Venous thromboembolism in the patient with cancer

Focus on burden of disease and benefits of thromboprophylaxis


  • Gary H. Lyman MD, MPH

    Corresponding author
    1. Comparative Effectiveness and Outcomes Research Program, Duke University and the Duke Comprehensive Cancer Center, Durham, North Carolina
    • Comparative Effectiveness and Outcomes Research Program, Duke University and the Duke Comprehensive Cancer Center, 2424 Erwin Road, Suite 205, Durham, NC 27705
    Search for more papers by this author
    • Fax: (919) 681-7488


Venous thromboembolism (VTE) is a significant cause of morbidity and mortality in patients with cancer. The risk of VTE varies over the natural history of cancer, with the highest risk occurring during hospitalization and after disease recurrence. Patient and disease characteristics are associated with further increased risk of VTE in this setting. Specific factors include cancer type (eg, pancreatic cancer, brain cancer, lymphoma) and the presence of metastatic disease at the time of diagnosis. VTE is a significant predictor of increased mortality during the first year among all types and stages of cancer, with metastatic disease reported to be the strongest predictor of mortality. VTE is also associated with early death in ambulatory patients with cancer. These data highlight the need for close monitoring, prompt treatment, and appropriate preventive strategies for VTE in patients with cancer. The American Society of Clinical Oncology and the National Comprehensive Cancer Network have issued guidelines regarding the prophylaxis and treatment of patients with cancer. This review summarizes the impact of VTE on patients with cancer, the effects of VTE on clinical outcomes, the importance of thromboprophylaxis in this population, relevant ongoing clinical trials examining the prevention of VTE, and new pharmacologic treatment options. Cancer 2011. © 2010 American Cancer Society.

Venous thromboembolism (VTE), which includes deep vein thrombosis (DVT) and pulmonary embolism (PE), is a major complication of cancer and one of the leading causes of death among cancer patients.1, 2 Overall, approximately 20% of all VTE cases occur in patients with cancer.3 In addition, VTE affects up to 20% of patients with cancer before death and has been reported in up to half of cancer patients at the time of postmortem examination, highlighting the fact that the true extent of this complication may be underestimated.4, 5 Cancer-associated VTE has important clinical and economic consequences, including increased morbidity resulting from hospitalization and anticoagulation use, bleeding complications, increased risk of recurrent VTE, and cancer treatment delays.6 In 1 analysis, Prandoni et al reported that patients with cancer and VTE were approximately 4 times more likely to develop recurrent thromboembolic complications and twice as likely to develop major bleeding while receiving anticoagulant treatment than those without malignancy.7 The occurrence of VTE in patients with cancer may interfere with planned chemotherapy regimens, worsen patient quality of life,8 and lead to the increased consumption of healthcare resources compared with patients without cancer who experience VTE. In a retrospective study of medical records from 529 cancer patients, the mean hospitalization cost for DVT was $20,065 per episode (2002 US$)9 compared with a cost of $7712 to $10,804 per episode in a general medical population with VTE.10

VTE is also associated with increased mortality in cancer patients. A retrospective study by Khorana et al found that in-hospital mortality was 2-fold to 5-fold more common in neutropenic cancer patients hospitalized with thromboembolism compared with those without thromboembolism.11 Similarly, Chew et al determined that the diagnosis of VTE was a significant predictor of increased mortality during the first year among all cancer types examined, with hazard ratios (HRs) ranging from 1.6 to 4.2 (P < .01).12 The strongest predictor of death in this analysis was metastatic disease at the time of cancer diagnosis, with an HR ranging from 1.8 to 49.0 (P < .001). In addition, stratified analyses demonstrated that VTE was associated with an increased risk of death for patients with all stages and cancer types, with a median overall relative risk (RR) of 3.7 (Table 1).12 A prospective study of patients initiating new chemotherapy (median follow-up, 75 days) found that VTE accounted for 9.2% of deaths.1 In addition, a VTE diagnosis has been associated with an approximately 2-fold increased risk of death within 2 years in patients with breast cancer.13

Table 1. Effect of VTE on Mortality Risk Within 1 Year of Diagnosis in Patients With Different Cancer Types Stratified by Cancer Stage
 HR by Stage
  • VTE indicates venous thromboembolism; HR, hazard ratio; NA, not applicable.

  • a

    P < .001.

  • b

    P < .01.

  • c

    P < .05.

  • Reproduced with permission from Chew HK, Wun T, Harvey D, Zhou H, White RH. Incidence of venous thromboembolism and its effect on survival among patients with common cancers. Arch Intern Med. 2006;166:458-464. Copyright © 2006 American Medical Association. All rights reserved.12

Non-Hodgkin lymphoma3.2a2.0b2.3a

Taken together, these data highlight the need for close monitoring, prompt treatment, and appropriate preventive strategies for VTE in patients with cancer. This review will describe the substantial impact of VTE on patients with cancer, the effects of VTE on clinical outcomes, the importance of thromboprophylaxis in this population, relevant ongoing clinical trial data, and new pharmacologic treatment options for the prevention of VTE.


In addition to the overall increased risk of VTE among patients with cancer, VTE risk is especially high among certain subgroups, such as hospitalized patients, those undergoing active antineoplastic therapy, and those with metastatic disease.14 Cancer patients undergoing major surgery are also at an increased risk of VTE.15, 16 Other factors that have been associated with increased risk include patient characteristics such as advanced age, ethnicity, and gender; cancer-related factors including cancer type and disease stage; the presence of specific biomarkers such as tissue factor and D-dimer; and factors related to systemic treatment such as type of therapeutic agent (Table 2).14

Table 2. Risk Factors for VTE in Patients With Cancer
CategoryRisk Factor
  1. VTE indicates venous thromboembolism.

