Real-world management and outcomes of venous thromboembolism in patients with cancer and limited life expectancy in the direct oral anticoagulant era: insights from the COMMAND VTE Registry-2 Authors: Takahiro Kuno, Norimichi Koitabashi, Yugo Yamashita, Takeshi Morimoto, Yoshiaki Ohyama, Noriaki Takama, Masaru Obokata, Ryuki Chatani, Kazuhisa Kaneda, Yuji Nishimoto, Nobutaka Ikeda, Yohei Kobayashi, Satoshi Ikeda, Kitae Kim, Moriaki Inoko, Toru Takase, Shuhei Tsuji, Maki Oi, Takuma Takada, Kazunori Otsui, Jiro Sakamoto, Yoshito Ogihara, Takeshi Inoue, Shunsuke Usami, Po-Min Chen, Kiyonori Togi, Seiichi Hiramori, Kosuke Doi, Hiroshi Mabuchi, Yoshiaki Tsuyuki, Koichiro Murata, Kensuke Takabayashi, Hisato Nakai, Daisuke Sueta, Wataru Shioyama, Tomohiro Dohke, Ryusuke Nishikawa, Hideki Ishii, Takeshi Kimura Published: 2025 ABSTRACT AbstractBackgroundOptimal management strategies for venous thromboembolism (VTE) in patients with cancer and limited life expectancy (LLE) could still be challenging in the current era of widespread use of direct oral anticoagulants (DOACs). Real-world treatment patterns and outcomes in patients with LLE were investigated using a nationwide registry in Japan.MethodsThe COMMAND VTE Registry-2 is a multicenter registry that enrolled 5,197 consecutive acute symptomatic VTE patients from 31 centers in Japan between January 2015 and August 2020. In this cohort, baseline characteristics, anticoagulation strategies, and long-term clinical outcomes were compared between cancer patients with and without LLE. The main outcomes were the one-year cumulative incidences of VTE recurrence, major bleeding, and all-cause death. Incidences of VTE recurrence and major bleeding were estimated using a competing risk analysis, with death as the competing event.ResultsThe study population consisted of 222 cancer patients with LLE (14.7%) and 1,285 patients without LLE (85.3%). Patients with LLE were older (mean age 69.9 vs. 67.8 years, P = 0.03) and received oral anticoagulants less frequently (80.6% vs. 92.1%, P < 0.001) than patients without LLE. Accounting for the competing risk of death, the one-year cumulative incidence of major bleeding was not significantly different between the two groups (11.0% vs. 10.5%, Gray’s test P = 0.91), whereas the incidence of VTE recurrence showed a trend to be lower in the LLE group (2.9% vs. 4.3%; P = 0.059). The cumulative one-year incidence of all-cause death was higher in the LLE group (94.1% vs. 35.4%, P < 0.001).ConclusionCancer patients with LLE were less likely to receive oral anticoagulants. Patients with LLE had a significantly higher one-year mortality rate than those without LLE. Accounting for the competing risk of death, there were no significant differences in the cumulative incidence of VTE recurrence or major bleeding between the two groups.Supplementary InformationThe online version contains supplementary material available at 10.1186/s40959-025-00431-5.Keywords: Venous thromboembolism, Cancer, Limited life expectancy, Direct oral anticoagulants, Competing risk analysis, Palliative care, Real-world data, Registry Introduction Venous thromboembolism (VTE), including deep vein thrombosis (DVT) and pulmonary embolism (PE), is a common and potentially fatal complication in patients with cancer [1]. Patients with active cancer were reported to have an approximately sevenfold higher incidence of VTE than those without malignancy [2], and thrombosis remains the second leading cause of death in patients receiving chemotherapy [3]. VTE is associated not only with cancer-related hypercoagulability, but also with treatment-related factors, including newer anticancer therapies such as molecular-targeted agents, immune checkpoint inhibitors, and chimeric antigen receptor-T cell therapy, which