The Washington Manual of Oncology, 3 Ed.

Cancer and Thrombosis

Kristen Sanfilippo • Tzu-Fei Wang

 I. INTRODUCTION. Cancer patients have a 5- to 7-fold increased risk of developing venous thromboemboli (VTE) compared with noncancer patients (JAMA 2005;293:715), with cancer-associated VTE accounting for 20% to 30% of all VTEs (J Thromb Haemost 2007;5:692). Patients with cancer-associated VTE had decreased survival compared with their thrombosis-free counterparts matched by cancer type and stage. When treated with anticoagulation, cancer patients have higher rates of major bleeding events compared with patients with VTE without cancer (Blood 2002;100:3484).

1. PATHOPHYSIOLOGY AND RISK FACTORS. A clearer understanding of the complex pathophysiology of cancer-associated thrombosis is beginning to emerge. In response to cancer-induced inflammatory cytokines, tissue factor (TF) or cancer procoagulants are aberrantly expressed on the surface of cancer cells, monocytes, and endothelial cells, promoting a hypercoagulable state, angiogenesis, and tumor metastasis. Secondly, decreased patient activity due to the disease or therapy leads to increased venous stasis. In addition, vascular injury from surgery, chemotherapy, radiation, and central venous catheters (CVCs) are major risk factors for VTE, completing Vichow’s triad. Thus, cancer is a heterogeneous disease, with varying VTE risk based on cancer subtype and stage, patient-associated factors, and cancer-specific therapy.
2. Type of cancer. A higher VTE risk has been noted in biologically aggressive malignancies with early metastatic potential and short overall survivals. Cancers associated with the highest risk of VTE include lung, pancreas, brain, ovary, and hematologic malignancies (myeloma, lymphoma, leukemia) (JAMA 2005;293:715).
3. Stage of cancer. As expected, more advanced stage is associated with higher risk (JAMA 2005;293:715).
4. Timing related to cancer diagnosis. Patients are at highest risk for VTE in the time period immediately following cancer diagnosis. This is hypothesized to be due to the presence of the largest disease burden; in addition, this period correlates with initiation of therapy (i.e., chemotherapy/surgery). In a study by Blom et al., VTE risk was highest in the 3 months after diagnosis (53-fold risk) and decreased over time (JAMA2005;293:715).
5. Patient-associated risk factors. These are similar to those in noncancer patients and include age, race, obesity, presence of medical comorbidities, surgery, history of VTE, presence of hereditary thrombophilia, and leukocytosis and/or thrombocytosis.
6. Cancer treatment. Recent surgery is a well-documented risk factor for VTE in both cancer and noncancer patients. Pathophysiology is attributed to direct vascular damage, prolonged immobility, and presence of an inflammatory state. It should be assumed that all patients receiving chemotherapy or hormonal therapy are at increased risk of VTE. Plausible mechanisms for this prothrombotic state include decreased activity of physiologic anticoagulants, release of procoagulants from apoptotic cancer cells, and drug-induced injury to endothelial cells. Some of the specific therapies that have been associated with increased risk of VTE in randomized, prospective trials include cisplatin therapy, hormonal therapy, antiangiogenic agents, erythrocyte-stimulating agents (ESAs), and immunomodulatory agents (i.e., lenalidomide and thalidomide).

 The knowledge of these potential risk factors for VTE in cancer patients can guide clinicians to identify patients with increased risk of VTE.

 Several risk prediction models have been generated to identify cancer patients at greatest risk of developing VTE. The 2013 American Society of Clinical Oncology (ASCO) VTE guidelines recommend the use of the model proposed by Khorana et al., given that it has been validated in a large population of cancer patients (Blood 2008;111:4902). The model was generated using a cohort of 4,066 ambulatory cancer patients initiated on chemotherapy. Several important risk factors for VTE were identified in multivariate analysis, including site of cancer, prechemotherapy platelet and leukocyte counts, hemoglobin or use of red cell growth factors, and body mass index (BMI). Each risk factor was assigned a corresponding score in the point system. Patients were then divided into three groups: high risk ($3 points), intermediate risk (1 to 2 points), and low risk (0 points), with VTE rates of 6.7%, 2%, and 0.3%, respectively, over a 2.5-month period. While the clinical utility of such models remains poorly defined beyond patient education, incorporation of these tools in the future will allow identification of patients at highest risk of VTE and potential consideration for thromboprophylaxis.