  2. Reprinted from Khorana AA, Rao MV. Approaches to risk-stratifying cancer patients for venous thromboembolism. ThrombRes. 2007;120(suppl 2):S41-S50. Copyright © 2007, with permission from Elsevier.14

Patient characteristics• Advanced age
• Gender
• Ethnicity
 ○ African American, higher
 ○ Asian, lower
Cancer-related factors• Cancer site
 ○ Brain
 ○ Pancreas
 ○ Kidney
 ○ Stomach
 ○ Bladder
 ○ Gynecologic
 ○ Lung
 ○ Blood
• Advanced stage
• Initial period after diagnosis
Biomarkers• Increased platelet count prior to chemotherapy
• D-dimer
• Tissue factor expression in tumor cells
Treatment-related factors• Major surgery
• Hospitalization
• Cancer therapy
• Chemotherapy or hormonal therapy
• Antiangiogenic and immunomodulatory agents
 ○ Bevacizumab
 ○ Thalidomide and lenalidomide
• Erythropoiesis-stimulating agents

The presence of metastatic disease is strongly associated with an increased risk of VTE. An analysis of the California Cancer Registry found that the incidence of VTE varied with cancer type, but regardless of cancer type, the incidence was highest among patients initially diagnosed with metastatic disease.12 Among patients with concurrent VTE, 56% had metastatic disease compared with 21% of patients without concurrent VTE (P < .001). Conversely, patients with metastatic disease at the time of cancer diagnosis had a 1.4-fold to 21.5-fold higher risk of thromboembolism than patients with localized disease for all cancer types analyzed.12

The risk of VTE varies over the natural history of cancer, with the highest risk occurring during hospitalization and after the development of metastatic disease (Fig. 1).17 In 1 study of patients with non-Hodgkin lymphoma and VTE, thrombosis was present at the time of diagnosis in 37%, occurred during the first chemotherapy cycle in 22%, and occurred overall within the first 3 cycles in 82% of patients.18 Another study found that the incidence of thromboembolism was higher during the first year of follow-up than during the second year for all types and stages of cancer, with the exception of localized pancreatic cancer.12 Similarly, in a study by Alcalay et al of patients with regional stage colon cancer, the 2-year cumulative incidence of VTE was 3.1%, but the incidence decreased significantly over time from 5.0% during the first 6 months to 1.4% from 6 months to 1 year. During the second year, the incidence had decreased further to 0.6%.19

Figure 1.

The risk of venous thromboembolism (VTE) varies over the natural history of cancer. Reproduced with permission from Rao MV, Francis CW, Khorana AA. Who's at risk for thrombosis? Approaches to risk stratifying cancer patients. In: Khorana AA, Francis CW, eds. Cancer-Associated Thrombosis: New Findings in Translational Science, Prevention, and Treatment. New York, New York: Informa Healthcare USA, Inc; 2007:169-192. ©2007 Informa Healthcare.17

VTE Risk and Cancer Type

The incidence of VTE may be closely associated with characteristics of tumor biology (not only the extent of metastatic spread, but primarily the rate of growth and spread of the cancer), suggesting that specific cancer types are associated with an increased risk of VTE.

Sites of cancer with the highest rates of VTE include the pancreas (8.1%), kidneys (5.6%), ovaries (5.6%), lungs (5.1%), and stomach (4.9%).20 Among the hematologic malignancies, myeloma (5%), non-Hodgkin lymphoma (4.8%), and Hodgkin disease (4.6%) were reported to have the highest rates of VTE.20 One retrospective record review estimated a cumulative frequency of VTE in patients with diffuse large B-cell lymphoma of 12.8%.18

In an analysis of data from the National Hospital Discharge Survey, the highest incidence of VTE among 19 cancer types included in the analysis occurred in patients with pancreatic cancer (4.3%), whereas the lowest evaluable incidence was noted in patients with bladder cancer (1.0%).21 In neutropenic cancer patients hospitalized with thromboembolism, Khorana et al reported that the sites of cancer with the highest percentage of patients with VTE were the pancreas (12.1%), brain (9.5%), and endometrium or cervix (9%).11 The risk in hospitalized patients with hematologic disorders was also high; patients with non-Hodgkin lymphoma and leukemia accounted for greater than one-third of all patients with venous events.11 Similarly, a large retrospective cohort study using the discharge database of the University HealthSystem Consortium (N = 1,015,598 cancer patients)20 found that 4.1% of patients were diagnosed with VTE. Factors associated with increased risk included black ethnicity and the use of chemotherapy.

VTE Risk and Systemic Cancer Therapy

Many cancer therapies (including surgery, chemotherapy, and hormonal therapy) appear to place patients with cancer at further increased risk of VTE. This appears to also be true of several newer cancer treatments, such as the antiangiogenesis agents thalidomide, lenalidomide, and bevacizumab.4 The use of thalidomide, an immunomodulatory agent with antiangiogenic activity, has been associated with an increased risk of VTE when used concomitantly with chemotherapy or dexamethasone in patients with multiple myeloma.4 In a study presented at the annual meeting of the American Society of Hematology in 2008, Gray et al characterized the incidence of VTE in 3977 patients with multiple myeloma as well as a variety of solid tumors in a meta-analysis of 17 randomized controlled trials.22 The overall incidence of VTE in the study was 11.7%, and patients treated with thalidomide were found to be at more than a 2-fold increased risk of VTE compared with controls (P < .001). The risk was especially high in patients with multiple myeloma, with approximately 15% of patients experiencing VTE and having a 3-fold increased risk compared with control patients not receiving thalidomide.22

Lenalidomide, a structural analog of thalidomide, was not found to be associated with an increased risk of VTE in a postmarketing survey of patients with myelodysplastic syndromes. In this survey, the observed risk of VTE was increased in patients treated with lenalidomide and erythropoiesis-stimulating agents (ESAs); there was no increase in VTE risk observed in patients treated with lenalidomide without ESAs.23 According to the American Society of Clinical Oncology (ASCO) recommendations for VTE prophylaxis and treatment in patients with cancer, patients receiving thalidomide or lenalidomide with chemotherapy or dexamethasone warrant prophylaxis with low molecular weight heparin (LMWH) or adjusted-dose warfarin (international normalized ratio [INR] of approximately 1.5).4 As more agents with antiangiogenic activity become indicated for the treatment of cancer, it will be important to consider this risk in cancer patients, especially in those already at increased risk from other factors.

