Heparin is not absorbed by the oral route, requiring parenteral administration. It is principally metabolized in the liver by heparinase. LMWHs have a higher and more predictable and efficient bioavailability than unfractionated heparin (Table 2). Once absorbed into the blood, serum levels of LMWH remain steady and persist longer than unfractionated heparin. Like unfractionated heparin, LMWH is not well absorbed by muscle, fat, and lymphatic tissue. Despite its smaller size LMWH does not cross the placenta and can safely be used during pregnancy (FDA Pregnancy Category B, Dulitzki 1996). Clearance of LMWH is not dose dependent and occurs through renal mechanisms.

Table 2. Properties of Low Molecular Weight Heparins Compared to Unfractionated Heparin

The longer half-life and better bioavailability of LMWH have been attributed to their decreased affinity and binding to endothelium, macrophages, and unfractionated heparin binding proteins.1,2 The latter are acute phase reactants that increase in response to stress. Binding to heparin-binding proteins alters the bioavailability of unfractionated heparin to a greater degree than that of low molecular weight heparin. The lower incidence of bleeding in patients treated with LMWH is thought to be from its reduced binding to platelets, endothelium, and Von Willebrand factor.3 Because of their predictable antithrombotic response and longer bioavailability, LMWH can be used safely without daily anticoagulation monitoring in the majority of patients. Patients with renal insufficiency and those with body weight less than 50 kg or more than 80 kg4 have variable pharmacokinetics and so plasma antifactor Xa concentrations should be monitored (Table 3).

Table 3. Desirable Properties of Low Molecular Weight Heparins and Possible Reasons

Current approved indications and other therapeutic uses of LMWHs are listed in Table 4 and discussed in detail below. Evidence for the use of LMWH for prophylaxis or treatment of thrombosis is presented. In addition, the benefits, limitations, and economic impact of this therapy are discussed.

Table 4. FDA Approved Indications and Dosage of Low Molecular Weight Heparins

PROPHYLAXIS FOR ARTERIAL OR VENOUS THROMBOSIS

Deep venous thrombosis and pulmonary embolism in general surgical procedures. DVT and PE occur in approximately 20% to 30% of patients undergoing general surgical procedures without any prophylaxis.5 The incidence varies by surgical type and patient-specific risk factors. Unfractionated heparin given perioperatively reduces the incidence of DVT and fatal PE by 70% and 50%, respectively.6 Several studies have been carried out to evaluate the role of LMWH in reducing the incidence of DVT and PE in patients undergoing general surgical procedures.

In a large group of patients undergoing general surgery Boneu and associates studied the efficacy of LMWH in preventing development of thromboembolic complications. One-thousand-three-hundred- fifty-one patients were randomized to receive either 5000 IU subcutaneously of unfractionated heparin or 1750 anti-Xa IU of LMWH (Reviparin-sodium, Knoll, Mt. Olive, NJ) once a day for at least 6 days. Both groups were comparable in terms of type of surgical procedures performed and thromboembolic risk factors. A large number ([lt] 50%) of patients in this study had cancer. In followup analysis the incidence of DVT and PE was 4.8% in the LMWH group and 4.4% in the unfractionated heparin group. This difference was not statistically significant. LMWH was well tolerated with fewer bleeding complications and wound hematomas compared with the unfractionated heparin group.7

In another large study involving patients undergoing major abdominal surgery Kakkar and coworkers established the safety of using LMWH (Fragmin, Pharmacia & Upjohn, Kalamazoo, MI) in the perioperative period. Three-thousand-eight-hundred-nine patients were enrolled in this study. The incidence of perioperative bleeding was 3.6% in the LMWH group and 4.8% in the unfractionated heparin group.8 These results were validated in a meta-analysis by Nurmohamed and colleagues.9 These observations indicate the safety and efficacy of LMWH in the perioperative prophylaxis of venous thrombosis. Cost-effectiveness of LMWH in this setting has not been established.

Orthopaedic surgery. The LMWHs have been studied extensively in orthopaedic surgery patients. Because orthopaedic surgery involves procedures in close proximity to veins, immobilizes the joint, and leaves patients unable to ambulate for varying lengths of time, the risk for venous thrombosis is very high. Of the orthopaedic procedures, those involving the hip and knee carry the highest risk for DVT.