 III. VENOUS THROMBOSES AND OCCULT CANCER. About 20% to 30% of all newly diagnosed VTE are cancer associated. The majority of these cases will present with thrombosis following diagnosis of an established malignancy. However, a significant percentage of patients with seemingly idiopathic VTE will subsequently be diagnosed with cancer. This association has raised the question of the clinical benefit of cancer screening in patients who present with idiopathic VTE.

 In a landmark study by Prandoni et al. (N Engl J Med 1992;327:1128), 260 consecutive outpatients with objectively diagnosed deep vein thrombosis (DVT) were followed for 2 years. Development of cancer in patients was compared between the idiopathic (n = 153) and secondary DVT (n = 107) groups. History, physical, and routine laboratory testing were performed at VTE diagnosis in each group, identifying 3.3% (n = 5) of patients in the idiopathic (2 lung cancers, 1 multiple myeloma, 1 chronic lymphocytic leukemia, and 1 osteosarcoma) compared to no patients in the secondary DVT group with an underlying malignancy. During the 2-year follow-up in the remaining patients, symptomatic malignancies were diagnosed in 11 of 145 patients (7.6%) with idiopathic DVT compared to 2 of 105 patients (1.9%) with secondary DVT. The majority of malignancies (77%) were diagnosed in the first 12 months of follow-up, with all cases diagnosed by 18 months. A similar association between occult malignancy and idiopathic VTE is further supported by findings from retrospective studies using hospital discharge administrative databases.

 Given the association, oncologists may be asked to evaluate patients with idiopathic VTE for occult cancer. Limited prospective data are available for guidance. In the SOMIT trial, 233 patients with newly diagnosed, idiopathic VTE were randomized to receive cancer screening (abdominal/pelvic ultrasound or computed tomography [CT], endoscopy, colonoscopy, hemoccult testing, sputum cytology, serum tumor markers, mammogram, and pelvic exam with cytology for women and prostate ultrasonography for men) or routine follow-up (J Thromb Haemost 2004;2:884). At baseline, all patients underwent history, physical, and routine laboratory tests, during which 32 cancers were diagnosed (14%). An additional 13 of 99 patients (13%) in the screening group were diagnosed with cancer based on the additional procedures. During the 2-year follow-up, one cancer was diagnosed in the screening arm compared to 10 cases in 102 patients in the control arm, with no significant difference in mortality between the arms (2% vs. 3.9%, respectively). However, the study was closed early due to poor accrual, failing to meet a target enrollment of 1,000 patients. A cost-effective analysis of the trial concluded that abdominal/pelvic CT was the most cost-effective test, and tumor markers were associated with high false-positive rates generating additional unnecessary testing.

 In a second, large prospective cohort study (J Thromb Haemost 2004;2:876), the impact of cancer screening was also assessed in new, idiopathic VTE. In this trial, 864 patients were initially evaluated with history, physical (including rectal), breast and pelvic exam in women, routine laboratory tests (in addition to erythrocyte sedimentation rate and serum protein electrophoresis), and chest X-ray (CXR), at which time a total of 34 cancers were detected, the majority of which were limited in stage (61%). Patients who did not have a cancer diagnosed in step 1 (n = 830) underwent a “limited” workup consisted of an abdominal/pelvic ultrasound, Carcinoembryonic antigen (CEA), and Prostate-specific antigen (PSA) in men and CA-125 in women, revealing an additional 13 cancers. The remaining patients (n = 817) were followed for 12 months for the occurrence of cancer. During the 12-month follow-up, 14 additional malignancies were diagnosed, of which 14% were limited stage for a total of 61 malignancies. The results of this study suggest that in adults with idiopathic VTE, more than 50% of underlying occult malignancies can be detected with a limited initial evaluation, followed by age/sex appropriate cancer screening. Patients diagnosed at the time of presentation tend to be diagnosed with earlier stage cancers compared with those diagnosed during follow-up. Despite diagnosis at early stage, data have not yet indicated a proven survival benefit for cancer screening in patients with idiopathic VTE.