Bevacizumab is a monoclonal antibody directed toward vascular endothelial growth factor that has an antiangiogenic effect. Currently, its role in the prevention of VTE is controversial. Bevacizumab has demonstrated a survival benefit in combination with chemotherapy in patients with colorectal cancer and with nonsquamous cell lung cancer.24 Scappaticci et al conducted a post hoc analysis of pooled data from randomized controlled trials evaluating combination treatment with bevacizumab and chemotherapy versus chemotherapy alone in 1745 patients with colorectal, breast, or nonsmall cell lung cancer. Compared with chemotherapy alone, bevacizumab was associated with a 2-fold increase in arterial thromboembolic events (P = .031) but was not found to be associated with an increased risk of venous thromboembolic events.24 These data are in contrast to a recent systematic review and meta-analysis by Nalluri et al, which included a total of 7956 patients with a variety of advanced solid tumors from 15 randomized controlled trials. Results indicated that bevacizumab was associated with an increased risk of VTE, with an RR of 1.33 (95% confidence interval [95% CI], 1.13-1.56 [P < .001]) compared with controls.25

In addition to antineoplastic therapies, certain supportive care measures used in cancer treatment appear to increase the risk of VTE. The use of epoetin α and darbepoetin α for managing anemia in patients undergoing cancer treatment has been associated with thromboembolic complications. Bohlius et al conducted a meta-analysis of 35 studies and reported that treatment with epoetin or darbepoetin increased the risk of thromboembolic events by approximately 67% compared with control patients not receiving these agents (RR, 1.67; 95% CI, 1.35-2.06).26

In addition, red blood cell transfusions may increase the risk of VTE. One study of patients receiving transfusions reported that 7.2% of patients developed VTE and 5.2% developed arterial thromboembolism compared with 3.7% and 3.0%, respectively, of patients who did not receive transfusions. Transfusions were also found to be associated with an increased risk of in-hospital mortality (odds ratio, 1.34 [95% CI, 1.29-1.38]).27

Clinical Risk Model for Chemotherapy-Associated VTE

Recently, a simple model for predicting chemotherapy-associated VTE was developed and validated to assist in the assessment of VTE risk in ambulatory cancer patients undergoing chemotherapy.28 A total of 2701 patients in the derivation cohort and 1365 patients in the validation cohort were included. Five clinical and laboratory parameters were found to independently predict symptomatic VTE in cancer patients starting a new chemotherapy regimen.28 These parameters were combined into a risk assessment model that allowed the classification of patients into 3 groups based on risk factors: 1) site of cancer (very high risk: stomach and pancreas; high risk: lung, lymphoma, gynecologic, bladder, and testicular); 2) prechemotherapy platelet count of ≥350 ×109/L; 3) hemoglobin levels <100 g/L or the use of red cell growth factors; 4) prechemotherapy leukocyte count >11 ×109/L; and 5) body mass index ≥35 kg/m2.28 VTE risk score categories using this model have been found to correlate with the development of VTE and with overall survival in patients with cancer who are undergoing chemotherapy.29


Key Clinical Trials of Pharmacologic Agents

Pharmacologic prophylactic options for VTE consist of unfractionated heparin (UFH), the class of LMWHs, fondaparinux (an indirect inhibitor of activated factor Xa), and the vitamin K antagonists.4, 30 Several novel agents are also currently in development. The pharmacologic anticoagulant agents currently being evaluated in cancer patients in phase 2 or phase 3 trials are provided in Table 3. The results of selected key studies of pharmacologic anticoagulants in cancer patients are discussed below, and an overview of recently published clinical studies is presented in Tables 4 through 6.31-55

Table 3. Pharmacologic Anticoagulant Agents Being Evaluated in Phase 2 or 3 Clinical Trials in Cancer Patientsa
AgentClass/MOARouteTitleNCT Reference
Phase 3
  • MOA indicates mechanism of action; NCT, National Clinical Trial; LMWH, low molecular-weight heparin; SC, subcutaneous; SOC, standard of care; ULMWH, ultra-low molecular weight heparin; VKA, vitamin K antagonist.

  • a

    Search of Web site on August 21, 2009. Search terms used were: “venous thromboembolism,” “thromboprophylaxis,” “thrombosis,” “phase II,” and “phase III.” Conditions searched for: cancer. Completed studies and studies actively recruiting participants were included.

Bemiparin vs placeboLMWHSCCANBESURE Study (Cancer, Bemiparin and Surgery Evaluation)NCT00219973
Dalteparin vs SOCLMWHSCA Study of Dalteparin Prophylaxis in High-Risk Ambulatory Cancer PatientsNCT00876915
Dalteparin vs SOCLMWHSCDalteparin in Preventing Blood Clots in Patients With Lung CancerNCT00519805
Dalteparin vs placeboLMWHSCDalteparin Low Molecular Weight Heparin for Primary Prophylaxis of Venous Thromboembolism in Brain Tumour PatientsNCT00135876
Gemcitabine with or without dalteparinLMWHSCGemcitabine With or Without Dalteparin in Treating Patients With Unresectable or Metastatic Pancreatic CancerNCT00031837
Gemcitabine or capecitabine with or without dalteparinLMWHSCGemcitabine With or Without Capecitabine and/or Dalteparin in Treating Patients with Metastatic Pancreatic CancerNCT00662688
Chemotherapy with or without enoxaparinLMWHSCChemotherapy With or Without Enoxaparin in Pancreatic Cancer (PROSPECT)NCT00785421
EnoxaparinLMWHSCEnoxaparin Thromboprophylaxis in Cancer Patients With Elevated Tissue Factor Bearing MicroparticlesNCT00908960
Enoxaparin vs intermittent pneumatic compressionLMWHSCJapanese Efficacy and Safety Study of Enoxaparin in Patients With Curative Abdominal Cancer SurgeryNCT00723216
Chemotherapy with or without enoxaparinLMWHSCOverall Survival of Inoperable Gastric/GastroOesophageal Cancer Subjects on Treating With LMWH + Chemotherapy (CT) vs Standard CT (GASTRANOX)NCT00718354
Fondaparinux with or without inferior vena cava filterIndirect factor Xa inhibitorSCAnticoagulation and Inferior Vena Cava Filters in Cancer Patients With a Venous ThromboembolismNCT00423683
Semuloparin vs placeboULMWHSCEvaluation of AVE5026 in the Prevention of Venous Thromboembolism in Cancer Patients Undergoing Chemotherapy (SAVE-ONCO)NCT00694382
TinzaparinLMWHSCEffect of Low Molecular Weight Heparin: Tinzaparin in Lung Tumours (TILT)NCT00475098
Tinzaparin vs warfarinLMWH/VKASCLong-Term innohep® Treatment Versus a Vitamin K Antagonist (Warfarin) for the Treatment of Venous Thromboembolism (VTE) in CancerNCT01130025
Phase 2
Apixaban vs placeboDirect factor Xa inhibitorOralA Phase 2 Pilot Study of Apixaban for the Prevention of Thromboembolic Events in Patients With Advanced (Metastatic) CancerNCT00320255
Combination chemotherapy with wafarinVKAOralCombination Chemotherapy Plus Warfarin in Treating Patients With Prostate CancerNCT00014352
Gemcitabine with or without dalteparinLMWHSCGemcitabine With or Without Dalteparin in Treating Patients With Locally Advanced or Metastatic Pancreatic CancerNCT00462852
Dalteparin and warfarinLMWH/VKASC/OralThe Catheter Study: Central Venous Catheter Survival in Cancer Patients Using Low Molecular Weight Heparin (Dalteparin) for the Treatment of Deep Vein ThrombosisNCT00216866
DalteparinLMWHSCFragmin in Ovarian Cancer: Utility on Survival (FOCUS)NCT00239980
DalteparinLMWHSCTreatment of Blood Clots in Children With CancerNCT00952380
EnoxaparinLMWHSCIdentification and Treatment of Clinically Silent Catheter-Related Deep Vein Thrombosis in Children With CancerNCT00633061
FondaparinuxIndirect factor Xa inhibitorSCFondaparinux in Preventing Blood Clots in Patients Undergoing Surgery for Gynecologic CancerNCT00381888
TinzaparinLMWHSCTinzaparin for Primary Treatment and Extended Secondary Prophylaxis of Venous Thromboembolism in Patients with CancerNCT00981903
TinzaparinLMWHSCTinzaparin in Treating Patients with Metastatic Kidney Cancer That Cannot Be Removed by SurgeryNCT00293501
Table 4. Recent Studies of Pharmacologic Anticoagulants in Medical Patients With Cancer
StudyPatient PopulationTreatmentsPrimary OutcomeResultSignificance LevelBleeding RatesSignificance LevelLength of TreatmentSetting
  1. QD indicates every day; VTE, venous thromboembolism; NR, not reported; INR, international normalized ratio; BID, twice daily; DVT, deep vein thrombosis; PE, pulmonary embolism; HR, hazard ratio; 95% CI, 95% confidence interval.

