Translate this page into:
The effect of anticoagulants on preventing venous thromboembolism in cancer patients: A systematic review and meta-analysis
Correspondence to Dan Zhu; zdan1004@163.com
[To cite: Ma J, Ke X, Zhu D. The effect of anticoagulants on preventing venous thromboembolism in cancer patients: A systematic review and meta-analysis. Natl Med J India 2025;38:268–74. DOI: 10.25259/NMJI_878_2023]
Abstract
Background
We aimed to evaluate the efficacy and safety of anticoagulant therapy, especially direct oral anticoagulant (DOAC) therapy, for primary prevention of venous thromboembolism (VTE), including deep vein thrombosis and pulmonary embolism (PE), in non-bedridden cancer patients undergoing chemotherapy.
Methods
English and Chinese literature published from the inception of the databases up to December 2022 was extracted from five databases: PubMed, Embase, MEDLINE, Wanfang Medical and CNKI. Statistical analysis was done using Stata/SE 16.0 software.
Results
Eight randomized controlled trials were included in this meta-analysis. The results showed that DOACs were effective in preventing symptomatic PE in cancer patients (Log risk ratio [RR]=–0.89, 95% confidence interval [CI]: –1.73, –0.05). Moreover, compared to the control group, DOACs did not increase the risk of major bleeding (Log RR 0.55, 95% CI –0.20, 1.30).
Conclusion
The meta-analysis indicates that in cancer patients, DOACs have a significant effect on the prevention of PE, with good safety profile and no increased risk of major bleeding.
INTRODUCTION
Cancer-associated thrombosis (CAT) is a condition that physicians and oncologists are increasingly recognizing in association with the treatment of venous thromboembolism (VTE).1-5 It is currently estimated that the annual incidence of VTE in cancer patients is 0.5%, compared to 0.1% in the general population.6 Active cancer accounts for 20% of the total incidence of VTE.6 Notably, VTE is the second leading cause of death in cancer patients after disease progression, and cancer is the most common cause of death in patients with VTE.5,7,8
Studies have shown that direct oral anticoagulants (DOACs) are effective in the treatment of cancer-associated VTE,5,7,9–11 with efficacy comparable to that of traditional low-molecular-weight heparin.11 However, other studies have found that in preventive treatment for increased risk of VTE in cancer patients, the bleeding risk of DOACs is slightly increased.2,9 The ease and precision of DOAC administration make them an attractive choice for anticoagulation therapy.3 However, further systematic assessment of their efficacy and safety risks is still needed.
Therefore, this meta-analysis primarily evaluates the efficacy and safety of anticoagulants, especially DOAC, for primary prevention of VTE in ambulatory cancer patients receiving chemotherapy. In addition, we also compare the preventive effects of other anticoagulant measures (such as vitamin K antagonists and warfarin).
METHODS
Literature search strategy
This meta-analysis was done in accordance with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines. We searched for English and Chinese literature published from database inception to December 2022 in 5 databases: PubMed, Embase, MEDLINE, Wanfang Medical, and CNKI. Our search terms included anticoagulants, anti-neoplastic agents, antithrombins, neoplasms, randomized controlled trials (RCTs), VTE and warfarin, among others. In addition, we inspected the titles and abstracts of articles to determine whether they could be included in the study and reviewed the full texts of candidate articles for further confirmation. We also identified other relevant studies from the reference lists of key studies. We screened studies for inclusion and exclusion according to the following criteria:
Inclusion criteria
Study type: RCTs
Subjects: Cancer outpatients receiving chemotherapy
Intervention: Among non-bedridden cancer patients who were about to undergo chemotherapy, the experimental group received anticoagulant drugs (including DOAC, vitamin K antagonists and warfarin) for intervention, while the control group received a placebo for intervention
-
Outcomes: The main outcome was VTE (including deep vein thrombosis [DVT] and pulmonary embolism [PE]), and our study also included the occurrence of major bleeding as an outcome indicator.
Symptomatic VTE, PE and DVT: Objective verification of DVT by Doppler (compression) ultrasound or venography, and objective verification of PE by spiral computed tomography, ventilation/perfusion lung scan or pulmonary angiography
Major bleeding: Typically defined as excessive bleeding associated with a decrease in haemoglobin of >2 g/dl, or resulting in the transfusion of 2 or more units of packed red blood cells or whole blood.
