Despite an increase in the incidence of cancer in an ageing population around the world, survival of cancer patients is improving due to the introduction of novel therapies, improvements in existing treatment strategies and cancer prevention programmes. The annual incidence of VTE in patients with cancer ranges from 0.5 to 20% depending on the type of cancer, disease stage and associated treatment regimens, and is predicted to increase in the future [
18], while VTE incidence in the general population remains unchanged [
19]. Moreover, the risk of VTE is higher in more advanced stages of cancer, or with a metastatic disease, compared with less advanced cancer [
20]. The association between VTE and cancer is also bidirectional, with 20% of all patients with VTE diagnosed with cancer, typically at the time that, or soon after, VTE is diagnosed [
21‐
24]. The mortality rate in patients with cancer and concurrent VTE is 3-fold higher than that in cancer patients without VTE, and VTE is the second most common cause of death, after the malignancy itself, in patients with cancer [
25]. Therefore, the primary and secondary prevention of VTE presents an important unmet need in this patient population.
Epidemiology of VTE in patients with cancer
Current guidelines highlight lung, pancreatic, ovarian, gastrointestinal and brain tumours, as well as haematological malignancies, as those associated with the highest risk of thrombosis; lymphoid, gynaecological (other than ovarian) and bladder tumours as those carrying high risk of thrombosis; and breast, head and neck, and prostate cancers presenting with a lower risk of VTE [
18,
26‐
28].
The incidence of VTE varies not only according to the type of cancer, but also within cancer types, specifically lung, ovarian, oesophageal, gastric, pancreatic and brain cancers, and chronic lymphocytic leukaemia. The incidence of VTE is particularly high in lung, pancreatic and brain cancers [
29‐
31], and a paucity of data on how to manage these patients was noted by the experts who were interviewed, especially when patients are undergoing chemotherapy. For example, in a study of patients with lung cancer who received cancer treatment with curative intent, the cumulative incidence of VTE after 1 year reached 13.5% [
29], while another study that examined VTE incidence in patients with non-small-cell lung cancer reported 6-month and 2-year rates of VTE of 4.2 and 6.4%, respectively [
32]. In general, VTE incidence in lung cancer varies widely, from 1.3 to 21.5% [
33‐
36], but the presence of VTE is consistently associated with a worse overall survival in such patients [
36‐
38].
A systematic review and meta-analysis of VTE incidence rates reported a 2-year cumulative incidence rate of 11.2% in advanced pancreatic cancer [
30], while in a more recent study the incidence of symptomatic and asymptomatic VTE was 16.5% in patients with pancreatic cancer [
39]. In a meta-analysis of patients with brain tumours, the risk of VTE was found to be significantly related to glioma (risk ratio [RR] = 1.68,
P < 0.001), high-grade glioma (RR = 1.70,
P < 0.001) and glioblastoma multiforme (GBM) (RR = 1.74,
P < 0.001) [
31]. Overall data on risk for VTE according to the type of cancer and staging will guide clinical decisions on the use of thromboprophylaxis.
Clinical practice guidelines provide evidence-based recommendations for the primary prevention of VTE in patients with multiple myeloma (MM) who are undergoing treatment with cytotoxic chemotherapy and immunomodulatory imide drugs (IMiD) [
40]. This evidence is based on the association of MM with VTE. At baseline the VTE rate is 3–4% in these patients, and this is further increased with exposure to risk factors, including treatment with high-dose dexamethasone, cytotoxic chemotherapy (doxorubicin), IMiDs (thalidomide and lenalidomide), erythropoiesis-stimulating agents, reduced mobility, fractures, and personal or family history of thrombosis [
40,
41].
Chemotherapeutic regimens also contribute to the development of VTE, as the annual incidence of VTE in cancer patients treated with chemotherapy is 1 in 200, while in the general population it is 1 in 855 [
42]. Outpatients receiving chemotherapy have a 6.5-fold increase in the risk of VTE compared to patients not treated with cytotoxic therapy [
43], and the second most common cause of death in patients receiving outpatient chemotherapy is VTE [
44].
In summary, expert opinion and a number of studies highlight the importance of identifying high-risk patients within cancer populations, who may benefit from thromboprophylaxis to prevent VTE and, in turn, decrease VTE-associated morbidity and mortality.
