Venous thromboembolism (VTE), comprising deep vein thrombosis (DVT) and pulmonary embolism (PE), is one of the principal causes of morbidity and mortality in surgical patients [
1]. The development of post-surgical VTE is associated with significantly higher rates of hospital readmission, VTE recurrence, and a greater than 3-fold increase in mortality [
2]. Cancer patients are at additional risk of VTE [
3], and following cancer surgery VTE is the most common cause of death at 30 days [
4].
The probability of diagnosing concomitant cancer is up to 10-times greater for cases of idiopathic VTE compared with those where the pre-disposing risk factor is known [
5,
6]. In a recent systematic review, the reported 12-month prevalence for cancer following VTE was 10.0% (95% confidence interval [CI]: 8.6-11.3) in patients with idiopathic VTE compared with 2.6% (95% CI: 1.6-3.6) in those with a provoked VTE [
7]. Overall, around 10-20% of all non-cancer patients who present with idiopathic VTE develop cancer over the following 3 years [
8]. The association between VTE and cancer is so pronounced that some researchers have argued that patients with idiopathic VTE be screened for occult cancer [
7].
The risk of VTE in cancer patients varies according to disease-specific factors such as the location, stage, and type of the malignancy [
3,
9]. In addition, cancer patients typically present with a number of co-morbid conditions that predispose individuals to thrombosis, such as older age and frequent hospitalization [
10,
11]. Cancer patients undergoing surgery have up to twice the risk of DVT and three-times the risk of PE as non-cancer patients undergoing similar operations [
12,
13]. VTE risk is further increased by cancer therapies, with significant increases in VTE in cancer patients treated with chemo- and hormonal-therapy [
11,
14,
15]. There is, therefore, a clear need for thromboprophylaxis in surgical cancer patients which is supported by current guidelines [
1,
16‐
19]. This review summarizes the relative merits of the low-molecular-weight heparins (LMWHs) and the other principal anticoagulants used for thromboprophylaxis in this high-risk population.
Pathophysiology of VTE and cancer
Although the relationship between cancer and VTE was first recognized almost 150 years ago [
20], the molecular basis of this association has only recently been investigated. Studies have shown a complex pathophysiology involving perturbation of multiple components within the coagulation and fibrinolytic pathways. The association between cancer and VTE works both ways, with cancer inducing a hypercoagulable state and the pro-thrombotic changes in turn facilitating cancer growth and metastasis [
21]. Cancer cells have been shown to aberrantly express several components involved in coagulation. For example, tissue factor, a key activator of the coagulation cascade, is expressed on endothelial cells, monocytes and, most importantly, on tumor cells themselves and is thought to play a pivotal role in cancer-induced hypercoagulability [
21]. In addition, cancerous cells can produce a cysteine proteinase, termed cancer procoagulant, which directly cleaves factor X to Xa leading to the generation of thrombin and thrombus formation [
22]. von Willebrand factor (vWF) promotes platelet adhesion during thrombus formation and elevated vWF levels have been detected in various cancers [
23]. Aberrant expression of glycoprotein IIb/IIIa receptors, which are involved in platelet activation and adhesion, and serve to promote and stabilize thrombi, is also observed on tumor cells [
24].
Cancer is also associated with disturbances of the fibrinolytic system. Plasmin, which breaks down fibrin clots, is produced from its precursor molecule plasminogen in response to plasminogen activator or urokinase-type plasminogen activator, and is inhibited by plasminogen activator inhibitor [
25]. However, deregulation of these factors is observed in cancer patients, resulting in disruptions to the normal process of clot lysis [
25]. Cancer patients may, therefore, possess abnormal expression of a number of factors which are crucial for normal hemostasis, resulting in a general state of hypercoagulability. In addition to their hemostatic effects, many of these molecules are also involved in other physiological systems, most notably angiogenesis. Many components of the coagulation cascade are also involved in tumor neovascularization, tumor cell growth, and metastasis [
21]. This has particular relevance for the anti-neoplastic effects of the various antithrombotics as discussed later.