 In summary, 20% to 30% of all newly diagnosed idiopathic VTEs are cancer associated. While a significant proportion of cancers are known at the time of VTE diagnosis, the majority of the remaining cases will be established between presentation and the following 12 months. Aggressive screening for occult cancers in asymptomatic patients with idiopathic VTEs has not been associated with improvement in survival and is thus not recommended. However, limited evaluation with history, physical, routine labs, and age/gender appropriate cancer screening is a reasonable strategy.

 IV. PREVENTION OF VTE IN PATIENTS WITH CANCER. Cancer patients have a sevenfold increased risk of VTE compared to persons without cancer, with risk highest in the first 3 months after cancer diagnosis (53-fold increased risk) (JAMA 2005;293:715). One of the main reasons for this high risk of VTE is due to surgical, chemotherapy, and hormonal interventions, and can be reduced by employing prevention strategies.

1. Prophylaxis against VTE in patients with cancer in the perioperative setting. The incidence of VTE after cancer surgery may be as high as 50% without prophylaxis. Although mechanical prophylaxis can reduce postoperative VTE by approximately 50%, it is inferior to anticoagulant prophylaxis.

 Prospective studies comparing prophylactically dosed low molecular weight heparin (LMWH) versus unfractionated heparin (UFH) for 7 to 10 days postoperatively in cancer patients have consistently shown equivalent efficacy and safety with VTE rates of approximately 15% and major bleeding incidents of approximately 4%. These findings formed the basis for recommendations that cancer patients should receive postoperative VTE prophylaxis for 7 to 10 days following major surgery.

 There are compelling data to support extended DVT prophylaxis beyond 10 days following some types of cancer surgery (i.e., abdominal and pelvic). In the ENOXACAN II study, patients were randomized to receive prophylactic enoxaparin for 6 to 10 (routine prophylaxis) versus 25 to 31 days (extended prophylaxis) after curative open surgery for abdominal or pelvic malignancies (N Engl J Med 2002;346:975). The incidence of DVT was significantly different: 12% for routine and 4% for extended prophylaxis, while bleeding complications were comparable (3.6% and 4.7% respectively). On the basis of this, the 2012 American College of Chest Physicians (ACCP) guidelines recommend extended DVT prophylaxis with LMWH for patients undergoing abdominal/pelvic open surgery for cancer (Chest 2012;141:e227S).

 There is limited consensus on prophylactic regimens following minor or less invasive surgeries. Patients undergoing laparoscopic colorectal, urologic, and gynecologic cancer surgeries have documented increased risk for VTE complications. Thus, 7 to 10 days of postoperative DVT prophylaxis with UFH or LMWH is reasonable until future studies determine the ideal prophylaxis schedule for this population.

 Lastly, patients who undergo surgery for central nervous system neoplasms have one of the highest rates of postoperative VTE and lowest tolerance for bleeding complications. Several prospective randomized trials have validated the safety and efficacy of VTE prophylaxis with UFH and LMWH starting approximately 24 hours after surgery (Br J Haematol 2004;128:291–302). These findings are supported in the ACCP guidelines recommending pharmacologic prophylaxis following craniotomy in cancer patients (Chest 2012;141:e227S).

1. Prophylaxis against VTE in hospitalized cancer patients. All hospitalized oncology patients admitted for management of acute medical conditions or cancer therapy should be considered for mechanical or pharmacologic VTE prophylaxis. Unfortunately, there is a lack of data supporting these recommendations that are primarily based on studies in general hospitalized, acutely ill medical patients. Several prospective, randomized VTE prevention studies have confirmed the efficacy and safety of LMWH and UFH prophylaxis in acutely ill medical patients; however, cancer patients have comprised only 5% to 14% of these populations. A subset analysis of cancer patients in the MEDENOX (N Engl J Med 1999;341:793) study detected a 50% reduction in VTE risk in patients receiving enoxaparin versus placebo.

 On the basis of the available data, the 2012 ACCP guidelines recommend UFH or LMWH for VTE prophylaxis in hospitalized cancer patients (Chest 2012;141:e227S). Temporary risk factors for bleeding complications, including invasive procedures or thrombocytopenia, may require interruption of anticoagulant prophylaxis. Substitution of mechanical devices until the bleeding risk is resolved should be pursued, and should not lead to complete avoidance of UFH or LMWH. Ambulation should also be conscientiously encouraged during hospitalization.

