa activity of LMWHs are therefore 2-4 times higher than that of UFH. Anti-Xa activity appears to be a major, but probably not the sole, determinant of the anticoagulant activity of LMWHs.2^,3

Pharmacokinetics

After subcutaneous (s.c.) injection, the shorter chain LMWHs are better absorbed than UFH, and bind less to proteins in plasma and in the endothelial wall.4 As a result, the bioavailability of LMWHs, as measured by anti-Xa activity in plasma, is around 90% at all s.c. doses, while that for UFH increases with the dose and is only about 10-30% at prophylactic doses.4–6 After s.c.injection, the plasma half-life of LMWHs (as measured by anti-Xa activity) is around 4 hours, sufficient to give an effective anticoagulant action with once-daily injection at currently recommended doses; that of UFH is 1.5 hours,2 so doses of UFH must be given more frequently. With UFH, variation in the anticoagulant effect between and within individuals means that regular monitoring is needed to adjust the therapeutic dose; this is not usually necessary with LMWHs.

UFH is partly metabolised in the liver and partly eliminated by the kidney; clearance is dose-dependent. Elimination of LMWH is mainly by the kidneys; clearance is less dose-dependent but may be reduced in patients with renal failure.4

The LMWHs differ slightly in their structure, in-vitro activity (e.g. anti-Xa:anti-IIa activity) and pharmacokinetics but there is no evidence that they differ significantly in their clinical effects for a given anti-Xa activity. All LMWH preparations licensed in the UK are calibrated against an International Standard for LMWH,7 with the potency of each preparation expressed in International Units (IU) for anti-Xa activity.

Use for surgical prophylaxis

General and gynaecological surgery

Many randomised double-blind studies have shown that, in patients undergoing major abdominal, gynaecological or chest surgery, prophylaxis using the currently available LMWH preparations is at least as effective as UFH (5000 units s.c. every 8 or 12 hours) in preventing DVT (detected by routine screening).8–11 Many patients in these trials had additional risk factors for venous thromboembolism, such as malignancy, previous DVT or obesity. In one meta-analysis of 17 randomised studies (involving a total of 6878 patients), both the incidence of DVT (5.3% with LMWH vs. 6.7% with UFH) and of PE (0.31% vs. 0.70%) were significantly lower in patients receiving LMWH than in those given UFH.12 However, these differences disappeared when the analysis was limited to the 8 studies with the soundest methodology. In a second, larger meta-analysis (25 studies, 9683 patients) there was no significant advantage in favour of LMWH.13 Neither analysis found any significant difference between the LMWHs and UFH in the incidence of major bleeding. One later study, in 3809 patients undergoing abdominal surgery, found that wound haematoma, severe bleeding, or re-operation to control bleeding were significantly more common in those receiving UFH (5000 units s.c. twice daily) than in those given prophylaxis with dalteparin (2500 IU s.c. once daily).11

Orthopaedic surgery

After major orthopaedic surgery, such as hip or knee replacement, DVT can be demonstrated in 50-65% of patients if no anti-thrombotic measures are used.4 A meta-analysis of six randomised controlled trials involving a total of 1294 patients, mostly undergoing elective hip replacement, found that treatment with LMWH reduced the incidence of DVT (detected by routine venographic screening) more effectively than UFH (13.8% vs. 21.2%).12 A second meta-analysis (14 studies, 2692 patients) found a non-significant trend in favour of LMWH (17.8% vs. 19.35%),13 while a further, recent meta-analysis (29 studies, about 7000 patients) again found significant advantage with LMWH (p<.05).14 In these analyses, the incidence of fatal and non-fatal PE was generally lower during prophylaxis with LMWH than with UFH but the event rate was usually too low to demonstrate a significant difference. None of the analyses found a significant difference between LMWH and UFH in the incidence of major bleeding.12–14 However, in the most recent meta-analysis, which also included some dose-ranging studies, the incidence of minor bleeding (haematoma of the wound or injection site) was significantly lower in patients receiving LMWH than in those given UFH.14