Risk assessment models and biomarkers for prediction of primary and recurrent VTE in patients with cancer
Guidelines from the British Committee for Standards in Haematology [
45] recommend the Khorana risk score (KS) as a tool for categorising patients into very high VTE risk patients, such as those with gastric and pancreatic cancer, and high VTE risk patients, such as those with lung, gynaecological, bladder or testicular cancer, or lymphoma [
45]. However, more recent evidence has shown that a high-risk score does not necessarily predict presence of VTE in patients with lung cancer, although it does associate with all-cause mortality [
46]. Moreover, the KS has insufficient precision in stratifying patients with lung cancer who are receiving chemotherapy into high- and low-risk groups [
47], and those with pancreatic cancer into high- and intermediate-risk groups [
48,
49]. The latter high-risk patient group, however, could be identified by combining the KS or CONKO scores with an activated partial thromboplastin time [
48]. In addition, the predictive value of the KS in cancer-associated thrombosis was higher when the analysis included platinum-based chemotherapy and the presence of distant metastases [
50]. Another VTE risk assessment tool, the Ottawa score, considers the type of cancer, disease stage, gender and history of thrombosis, but it could not adequately predict recurrent VTE in patients receiving anticoagulants [
51]. The risk of VTE fluctuates throughout a patient’s disease course, and the type of cancer, stage and therapeutic regimens will have an impact on the level of VTE risk [
52]. The experts recognise that there is an unmet need for improving VTE risk assessment tools for identifying cancer patients at risk, and for balancing those risks against anticoagulant-induced bleeding.
Current risk-assessment-based decisions that usually rely on clinical parameters as potential VTE biomarkers, including D-dimer and other biomarkers, are unlikely to have a significant impact on routine clinical practice in the foreseeable future [
25]. However, a recent systematic review on biomarkers for prediction of thromboembolism in lung cancer demonstrated that D-dimer and epidermal growth factor receptor mutation were the most reproducible predictors of thromboembolism in this patient population [
53]. Circulating tissue factor emerged as a potential biomarker in its highest quartile, where it was associated with the highest VTE recurrence rate in patients with cancer who were receiving anticoagulants, but this marker is not widely used in clinical practice [
54].
Taken together, and with the experts advocating for personalised treatment based on risk-assessment models, these data demonstrate an urgent need to develop practical, realistic and useful risk assessment tools, which would be able to stratify cancer patients into high-, intermediate- and low-risk primary and recurrent VTE groups eligible for targeted thromboprophylaxis approaches.
Primary prevention of VTE in patients with cancer
Primary prophylaxis in patients with cancer should be considered, as the experts agreed on the importance of improving prevention of VTE, which presents a challenge in terms of recurrent thrombosis and clinically relevant bleeding. It was acknowledged that significant improvements in inpatient thromboprophylaxis have occurred over recent years; however, beyond hospitalisation the benefits of extended prophylaxis remain less well-defined and require further investigation as patients may remain at risk of VTE after hospital discharge. Current guidelines advise against routine thromboprophylaxis with low-molecular-weight heparin (LMWH) in ambulatory patients with active cancer receiving systemic anticancer therapy, such as adjuvant hormonal therapies or chemotherapy [
45,
55,
56]. Primary prophylaxis may be indicated in specific subpopulations only, including those with pancreatic or lung cancer, ambulatory patients at high thrombotic risk [
12,
27] and those receiving chemotherapy for prolonged periods of time [
57]. However, antithrombotic therapy that extends beyond 6 months is controversial, but may be advised for patients with metastatic disease receiving chemotherapy, immunochemotherapy or radiotherapy [
58]. It is also recommended that patients with cancer and reduced mobility who are admitted to hospital, or who are treated with thalidomide and lenalidomide combined with steroids or other systemic anticancer therapies, should receive prophylaxis with LMWH, unfractionated heparin (UFH) or fondaparinux [
7,
59]. In at-risk patients, VTE prophylaxis should begin as soon as VTE risk has been identified [
12], but administration of anticoagulants should consider comorbidities associated with cancer, the risk of bleeding and patient preferences [
27].