1. Prophylaxis in ambulatory cancer patients. Given the high risk of developing VTE in cancer patients, numerous investigators have assessed the efficacy of thromboprophylaxis. One smaller trial randomized patients with metastatic breast cancer receiving chemotherapy to warfarin 1 mg/day for 6 weeks followed by dose adjustment to an international normalized ratio (INR) target of 1.3 to 1.9 versus placebo showed an 85% reduction of symptomatic VTE in the treatment arm (incidence 0.16%/month vs. 0.7%/month, respectively) without a significant increase in bleeding (Lancet 1994;343:886). Two larger trials, the PROTECHT trial and the SAVE-ONCO trial, also showed a risk reduction in the incidence of VTE with thromboprophylaxis (N Engl J Med 2012;366:601) (Lancet Oncol 2009;10:943). In the PROTECHT trial, 1,166 patients with metastatic or locally advanced cancer were randomized in a 2:1 manner to prophylaxis with nadroparin versus placebo, starting on day 1 of chemotherapy to a maximum of 4 months. Patients receiving nadroparin had a significant reduction in the risk of VTE with an incidence of 1.4% compared to 2.9% in the placebo arm. The SAVE-ONCO trial was the largest of the three trials, randomizing 3,212 patients with metastatic or locally advanced cancer to prophylaxis with semuloparin versus placebo, starting on day 1 of chemotherapy to a maximum of 4 months. The primary outcome was development of any VTE or VTE-related death, and the study revealed a 64% risk reduction in patients receiving semuloparin (HR 0.36; 95% CI 0.21 to 0.60).

 Despite the efficacy and safety of primary VTE prophylaxis in the cancer population, routine primary VTE prophylaxis with anticoagulants during chemotherapy is not currently recommend. This is likely in part due to the overall low occurrence of VTE in the available studies, and the concern of increased risk of bleeding in cancer patients receiving chemotherapy. However, guidelines do recommend consideration of LMWH prophylaxis on a case-by-case basis in selected cancer patients with highest risk of VTE, accompanied by a thorough risk–benefit description. Additionally, patients with multiple myeloma receiving thalidomide- or lenalidomide-based therapy and/or dexamethasone have been documented to have an exceedingly high risk of VTE. Therefore, the guidelines do recommend pharmacologic prophylaxis with either aspirin 81 mg or prophylactic dose of LMWH in this unique population.

1. CVC AND THROMBOSIS. Percutaneous inserted CVCs and port-a-caths provide reliable venous access for blood collection, as well as administration of chemotherapy, medications, and blood components to cancer patients. However, it is associated with several complications, one of which is thrombosis (including occluded catheter lumen, external catheter fibrin sheath, and partial or occlusive venous thrombus). The reported frequencies of symptomatic upper extremity DVT associated with CVC in cancer patients not using prophylactic anticoagulants decreased from approximately 38% (Ann Intern Med 1990;112:423) in the 1990s to around 4% in more recent years (J Clin Oncol 2005;23:4057), most likely reflecting refinements in catheter materials and insertion techniques. Although earlier randomized control trials (RCTs) have demonstrated a significant reduction in venography-confirmed upper extremity DVTs with either warfarin 1.0 mg/day or dalteparin 2,500 IU/day, further RCTs have failed to show a significant reduction of asymptomatic or symptomatic CVC-associated DVTs with warfarin 1.0 mg/day, enoxaparin 40 mg/day, or dalteparin 5,000 IU/day compared with placebo. On the basis of the results of these contemporary studies, the 2012 ACCP guidelines do not recommend routine anticoagulant DVT prophylaxis for cancer patients with CVCs. Patients with hematologic malignancies treated with intensive chemotherapeutic regimens appear to have a 2- to 4-fold higher risk of CVC-associated symptomatic DVTs compared to patients receiving outpatient chemotherapy for solid tumors. However, studies evaluating the risks and benefits of DVT prophylaxis in the population with hematologic malignancies have not been published, and routine anticoagulation is not recommended. Although several cohort studies have shown an increased relative risk of CVC-associated DVT in cancer patients who are also heterozygous for factor V Leiden or prothrombin gene G20210A mutation, routine screening for these inherited thrombophilia risk factors in cancer patients without a VTE history is not recommended.