Risk of venous thromboembolism persists for several weeks after hip replacement. In two randomised, double-blind, placebo-controlled trials (total of 406 evaluable patients), extended prophylaxis with enoxaparin for 4-5 weeks after the operation reduced the incidence of late DVT.15^,16 In the 3 weeks after discharge from hospital, DVT (detected by routine venogram) occurred in 7.1% of those assigned to enoxaparin and 19.3% of those given placebo in one study;15 of these, 3.5% and 8.0%, respectively had symptoms or clinical signs of DVT.15 Similar benefit has been reported with extended prophylaxis (35 days) with dalteparin.17

The implications of these findings for orthopaedic practice are not clear-cut: the clinical importance of DVT detected by routine screening in the setting of a clinical trial is not known; and in the absence of prophylaxis, the incidence of fatal PE after hip replacement is probably much lower now (0.1-0.2%) than previously reported (1-3%), because of shorter operations, earlier mobilisation and wider use of regional anaesthesia.18 A very large trial would be needed to investigate whether anticoagulant prophylaxis reduces mortality after major orthopaedic surgery, and there are very few data on whether it can reduce the incidence of post-phlebitic limb. Nevertheless, in the UK, over 80% of orthopaedic surgeons now use anticoagulant prophylaxis (usually LMWH) with hip replacement, either alone or with mechanical measures such as graduated stockings, calf stimulation/compression or use of a foot pump.19 Studies are still needed to clarify the optimum time (pre-operative or postoperative) for starting LMWH.

Prophylaxis after hip fracture

No randomised, double-blind trial has adequately investigated the role of LMWH versus UFH in patients undergoing emergency surgery for hip fracture; small double-blind and open studies have given conflicting results.

Treatment of DVT

Initial treatment for patients with DVT has until recently required admission to hospital and continuous i.v. infusion (or 12 hourly s.c. injection) of UFH, in a dose adjusted according to the activated partial thromboplastin time (APTT), until the patient has been stabilised on warfarin (usually about 5-7 days).

Studies in hospital

Three large randomised studies (double-blind or with independent blinded assessment of outcome), involving a total of 736 inpatients with proven DVT,20–22 and several smaller studies, have shown that initial treatment with LMWH is at least as effective and safe as conventional i.v. UFH. Patients assigned to LMWH received a fixed, weight-adjusted dose (tinzaparin,20 nadroparin,21 enoxaparin22) by s.c. injection for 6-10 days. Warfarin was started after 2-10 days and continued for at least 3 months. The pooled results from these three studies20–22 showed that initial treatment with LMWH reduced the incidence of symptomatic recurrent venous thromboembolism more effectively than UFH during the first 15 days of therapy (0.8% vs. 3.2%; p=0.02) and over the period of anticoagulation as a whole (2.9% vs. 6.4%; p=0.006).23 Meta-analyses of these and other studies also found lower mortality in patients given LMWH, relative to those given UFH, chiefly because of unexplained better survival in the subgroup of patients with cancer.23^,24 In most studies conducted, the incidence of major bleeding did not differ between the treatments,3 but two meta-analyses have found significant advantage from use of LMWH.23^,24 This is due to the inclusion of one particular trial20 which, exceptionally, showed a very significant reduction (p=0.006) in the incidence of major haemorrhage with LMWH compared with UFH.

Treatment with LMWH outside hospital

In two large studies, a total of 900 patients with proximal DVT were randomised to receive UFH, given in hospital by i.v. infusion, or a fixed, weight-adjusted s.c. dose of LMWH (enoxaparin25 or nadroparin26), administered wholly or partly at home by the patient, a nurse or a relative. All patients started warfarin within 48 hours of randomisation. Up to two-thirds of those screened were excluded from the studies because of co-existing disorders that required hospital admission or increased the risk of bleeding, short life expectancy or geographic inaccessibility.25 A third study randomised 1021 patients with venous thromboembolism (27% with PE) to i.v. UFH or s.c. LMWH.27 In this trial, feasibility of home treatment was not considered in assessing eligibility, but 42% of those with DVT were treated wholly or partly at home.

