To the practicing clinician, cancer remains the most significant acquired risk factor for the development of VTE, with an annual incidence of 1 in 200, ultimately affecting at least 15% of this population (6). VTE in patients with underlying malignancy as opposed to those without cancer can be particularly more serious given their increased likelihood of VTE recurrence, risk of major bleeding complications from anticoagulants, and their reduced survival from such events. Prandoni et al. for instance, reported that patients with cancer and VTE were approximately four times more likely to develop recurrent thromboembolic complications and twice as likely to develop major bleeding while on anticoagulation when compared to patients without underlying malignancy (7). In a retrospective analysis, Khorana et al. found that inhospital mortality was two- to fivefold higher among neutropenic cancer patients hospitalized with thromboembolism as compared to those without thromboembolism. Chew and colleagues analyzed the effect of VTE on survival between cancer patients and found that a diagnosis of thromboembolism was associated with reduced survival rates during the first year, regardless of the type of cancer studied [hazard ratio (HR) 1.6–4.2, P > 0.1] (8). In addition to its human cost, VTE in cancer patients confers additional economic burden. Of cancer patients who develop deep vein thrombosis (DVT), the mean cost of hospitalization in 2002 was US $20,065 (9) compared to an average between $7712 and $10,804 for a similar episode in the general population (10).

While the exact mechanism of VTE in cancer patients is unclear, several theories seem to bear credence. It is recognized, for example, that tissue factor (TF) (which initiates the coagulation cascade) is itself expressed in a variety of malignancies and released into the circulation, suggesting its potential role in cancer-related VTE (8). One study was able to demonstrate a consistent relationship between cell surface expression of TF and prothrombotic potential across a range of sites, including breast, colorectal, and pancreatic tumor cell lines (11). Other observations have lent themselves to other theories. Falanga and Gordon (12), for example, described a cysteine protease that directly activates factor X in the absence of factor VII while Denko and Giaccia had proposed that tumor cell hypoxia stimulates production of procoagulant and angiogenic factors (13). Other theories based on animal models have raised the possibility of oncogene activation to explain the manifestations of Trousseau’s syndrome (14). Yet others describe release of mucins and their interaction with L- and P-selectins particularly in patients with mucinous adenocarcinomas (15, 16). It would not be unreasonable, therefore, to assume that some of these pathways operate and overlap in ways that ultimately predispose the cancer patient to thromboembolic events.

It is difficult to directly compare the rates of cancer-related VTE among patients as the studies vary in their study periods, the methods employed in detecting and reporting VTE, the patient populations, and their follow-up periods. Additionally, with the temporal rise in the incidence of VTE, newer studies seem to report higher rates than those that are less recent (17, 18). Nevertheless, there is broad agreement in the literature with regards to most risk factors for cancer-related VTE. These can be broadly divided into three categories: cancer-related factors, treatment-related factors, and patient-related factors.

In order to reduce the disease burden of VTE, it is important to identify cancer patients at high risk for VTE and who may, therefore, benefit from thromboprophylaxis. Conversely, identifying patients at low risk for VTE may allow us to determine those patients in whom prophylactic anticoagulation can be foregone in order to mitigate the risk of iatrogenic bleeding. Though both intuitive and appealing, the utility of such an approach has had conflicting results (68–70). Recently, a validated risk model for use in the ambulatory setting was published to identify cancer patients whom are at high risk for VTE (Table 2) (62). Five predictive variables were identified in the development cohort, before initiation of chemotherapy. Rates of VTE in the development cohort were 0.8% in the low-risk category (score = 0), 1.8% in the intermediate risk category (score = 1–2), and 7.1% in the high risk category (score = 3 or above), while rates in the validation cohorts were 0.3, 2, and 6.7%, respectively (over a median period of 2.5 months, C statistics = 0.7 for both cohorts).

The model has a negative predictive value of 98.5% at the cutoff point for high risk (score ≥ 3). This score is being used to define cancer patients at high risk for VTE in a study on thromboprophylaxis.¹ Additionally, this model is being assessed for its use in clinical practice as well.²

In surgical oncology patients without thromboprophylaxis, the incidence of lower extremity DVT, as shown by venography, ranges from 40 to 80% while that of proximal DVT varies between 10 and 20%. The risk of fatal postoperative PE associated with cancer surgery is about four times higher in comparison to non-oncologic surgery (71). Strong risk factors associated with VTE in this setting have been described above (treatment-related risk factors). One commonly used prophylactic regimen consists of a pre-operative dose of subcutaneously administered Heparin, followed by scheduled dosing 12–24 h postoperatively. Typically, unfractionated Heparin (UFH) is given two to three times a day and low molecular weight Heparin (LMWH) is injected once daily. International guidelines direct the use of anti-thrombotic prophylaxis for at least 7–10 postoperative days in patients with cancer surgery (5, 72–74).