Exclusion criteria
Duplicate articles or those without full text
Non-randomized controlled group experimental studies
Studies with missing or incorrect data that could not be completed or corrected
Studies lacking the outcome indicators required for this meta-analysis
Letters, case reports, reviews, practice guidelines, etc.
Patients with other underlying diseases were included in the study population
All animal experiments.
Data extraction
Two authors independently extracted the data from the included studies into a standardized table, and any disagreements were resolved by consensus or by involving a third author. We collected information on bias risk, participant characteristics, intervention group and control group characteristics and outcomes. We extracted results from intention-to-treat analyses whenever possible. If we could not calculate the effect size, we contacted the trial authors to request more data.
Quality assessment
Two independent researchers assessed the quality of the literature based on the Cochrane risk of bias assessment tool provided by Review Manager 5.4 software. In case of disagreement, a consensus was reached through discussion with a third party.
Statistical methods
Data analysis was done using Stata/SE 16.0 software. We present the results as summary risk ratios (RRs) for binary outcomes, with 95% confidence intervals (CIs) for each estimate. Heterogeneity between studies was assessed using the Q test; if I2<50% and p>0.1, indicating low heterogeneity, a fixed-effects model was used; otherwise, a random-effects model was applied. Forest plots were used to present the statistical results of the meta-analysis. Publication bias test was done using Egger’s test.
RESULTS
Literature screening results
A total of 317 studies were retrieved from the 5 databases. After eliminating duplicate articles, 305 original studies were extracted. By browsing the titles, keywords and abstracts, 184 studies were identified as potentially relevant to the research topic. We further searched for the full text of these studies, obtaining the full text of 181 studies. Based on the inclusion and exclusion criteria, 173 studies were excluded. Finally, 8 RCTs12-19 were included in this meta-analysis (Fig. 1).
- Flowchart of literature screening
Characteristics of included studies
The 8 included RCTs,12–19 from 1981 to 2019, included a total of 2926 non-bedridden cancer patients who received chemo-therapy. Five studies12,13,15,16,18 reported symptomatic VTE, including DVT and PE, in a total of 2277 patients. Among them, there were 1173 cases in the experimental group and 1104 cases in the control group. The remaining 3 articles14,17,19 reported if there was an increased risk of major bleeding. In the use of anticoagulants, 3 studies13,15,16 used DOAC, 4 studies used vitamin K inhibitors12, 1 study18 used warfarin and the main characteristics of the other included studies are shown in Table 1.
| Study (year) | Design | Median duration of follow-up | Mean age (years) | Interventions | Primary outcomes |
|---|---|---|---|---|---|
| Carrier et al., 201915 | Double-blind (participant, care, investigator, outcomes assessor), parallel-assig nment RCT | 183 days in each group | 61 in whole study population; 61.2 (SD 12.4) in apixaban group; 61.7 (SD 11.3) in placebo group | Apixaban 2.5 mg twice daily for 6 months | Symptomatic or incidental VTE (DVT, PE, or both) at 6 months |
| Levine et al., 200913 | Randomized, double-blind, phase II trial; intention-to-treat analyses not reported | Not reported, maximum 114–121 days | 57 (range 41–67) in apixaban 5 mg group, 60 (range 39–76) in apixaban 10 mg group, 64 (range 25–86) in apixaban 20 mg group, and 59 (range 20–82) in placebo group | Factor Xa inhibitor, apixaban 5 mg, 10 mg, or 20 mg once daily orally | Major bleeding or clinically relevant non-major bleeding |
| Khorana et al., 201916 | Double-blind, randomized, placebo-controlled, parallel-group, multicenter phase IIIB study | Not reported | 63 (range 23–88) years overall; 63 (range 23–87) years in rivaroxaban group; 62 (range 28–88) years in placebo group | Rivaroxaban 10 mg once daily up to day 180 | Primary efficacy endpoint: composite of objectively confirmed symptomatic or asymptomatic lower extremity proximal DVT, symptomatic upper extremity, symptomatic lower extremity distal DVT, symptomatic or incidental PE, and VTE-related death |