Hull 201031Acutely ill medical patientsEnoxaparin, 40 mg QD or placeboVTEEnoxaparin, 2.5% (45/1818); Placebo, 4.2% (78/1867)P < .042Major bleeding: enoxaparin, 0.8%; placebo, 0.3%P < .0528 dProphylaxis
De Cicco 200932Cancer patients with a central vein catheterAcenocumarine, 1 mg QD or dalteparin, 5000 IU QD or no anticoagulant therapyCentral vein catheter-related thrombosisAcenocumarine, 21.9% (25/114); dalteparin, 40.0% (48/120); no treatment, 52.6% (60/114)Acenocumarine vs no treatment: P < .01; dalteparin vs no treatment: P = .05; acenocumarine vs dalteparin: P = .01Major bleeding: none observedNRAcenocumarine: 11 d; dalteparin, 8 dProphylaxis
Young 200933Cancer patients receiving chemotherapy via central venous cathetersFixed-dose warfarin, 1 mg QD or INR-adjusted warfarin QD or no warfarinCatheter-related thrombotic eventsFixed-dose warfarin, 7% (34/471); INR-adjusted warfarin, 3% (13/473); no warfarin, 6% (24/404)Warfarin vs no warfarin: P = .98; fixed-dose warfarin vs INR-adjusted warfarin: P = .002Major bleeding: fixed-dose warfarin, 1%; INR-adjusted warfarin, 3%; no warfarin, <1%Warfarin vs no warfarin: P = .07; INR-adjusted warfarin vs fixed-dose warfarin: P = .09Treatment continued until catheter removal or occurrence of thrombosisProphylaxis
Weber 200834Terminal cancerNadroparin, 2850–3800 IU/kg QD or no treatmentVTENadroparin, 10% (1/10); no treatment, 0% (0/10)P = 1.00Major bleeding: nadroparin, 10%; no treatment, 0%P = 1.00Treatment continued until deathProphylaxis
Robins 200835Glioblastoma multiformeDalteparin, 5000 IU QD with conventional radiotherapy vs control cohortSurvival timeMedian survival time in dalteparin-treated patients: 11.9 moP = .47 vs control cohortMajor bleeding: none reportedNR≤24 moProphylaxis
Niers 200736Hematologic malignancyNadroparin, 2850 IU QD vs placeboCatheter-related thrombosisNadroparin, 17% (7/41); placebo, 9% (4/46)P = .49Major bleeding: none reportedNR3 wkProphylaxis
Meister 200837Acute lymphoblastic leukemiaAntithrombin alone vs antithrombin plus enoxaparin, 0.75–1.2 mg/kg QDVTEAntithrombin alone, 12.7% (9/71); antithrombin plus enoxaparin, 0%P = .02Major bleeding: none reportedNR1–2 wk during chemotherapy induction and reinduction phasesProphylaxis
Icli 200738Advanced pancreatic cancerCombination chemotherapy plus nadroparin, 2850 IU QD vs combination chemotherapy aloneTreatment response rate; survivalResponse rate: nadroparin, 58.8% (20/34); no nadroparin, 12.1% (4/33). Median overall survival time: nadroparin, 13.0 mo; no nadroparin, 5.5 moResponse rate: P = .0001; survival time: P = .0001Treatment-related bleeding: none reportedNRUntil disease progressionProphylaxis
Miller 200639Patients with multiple myeloma or chronic lymphocytic leukemia treated with thalidomide-based therapiesWarfarin 1 or 2 mg QD vs historical studies with similar chemotherapy regimensVTEWarfarin, 5.9% (4/68); thalidomide plus doxorubicin, 27%; thalidomide plus epirubicin, 26%Warfarin regimen vs thalidomide plus doxorubicin: P = .034; warfarin regimen vs thalidomide plus epirubicin: P = .009Treatment-related bleeding: none reportedNR4 moProphylaxis
Deitcher 200640Patients with active cancer and acute VTEEnoxaparin, 1 mg/kg BID × 5 d, then 1 mg/kg QD thereafter or enoxaparin, 1 mg/kg BID × 5 d, then 1.5 mg/kg QD thereafter vs enoxaparin, 1 mg/kg BID × 5 d or until INR target achieved, then INR-adjusted warfarin thereafterRecurrent VTEEnoxaparin at 1 mg/kg, 3.4% (1/29); enoxaparin at 1.5 mg/kg, 3.1% (1/32); warfarin, 6.7% (2/30)NRMajor bleeding: enoxaparin at 1 mg/kg, 6.5%; enoxaparin at 1.5 mg/kg, 11.1%; warfarin, 2.9%NR180 dTreatment
Ruud 200641Children with active cancer and central venous linesINR-adjusted warfarin QD vs no prophylaxisCentral vein catheter-related VTEWarfarin, 48% (14/29); no prophylaxis, 36% (12/33)P = .44Bleeding rates NRNR6 moProphylaxis
Ikhlaque 200642Patients receiving thalidomide therapyLow-dose warfarin (1–2 mg/d) or high-dose warfarin (adjusted to INR 2–3) vs no prophylaxisDVTLow-dose warfarin, 2.7% (1/37); high-dose warfarin, 11.1% (2/18); no warfarin, 23.7% (18/76)P = .01 for any dose of warfarin vs no warfarinClinical bleeding: low-dose warfarin, 0%; high-dose warfarin, 22.2%; no warfarin, 0%NR≤14 moProphylaxis
Baz 200543Multiple myelomaAspirin, 81 mg QD initiated at the start of chemotherapy or aspirin, 81 mg QD initiated after the start of chemotherapy vs no aspirinVTEAspirin initiated at the start of chemotherapy, 19% (11/58); aspirin initiated after the start of chemotherapy, 15% (4/26); no aspirin, 58% (11/19)P ≤.002 for both aspirin groups vs no aspirinSignificant bleeding complications: none reportedNRMedian, 2 yProphylaxis
Karthaus 200644Cancer patients with central venous cathetersDalteparin, 5000 IU QD vs placeboCatheter-related complicationsDalteparin, 3.7% (11/294); placebo, 3.4% (5/145)P = .88Any bleeding event: dalteparin, 17.5%; placebo, 15%NR16 wkProphylaxis
Verso 200545Cancer patients with central venous cathetersEnoxaparin, 40 mg QD vs placeboDVT or clinically overt PEDVT: enoxaparin, 14.1% (22/155); placebo, 18.0% (28/155)P = .35Major bleeding: none reportedNR6 wkProphylaxis
Couban 200546Cancer patients with central venous cathetersWarfarin, 1 mg QD vs placeboCentral venous catheter-related thrombosisWarfarin, 4.6% (6/130); placebo, 4.0% (5/125)HR 1.20 (95% CI, 0.37–3.94)Major bleeding: warfarin, 0%; placebo, 2%P = .07Until catheter removal, death, or catheter-related thrombosisProphylaxis
Table 5. Recent Studies of Pharmacologic Anticoagulants in Surgical Patients With Cancer
StudyPatient PopulationTreatmentsPrimary OutcomeResultSignificance LevelBleeding RatesSignificance LevelLength of TreatmentSetting
  1. UFH indicates unfractionated heparin; Q, every; VTE, venous thromboembolism; OR, odds ratio; 95% CI, 95% confidence interval; QD indicates every day; DVT, deep vein thrombosis; NR, not reported; NS, not significant; CNS, central nervous system.