 Optimal treatment for symptomatic CVC-associated DVTs is controversial. While the risk of symptomatic pulmonary embolism (PE) appears to be low, it remains a potentially fatal complication. Because of the lack of high-quality evidence, expert opinions have been used to guide routine practices. It is now thought that catheters can remain in place as long as it is still needed, functional, and with no other compelling reason for removal such as infections. Catheter should be removed whenever it is not needed or not functional. It is recommended to continue anticoagulation as long as the catheter is present, until 3 months after removal of the catheter, unless if bleeding risk is unacceptable. Current studies are underway to investigate whether shorter duration of anticoagulation (4 to 6 weeks) is appropriate after catheter removal.

 VI. DIAGNOSIS AND TREATMENT OF VTE IN CANCER PATIENTS

1. Diagnosis of VTE. Although the combination of a low clinical assessment score and a negative quantitative D-dimer result has been validated for ruling out VTE in patients with a low risk of thrombosis, this strategy should be used with great caution in cancer patients suspected of VTE. In cancer patients, a higher underlying prevalence of VTE may produce an unacceptably high false-negative rate and low negative predictive value using the same decision rule. In addition, D-dimer is typically elevated in cancer patients in the absence of an underlying VTE, producing a high false-positive rate and a low positive predictive value for the detection of VTE. A clinical suspicion for VTE in cancer patients requires sensitive imaging studies, guided by the presenting signs and symptoms, to assess for VTE. The noninvasive imaging techniques, including lower extremity duplex ultrasonography, spiral chest CT, and ventilation–perfusion scan, are likely to have similar sensitivities and specificities for the detection of DVT and PE in cancer patients as have been reported in patients without underlying malignancies.

 CVC-associated upper extremity DVTs tend to be more centrally located, and ultrasonographic techniques have consistently been shown to have suboptimal sensitivities (ranging from 56% to 94%) with specificities of 100%. Therefore, although a positive upper extremity duplex color ultrasonographic study confirms a suspected CVC-associated DVT, a negative study in the setting of high suspicion requires more sensitive imaging techniques such as venography, CT, or magnetic resonance imaging to rule out a thrombus.

1. Therapy of VTE in patients with cancer
2. Special issues in patients with cancer. Three important issues related to treatment of cancer-associated VTE are the safety, type, and duration of anticoagulation therapy. The only absolute contraindication to therapeutic anticoagulation is active bleeding that cannot be rapidly and reliably controlled. However, the bleeding risk can be difficult to predict in a quantitative manner for each individual patient. Treatment of VTEs in patients with primary and metastatic brain tumors is particularly challenging, due to the potentially devastating complications of intracranial hemorrhage associated with anticoagulation therapy, the high risk of recurrent VTE, and the inadequacy of high-quality evidence to guide management. Therapeutic anticoagulation is considered by some oncologists and neurosurgeons to be an absolute contraindication in patients with highly vascular brain metastases or recent craniotomies. For patients in whom the benefit of anticoagulation therapy is judged to outweigh the risk, one approach involves cautious initiation of continuous infusion of UFH, and, if tolerated, to switch to other long-term anticoagulation therapy such as LMWH.

 Thrombocytopenia, either cancer related or secondary to chemotherapy, complicates anticoagulation therapy. While data to define an evidence-based, safe, minimal platelet count are not available, expert opinions and anecdotal clinical experiences support the safety of therapeutic anticoagulation when the platelet count is >50,000/µL. When the platelet count drops below this level, lower intensity or interruption of chronic anticoagulation should be considered.