| Levine et al. 199412 | Multicenter, double-blind, randomized, placebo-controlled trial; intention-to-treat analysis | 199 days (SD 126) in warfarin and 188 days (SD 137) in placebo | 57.1 (SD 10.2) in warfarin group; 56.1 (SD 10.9) in placebo group | Warfarin 1 mg daily for 6 weeks and then adjusted to maintain the INR at 1.3–1.9 | VTE and arterial thrombosis; major and minor bleeding |
| Palumbo et al., 200918 | Randomized, open-label, multicenter study; modified intention-to-treat analysis, including participant receiving >1 study dose | 24.9 months | 61 (range 55–66) in aspirin group; 60 (range 54–66) in warfarin group; 62 (range 55–66) in heparin group | Intervention 1: aspirin 100 mg/day Intervention 2: low-dose warfarin (1.25 mg/day) Intervention 3: LMWH (enoxaparin 40 mg/day) | Composite measure of a first episode of objectively confirmed symptomatic DVT, PE, arterial thrombosis, acute myocardial infarction or stroke, or sudden, otherwise unexplained death during the first 6 months from random assignment |
| Chahinian et al., 198914 | Multicenter, 3-arm RCT, type of analyses not reported | 36 months | Not reported | Warfarin to maintain a prothrombin 1.5 to twice the control values | Overall survival, failure-free survival, and cancer response (complete response, partial response, and objective response rate) to therapy |
| Maurer et al., 199719 | Multicenter RCT; intention-to-treat analyses not reported | 69 months in those still alive | aged >60 years: 57.6% | Warfarin 10 mg/day for the first 3 days and then at a dose to maintain the prothrombin time between 1.4 and 1.6 times the local institutional control standards | Overall survival and cancer response to therapy |
| Zacharski et al., 198117 | Multicenter RCT, type of analyses not reported | Not reported, maximum follow-up was approximately 95 weeks | 58.9 (SD not reported) in warfarin group: 59.8 (SD not reported) in control group | Warfarin at doses to prolong the prothrombin time to approximately 2 times the control value | Primary efficacy outcomes: survival and cancer response to treatment |
RCT randomized controlled trialINR international normalized ratio SD standard deviation VTE venous thromboembolism LMWH low molecular weight heparin DVT deep vein thrombosis PE pulmonary embolism
Quality assessment of included literature
Based on the Cochrane risk bias assessment tool included in the Review Manager 5.4 software, we evaluated the quality of literature separately. Two studies did not specify the randomization methods. In addition, there was a study where the blinding method was incomplete in terms of implementation. In terms of selective reporting of results, 1 study did not report all pre-specified outcome indicators. The results of the quality assessment of the included literature are shown in Figs. 2 and 3.
- Quality assessment of included studies
- Quality assessment of included studies
Results of meta-analysis results
The effect of anticoagulants on preventing symptomatic VTE in cancer patients. Five studies have reported the effectiveness of anticoagulants in preventing VTE, including 2277 patients. The meta-analysis showed that compared with the control group, neither DOAC (Log RR –0.84, 95% CI –1.75, 0.06) nor vitamin K inhibitors (Log RR –1.90, 95% CI –3.98, 0.18) nor warfarin (Log RR 0.41, 95% CI –0.30, 1.11) had the effect of preventing symptomatic VTE in cancer patients, as shown in Fig. 4.
- Forest plot of the effectiveness of anticoagulants in preventing symptomatic venous thromboembolism in cancer patients
Effect of anticoagulants on preventing symptomatic DVT in cancer patients. Five studies have reported the effectiveness of anticoagulants in preventing DVT, including 2277 patients. The meta-analysis showed that compared with the control group, neither DOAC (Log RR–0.56, 95% CI–1.21, 0.09) nor vitamin K inhibitors (Log RR–2.52, 95% CI–5.39, 0.35) nor warfarin (Log RR=0.56, 95% CI–0.29, 1.41) had the effect of preventing symptomatic DVT in cancer patients, as shown in Fig. 5.
- Forest plot of the effect of anticoagulants on preventing symptomatic deep vein thrombosis in cancer patients
Effect of anticoagulants on preventing PE in cancer patients. Five studies have reported the effectiveness of anticoagulants in preventing PE, including 2277 patients. The meta-analysis showed that compared with the control group, DOAC had the effect of preventing symptomatic PE in cancer patients (Log RR –0.89, 95% CI–1.73, –0.05), as shown in Fig. 6. However, vitamin K antagonists (Log RR 0.05, 95% CI –2.72, 2.81) and warfarin (Log RR 0.00, 95% CI –1.37, 1.37) did not have the effect of preventing symptomatic PE in cancer patients, as shown in Fig. 6.