Einstein 200847Gynecologic cancer surgeryDual prophylaxis with sequential compression devices alone or compression devices plus heparin, 5000 U Q 12 h or Q 8 hVTEDual prophylaxis with prolonged prophylaxis in high-risk patients resulted in a significant reduction in VTE rate from 6.5% (19/294) in 2005 to 1.9% (6/311) in 2006OR, 0.33 (95% CI, 0.12–0.88)Median blood loss: 2005: 250 mL; 2006: 200 mLP = .22Until hospital discharge, extended to 2 wk after hospital discharge in high-risk patientsProphylaxis
Shukla 200848Colorectal cancer surgeryDalteparin, 2500 IU QD × 6 d or no prophylaxisDVTNo DVT occurred in either groupNRNot specifiedNR6 dProphylaxis
Simonneau 200649Colorectal cancer surgeryNadroparin, 2850 IU QD vs enoxaparin, 40 mg QDVTENadroparin, 15.9% (74/464); enoxaparin, 12.6% (61/486)P = NSMajor bleeding: nadroparin, 7.3%; enoxaparin, 11.5%P = .0127–11 dProphylaxis
Perry 200950Patients with brain tumorsTinzaparin, 4500 IU QDSafety outcomesCNS hemorrhage in 5% (2/40)NRCNS hemorrhage: grade 1: 2.5%; grade 2: 2.5%NR12 moProphylaxis
Table 6. Recent Studies of Pharmacologic Anticoagulants in Ambulatory Cancer Patients
SudyPatient PopulationTreatmentsPrimary OutcomeResultSignificance LevelBleeding RatesSignificance LevelLength of TreatmentSetting
  1. VTE indicates venous thromboembolism; GI, gastrointestinal; QD, every day; INR, international normalized ratio; UFH, unfractionated heparin; NS, not significant.

Cini 201051Multiple myelomaThalidomide and dexamethasone or thalidomide and dexamethasone plus warfarinVTENo prophylaxis, 26.3% (5/19); warfarin, 10.6% (26/246)P = .095Major bleeding: none recordedNot specified120 dProphylaxis
Agnelli 200952Lung, GI, pancreatic, breast, ovarian, or head and neck cancerNadroparin, 3800 IU QD or placeboComposite of symptomatic venous or arterial thromboembolic eventsNadroparin, 2.0% (15/769); placebo, 3.9% (15/381)P = .02Major bleeding: nadroparin, 0.7%; placebo, 0%P = .18≤4 moProphylaxis
Lee 200553Patients with solid tumors and VTEDalteparin, 200 U/kg QD × 1 mo then 150 U/kg QD × 5 mo or dalteparin, 200 U/kg × 7 d then INR-adjusted coumarin derivative × 6 moAll-cause mortality at 12 moDalteparin, 20% (15/75); warfarin, 36% (26/75) in patients with no metastasesP = .03Not specifiedNot specified6 moProphylaxis
Hull 200654Patients with cancer and VTETinzaparin, 175 U/kg QD vs usual care (UFH plus warfarin)Recurrent VTE or death at 3 moRecurrent VTE: tinzaparin, 6% (6/100); usual care, 10% (10/100); death: tinzaparin, 20% (20/100); usual care, 19% (19/100)Recurrent VTE: P = NS; death: P = NSMajor bleeding: tinzaparin, 7%; usual care, 7%P = NS3 moTreatment
Romera 200955Patients with VTE including 28.6% (69/241) with cancerTinzaparin, 175 IU/kg QD or INR-adjusted acenocoumarolRecurrent VTE at 6 mo and 1 yCancer population: 6 mo: tinzaparin, 5.5% (2/36); warfarin, 9.1% (3/33); 1 y: tinzaparin, 5.5% (2/36); warfarin, 21.2% (7/33)6 mo: P = .58; 1 y: P = .06Major bleeding in total population: tinzaparin, 0.8%; warfarin, 2.5%P = .66 moTreatment