2. Role of inferior vena cava (IVC) filters. IVC filters are used to prevent development of a PE in patients with acute, lower extremity DVT when anticoagulation cannot be safely administered. Placement of an IVC filter is an option in cancer patients with acute lower extremity DVT and unacceptable bleeding risk or active bleeding. However, IVC filters are thrombogenic, and over time increase the risk of lower extremity DVT. Therefore, a retrievable filter should be used whenever possible, with reinitiation of anticoagulation and removal of the filter once the risk of acute bleeding has resolved.
3. Role of thrombolytic therapy. Indications for thrombolytic therapy in cancer patients with VTE are limited due to the increased risk of bleeding, especially in patients with brain tumors. Use should be considered for a PE causing severe hemodynamic instability, a DVT causing arterial insufficiency due to severe venous congestion, clinically significant extension of a thrombus despite therapeutic anticoagulation, and an occluded CVC that must be kept patent.
4. Initial treatment of VTE. Options for initial anticoagulation in cancer patients include continuous UFH, LMWH, and fondaparinux. The traditional approach was initial anticoagulation with UFH, using activated partial thromboplastin time (aPTT) for therapeutic monitoring, and concurrent warfarin until two therapeutic INRs (2 to 3) were obtained ≥24 hours apart, at which time UFH was stopped. Subgroup analyses of cancer patients initially treated with LMWH versus UFH for acute VTE showed similar efficacy for VTE recurrence. However, LMWH is now the preferred initial anticoagulant in cancer patients due to the ease of administration, ability to use in the outpatient setting, elimination of need for therapeutic monitoring, and lower risk of heparin-induced thrombocytopenia (HIT). In a study comparing twice-daily (1 mg/kg) versus once-daily (1.5 mg/kg) dosing of enoxaparin for acute VTE, the cancer patient subgroup had a nonsignificant higher VTE recurrence rate with once-daily dosing (12.2%) compared with twice-daily dosing (6.4%) (Ann Intern Med 2001;134:191). Limited data exist on the use of fondaparinux in cancer patients. Limitations include its long half-life (17 to 21 hours), 100% renal clearance, and inability to reverse. However, small studies have shown the safety of fondaparinux in patients with HIT. In summary, LMWH is the recommended treatment of choice for initial management of acute VTE in patients with cancer. UFH should be considered in patients with severe renal insufficiency (CrCl [creatinine clearance] <30 mL/minute), while fondaparinux can be considered for patients with active or a history of HIT.
5. Long-term anticoagulation for VTE in patients with cancer. While oral vitamin K antagonists (VKA) are the standard for treatment and secondary prevention of VTE in noncancer patients, they pose several dilemmas in cancer patients. Cancer patients have more thrombotic recurrences (fourfold higher) and bleeding complications (twofold higher) than patients without malignancies during long-term VKA therapy. It translates into cumulative recurrent VTE and bleeding incidences of 20% and 12%, respectively, after 1 year of anticoagulation therapy (Blood 2002;100:3484). Additionally, VKAs require frequent therapeutic monitoring, have a slow onset and offset of action, depend on adequate gastric absorption, and have numerous food and medication interactions.

 Several prospective, randomized, open-label studies have shown that LMWH is superior to VKAs in long-term management of VTE in cancer patients. The largest of these trials, the CLOT trial, randomized cancer patients to VKA versus dalteparin (LMWH) for 6 months following diagnosis of a first symptomatic VTE (N Engl J Med 2003;349:146). Patients receiving LMWH had a 52% reduction in VTE recurrence and no significant difference in major bleeding compared to those receiving VKA. A meta-analysis supported these findings when combing the results of eight prospective, randomized trials, reporting an overall 53% reduction in risk of recurrent VTE with LMWH compared to VKA therapy (HR 0.47; 95% CI 0.32 to 0.71) (Cochrane Database Syst Rev 2008:CD006650). The same meta-analysis found no difference in rates of bleeding or overall mortality between the two groups.

 Based on the aforementioned evidence, the 2012 ACCP, 2013 ASCO, and the National Comprehensive Cancer Network, all recommend chronic anticoagulation with LMWH when possible for treatment of DVT or PE in cancer patients.

 Novel oral anticoagulants (NOACs) such as rivaroxaban (Xarelto), dabigatran (Pradaxa), and apixaban (Eliquis) were approved for VTE and stroke prevention and VTE treatment in recent years. Among them, rivaroxaban is the only FDA-approved NOAC for VTE treatment in the general population currently. NOACs are attractive options in the cancer population, with their oral formulation, no need of monitoring, and less interactions with drug and diet. However, in all pivotal phase III trials leading to drug approval, cancer patients only constituted a small proportion of the enrollees. In addition, NOACs have not had head-to-head comparison with LMWH as the treatment of VTE, the superior and standard therapy for cancer patients. Given the lack of high-quality data, at the current time, major guidelines including ACCP, ASCO, and NCCN all recommend against routine use of NOAC for the treatment of VTE in the cancer population, and advocate further dedicated studies.