Thromboprophylaxis in cancer patients
Several evidence-based guidelines are available that delineate appropriate anticoagulation regimens for VTE treatment, primary and secondary prophylaxis, and long-term anticoagulation in cancer patients [
1,
16‐
19]. However, despite the existence of these guidelines and the high-risk of VTE in cancer patients, up to 75% of cancer patients do not receive appropriate prophylaxis [
26]. Cancer patients are significantly less likely than non-cancer patients to have received thromboprophylaxis prior to DVT occurrence [
27]. In a sub-group analysis of the Epidemiologic International Day for the Evaluation of Patients at Risk for Venous Thromboembolism in the Acute Hospital Care Setting (ENDORSE) [
28] worldwide study comprising 1,767 cancer patients undergoing abdominal, gynecological or urological surgery, 27.7% of patients did not receive appropriate thromboprophylaxis [
29].
The use of thromboprophylaxis in cancer patients is complicated by the fact that although they are at an increased risk of VTE, they are also at an increased risk of bleeding [
30,
31]. Given that the risk of VTE outweighs the risk of bleeding in most cancer patients [
30,
31], the use of antithrombotic agents which provide stable anticoagulation while minimizing bleeding complications is especially important in this high-risk population. In addition, the available guidelines are not always consistent in their recommendations, often give several drugs equally weighted recommendation, and do not always specify the appropriate treatment duration [
1,
16‐
19]. Given these ambiguities, it can be difficult for oncologists to make informed decisions about which anticoagulant and treatment regimen would be of most benefit to their patients.
Choosing the appropriate thromboprophylactic agent
The principal role of the antithrombotics in surgical patients is to provide effective anticoagulation over the course of the increased risk of VTE (during and post-surgery) with the minimum of adverse events such as post-operative bleeding. For cancer patients, there is also increasing evidence that antithrombotics may possess anti-neoplastic effects and are potentially associated with a reduced incidence of cancer and increased survival times when given for long-term prophylaxis [
32]. Such findings reinforce the importance of thromboprophylaxis in oncology patients.
When deciding which drug to prescribe, each thromboprophylactic agent can be assessed in terms of three main properties; their efficacy and safety profiles, practicality of use, and cost-effectiveness. The relative merits of the major anticoagulants according to these criteria, as well as any possible anti-neoplastic effects are discussed below.
Practicality of use
The use of UFH is complicated by its narrow therapeutic range, resulting in frequent monitoring of activated partial thromboplastin time or anti-Xa levels, and dose adjustments may be required during treatment [
66]. UFH has sub-optimal bioavailability and accordingly when it is given subcutaneously it requires much larger doses than when given intravenously to achieve equivalent anticoagulation levels [
67]. The heterogonous assortment of molecules that make up UFH result in a highly polygamous mixture capable of binding multiple plasma proteins, macrophages, and endothelial cells [
68]. Accordingly, resulting anticoagulation responses to UFH can vary widely. In contrast, the LMWHs have higher bioavailability [
69] and, because of their reduced plasma protein binding, have a more predictable pharmacokinetic profile [
70] and a longer half-life [
68]. Unlike UFH which is hepatically cleared, the LMWHs are renally excreted. As such, monitoring may be required in certain patient populations, including the morbidly obese, those with severe renal impairment, and pregnant women [
68]. However, in the majority of patients, the pharmacokinetic advantages of the LMWHs mean they can be effectively administered without the need to monitor the anticoagulant effect, both in inpatient and outpatient settings.
Outpatient use
Although warfarin has historically been the mainstay of long-term thromboprophylaxis, its use is complicated by its narrow therapeutic window and the difficulty in maintaining appropriate levels of anticoagulation [
71]. Warfarin is affected by numerous interactions with a wide-range of drugs, nutritional supplements and herbal remedies, and its efficacy can be affected by vomiting and diarrhea, all of which are common in cancer patients. Intolerance, hypersensitivity, and resistance to warfarin leading to treatment failure have all been reported [
71]. The effectiveness of warfarin is especially compromised in cancer patients [
72]. Compared with matched controls, cancer patients on warfarin spent significantly less time inside the target international normalized ratio (INR) range (both supra- and sub-therapeutic, 54% vs. 64%, respectively; p < 0.001), had more variable INR values (p < 0.001), and had more thrombotic events compared with matched non-cancer patients (p < 0.001) [
72]. Warfarin use in cancer patients undergoing chemotherapy results in extra utilization of hospital resources, especially through increased day visits associated with warfarin monitoring and resulting laboratory costs [
73]. A further potential limitation of warfarin therapy is that it has a slow onset and long duration of action (half-life 36-42 hours) [
74], which can represent a problem if anticoagulation needs to be interrupted quickly for invasive procedures, as is frequently the case in cancer patients.