- Forest plot of the effect of anticoagulants on preventing symptomatic pulmonary embolism in cancer patients
Anticoagulant drugs increase the risk of major bleeding. Eight studies have reported the effect of anticoagulants on increasing major bleeding, including 2926 patients. The meta-analysis showed that compared with the control group, neither DOAC (Log RR 0.55, 95% CI –0.20, 1.30) nor the use of vitamin K inhibitors (Log RR 1.33, 95% CI 0.05, 2.62) or warfarin (Log RR –1.95, 95% CI –0.490, 1.01) had the effect of increasing primary bleeding in cancer patients, as shown in Fig. 7.
- Forest plot of anticoagulant drugs increasing the risk of major bleeding
Publication bias
Egger’s test on the meta-analysis of the effectiveness of DOAC in preventing VTE, DVT, PE and major bleeding showed that there was a certain publication bias (p=0.04) between studies on the prevention of DVT by DOAC, while there was no publication bias between studies on the prevention of other conditions (VTE: p=0.06; PE: 0.31; major bleeding: p=0.38). Egger’s test was done on the meta-analysis of vitamin K inhibitors increasing major bleeding, and there was no publication bias (p=0.70).
DISCUSSION
Cancer often leads to VTE, which manifests as DVT, PE or both.5,7,11,20 Compared to cancer patients without thrombosis, the mortality rate for cancer patients with VTE increases by at least 2-fold, which can be attributed to the development of fatal PE or a more aggressive disease in VTE patients.1,3,4,11,21,22
VTE in cancer patients may be difficult to identify due to specific symptoms, which can overlap and be confused with symptoms caused by the underlying cancer disease process or cancer treatment.20 VTE has a significant morbidity due to the need for hospitalization and increased risk of recurrent VTE or bleeding complications during anticoagulation therapy.5,22 The occurrence of symptomatic or incidental VTE may delay the implementation of cancer treatments such as chemotherapy, negatively impacting morbidity and potential mortality rates. Moreover, the occurrence of VTE adds further emotional stress on patients and their families, negatively affecting their quality of life.2 Finally, the costs associated with VTE management can be quite substantial due to drug and hospitalization-related expenses.23
The total burden of VTE in cancer patients is steadily increasing due to an aging population, prothrombotic anticancer treatments and growth in the cancer population.2 In addition, more cancer patients are being diagnosed with VTE incidentally during imaging studies required for routine monitoring during baseline staging, treatment response assessment or nonanticancer treatment periods.3 Providing widespread primary thromboprophylaxis for ambulatory cancer patients undergoing chemotherapy may help prevent VTE.4
Several measures have been proven effective in preventing venous thrombosis formation in cancer patients.2,4–7,9,11,20,21,23 Among them, DOACs offer a new choice for anticoagulation therapy with advantages such as predictable pharmacological effects and no need for routine coagulation monitoring.1,7,21
However, existing studies on the effectiveness of DOACs in preventing venous thrombosis in cancer patients are contradictory,1,2,4,7,8,10,11,22 and the efficacy of thromboprophylaxis needs to be balanced with the associated risk of bleeding complications.
Our study results show that DOACs are effective in preventing symptomatic PE in cancer patients (Log RR –0.89, 95% CI –1.73, –0.05). Moreover, compared to the control group, DOACs do not increase the risk of major bleeding (Log RR 0.55, 95%CI –0.20, 1.30). Among other anticoagulant preventive measures, vitamin K antagonists increase the risk of major bleeding (Log RR 1.49, 95% CI 0.64, 2.34).
Our study has some shortcomings. The first is that not all cancer patients need anticoagulant drug therapy. Patients undergoing anticoagulation therapy are themselves at high risk for thrombosis, and the results of this study are not applicable to all tumour patients, only to VTE high-risk tumour patients. The applicable window of the conclusions of this paper is narrow. Second, there may be some bias in the studies included in this paper. In drug research, researchers may subjectively prefer to obtain positive conclusions, which can affect the reliability of the results of this study. Finally, there are a few observational indicators included in this study, and more indicators need to be included to evaluate the therapeutic effect and safety comprehensively.
Therefore, DOACs have a certain effect on the prevention of PE in cancer patients with a high risk of VTE, and their safety is good without increasing the risk of major bleeding.
Conflicts of interest
None declared
References
- Anticoagulant therapy for cancer-associated thrombosis/: A cost-effectiveness analysis. Ann Intern Med. 2023;176:1-9.
- [CrossRef] [PubMed] [Google Scholar]
- Cancer-associated thrombosis in pediatric patients. Thromb Res. 2020;191(Suppl 1):S22-S25.