In patients with metastatic disease and high risk of bleeding, LMWH is the preferred option over other anticoagulants, while vitamin K antagonist (VKA) should be avoided [
6,
55]. However, in patients with renal failure, LMWH is not routinely recommended, as it increases the risk of major bleeding [
57]. In patients with MM at high risk of VTE, full-dose LMWH or adjusted-dose warfarin (targeted international normalised ratio ∼1.5) are recommended, in contrast to MM patients with low-risk factors who are advised low-dose aspirin [
7,
57]. A recent systematic review on MM patients who were treated with lenalidomide-based therapy and/or dexamethasone showed a reduction in VTE risk for patients on LMWH (1.4%) compared to patients on aspirin (10.7%) [
60].
Clinical practice guidelines recommend LMWH for thromboprophylaxis in low-bleeding-risk patients with locally advanced or metastatic pancreatic cancer or lung cancer treated with systemic anticancer therapy [
7]. The benefits of LMWH in reducing VTE risk were also demonstrated in the CONKO-004 trial, where the rates of symptomatic VTE reduced by 8.7% in patients with advanced-stage pancreatic cancer receiving primary prophylaxis with enoxaparin, compared to patients not receiving prophylaxis [
61]. Fondaparinux should be considered in patients with previous heparin-induced thrombocytopenia (HIT) [
6]. The role of direct oral anticoagulants (DOAC) for the prevention of VTE in cancer patients is uncertain [
27]. Recent results from the CASSINI trial demonstrated the safety and efficacy of thromboprophylaxis with rivaroxaban in patients with cancer [
62]. The study examined this DOAC for VTE prevention in ambulatory patients with various cancers and found that VTE and VTE-related deaths were significantly reduced during the on-treatment period, and major bleeding was low. However, the AVERT trial, which examined the safety and efficacy of apixaban to prevent VTE in high-risk cancer patients, found that although lower rates of VTE were observed in the apixaban group compared with the placebo group, major bleeding rates were much higher [
63]. These contradictory results suggest that more trials are needed to consolidate future recommendations on the use of DOACs in this patient population [
62].
Secondary prevention and treatment of VTE
In cancer patients who develop recurrent VTE despite appropriate anticoagulant therapy, expert opinion guidance suggests three treatment options, which include increasing the LMWH dose by 20–25%, switching therapy from VKA to LMWH, or inserting an inferior vena cava (IVC) filter in combination with anticoagulation therapy [
7]. If a patient with cancer develops DVT or PE, treatment guidelines recommend initiating treatment with a once-daily LMWH regimen, with suggested pharmacological alternatives of fondaparinux for patients with ongoing or prior HIT, or UFH for patients with severe renal insufficiency or who are dialysis-dependent. An IVC filter should only be considered in selected patients with an absolute contraindication to anticoagulant therapy, given the prothrombotic stimulus of such foreign bodies [
6,
7]. Finally, thrombolytic therapy should only be considered in patients with clinically massive DVT or PE, and with caution given the increased bleeding risk of patients with cancer [
7]. A minimum of 3 months’ anticoagulant therapy is recommended, with LMWH preferred over VKAs and with consideration given to at least 6 months of treatment, especially in patients who are receiving cancer treatment or with metastatic disease [
7,
59,
64,
65]. However, the qualitative interviews highlighted a lack of consensus among physicians regarding treatment after the initial 6-month period. Indeed, the DALTECAN study examined the efficacy and safety of up to 12 months’ treatment with dalteparin, and found that VTE recurrence and bleeding were clustered during the initial month after diagnosis, thereby supporting the long-term safety of LMWH therapy [
66]. After 6 months’ treatment, the need for ongoing anticoagulant therapy should be reassessed based on a risk versus benefit assessment in conjunction with patient values and preference [
12].
The lack of published data on the benefits of VTE reduction through thromboprophylaxis versus the risks of bleeding was highlighted by the experts as one of the reasons why guideline recommendations are not routinely followed by physicians. Patients with GBM treated with lifelong anticoagulation have a reduced rate of recurrent VTE but, despite these findings, thromboprophylaxis is underutilised in such patients [
67]. According to the experts interviewed, this may be partly due to differences in the views of oncology specialists regarding the need for treatment of established VTE compared to the role of primary thromboprophylaxis. In contrast, another study demonstrated that patients with cancer treated with anticoagulant therapy suffered a 3-fold higher incidence of intracranial haemorrhage [
68]. The differences in the rate of VTE recurrence incidence and major bleeding events were linked to the type of cancer, as a study demonstrated that the rates of VTE recurrence and major bleeding events during the course of anticoagulation were similar in patients with breast or colorectal cancer, whereas a 2-fold higher rate of thromboembolic recurrences than the rate of major bleeding events was identified in patients with lung cancer, and a lower rate of VTE than bleeding was recorded in patients with prostate cancer [
19]. These data underscore the importance of considering the type of cancer and associated comorbidities in order to weigh the potential benefits and risks of anticoagulant prophylaxis.