The principal practical advantage of the LMWHs in the outpatient setting, both for extended-duration primary prophylaxis and long-term secondary prophylaxis, is the lack of a need to monitor anticoagulation in the majority of patients. Patient compliance with LMWHs is high in both non-cancer [
75] and cancer populations [
76,
77], with the majority of patients being comfortable with self-administration. Outpatient thromboprophylaxis in cancer patients with SC LMWH is associated with a good safety profile and a high level of compliance [
77]. In a study of VTE treatment in cancer patients, LMWH was associated with improved quality-of-life over warfarin, primarily on the basis of reduced blood tests and increased optimism regarding therapy [
78]. Similarly in a pharmacoeconomic analysis using data from the Comparison of Low Molecular Weight Heparin Versus Oral Anticoagulant Therapy for Long Term Anticoagulation in Cancer Patients With Venous Thromboembolism (CLOT) trial which compared the LMWH dalteparin and warfarin for the long-term anticoagulation of cancer patients with DVT, dalteparin was the preferred treatment in 96% (23/24) of respondents and was associated with a gain of 0.157 quality-adjusted life years (QALY) [
79].
Overall, therefore, LMWHs offer considerable advantages over warfarin in cancer patients in terms of practicality of use as well as in efficacy.
Cost comparisons
The costs involved in managing VTE in general are considerable. In the US managing DVT alone is estimated to cost around $1.5 billion annually [
80]. The yearly direct costs for treating an individual VTE episode are high; $10,804 for a DVT event and $16,644 for a PE event [
81]. Furthermore, the long-term sequelae of VTE such as post-thrombotic syndrome can further increase costs [
82]. Accordingly, the expenditure associated with the provision of appropriate anticoagulation following surgery can be offset by the savings achieved by averting the costs of VTE management.
A number of studies have indicated that the LMWHs are economically superior to UFH both for DVT treatment [
80,
83,
84], and for bridging to long-term anticoagulation [
85‐
87], and they are at least non-inferior or superior to warfarin in preventing VTE following orthopedic surgery [
88,
89]. However, there are limited data regarding the economics associated with thromboprophylaxis following cancer surgery. Data from the CLOT trial on long-term prophylaxis in cancer patients with DVT showed that overall costs were lower with warfarin than dalteparin (CA $2,003 vs. $4,262, respectively; p < 0.001), primarily due to reduced drug-acquisition costs [
79]. However, when patient quality of life was also included in the analysis dalteparin therapy was associated with a cost of around CA $13,800 per QALY gained [
79], far below the $50,000 cost per QALY considered to be economically acceptable [
90].
Anticoagulants may be associated with increased survival in cancer patients
Three major studies have indicated that the LMWHs may be associated with a survival benefit in cancer patients that could not be directly linked to a reduction in VTE incidence [
91‐
93]. In the Malignancy and Low Molecular Weight-Heparin Therapy (MALT) trial, cancer patients were randomly assigned to 6-weeks of either the LMWH nadroparin (
n = 148) or placebo (
n = 154). At 12 months the overall HR for death was 0.75 (95% CI: 0.59-0.96) with a median survival of 8.0 months in the nadroparin group compared with 6.6 months in the placebo group [
91]. Similarly, The Fragmin Advanced Malignancy Outcome Study (FAMOUS) compared the LMWH dalteparin given for 1-year with placebo in cancer patients. The Kaplan-Meier survival estimates for 1, 2, and 3 years after randomization were not different between the dalteparin and placebo groups (p = 0.19). However, in an analysis not planned
a priori, a sub-group of patients who were alive at 17 months, experienced significantly improved survival estimates at 2- and 3-years following randomization with dalteparin versus placebo (78% vs. 55% and 60% vs. 36%, respectively; p = 0.03) with no increase in major bleeding rates [
92]. Notably, these effects were observed long after dalteparin was discontinued, suggesting the survival benefit is not dependent on VTE prophylaxis.
In the CLOT trial, over 600 patients with cancer and VTE were randomized to receive 6-months of warfarin or dalteparin therapy. A survival benefit for LMWH over warfarin was observed in patients with non-metastatic cancer, with a 20% mortality rate in the dalteparin group compared with 36% with warfarin (HR 0.50; 95% CI: 0.27-0.95; p = 0.03). However this benefit was not maintained in patients with metastatic cancer (72% vs. 69%, respectively; HR 1.1; 95% CI: 0.87-1.4; p = 0.46) [
93].
