Introduction
Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), is a common complication among patients with cancer and is one of the most common causes of increased morbidity and mortality [
1]. The risk factors for VTE are multifactorial and include cancer itself, chemotherapeutic agents, and patient-specific characteristics [
2], such as female sex, older age, and comorbidities like diabetes and atherosclerosis [
3].
The management of VTE in patients with cancer is challenging due to an increased risk of bleeding and VTE recurrence [
4]. Low-molecular-weight heparin (LMWH) has been the gold standard treatment for cancer-associated VTE (CA-VTE) and is recommended over vitamin K antagonist (VKA) on the basis of evidence from several randomized controlled trials (RCTs) [
5,
6]. However, considering the patient preference, availability of oral anticoagulants, cost, and the risk of thrombocytopenia, the use of this agent might not be the best option in cancer patients with VTE [
7,
8].
Direct oral anticoagulants (DOACs), including apixaban, edoxaban, rivaroxaban, betrixaban, and dabigatran, have been approved for use in medical patients for VTE prophylaxis or treatment [
9‐
11]. Since the original RCTs evaluating DOACs for VTE management included a limited number of cancer patients, separate studies were designed to investigate the use of DOACs specifically for CA-VTE. Since then, multiple RCTs have been conducted among cancer patients to evaluate the safety and efficacy of DOACs for thromboprophylaxis and treatment of CA-VTE [
7,
12‐
17].
The PHACS trial, which assessed the use of dalteparin for the prevention of CA-VTE, provided inconclusive results because patients were noncompliant with treatment (once-daily subcutaneous injections) [
18]. On the other hand, the AVERT trial concluded that apixaban significantly reduced the incidence of VTE in cancer patients, as opposed to the CASSINI trial, which showed that rivaroxaban did not lower the incidence of VTE events during the study period [
12,
13]. Later on, The International Initiative on Thrombosis and Cancer (ITAC) and the American Society of Hematology (ASH) updated their guidelines to recommend the use of rivaroxaban or apixaban for thromboprophylaxis in ambulatory patients with cancer [
19], and the American Society of Clinical Oncology (ASCO) guidelines stated that apixaban, rivaroxaban, or LMWH may be offered to cancer patients for thromboprophylaxis [
20]. Regarding the management of CA-VTE, apixaban, edoxaban, and rivaroxaban were found to be noninferior to LMWH [
15‐
17]. Recently, the ASCO and ITAC guidelines have added rivaroxaban and edoxaban as the first-line agents next to LMWH for the treatment of CA-VTE [
20,
21]. Furthermore, the 2021 ASH guideline recommended apixaban, rivaroxaban or LMWH for initial treatment of CA-VTE [
19]. The latest American College of Chest Physicians (ACCP) guidelines also recommend DOACs over LMWH for the initiation and treatment of CA-VTE [
22].
The aim of this systematic review and meta-analysis was to assess the efficacy and safety of using DOACs for thromboprophylaxis and treatment of CA-VTE. We also summarized the recommendations from guidelines regarding the use of DOACs in patients with cancer for the management of CA-VTE.
Discussion
This systematic review and meta-analysis evaluated the efficacy and safety of using DOACs compared with LMWH or placebo for thromboprophylaxis and LMWH for CA-VTE treatment. By pooling the data from RCTs and post-hoc analyses of RCTs, more than 5300 patients from eight trials were analyzed. The use of DOACs was associated with about 40% lower risk of VTE recurrence and symptomatic VTE when used for CA-VTE treatment and thromboprophylaxis, respectively. However, when DOACs were used for thromboprophylaxis, the risk of major bleeding was twice higher when compared with placebo or LMWH. When thromboprophylaxis data were analyzed separately based on the comparator (i.e., placebo or LMWH), no difference was found between DOACs and LMWH in both efficacy and safety outcomes (Supplementary Fig.
S11), while in studies that use placebo as a comparator, DOACs showed a significant reduction in VTE events with no difference in the other efficacy and safety outcomes (Supplementary Fig.
S12). On the other hand, when DOACs were used for CA-VTE treatment, the risk of major bleeding was not significantly higher than that for LMWH.
The use of DOACs for thromboprophylaxis in patients with cancer reduced the risk of symptomatic VTE events (RR = 0.58; 95%CI 0.37, 0.91). This result is in contrast to a previous meta-analysis, in which patients with cancer who received DOACs for thromboprophylaxis showed a nonsignificant reduction in the risk of symptomatic VTE (RR = 0.57; 95%CI 0.29, 1.12) [
29]. This discrepancy is mainly due to the fact that the previous meta-analysis included only the AVERT and CASSINI trials. In the present study, the efficacy of DOACs was compared against placebo in patients who mostly had a Khorana score of two or higher [
12,
13,
29]. However, the AVERT trial utilized a modified Khorana risk score, which led to the inclusion of patients with more types of cancer [
13]. Even though the Khorana scores for patients included in the subgroup analysis of the MAGELLAN and APEX trials were not reported [
14,
24], the addition of data from their subgroup analyses provided significant evidence supporting the use of DOACs for thromboprophylaxis to reduce the risk of VTE in patients with cancer. The majority of those trials included high-risk patients with cancer, including GI, lung, pancreatic, and metastatic cancer. In contrast, a limited number of patients with brain malignancies were included only in the AVERT trial, which may weaken the evidence supporting the use of DOACs for thromboprophylaxis in this population.