Key Studies in Surgical Cancer Patients

LMWHs, including enoxaparin, dalteparin, and tinzaparin, are available in the United States for use in thromboprophylaxis.56 ENOXACAN I and II were large randomized trials evaluating enoxaparin thromboprophylaxis in cancer patients. In ENOXACAN I, enoxaparin was compared directly with UFH for its ability to prevent DVT in 631 patients undergoing elective cancer surgery.57 Overall, 16.5% of patients developed thromboembolic complications, with no statistically significant difference noted between the 2 groups. There were also no significant differences noted with regard to bleeding events, other complications, and mortality. The ENOXACAN II study evaluated the duration of prophylaxis for VTE with enoxaparin in cancer patients after surgery for cancer. Enoxaparin was given for approximately 1 week (range, 6-10 days), and patients were thereafter randomized to receive enoxaparin or placebo for an additional 21 days, for a total treatment duration of approximately 1 month. Patients receiving enoxaparin for 1 month were found to have a significantly reduced incidence of thrombosis compared with those receiving enoxaparin given for 1 week followed by placebo.58 The rates of VTE were 12.0% in the placebo group and 4.8% in the enoxaparin group, corresponding to a reduction in RR of 60% (P = .02). There were no significant differences noted with regard to the rates of bleeding or other complications during the study.58

Key Studies in Hospitalized Cancer Patients

Thromboprophylaxis has been shown to decrease DVT, specifically in high-risk hospitalized patients. Key trials include a study comparing enoxaparin with placebo for the prevention of VTE in acutely ill medical patients (MEDENOX).59 In that study, prophylactic treatment with a dose of 40 mg per day of subcutaneous enoxaparin safely reduced the risk of VTE in patients with acute medical illnesses including cancer, with no difference in the rates of adverse events reported between the active comparator and placebo. Similarly, dalteparin at a dose of 5000 IU once daily was shown in the PREVENT (Prevention of Recurrent Venous Thromboembolism) trial to reduce the risk of VTE in acutely ill medical patients, with a low overall incidence of major bleeding.60 Comparable results to LMWH have been reported with fondaparinux in the ARTEMIS trial, in which fondaparinux (at a dose of 2.5 mg subcutaneously for 6-14 days) was found to be effective in preventing symptomatic and asymptomatic VTE in older acute medical patients.61 VTE was detected in 5.6% (18 of 321 patients) of patients treated with fondaparinux and 10.5% (34 of 323 patients) of patients given placebo, with a RR reduction of 46.7% (95% CI, 7.7% to 69.3%). Symptomatic VTE occurred in 5 patients in the placebo group and none of the patients in the fondaparinux group (P = .029). The frequency of major bleeding was similar for both fondaparinux and placebo, with major bleeding occurring in 1 patient (0.2%) in each group.61

Key Studies in Ambulatory Cancer Patients

Several randomized controlled trials of thromboprophylaxis in ambulatory cancer patients have been reported.62 The PROSPECT-CONKO 004 study (a prospective, randomized trial in patients with pancreatic cancer undergoing chemotherapy and also receiving enoxaparin) compared concomitant treatment using enoxaparin with no anticoagulation in 312 patients. Within the first 12 weeks, enoxaparin at a dose of 1 mg/kg/day was found to be associated with a RR reduction in the incidence of clinically relevant VTE of 65% (from 14.5% to 5%). Preliminary data demonstrated no differences between the observational and enoxaparin groups with regard to the secondary endpoints of time to disease progression (19 weeks vs 22 weeks, respectively) and overall survival (29 weeks vs 31 weeks, respectively). In addition, there was no increased risk of bleeding events noted with the use of enoxaparin in this setting (observational, 9.9% and enoxaparin, 6.3%; P value not significant).63

In the FAMOUS (Fragmin Advanced Malignancy Outcome Study) trial, dalteparin at a dose of 5000 IU daily was not found to have a significant impact on the risk of VTE compared with placebo in patients with advanced cancer.64 Dalteparin has also been compared with placebo in 186 patients with newly diagnosed malignant glioma (PRODIGE).65 Patients received dalteparin subcutaneously once daily for 6 months, starting within the first month of surgery. Twenty-two patients developed VTE during the first 6 months: 9 patients receiving dalteparin and 13 receiving placebo (HR, 0.51, 95% CI, 0.19-1.4 [P = .29]). Over 12 months there were 5 (5.1%) major bleeding events with dalteparin (all intracranial), and 1 (1.2%) with placebo (HR, 4.2, 95% CI, 0.48-36 [P = .22]). Survival was comparable between treatment arms. A randomized controlled trial of dalteparin prophylaxis in patients with solid tumors performed by Sideras et al found no survival benefit in 141 patients with advanced cancer who were treated with daily injections of 5000 U of dalteparin compared with placebo.66 A randomized controlled clinical trial of dalteparin in patients with advanced pancreatic cancer (the UK FRAGEM study) reported a significant reduction in the risk of VTE (RR, 0.38; 95% CI, 0.17-0.84 [P < .02]).67 Although dalteparin was administered at weight-adjusted doses of 200 IU/kg/day for 4 weeks followed by 150 IU/kg/day for 8 additional weeks, no increase in major bleeding was observed.

The LMWH nadroparin was found to reduce the incidence of thromboembolic events in ambulatory cancer patients receiving chemotherapy.52 The PROTECHT (Prophylaxis of Thromboembolism During Chemotherapy) trial was a randomized, double-blind, placebo-controlled study designed to evaluate the efficacy of nadroparin versus placebo for prophylaxis of thromboembolic events in 1166 patients receiving chemotherapy for advanced cancer. Patients had metastatic or locally advanced lung, breast, gastrointestinal, ovarian, or head and neck cancer with an Eastern Cooperative Oncology Group performance status ≤2. Of the 769 patients treated with nadroparin, 2.0% had a thromboembolic event compared with 3.9% of patients receiving placebo (P = .02), although the difference for VTE did not reach statistical significance. The incidence of minor bleeding in the nadroparin group (<8%) was comparable to that of the placebo group.