Although some studies have suggested that warfarin may also improve survival in cancer patients [
94,
95] and reduce the incidence of cancer [
96], a meta-analysis of 11 studies comparing mortality with the LMWHs versus warfarin demonstrated that although the LMWHs increased survival (RR 0.877; 95% CI: 0.789-0.975; p = 0.015) warfarin did not (RR 0.942; 95% CI: 0.854-1.040; p = 0.239) [
32]. Furthermore, patients receiving warfarin therapy also had a significant increase in the risk of major bleeding (RR 2.979; 95% CI: 2.134-4.157; p < 0.0001) whereas those receiving LMWH did not (RR 1.678; 95% CI: 0.861-2.269, p = 0.128). In a systematic review of the literature, heparin (UFH or LMWH) was associated with a survival benefit in cancer patients (HR 0.77, 95% CI 0.65-0.91) without significantly increasing the risk of bleeding (RR 1.78, 95% CI 0.73-4.38) [
97]. When analyzed by subgroups however, a statistically significant survival benefit was observed in patients with limited small-cell lung cancer (SCLC) (HR 0.56, 95% CI 0.38-0.83), but was not seen in patients with more extensive SCLC (HR 0.80, 95% CI 0.60-1.06) or patients with advanced disease (HR 0.84, 95% CI 0.68-1.03) [
97].
Thus it appears that the LMWHs may be associated with improved survival in certain cancer populations. However, more studies are needed to fully characterize this effect and how it is affected by different cancer locations, types, and disease stage. Accordingly, current evidence-based guidelines delineating appropriate thromboprophylaxis and VTE-treatment in cancer patients do not recommend the use of primary thromboprophylaxis to try to improve survival in cancer patients, and use of a LMWH for this indication would be off-label [
1,
17,
19].
Findings suggest that the improvements in survival seen with the LMWHs in cancer patients do not simply result from a decrease in the incidence of VTE, but also from potential anti-neoplastic properties of heparins. Heparin and its derivatives possess mucopolysaccharide chains similar to cell-surface and extra-cellular matrix molecules, raising the possibility that UFH and LMWHs can modulate how cells interact with their environment, enzymes, and cell-signaling molecules, and so affect malignant cell growth [
98]. In vivo evidence suggests that the anti-metastatic effects of heparins depend upon P-selectin-mediated binding via their polysaccharide chains rather than their antithrombotic activity [
99]. Accordingly fondaparinux, which lacks a polysaccharide chain, did not inhibit metastasis at clinically relevant anticoagulation levels in this model [
99]. Similarly, UFH, LMWHs and oligosaccharide truncates of heparin have been shown to inhibit tumor growth and metastasis in vivo [
100].
Heparin and oligosaccharide truncates of heparin have also been shown to inhibit angiogenesis [
101]. Studies have demonstrated UFH and LMWH have dose-dependent antiangiogenic effects that are mediated via release of endothelial tissue factor pathway inhibitor, which are independent of their antithrombotic activity [
102]. Furthermore, heparins can directly affect the immune system by their inhibitory effects on extravasation of leukocytes and the complement system, or by enhancing the susceptibility of cancer cells to immunologic attacks [
103]. Consequently, it is likely the proposed anti-neoplastic effects of heparin and the LMWHs are a combination of direct anti-neoplastic, antiangiogenetic, and immunomodulatory effects, as well as indirect effects resulting from their pleiotropic action on the coagulation system.
Each LMWH has a particular structural profile which in turns gives it specific pharmacokinetic and pharmacodynamic properties [
104,
105]. Structural differences between the LMWHs such as in the molecular weight, molecule length, end-group composition, carboxyl-to-sulfate group ratio and the proportion of anti-Xa binding domains have been shown to affect the biological activity of the resulting molecule [
104,
105]. It is possible therefore that the LMWHs possess different anti-metastatic properties to one another. However, it is unclear at present to what extent the structural heterogeneity between the LMWHs translates into clinical differences in the drug's anti-metastatic effect. Current research is investigating separating the anticoagulant and anti-metastatic properties of heparin molecules for use in cancer patients [
106].
The complex mechanisms associated with improvements in survival of cancer patients treated with heparins are of relevance to the new generation of oral anticoagulants which are under development [
107]. In an attempt to separate antithrombotic and bleeding effects, agents have been designed to inhibit specific proteins within the coagulation cascade. However, these new drugs lack the polypharmacological actions of the UFH and LMWHs which are thought to be involved in anti-neoplastic effects, and accordingly it is likely that they will also have concurrent reductions in their anti-neoplastic activity.