For the past two decades, LMWHs have been the preferred first-line treatment option for the management of cancer-associated VTE. This is supported by results from the qualitative survey, which found that 75% (33/44) of physicians interviewed use LMWH as standard of care for thrombosis treatment. Several seminal studies have shown superior efficacy and safety of LMWHs over VKA. The CANTHANOX study reported 10.6% more patients experiencing one combined major outcome event, such as major bleeding or recurrent VTE within a 3-month period, in the warfarin group compared to the fixed-dose enoxaparin group [
69]. The CLOT study demonstrated that a weight-adjusted dose of dalteparin was more effective in reducing the probability of recurrent VTE compared to a warfarin derivative over a 6-month treatment period [
64]. Similarly, the LITE study found a greater number of VTE episodes in the VKA than the tinzaparin treatment group at 3 and 12 months, with largely similar minor bleeding complications [
70], whereas the more recent CATCH study demonstrated a similar rate of recurrent VTE over a 6-month period with tinzaparin compared to warfarin, but a lower rate of relevant non-major bleeding was noted in the tinzaparin group [
71]. The cumulative probability of being VTE-free at 6 months, as demonstrated by the ONCENOX study, was higher for the group receiving enoxaparin than for the group where treatment with enoxaparin preceded that with warfarin [
72]. In summary, LMWHs seem to be preferred anticoagulants over VKAs.
More recently, DOACs have emerged as a potential alternative first-line treatment option for cancer-associated VTE, but with caveats. In general, the type of anticoagulant administered should be tailored according to patient and cancer type characteristics. The expert discussions and interviews highlighted that the advantages of LMWHs over DOACs include the ease of adapting the dose to the patient’s body weight and anticoagulation need, no drug–drug interactions related to chemotherapy regimens, and flexibility around procedures and other clinical situations (e.g., thrombocytopenia) that require treatment interruption or dose reduction. However, the experts agreed that prolonged drug administration through subcutaneous injection was the most common disadvantage associated with LMWH treatment, followed, in certain countries, by the relatively high cost of LMWH. Currently, DOACs are used for treatment of VTE in patients with stable cancer who are not receiving anticancer therapy and when VKAs are unavailable [
59]. Some published studies, specifically the AMPLIFY trial [
73] and the Hokusai-VTE trial [
74], demonstrate that DOACs have similar efficacy to that of LMWHs or warfarin. Nevertheless, the authors agreed that bleeding risk associated with DOACs should be addressed, as bleeding may be a more frequent cause of death than fatal VTE. The HOKUSAI-CANCER study compared dalteparin with edoxaban over a 12-month period and demonstrated a comparable rate of VTE recurrence in both groups, although a higher rate of major bleeding was observed in the edoxaban group [
75]. In addition, in the recent SELECT-D study, patients treated with rivaroxaban displayed a lower cumulative VTE recurrence rate than those treated with dalteparin, but had a higher rate of major and clinically relevant non-major bleeding events in the 6-month period [
65].
The experts noted that the administration of LMWHs and DOACs may become interchangeable, as patients with cancer have a complex clinical course and receive many different therapies; for example, it is possible to envisage that LMWHs will be used during hospitalisation and DOACs in out-of-hospital periods. The cost was considered to be one of the major reasons for insufficient adherence to guideline recommendations for the use of LMWHs. Among other reasons for the suboptimal use of LMWHs are inconvenience of LMWH injections and insufficient awareness of care givers regarding the importance of secondary prevention of VTE.
In conclusion, it is important to consider cancer site, stage of the disease and anticancer treatments given to patients to ensure the choice of an optimal anticoagulant and its dosage for secondary prevention of VTE.