Total hip replacement is associated with a 70% incidence of DVT in the absence of prophylaxis. An elderly patient population and the proximity of the surgery to the major veins are likely contributing factors to the high incidence of thrombosis. Twenty percent of thrombi in this population involve the proximal veins, with a corresponding increased risk of pulmonary embolus. Traditionally, unfractionated heparin has been used perioperatively, and more recently, warfarin has been used to prevent DVT postoperatively in hip replacement patients.

Prophylactic use of unfractionated heparin has been shown to reduce the risk of developing DVT during total hip replacement surgery.10 Recent studies demonstrated that LMWH might be a better alternative than unfractionated heparin in preventing DVT as they are equally safe and effective and are easy to administer.11 These conclusions were also reached in a recently published metaanalysis of 24 studies evaluating LMWH and unfractionated heparin or warfarin in the prophylaxis of DVT.12 Three additional trials compared the effectiveness of LMWH to low dose warfarin in hip replacement patients. In these trials, LMWH demonstrated comparable efficacy in preventing DVT and PE.13,14 The largest of the 3 studies reported by Hull and associates15 included 1207 patients randomized to receive LMWH (Logiparin, Novo/Nordisk Pharm. Inc., Princeton, NJ) or adjusted dose warfarin. On followup venograms, 37.4% of patients in the warfarin group and 31.4% in the LMWH group had evidence of deep venous thrombosis. The risk reduction was statistically significant in favor of LMWH, but there was a higher incidence of bleeding complications in the group of patients treated with LMWH, 1.2% versus 2.8%. Again, LMWH offers the advantage of minimal monitoring requirement.15

The incidence of DVT after total knee replacement surgery, even with the use of prophylaxis, is as high as 30%. With prophylactic, perioperative use of unfractionated heparin or low dose warfarin, the incidence of DVT is decreased by 71%.16 Several studies have demonstrated the superior efficacy of LMWH in preventing the development of DVT in patients undergoing total knee replacement surgery, compared with placebo,16 unfractionated heparin,17 or warfarin.13,15,16,18 In one study, use of LMWH (Logiparin) was associated with increased incidence of bleeding compared with warfarin.15 This was thought to be related to the more rapid onset of anticoagulation achieved with low molecular unfractionated heparin compared with warfarin. It was suggested from the above observation that initiation of LMWH in the prophylaxis of DVT be delayed for at least 12 hours after knee replacement surgery. It should be noted that nonpharmacologic measures such as pneumatic compression devices are not as effective as LMWH.

Spinal cord injury. The use of LMWH has also been evaluated in patients with spinal cord injury. Trauma to the spinal cord can leave patients with complete motor paraplegia placing them at risk for developing deep venous thrombosis. The incidence of DVT in patients with spinal cord injury is 70% and fatal PE develops in 3% to 5%.19 Because of the high incidence and severity of these complications, various measures have been used to prevent thromboembolic complications in this group of patients. Elastic stockings20 and pneumatic compression devices21 either alone or in combination have demonstrated marginal benefit. Unfractionated heparin at a dose of 5,000 IU given subcutaneously twice a day is ineffective in up to 31% of patients with deep venous thrombosis. When titrated to keep the aPTT at the upper limit of normal, unfractionated heparin is more effective in preventing thromboembolic complications but is associated with a 24% risk for bleeding complications.22

In a prospective randomized study in a small group of patients with complete motor paralysis, 21 patients were randomized to receive unfractionated heparin 5,000 IU 3 times a day and 20 patients to receive LMWH (Logiparin) at 3,500 antifactor Xa units once a day. Followup at 8 weeks showed a 34% incidence of bleeding complications in the unfractionated heparin group. In addition, 5 of the 21 patients had thromboembolic complications including 2 cases of fatal pulmonary embolism. Patients randomized to LMWH therapy had no thromboembolic or bleeding complications. This study affirms the safety and efficacy of LMWH in preventing thromboembolic complications. However, the unusually low incidence of thromboembolic complications and bleeding episodes in the LMWH group is likely from the selection criteria and small sample size used in this study.23 Two additional studies have also reported the safety and efficacy of LMWH in patients with spinal cord injury. These studies also demonstrated a lower incidence of bleeding complications among LMWH treated patients compared to those given unfractionated heparin.24,25