Certoparin, another LMWH, has been evaluated in 2 double-blind, placebo-controlled studies (TOPIC-1 and TOPIC-2) that randomized patients with advanced breast cancer (N = 353) or nonsmall cell lung cancer (N = 547) to receive certoparin at a dose of 3000 U daily or placebo for the prevention of chemotherapy-associated VTE.68 The overall rate of symptomatic and asymptomatic thrombosis in breast cancer patients was 4% for certoparin and 3.9% for placebo. Rates of major bleeding complications over 6 months of therapy were 1.7% for certoparin and 0% for placebo. Rates of thrombosis were higher in patients with nonsmall cell lung cancer compared with rates in breast cancer patients and demonstrated a trend toward a reduction in thrombosis with certoparin (4.5% vs 8.3% for placebo; P = .07). Certoparin was especially effective in patients with stage IV disease (3.5% vs 10.1% for placebo; P = .03).

Taken together, the findings of these trials of LMWHs in the ambulatory cancer care setting suggest that these agents have the most benefit in patients at high risk of VTE, such as those with pancreatic cancer. A systematic review and meta-analysis of 8 randomized controlled trials enrolling ambulatory cancer patients indicated a favorable benefit-to-risk ratio for the use of thromboprophylaxis in patients with advanced pancreatic cancer.62

Impact of Anticoagulants on Cancer Patient Survival

Anticoagulants have been postulated to improve survival in cancer patients.69 In a systematic review identifying 11 randomized controlled trials, anticoagulants (particularly LMWH) demonstrated significantly improved survival at 1 year in cancer patients without VTE while increasing the risk for bleeding complications.70 Improved survival with anticoagulation may be dependent on tumor type and disease stage. The meta-analysis of randomized controlled trials of prophylactic LMWHs in ambulatory cancer patients found no evidence of a survival benefit with the use of these agents in this setting.62 Given the limitations of available data, the use of anticoagulants as antineoplastic therapy cannot be recommended until additional randomized controlled trials have been conducted. Several trials are currently ongoing to test the effects of LMWH on survival in patients with cancer.71

Current ASCO and National Comprehensive Cancer Network Guideline Recommendations

ASCO4 and the National Comprehensive Cancer Network (NCCN),30 among other professional organizations, have developed guidelines for VTE prophylaxis and treatment in patients with cancer. As summarized in these guidelines, the primary goal of thromboprophylaxis in patients with cancer is to prevent VTE, including PE and early death from these complications. Both guidelines support the use of pharmacologic VTE prophylaxis in hospitalized cancer patients unless contraindications to prophylactic anticoagulation are present. It must be acknowledged, however, that these recommendations are based on studies of seriously ill medical patients, only a small subgroup of which were actually cancer patients. Although guideline panels and most clinicians have found it reasonable to extrapolate the results of these studies to the cancer population, more direct evidence on the risk of VTE in hospitalized cancer patients is needed.

In addition to cancer patients hospitalized for medical care, prophylaxis should include cancer patients undergoing major surgery and those with cancer and established VTE to prevent recurrence of thromboembolic events. According to the ASCO guidelines, low-dose UFH or LMWH is the recommended prophylaxis in patients undergoing laparotomy, laparoscopy, or thoracotomy.4 Prophylaxis should be initiated before surgery or as early as possible in the postoperative period and be continued for at least 7 to 10 days after surgery. Prophylaxis may be prolonged for up to 4 weeks in obese patients, patients undergoing major abdominal or pelvic surgery for cancer, and patients with a history of VTE. Mechanical methods of VTE prophylaxis may be used with pharmacologic anticoagulation but should not be used alone except in patients with active bleeding, for whom the medications are contraindicated.4 Treatment with LMWH is preferred in cancer patients with established VTE for the initial 5 to 10 days of treatment and should be given up to 6 months or longer to prevent VTE recurrence. Vitamin K antagonists with a targeted INR of 2 to 3 are acceptable for extended secondary prophylaxis when LMWH is not available. Indefinite anticoagulant therapy should be considered for patients with active cancer.

Routine thromboprophylaxis is currently not recommended in ambulatory patients with cancer who are receiving systemic chemotherapy because of the lower risk of VTE in this setting along with an increased risk of major bleeding in these patients.4 However, ASCO guidelines recommend anticoagulation for VTE prophylaxis, specifically in patients receiving thalidomide or lenalidomide adjunctively with chemotherapy or dexamethasone because of the high risk of thrombosis associated with these treatment regimens.4 The evaluation of various biomarkers to enhance clinical prediction tools for the identification of cancer patients at increased risk of VTE who may benefit from thromboprophylaxis is an area of active investigation.72-74 Current studies of VTE prophylaxis in ambulatory cancer patients at high risk for VTE based on cancer type (eg, pancreatic cancer or risk model evaluation) may lead to future recommendations for prophylaxis in such settings.

The NCCN guidelines recommend LMWHs, fondaparinux, or UFHs for the acute treatment of VTE while the diagnosis and risk are being assessed, with LMWHs preferred in patients who are expected to receive chronic anticoagulation therapy. Warfarin can be used in patients requiring chronic anticoagulation but should be initiated in a 5-day to 7-day transition period with the LMWH, fondaparinux, or UFHs and be monitored to INR.30 The guidelines state that LMWHs such as enoxaparin, dalteparin, and tinzaparin are commonly considered therapeutically equivalent, but each has distinct pharmacokinetics and few clinical studies to date have directly compared the clinical effects of these agents.30 LMWH as monotherapy (without warfarin) is recommended for the treatment of proximal DVT or PE and the prevention of recurrent VTE in patients with advanced or metastatic cancer.30 Indefinite anticoagulation should be considered if cancer is active or important risk factors are persistent.30