Enoxaparin and Cancer (ENOXACAN) II was the first study to demonstrate the benefits of extended thromboprophylaxis with enoxaparin in reducing postoperative VTE among cancer patients undergoing surgery (75). These benefits have been demonstrated across a wide range of patient profiles among oncologic patients, including those undergoing abdominal, pelvic, and thoracic surgery. Rasmussen et al. showed that prolonged prophylaxis with LMWH significantly reduces the risk of VTE compared with prophylaxis given for 7–10 days, without an increase in bleeding complications (76). The overall incidence of VTE was 14.3% (95% CI, 11.2–17.8%) in the control group as compared to 6.1% (95% CI, 4.0–8.7%) in the patients receiving out-of-hospital LMWH (76). The ESMO and AIOM guidelines (73, 74) recommend extended prophylaxis for all patients undergoing elective cancer surgery while the ASCO panel (5) recommends extended prophylaxis for up to 4 weeks in patients undergoing major abdominal or pelvic cancer surgery in the presence of strong risk factors (please see above).

If pharmacologic prophylaxis is contraindicated, mechanical prophylaxis, such as continued use of intermittent pneumatic compression devices with or without compression stockings, should be employed. If compression stockings are used, it is important to ensure that they are of the appropriate size in order to be effective in preventing DVT (77).

The incidence of asymptomatic CVC-related DVT has been reported to be about 20% while that of overt DVT of the upper extremities ranges between 2 and 4% (71). Thromboprophylaxis for CVC-related thrombosis is controversial. Both LMWH and Warfarin have been found to be safe and effective in these patients (68, 70, 78, 79). International guidelines, however, do not recommend routine prophylaxis for this indication (5, 73, 74).

In one of the earliest studies examining this topic, Levine et al. reported that low-dose Warfarin is safe and effective for VTE prevention in stage IV breast cancer patients receiving chemotherapy, the relative risk reduction vs. placebo being 85% (80). In the more recent PROTECHT study, Nadroparin use was found to be associated with a statistically significant 50% relative risk reduction in thromboembolic events among cancer patients in the ambulatory setting receiving chemotherapy for advanced cancers of the breast, lung, gastrointestinal tract, head/neck region, ovary, and pancreas. Fifteen (2.0%) of 769 patients treated with Nadroparin and 15 (3.9%) of 381 patients treated with placebo had a thromboembolic event (single-sided P = 0.02) (81).

Thromboprophylaxis with Certoparin; a LMWH, primarily active against factor Xa, was evaluated in the TOPIC-1 and TOPIC-2 studies that examined patients with advanced breast cancer (TOPIC-1) and stage III/IV non-small cell lung cancer (TOPIC-2), respectively (69). VTE occurrence was not different between the treatment groups in TOPIC-1 (4% treated with Certoparin vs. 4% receiving placebo, OR = 1.02, 95% CI, 0.3–3.48) and TOPIC-2 [which showed a non-significant trend toward efficacy in lung cancer group (4.5 vs. 8.3%, OR = 0.52, 95% CI, 0.23–1.12)].

More recently, Semuloparin, a novel ultra LMWH, was tested for its safety and efficacy in preventing VTE among patients with advanced solid cancers receiving chemotherapy (82). In a double-blind, multicenter trial, patients were randomly assigned to receive subcutaneous Semuloparin 20 mg once daily, or placebo. The primary efficacy outcome was the composite of any symptomatic DVT, any non-fatal PE, and VTE-related death. VTE occurred in 20 of 1,608 patients (1.2%) receiving Semuloparin, compared with 55 of 1,604 patients (3.4%) on placebo (HR: 0.36; 95% CI, 0.21–0.60). Major bleeding rates of patients receiving Semuloparin and placebo rates were similar, occurring in 1.2 and 1.1%, respectively (HR: 1.05; 95% CI, 0.55–1.99).

International guidelines do not recommend routine prophylaxis of cancer patients in the ambulatory setting who receive anti-neoplastic agents (5, 72–74). In this setting, thromboprophylaxis is only recommended for patients with Multiple Myeloma on Thalidomide or Lenalidomide-based combination chemotherapy. Table 3 summarizes the commonly used anticoagulant regimens for VTE prophylaxis.

Patients with acute leukemia and hematopoietic stem cell transplant (HSCT) recipients present a special challenge in this regard as both groups are at great risk for both thrombotic and bleeding events. The incidence of VTE in patients with acute leukemia, for example, ranges from 1.7 to 12% (83). Thrombocytopenia experienced by many of these patients as a result of chemotherapy places them at great risk for bleeding complications. In one retrospective study, of 1,514 HSCT transplant recipients, 4.6% patients developed symptomatic VTE within 180 days of HSCT (84), while 3.6% of patients had had a fatal bleeding episode. Randomized clinical trials are required in order to devise the best regimen for VTE prophylaxis in this high-risk patient population.