Trauma. Trauma patients constitute a special high-risk population for the development of deep venous thrombosis. The risk for DVT in this patient population more closely resembles the orthopaedic surgery population than the general surgery population. In a prospective study Geerts and coworkers26 reported the incidence of DVT in trauma patients to be 58%, of which proximal DVT comprised 18%. Literature reports of the incidence of PE in trauma patients ranges from 2% to 22%.27,28 Identifiable risk factors for thromboembolic complications among trauma patients include older age, need for blood transfusion, fracture of a lower extremity, fracture of long bones, and spinal fracture. Nonpharmacologic methods of prophylaxis such as pneumatic compression devices have limited benefit in this population and many trauma patients cannot use such devices because of long bone fractures.29-31 Oral anticoagulants are effective, but the onset of action is delayed and patients anticipating surgical procedures are not suitable candidates for the oral anticoagulants because of the prolonged duration of action of these drugs. Inferior vena cava filters are another alternative for some patients with anticipated longterm immobilization and contraindication to oral anticoagulation. However, inferior vena cava filters do not protect patients from developing DVT, and not all of them have been fully evaluated for longterm safety.32,33

In a prospective, randomized, double blind study Geerts and colleagues26 enrolled 344 patients with severity of injury score greater than or equal to 9 with no evidence of intracranial bleeding. Primary endpoints were development of DVT on contrast venography performed on or before day 14 of randomization. Of the 344 patients randomized, 265 patients had analyzable venograms. Sixty patients (44%) in the unfractionated unfractionated heparin group and 40 (31%) in the LMWH (Enoxaparin) group developed deep venous thrombosis, of which 15% and 6%, respectively, involved proximal veins. The risk reduction with LMWH for all venous thrombosis was 30%, but for proximal deep venous thrombosis, it was 58%. The incidence of major bleeding was higher in the LMWH group, but the difference did not attain statistic significance. These authors concluded that LMWH was more effective than unfractionated heparin in preventing deep vein thrombosis in trauma patients. Despite the greater efficacy of LMWH seen in this study, a significant proportion of patients developed DVT despite prophylaxis. These results indicate that 44% of patients in the unfractionated heparin group and 31% of patients in the LMWH group developed DVT while receiving prophylaxis. Most of these thrombi developed in distal veins. The clinical relevance of distal DVT is unclear, because patients are not at immediate risk for development of pulmonary emboli. However, these clots can propagate and embolize once the patient becomes ambulatory.8 The authors speculated that the efficacy of LMWH could be improved by initiating therapy early and combining LMWH with pneumatic compression devices when possible.

Medical systemic conditions. Elderly patients admitted to the hospital with various medical illnesses are at risk of developing thromboembolic complications. Epidemiologic studies have noted an incidence of DVT in the range of 10% to 26%.34 The highest incidence (42%) was noted in patients admitted with ischemic stroke with either paresis or paralysis of the lower extremity.35 The role of unfractionated heparin in preventing thromboembolic complications in medically ill patients is well documented.36

In a prospective randomized study Bergmann and associates37 compared the efficacy of LMWH with that of unfractionated heparin in preventing thromboembolic complications in patients admitted with various medical illnesses. The incidence of DVT and PE in LMWH and unfractionated heparin groups were comparable, 4.8% and 4.6% respectively. There was no significant difference in bleeding complications between groups.

Because patients admitted with stroke carry a high risk of developing thromboembolic complications, LMWH was tested against placebo38 and found to be superior in preventing thromboembolic compilations. In a more recent study a low molecular weight heparinoid (Organon Inc., West Orange, NJ) was also found to be superior to unfractionated heparin. There was a 9% incidence of thromboembolic complications in the low molecular weight heparinoid group compared with 31% in unfractionated heparin group.39 The incidence of bleeding was 2% in each group.

Thus, LMWHs are safe and effective alternatives to unfractionated heparin in preventing thromboembolic complications in medically ill patients admitted to the hospital, and superior to placebo in patients admitted with ischemic stroke.