There are few data regarding the impact of thrombosis on quality of life in cancer patients. Likewise, the impact of VTE on the delivery of optimal cancer treatment has received little attention. The prospective international Prospective Registry of Cancer and Events Involving Thromboembolism (PERCEIVE) Registry is designed to study the extent to which VTE complicates the course of common solid tumor malignancies and subsequent clinical outcomes.75 In addition, a prospective, randomized clinical trial will compare the safety and efficacy of LMWH prophylaxis (dalteparin) versus no treatment in reducing VTE in high-risk ambulatory cancer patients initiating chemotherapy.76

Representatives of the major international guidelines panels have recently issued a call to action for improved treatment and prevention strategies as well as a sustained research effort to further our understanding of the relation between cancer and thrombosis to reduce the burden of VTE and its consequences in patients with cancer.77

New Pharmacologic Options for the Treatment and Prevention of VTE

The evaluation of new anticoagulants is important to enhance the treatment options available for patients with cancer. New agents for the pharmacologic treatment and prevention of VTE include the parenteral agents bemiparin and semuloparin, as well as the oral agents rivaroxaban and apixaban. It is anticipated that oral agents may provide greater convenience of administration, whereas parenteral agents continue to be more suitable in the hospital setting for patients undergoing active cancer treatment, as well as for some patients with advanced malignancy.


Bemiparin, a LMWH with anti-factor Xa/anti-factor IIa activity,78 has been studied for the prevention of VTE with prolonged use in cancer patients undergoing abdominal or pelvic surgery.79 In the CANBESURE (Cancer, Bemiparin and Surgery Evaluation) study, Kakkar et al randomized 703 cancer surgery patients to receive once-daily subcutaneous injections of bemiparin at a dose of 3500 IU (with the first dose administered 6 hours after surgery) for approximately 1 week. Patients were then randomized to receive bemiparin or placebo for an additional 3 weeks. Major VTE (composite of proximal DVT, nonfatal PE, and VTE-related deaths) occurred in 0.4% of patients in the bemiparin group compared with 3.3% in the placebo group (RR reduction, 87.9%; 95% CI, 98.5%, 4.0% [P = .016]). Bemiparin was found to significantly reduce the rate of major VTE without significantly increasing the risk of hemorrhagic complications compared with 1 week of bemiparin prophylaxis and subsequent placebo.79


Semuloparin, another parenteral agent, is a subcutaneous ultra-LMWH that acts as a factor Xa inhibitor with residual anti-factor IIa activity.80, 81 Semuloparin is being studied for VTE prevention in patients with cancer as well as in patients undergoing major abdominal or orthopedic surgery.82 The dose response of semuloparin was recently examined in patients undergoing total knee replacement surgery (TREK study).80 There was a significant dose response across the 5 semuloparin doses tested, with the incidence of VTE ranging from 5.3% (dose of 60 mg/day) to 44.1% (dose of 10 mg/day) for semuloparin. The 3 highest doses of semuloparin (20 mg/day, 40 mg/day, and 60 mg/day) were found to be significantly more effective at reducing confirmed VTE compared with 40 mg/day of enoxaparin (used as calibrator), reducing the risk of VTE by 58%, 61%, and 85%, respectively. Six patients in the semuloparin groups (4 in the 60-mg group, 1 in the 40-mg group, and 1 in the 20-mg group) experienced major bleeding compared with none in the enoxaparin calibrator group. The 20-mg dose was selected for further investigation and currently is being studied in several ongoing phase 3 trials. Two of these trials are studying semuloparin use in cancer patients. The SAVE-ONCO (Evaluation of AVE5026 in the Prevention of Venous Thromboembolism in Cancer Patients Undergoing Chemotherapy) trial is evaluating semuloparin for the prevention of VTE in cancer patients undergoing chemotherapy (NCT00694382).83 The SAVE-ABDO (Evaluation of AVE5026 as Compared to Enoxaparin for the Prevention of Venous Thromboembolism in Patients Undergoing Major Abdominal Surgery) trial (NCT00679588) is evaluating semuloparin compared with enoxaparin for the prevention of VTE in patients undergoing major surgery of the abdomen and/or pelvis, and includes patients undergoing cancer surgery.84


Rivaroxaban is an oral direct inhibitor of factor Xa85 and currently is being studied for the prevention of DVT and PE in patients undergoing hip or knee replacement (Regulation of Coagulation in Major Orthopedic Surgery Reducing the Risk of DVT and PE [RECORD] 1-4 trials).86 The Venous Thromboembolic Event Prophylaxis in Medically Ill Patients (MAGELLAN) trial will evaluate whether extended therapy with oral rivaroxaban can prevent blood clots in the leg and lung that can occur in patients hospitalized for acute illness (including active cancer patients); results will be compared with a standard regimen of enoxaparin.87


Apixaban is another oral direct inhibitor of factor Xa. As demonstrated in an interim analysis of a phase 2 study, apixaban was found to be well tolerated in patients with metastatic cancer. The incidences of major bleeding and thrombosis among 125 patients were very low (major bleeding reported in 2 patients receiving apixaban at a dose of 20 mg and 1 patient receiving placebo; thrombosis was reported in 3 cases in the placebo group).88


VTE is a common complication of cancer and cancer treatment and is associated with considerable morbidity and mortality. Hospitalized medical and surgical patients with cancer are at an increased risk of VTE and should be considered for pharmacologic prophylaxis if no contraindication to anticoagulation is present.

Patients with cancer who are treated for documented VTE should be considered for continued anticoagulation, preferably with LMWH, for up to 6 months or longer in the presence of active malignancy. Routine thromboprophylaxis in ambulatory patients with cancer is not currently recommended. Nevertheless, many ambulatory cancer patients are also at an increased risk of thrombosis. Although results from randomized controlled trials are still needed, thromboprophylaxis may be considered in selected high-risk patients, such as those with multiple myeloma who are receiving thalidomide or lenalidomide plus chemotherapy. Consideration of prophylactic anticoagulation in patients with cancer must always balance the risk of VTE with the increased risk of bleeding. Improved methods for the identification of ambulatory patients with cancer who are at an increased risk of VTE, including assessing clinical risk factors and using biomarkers, are currently under investigation and should enable safe, effective, and targeted thromboprophylaxis.


Dr. Lyman is supported by grants from the National Heart, Lung and Blood Institute (1R01HL095109-01) and the National Cancer Institute (RC2CA148041-01). Editorial support for this article was provided by Peloton Advantage LLC and funded by Sanofi-Aventis US. However, the author had complete independence in defining content and in all editorial decisions with respect to this article.