Bemiparin is a LMWH with anti-factor Xa/anti-factor IIa activity that has been studied for VTE prophylaxis in cancer patients undergoing abdominal or pelvic surgery (109). In the CANBESURE study, 703 cancer patients undergoing surgery were randomized to receive 3,500 IU of Bemiparin subcutaneously daily for 8 days. They were then randomized to receive Bemiparin or placebo for an additional 20 days. Major VTE occurred in 0.8% of patients in the Bemiparin group compared with 4.6% in the placebo group (relative risk reduction 82.4%; 95% CI, 21.5–96.1%; P = 0.010). The study authors concluded that 4 weeks of Bemiparin use (when compared to 1 week of Bemiparin) significantly reduced the rate of major VTE without an associated increase in bleeding risk in cancer patients undergoing surgery.

Semuloparin is a subcutaneous ultra LMWH that acts as a factor Xa inhibitor with residual factor IIa activity (110). The TREK study evaluated the dose–response profile of Semuloparin in patients undergoing total knee replacement (110). A significant dose–response correlation was observed across five tested doses of Semuloparin and the incidence of VTE ranged from 5.3% (at a dose of 60 mg/day) to 44.1% (at a dose of 10 mg/day). A similar dose–response effect was observed for incidents of major bleeding (P = 0.0231).

It was concluded that a dose between 20 and 40 mg/day provides an adequate benefit-to-risk profile. Agnelli et al. evaluated Semuloparin for thromboprophylaxis in cancer patients receiving chemotherapy (82).

Fondaparinux is a synthetic pentasaccharide that acts by binding to antithrombin and increasing its inhibitory effect against factor Xa by a factor of ~300. Turpie et al., in a meta-analysis of four multicenter, randomized, double-blind trials, evaluated the safety and efficacy of Fondaparinux in comparison with Enoxaparin (111). The incidence of VTE was reduced by half from 13.7% with LMWH to 6.8% with Fondaparinux. The US Food and Drug Administration has approved Fondaparinux as a substitute for Heparin or LMWH in the initial treatment of VTE. Additional studies are needed to evaluate the safety and efficacy of this agent in patients with malignancy.

Rivaroxaban (109, 112–115) and Apixaban (116–118) are two promising oral inhibitors of factor Xa that have been predominantly studied for prevention of VTE in patients undergoing orthopedic surgery.

In one study, extended duration prophylaxis with Rivaroxaban (10 mg once daily for 35 days) was compared with shorter duration prophylaxis with Enoxaparin (40 mg once daily for 10 days) in acutely ill medical patients, including patients with cancer (MAGELLAN trial) (119). The primary efficacy outcomes were symptomatic VTE, VTE-related death, or asymptomatic proximal DVT detected by routine compressive ultrasonography. Rivaroxaban was associated with a reduction in the risk of venous thrombosis compared to 10 days of treatment with Enoxaparin. Bleeding rates were significantly increased with this new factor Xa inhibitor.

In an interim analysis of a phase II trial, Levine et al. noted that Apixaban was well tolerated by patients with metastatic cancer. The incidence of major bleeding and thrombosis were very low among 125 patients (two patients receiving Apixaban at a dose of 20 mg and one patient in the placebo group developed major bleeding). Thrombosis was reported among three cases in the placebo group (120).

Edoxaban, another oral direct factor Xa inhibitor has been recommended as an alternative agent to treat VTE in cancer patients by the latest Chest guidelines (Grade 2C) (121). In an indirect comparison with apixaban, rivaroxaban, and dabigatran; edoxaban did not show statistically significant difference in risk of recurrent VTE or all – cause mortality (122). In addition, for the composite end – point of major or clinically relevant non-major bleeding, the relative risk for apixaban vs. edoxaban was 1.50 (95% CI, 1.17–1.92, P = 0.001), was 1.15 (95% CI, 0.95–1.39, P = 0.16) for rivaroxaban vs. edoxaban and 1.31 (95% CI, 1.02–1.68, P = 0.04) for edoxaban vs. dabigatran. However, this analysis did not look at the cancer patients. Therefore, edoxaban appears a safer alternative to Warfarin except patients with creatinine clearance >95 ml/min in whom it had shown to increase the risk of ischemic stroke. It does carry the benefit of once a day dosing as compared to its contemporary agents.

Ximelagatran was the first oral direct thrombin inhibitor studied in clinical trials but its development was discontinued because of potentially severe hepatotoxicity. Dabigatran etexilate is the most developmentally advanced oral direct thrombin inhibitor. In randomized, double-blinded trials, it has been shown to be non-inferior to LMWH in reducing the risk of major VTE in patients undergoing orthopedic surgery.