6. Duration of anticoagulation in cancer patients with VTE. The duration of anticoagulation in cancer patients must be individualized. In general, if cancer treatment leads to cure or durable remission, 3 to 6 months is appropriate. In the setting of ongoing active malignancy, the risk of recurrence is much higher, and consideration of long-term anticoagulation is appropriate and often used.
7. Treatment of incidental VTE. Incidental VTEs (VTE found incidentally on imaging studies without symptoms) are common in the cancer population, given the baseline high risk of thrombosis and the frequency of imaging studies for cancer staging. Several retrospective studies showed that 35% to 50% of DVTs and PEs were incidentally discovered in cancer patients. The need of treatment for incidental VTEs has been questioned, given the lack of symptoms. However, many studies have revealed that rates of VTE recurrence, bleeding, and mortality are similar in cancer patients with incidental VTE compared to those with symptomatic VTE. Therefore, expert consensus recommends the same treatment for incidental VTE as compared with symptomatic VTE.
8. Thrombosis despite anticoagulation. Cancer patients commonly have recurrent VTE despite anticoagulation. Approximately 10% to 17% of cancer patients on warfarin and 6% to 9% on LMWH will have recurrent VTE (N Engl J Med 2003;349:146). There is no standard therapy for patients with recurrent VTE despite anticoagulation, since no high-quality evidence is available. In this setting, it is important to first confirm compliance and rule out HIT. In the case of true failure of therapeutic anticoagulation, various treatment strategies have been employed, including switching anticoagulation from warfarin to LMWH (if the patient were to be on warfarin at the time of VTE recurrence), switching to a different type of LMWH (if the patient were to be on LMWH at the time of VTE recurrence), increasing the dose of LMWH, placement of IVC filter, addition of antiplatelet agents such as aspirin, or a combination of these methods. Studies are needed for optimal management of this group of challenging patients.

SUGGESTED READINGS

Agnelli G, George DJ, Kakkar AK, et al. Semuloparin for thromboprophylaxis in patients receiving chemotherapy for cancer. N Engl J Med 2012;366:601–609.

Agnelli G, Gussoni G, Bianchini C, et al. Nadroparin for the prevention of thromboembolic events in ambulatory patients with metastatic or locally advanced solid cancer receiving chemotherapy: a randomised, placebo-controlled, double-blind study. Lancet Oncol 2009;10:943–949.

Blom JW, Doggen CJ, Osanto S, et al. Malignancies, prothrombotic mutations, and the risk of venous thrombosis. JAMA 2005;293:715–722.

den Exter PL, Hooijer J, Dekkers OM, et al. Risk of recurrent venous thromboembolism and mortality in patients with cancer incidentally diagnosed with pulmonary embolism: a comparison with symptomatic patients. J Clin Oncol2011;29:2405–2409.

Falanga A. Thrombophilia in cancer: semin thrombosis. Hemostasis 2005;31:104–110.

Geerts WH, Pineo GF, Heit JA, et al. Prevention of venous thromboembolism: the Seventh ACCP Conference on Antithrombotic and Thrombolytic Therapy. Chest 2004;126:338S–400S.

Khorana AA, Kuderer NM, Culakova E, et al. Development and validation of a predictive model for chemotherapy-associated thrombosis. Blood 2008;111:4902–4907.

Lee AY, Levine MN, Baker RI, et al. Low-molecular-weight heparin versus a coumarin for the prevention of recurrent venous thromboembolism in patients with cancer. N Engl J Med 2003;349:146–153.

Lee AY. Management of thrombosis in cancer: primary prevention and secondary prophylaxis. Br J Haematol 2004;128:291–302.

Lyman GH, Khorana AA, Kuderer NM, et al. Venous thromboembolism prophylaxis and treatment in patients with cancer: American Society of Clinical Oncology clinical practice guideline update. J Clin Oncol 2013;31:2189–2204.

Piccioli A, Lensing AW, Prins MH, et al. Extensive screening for occult malignant disease in idiopathic venous thromboembolism: a prospective randomized clinical trial. Journal of thrombosis and haemostasis. JTH 2004;2:884–889.

Prandoni P, Lensing AW, Buller HR, et al. Deep-vein thrombosis and the incidence of subsequent symptomatic cancer. N Engl J Med 1992;327:1128–1133.

Prandoni P, Lensing AW, Piccioli A, et al. Recurrent venous thromboembolism and bleeding complications during anticoagulant treatment in patients with cancer and venous thrombosis. Blood 2002;100:3484–3488.