Arterial thrombosis

Vascular surgery. Several studies have established the efficacy of LMWH in patients with venous thrombosis, but little has been done in evaluating LMWH in patients undergoing arterial revascularization. The role of anticoagulation in preventing early graft thrombosis in the postoperative period in patients undergoing arterial revascularization remains unclear. There is no consensus regarding optimal dose, duration, and timing of administration of unfractionated heparin in patients undergoing arterial reconstructive surgery.40 But there are data available to support use of anticoagulants in a subgroup of patients treated with distal femoropopliteal bypass grafts.41,42

Samama and colleagues43 evaluated the role of LMWH in preventing early graft thrombosis. Two- hundred-one patients undergoing femorodistal reconstructive surgery were randomized into 2 groups. One hundred patients received enoxaparin and 101 patients received unfractionated heparin just before clamping and for 10 days postoperatively. Doses of unfractionated and LMWH were equipotent based on antifactor Xa activity. Sixty eight percent of patients had autologous venous graft, 24% PTFE graft, and 8% others. Graft patency was assessed at day 10 by angiography. There were no significant differences between the 2 groups in terms of risk factors, type of bypass surgery (90% of patients had below knee anastomosis), type of bypass material used, and duration of surgery. There were 30 patients with thrombosed grafts, of which 8 (8%) were in the group treated with LMWH and 22 (22%) in the group treated with unfractionated heparin. There were 12 major hemorrhages in each group. These results point toward better efficacy and comparable safety of LMWH in preventing early graft thrombosis.44 The authors concluded from the above data that there is a statistically significant difference between the effectiveness of LMWH compared with unfractionated heparin in preventing early graft thrombosis. The superiority of LMWH seen in this study is probably because of its greater bioavailability and steady anticoagulant effect.

These results are promising, but enhancing graft patency is an even greater challenge. Late graft failure occurs in up to 30% of patients45 and is secondary to intimal hyperplasia and thrombosis.45 It has been noted in experimental models that unfractionated heparin blocks smooth muscle cell proliferation,46,47 warranting evaluation of its role in preserving longterm patency of femorodistal bypass. In an open randomized trial Edmondson and associates48 reported 78% and 64% graft patency after 3 months of therapy with LMWH (Fragmin) or aspirin plus dipyridamole, respectively. The beneficial effect of LMWH was more pronounced in the subgroup of patients undergoing surgery for a salvage procedure. This study was criticized because all patients received LMWH during the first week after surgery and before randomization. It is apparent that LMWH may have a role in improving both early and longterm graft patency, but larger, prospective, randomized studies need to be conducted.

Coronary artery disease. Coronary artery disease remains a leading cause of morbidity and mortality. Aspirin, along with intravenous unfractionated heparin, has gained an undisputed role as the standard of treatment for patients with unstable angina or an evolving myocardial infarction.49 Researchers, encouraged by results of studies evaluating LMWH in other disease states, have studied the safety and efficacy of LMWH in patients with unstable angina and acute coronary syndromes.

Small studies have reported the benefit of LMWH (Fragmin) in improving treadmill exercise capacity in combination with exercise in patients with stable angina.50,51 A recently published study by Cohen and colleagues52 noted the superiority of LMWH (Enoxaparin) over unfractionated heparin in preventing myocardial infarction in patients with unstable angina. However, a significant number of patients in the unfractionated heparin group had subtherapeutic aPTTs. Thus, the difference in improvement in the LMWH group of patients is therefore questioned and interpretation of results is difficult.53 Nonetheless, LMWH remains a potential alternative to unfractionated heparin in patients with unstable angina, with the added benefits of ease of administration and lack of anticoagulation monitoring requirements. In a dose ranging study, patients with unstable angina receiving LMWH (Nadroparin, Sanofi, New York, NY) at a dose of more than 100 antifactor Xa units/kg, had a higher incidence of bleeding complications. It was recommended that the dose of LMWH be limited to 100 antifactor Xa units/kg or less in patients with unstable angina.54

Restenosis after percutaneous transluminal angioplasty (PTCA) remains a vexing problem and occurs in up to 40% of patients in the 6 months postprocedure. Because of its known anticoagulant efficacy, LMWH (Reviparin, Enoxaparin) was tested and found to have no impact on the incidence of restenosis after PTCA when given for a month after the revascularization.55,56 Even when combined with fish oil the efficacy remained poor.57 In this setting, available data do not support the use of LMWH as a viable therapeutic alternative to standard antiplatelet regimens.

The role of LMWH in patients undergoing cardiopulmonary bypass is also undefined. It has been reported to be safe in experimental animal models,58,59 however, when Fragmin was tested in small group of patients on cardiopulmonary bypass the incidence of bleeding and blood loss was high, and protamine was unsuccessful in reversing the anticoagulation of low molecular weight heparin. Thus, the role of LMWH in the setting of acute coronary events remains limited at this point to patients with unstable angina.