In these studies, which used independent, blinded assessment of outcome events over 3 months, there were no significant differences between the treatment groups in terms of symptomatic recurrence of thromboembolism, mortality or incidence of major bleeding, nor in the time to achieve anticoagulant control with warfarin. Patients randomised to LMWH spent an average of 3-5.4 fewer days in hospital than those receiving UFH.25–27

Treatment of pulmonary embolism

In a European study, 612 patients with symptoms and signs of PE were randomised to conventional therapy with i.v. UFH or to tinzaparin in a fixed, weight-adjusted s.c. dose, each for a mean of 7 days; warfarin was started on day 1-3.28 About two-thirds of all patients received therapeutic doses of UFH for a mean of 18 hours before randomisation. Patients with massive PE, requiring thrombolytic therapy or embolectomy, were excluded; nonetheless, 28% of randomised patients had clinical features of major embolism (cyanosis, syncope, acute right ventricular strain) and nearly half had 50% pulmonary vascular obstruction on perfusion scan. During the first 8 days, 9 patients in each group died or had recurrent thromboembolism or major bleeding; during 90 days of anticoagulant therapy, these events occurred in 18 (5.9%) patients (12 deaths) initially given tinzaparin and 22 (7.1%) of those given UFH (14 deaths); the differences were not significant.

Unwanted effects

Bleeding

The balance of evidence suggests that, during both prophylaxis12–14 and treatment of venous thromboembolism, the incidence of major haemorrhage (e.g. retroperitoneal or intracranial bleeding, bleeding necessitating re-operation or withdrawal of heparin, or causing a significant fall in haemoglobin) is about the same with currently recommended regimens of LMWH as it is with the corresponding UFH regimens.3 The incidence of minor bleeding is also similar.3 Bleeding is more common with higher levels of anti-Xa activity;29 in one study, it was most likely to occur if mean plasma levels of anti-Xa exceeded 0.8 IU/mL.30

Thrombocytopenia

Both UFH and LMWH can cause thrombocytopenia (platelet count below 100x10/L). The most severe manifestations occur with immune-mediated thrombocytopenia which typically develops after 5-10 days' treatment and is associated with heparin-dependent IgG antibodies and risk of severe venous and/or arterial thromboses.31 Immune-mediated thrombocytopenia occurs in about 2-3% of patients treated with UFH irrespective of the dose.32 In one trial, heparin-dependent antibodies developed less frequently with LMWH than with UFH (2.2% vs. 7.8%)32 but the exact incidence of immune-mediated thrombocytopenia in patients receiving LMWH is not known. Therefore, in any patient receiving UFH or LMWH, at whatever dose, a full blood count should be done on the 4th-6th day of treatment (repeated a few days later if heparinisation is continued). UFH or LMWH must be stopped immediately if thrombocytopenia is found at this stage in treatment, and the cause sought. Alternative anticoagulant therapy should be given: substitution of LMWH for UFH or for another LMWH is not safe as cross-reaction may occur. The heparinoid, danaparoid, is usually a suitable choice as cross-reactivity is rare.

Osteoporosis

The data available suggest that osteopenic effects may be less pronounced during long-term administration of LMWH than with UFH.33^,34 In one randomised trial in 80 older patients (mean age 68 years) with DVT, in whom warfarin was contraindicated, fewer vertebral fractures occurred in those given dalteparin for 3-6 months (1 fracture; 2.5%) than in those receiving UFH (6 fractures; 15%).34

Other effects

Local skin reactions at the injection site (erythema, induration, pruritus), or, rarely, systemic reactions (urticaria, angio-oedema, anaphylaxis) can occur. Abnormal liver function tests have been reported.

Dosage and administration

The manufacturers' recommended dosing schedules for prophylaxis and treatment are summarised in the Table (page 28). In patients at moderate risk of venous thromboembolism, prophylaxis with LMWH is given in a fixed, once-daily dose, beginning 1-2 hours before surgery and continued for 5-10 days or until the patient is mobile or the risk diminishes. The manufacturers of enoxaparin and dalteparin recommend a higher fixed dose in patients considered at high risk of venous thrombosis, beginning up to 12 hours before or on the evening before surgery; for tinzaparin, the higher dose is adjusted to the patient's weight.

For the treatment of DVT or PE, LMWH is given for 5-6 days and until the patient is established on warfarin (which should be begun on day 1), with an INR greater than 2.0 for at least 2 consecutive days. Dalteparin and tinzaparin are given once daily, enoxaparin every 12 hours, all three at a weight-adjusted dose.