Similar to previous systematic reviews and meta-analyses that assessed the use of DOACs for CA-VTE treatment [
30‐
32], our meta-analysis showed that DOACs were associated with a 38% risk reduction of VTE recurrence compared with LMWH. This rate was comparable to the rates reported in previous meta-analyses, which ranged from 32 to 55% [
30‐
32]. Mulder et al. found the risk of VTE recurrence in patients using DOACs to be nonsignificantly lower than that for the LMWH group (RR = 0.68; 95%CI 0.39, 1.17) [
30]. However, the most recent meta-analysis reported DOACs to be associated with a significantly lower risk of VTE recurrence compared with LMWH in patients with cancer [
32], even though it did not include data from the Caravaggio trial, one of the largest trials that demonstrated the efficacy of DOACs in preventing VTE recurrence in patients with CA-VTE. The overall risk reduction of VTE recurrence observed in our meta-analysis was mainly driven by the results from the ADAM VTE trial, in which the VTE recurrence was a secondary outcome that was barely observed [
7]. Moreover, the ADAM VTE was an open-label RCT that included a small sample of patients (
n = 300) compared with the Caravaggio trial (
n = 1155); however, the percentage of patients with upper GI cancer was similar in both trials [
7,
16].
In this meta-analysis, the use of DOACs for CA-VTE treatment or thromboprophylaxis was associated with a higher risk of bleeding than LMWH or placebo. The risk of major bleeding was nonsignificantly higher with DOACs when used for CA-VTE treatment (RR = 1.33, 95%CI 0.84, 2.11), while it was twice higher when DOACs were used for thromboprophylaxis (RR = 2.13, 95%CI 1.03, 4.42). Since the definition of major bleeding was almost consistent across the studies, the differences in the risk of major bleeding can be attributed to the heterogeneity of the cancer patient population. Compared with other trials, betrixaban in the APEX trial and rivaroxaban in the CASSINI trial were associated with a higher risk of major bleeding [
12,
14]. Nonetheless, the number of patients with GI and metastatic malignancies in the APEX and CASSINI trials was higher than in the AVERT trial, which might explain the increase in bleeding risk [
12‐
14]. Here, the risk of CRNMB did not differ between the DOACs and LMWH or placebo when used for thromboprophylaxis. However, the risk of CRNMB was significantly higher among patients receiving DOACs for CA-VTE treatment. This increase was mainly driven by the SELECT-D trial, as it reported a higher risk of CRNMB in patients receiving rivaroxaban than in those on dalteparin [
16]. It is also worth mentioning that the SELECT-D trial subsequently excluded patients with GI malignancies due to the increased risk of bleeding [
16].
Until recently, most of the previous guidelines recommended LMWH as the first-line therapy for CA-VTE treatment over VKAs and DOACs [
9,
33,
34]. This was due to insufficient evidence that supported the efficacy and safety of DOACs in patients with cancer, especially that DOAC approval trials only included a small proportion of patients with a history of active cancer [
34]. However, numerous RCTs and observational studies have been published recently that used DOACs for the treatment or prevention of CA-VTE [
7,
12,
13,
15‐
17,
25]. Also, several meta-analyses supported the lower risk of VTE recurrence and bleeding with DOACs as compared with LMWH and VKAs [
30‐
32]. This body of evidence from RCTs and meta-analyses prompted some of the large organizations to add DOACs as a viable option for CA-VTE treatment or thromboprophylaxis in their latest guidelines [
20,
21]. The ACCP, ASCO, ITAC, NCCN, and ASH have recently updated their guidelines to recommend the use of edoxaban, apixaban, or rivaroxaban as initial therapy for CA-VTE treatment in patients without gastric lesions or GI malignancies due to the increased risk of bleeding in these patients [
19‐
22,
28]. Clinical guidelines do not recommend the routine use of pharmacological thromboprophylaxis in all patients with cancer [
9,
10,
20,
21,
28]. This is mainly due to the relatively low risk of VTE in ambulatory patients with cancer and the increased risk of bleeding with the use of anticoagulants [
20,
21]. The use of validated VTE risk prediction tools such as the Khorana VTE risk assessment tool can help identify patients with cancer receiving chemotherapy who are at greater risk for VTE [
34]. The use of those tools prompted the development of newer guidelines supporting the use of DOACs in high-risk patients who are not at increased risk of bleeding. Thus, the current guidelines recommend the use of rivaroxaban and apixaban in patients at intermediate-to-high-risk of VTE, identified by cancer type (i.e., pancreatic) or a Khorana score of two or higher in the absence of active bleeding or not at high-risk of bleeding [
20,
21]. The convenience of oral administration of DOACs provides an easier, yet effective, therapeutic alternative for subcutaneous anticoagulation, which may increase patient adherence. However, caution should be warranted with the use of DOACs in patients with high-risk of bleeding and brain malignancies due to limited safety data.
To our knowledge, this is the first systematic review and meta-analysis that assessed the efficacy and safety of DOACs for both primary and secondary prevention of CA-VTE. Nevertheless, as any other meta-analysis, it is not free of limitations. There was a variation among the included trials in terms of patient population and follow-up duration, but this does not seem to have impacted the results as no significant heterogeneity was observed in the analysis. Although we included only high-quality RCTs to assess the efficacy and safety of DOACs for CA-VTE treatment, limited RCTs were available to assess the efficacy of DOACs for thromboprophylaxis. Therefore, data from subgroup or post-hoc analyses of large RCTs were included and the heterogeneity between these studies was low. The LMWH comparator included in our analysis was limited to dalteparin, as in all previous trials starting with the CLOT trial [
6]. Also, some included studies excluded patients with hematological, brain, and metastatic brain tumors, so caution is needed when generalizing our results to those populations. Finally, this analysis did not include all DOACs for both thromboprophylaxis and treatment of CA-VTE, since there were no studies that assessed the use of edoxaban for thromboprophylaxis, betrixaban for treatment, or dabigatran for both.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.