- [CrossRef] [PubMed] [Google Scholar]
- Antithrombin activity and association with risk of thrombosis and mortality in patients with cancer. Int J Mol Sci. 2022;23:15770.
- [CrossRef] [PubMed] [Google Scholar]
- Mechanisms and biomarkers of cancer-associated thrombosis. Transl Res J Lab Clin Med. 2020;225:33-53.
- [CrossRef] [PubMed] [Google Scholar]
- Platelets and cancer-associated thrombosis: Focusing on the platelet activation receptor CLEC-2 and podoplanin. Blood. 2019;134:1912-18.
- [CrossRef] [PubMed] [Google Scholar]
- Epidemiology of cancer-associated venous thrombosis. Blood. 2013;122:1712-23.
- [CrossRef] [PubMed] [Google Scholar]
- Cancer-associated venous thromboembolism. Nat Rev Dis Primers. 2022;8:11.
- [CrossRef] [PubMed] [Google Scholar]
- Treatment algorithm in cancer-associated thrombosis: Updated Canadian expert consensus. Curr Oncol Tor Ont. 2021;28:5434-51.
- [CrossRef] [PubMed] [Google Scholar]
- Cancer-associated thrombosis: A two-way street. Semin ThrombHemost. 2019;45:559-68.
- [CrossRef] [PubMed] [Google Scholar]
- Unanswered questions in cancer-associated thrombosis. Br J Haematol. 2022;198:812-25.
- [CrossRef] [PubMed] [Google Scholar]
- Cancer-associated thrombosis: An overview of mechanisms, risk factors, and treatment. Cancers (Basel). 2018;10:380.
- [CrossRef] [PubMed] [Google Scholar]
- Efficacy of double-blinding in a randomized controlled trial (RCT) of low dose warfarin to prevent thromboembolic disease in patients with metastatic breast cancer. Control Clin Trials. 1994;14:462.
- [CrossRef] [Google Scholar]
- A randomized phase II trial of a new anticoagulant, apixaban, in metastatic cancer. J Clin Oncol. 2009;27(Suppl 15):e20514.
- [CrossRef] [Google Scholar]
- A randomized trial of anticoagulation with warfarin and of alternating chemotherapy in extensive small-cell lung cancer by the cancer and leukemia group B. J Clin Oncol. 1989;7:993-1002.
- [CrossRef] [PubMed] [Google Scholar]
- Apixaban to prevent venous thromboembolism in patients with cancer. N Engl J Med. 2019;380:711-19.
- [CrossRef] [PubMed] [Google Scholar]
- Assessing full benefit of rivaroxaban prophylaxis in high-risk ambulatory patients with cancer: Thromboembolic events in the randomized CASSINI trial. TH Open. 2019;4:e107-e112.
- [CrossRef] [PubMed] [Google Scholar]
- Effect of warfarin anticoagulation on survival in carcinoma of the lung, colon, head and neck, and prostate. Final report of VA cooperative study #75. Cancer. 1981;53:2046-52.
- [CrossRef] [PubMed] [Google Scholar]
- A Phase III Study of Enoxaparin vs Aspirin vs Low-Dose Warfarin as Thromboprophylaxis for Newly Diagnosed Myeloma Patients Treated with Thalidomide Based-Regimens. Blood. 2009;114:492.
- [CrossRef] [Google Scholar]
- Randomized trial of chemotherapy and radiation therapy with or without warfarin for limited-stage small-cell lung cancer: A cancer and leukemia group B study. J Clin Oncol. 1997;15:3378-87.
- [CrossRef] [PubMed] [Google Scholar]
- Machine learning predicts cancer-associated venous thromboembolism using clinically available variables in gastric cancer patients. Heliyon. 2023;9:e12681.
- [CrossRef] [PubMed] [Google Scholar]
- Cancer-associated pathways and biomarkers of venous thrombosis. Blood. 2017;130:1499-506.
- [CrossRef] [PubMed] [Google Scholar]
- Survival implications of thrombus recurrence or bleeding in cancer patients receiving anticoagulation for venous thromboembolism treatment. Thromb Haemost. 2022;123:535-44.
- [CrossRef] [PubMed] [Google Scholar]
- Apixaban for the treatment of venous thromboembolism associated with cancer. N Engl J Med. 2020;382:1599-607.
- [CrossRef] [PubMed] [Google Scholar]