Monitoring and precautions

For most patients receiving prophylactic or therapeutic doses of LMWH, routine laboratory monitoring of anticoagulant activity is unnecessary.35 Monitoring may, however, be indicated if there are complications such as bleeding, extension or recurrence of thrombosis, or accidental overdose;35 a target plasma range of 0.5-1.0 IU anti-Xa activity per mL is usually advised during treatment, with peak levels measured 3-4 hours after s.c. injection. Tests used routinely to monitor UFH, such as the APTT and thrombin time, are unsuitable for monitoring LMWH.

The summaries of product characteristics (SPCs) also variously advise "caution", "close monitoring" or consideration of a lower and/or divided dose in patients at increased risk of bleeding, such as those with severe renal or hepatic insufficiency, a haemostatic defect or recent haemorrhage. However, targets for anticoagulant activity are better established for UFH than for LMWH:35 in difficult cases, where monitoring might be needed, use of UFH is probably preferable.

Contraindications

Although varying in detail, most SPCs advise that use of LMWH should be avoided in any patient with a generalised bleeding tendency, active peptic ulcer, cerebral haemorrhage, severe hypertension, recent injury or surgery to the CNS, eyes or ears, bacterial endocarditis, heparin-induced immune thrombocytopenia or known hypersensitivity to heparin. The SPC for dalteparin advises that, because of the risk of haemorrhage, regional anaesthesia should not be given in patients receiving the drug as treatment for established DVT.

Cost considerations

Prophylaxis

In patients considered at moderate risk of venous thromboembolism, prophylaxis with low-dose UFH is as effective as LMWH and costs less. In patients undergoing major orthopaedic surgery, LMWH appears more effective than UFH in preventing DVT, as detected by routine screening.12^,36 Cost analyses based on these findings have concluded that prophylaxis with enoxaparin37 or other LMWHs38 is cheaper overall than with UFH. We know of no such analyses for high-risk patients undergoing general surgery where any clinical advantage for LMWH appears marginal.

Treatment of DVT

For the initial treatment of DVT, the LMWH regimen is simpler for patients and their medical carers than i.v. UFH, and avoids the need for cannulation, a syringe-drive, monitoring of an infusion, and laboratory monitoring. An economic evaluation, based on a large double-blind study in Canada,19 calculated a net saving in all hospital costs from the use of tinzaparin rather than UFH.39 Treatment of selected patients outside hospital could, in theory, lead to further savings, but implementation of a policy of outpatient treatment needs careful planning. It requires agreement at local level on patient eligibility for outpatient treatment and on clinical responsibility for treatment decisions; arrangements need to be in place for access to investigations to confirm the diagnosis of DVT, for daily administration of LMWH, warfarin monitoring and dose adjustment, emergency admission if needed, and involvement of community nurses or specialist anticoagulant nurses.

Please click here to view a table showing the treatment regimens and costs.

Conclusion

Low molecular weight heparins (LMWHs) are an important addition to the range of treatments available for the prevention and treatment of venous thromboembolism. They simplify treatment and require less monitoring by blood tests than conventional unfractionated heparin (UFH), but patients who require careful titration of the anticoagulant effect, such as those with additional risk factors for bleeding, are probably better managed with UFH. Comparative data are at present too few to identify clinically important differences between currently available LMWH preparations. All cost more than UFH.

For prophylaxis in patients undergoing general surgery, any differences in clinical outcome between LMWHs and UFH are marginal. Only in patients undergoing major orthopaedic surgery does there appear to be significant clinical advantage from use of LMWH. For these patients, use of LMWH might also be more cost-effective. Extended prophylaxis with LMWH reduces the incidence of late deep vein thrombosis (DVT) in patients undergoing hip surgery; whether it should be used in all patients, or selected high-risk patients, is not clear.

For the initial treatment of established DVT, use of LMWH in a fixed, weight-adjusted subcutaneous dose is at least as effective and safe as intravenous therapy with UFH in a dose adjusted by laboratory monitoring. It is much simpler to use and probably reduces overall treatment costs. Outpatient treatment with LMWH is now possible for some patients with DVT, but their selection and the organisation of their care requires agreement at local level. In hospital, treatment with tinzaparin has been shown to be an effective alternative to UFH in patients with pulmonary embolism.