may further increase thrombotic risk in the current era [1, 4]. In addition, patients with cancer and limited life expectancy (LLE) are particularly susceptible to VTE due to cancer-related hypercoagulability and decreased mobility [1]. Although anticoagulation is essential for patients with cancer and LLE, it remains particularly challenging due to their high bleeding risk. Despite the clinical importance, existing guidelines offer no obvious recommendations due to a lack of supporting evidence for this population [5–7]. Recent randomized, controlled trials have demonstrated the efficacy of direct oral anticoagulants (DOACs) for cancer-associated VTE. However, patients with cancer and LLE, who have a high-bleeding risk, as well as a poor prognosis, were frequently excluded [8]. In fact, data on VTE patients with cancer and LLE in the current DOAC era are still limited. These data could be clinically relevant for understanding the current issues and unmet needs. Therefore, this study aimed to investigate real-world management and outcomes of VTE in patients with cancer and LLE. Clinical characteristics, treatment strategies, and outcomes including recurrent VTE, major bleeding, and all-cause death were compared between cancer patients with and without LLE. Furthermore, whether anticoagulation therapy was associated with different outcomes in cancer patients with LLE than in those without LLE was evaluated. These findings may help support individualized decision-making in palliative settings and inform future guidance on anticoagulation strategies in patients with cancer and LLE. Methods Study population The COMMAND VTE Registry-2 was a physician-initiated, multicenter, retrospective, cohort study that enrolled consecutive patients with acute symptomatic VTE objectively confirmed by imaging examination or autopsy at 31 centers in Japan between January 2015 and August 2020, after the introduction of DOACs for VTE in Japan [9]. The design of the registry has been previously reported in detail [9]. The relevant review boards or ethics committees of all 31 participating centers (Supplementary Appendix 1) approved the research protocol. Written, informed consent from each patient was waived because clinical information obtained from routine clinical practice was used, and none of the patients refused to participate in the study when contacted for follow-up. This method was concordant with the guidelines for epidemiological studies issued by the Ministry of Health, Labour and Welfare in Japan. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guideline. Data collection and definitions Baseline data were collected from hospital charts or databases according to prespecified definitions, using an electronic case report form in a web-based database system. The study investigators at each institution were responsible for data entry, and data were automatically checked for missing or contradictory input and values out of the expected ranges. Additional data monitoring was performed at the general office of the registry. Prespecified variables including demographics, comorbidities and prior bleeding, transient risk factors, VTE presentation and imaging-based confirmation, cancer characteristics (site, metastasis, and treatments), anticoagulation details (drug, dose, discontinuation), key laboratory tests (hemoglobin, platelet count, renal function, D-dimer), and clinical outcomes (recurrent VTE, International Society of Thrombosis and Hemostasis (ISTH) major bleeding, and all-cause death) were collected. Detailed definitions are provided in Table 1. Table 1Baseline characteristicsWith LLE(N = 222)Without LLE(N = 1285)P-valuePatients’ characteristic Age, ya,b,c69.9 ± 13.167.8 ± 12.70.03 Femalea,b,c135 (60.8%)709 (55.2%)0.12 Body weight, kg51.9 ± 10.657.5 ± 12.7 < 0.001 Body mass index, kg/m220.9 ± 3.822.7 ± 4.2 < 0.001 BMI ≥ 30 kg/m2 a,c6 (2.7%)68 (5.3%)0.09Comorbidity Hypertensionb,c70 (31.5%)505 (39.3%)0.03 Diabetes mellitusb,c34 (15.3%)203 (15.8%)0.86 Dyslipidemia29 (13.1%)272 (21.2%)0.005 Chronic kidney diseaseb,c47 (21.2%)229 (17.8%)0.23 Dialysis2 (0.9%)1 (0.1%)0.06 Heart failure3 (1.4%)61 (4.7%)0.02 Myocardial infarction4 (1.8%)23 (1.8%)1.00 Atrial fibrillation3 (1.4%)61 (4.7%)0.02 Chronic lung disease14 (6.3%)107 (8.3%)0.31 History of strokeb,c12 (5.4%)77 (6.0%)0.73 Varicose veinsa,c3 (1.4%)40 (3.0%)0.15 Liver cirrhosisb,c1 (0.5%)17 (1.3%)0.50 History of VTEa,c3 (1.4%)40 (3.1%)0.15 History of major bleedingb,c20 (9.0%)62 (4.8%)0.01 Autoimmune diseasea,b,c11 (5.0%)63 (4.9%)0.97 Transient risk factora,b,c69 (31.1%)439 (34.2%)0.37Cancer status On chemotherapy45 (20.3%)720 (56.0%) < 0.001 Scheduled to be operated2 (0.9%)165 (12.8%) < 0.001 Distant metastasis86 (38.7%)331 (25.8%) < 0.001 Receiving radiotherapy5 (2.3%)53 (4.1%)0.18Presentation PE with or without DVTa,b,c109 (49.1%)662 (51.5%)0.51 Hypoxemia54 (24.3%)257 (20.0%)0.04 Shock9 (4.1%)45 (3.5%)0.58 Cardiac arrest/collapse1 (0.5%)16 (1.2%)0.32 DVT only113 (50.9%)623 (48.5%)0.51 Proximal DVTa,b,c144 (64.9%)704 (54.8%)0.005Laboratory tests at diagnosis Anemiab,c187 (84.2%)976 (76.0%)0.007 Thrombocytopeniab,c24 (10.8%)104 (8.1%)0.18 eGFR (mL/min/1.73 m2)73.2 ± 42.172.7 ± 41.10.87 D-dimer (μg/mL)16.9 (9.3, 29.8)9.8 (4.6, 20.3) < 0.001 Hereditary thrombophiliaa,c3 (1.4%)23 (1.8%)1.00Categorical variables are presented as number and percentage values, and continuous variables are presented as mean ± standard deviation or median and interquartile range values based on their distributionsChronic lung disease was defined as persistent lung disorders such as asthma, chronic obstructive pulmonary disease, and restrictive lung diseases. History of major bleeding was diagnosed if the patient had a history of International Society of Thrombosis and Hemostasis (ISTH) major bleeding. Proximal DVT in the lower extremities was defined as venous thrombosis located in popliteal, femoral, or iliac veins. Anemia was defined as hemoglobin level < 13 g/dL for men and < 12 g/dL for women. Thrombocytopenia was defined as platelet count < 100 × 109/L. Hereditary thrombophilia included protein C deficiency, protein S deficiency, and antithrombin III deficiency. The detailed definitions of the patients’ characteristics are provided in Supplementary Appendix 2LLE limited life expectancy, BMI body mass index, VTE venous thromboembolism, PE pulmonary embolism, DVT deep vein thrombosis, eGFR estimated glomerular filtration rateaRisk-adjusting variables for the multivariable Cox regression model to estimate the risk for recurrent VTEbRisk-adjusting variables for the multivariable Cox regression model to estimate the risk for major bleedingcRisk-adjusting variables for the multivariable Cox regression model to estimate the risk for all-cause death Baseline characteristics Categorical variables are presented as number and percentage values, and continuous variables are presented as mean ± standard deviation or median and interquartile range values based on their distributions Chronic lung disease was defined as persistent lung disorders such as asthma, chronic obstructive pulmonary disease, and restrictive lung diseases. History of major bleeding was diagnosed if the patient had a history of International Society of Thrombosis and Hemostasis (ISTH) major bleeding. Proximal DVT in the lower extremities was defined as venous thrombosis located in popliteal, femoral, or iliac veins. Anemia was defined as hemoglobin level < 13 g/dL for men and < 12 g/dL for women. Thrombocytopenia was defined as platelet count < 100 × 109/L. Hereditary thrombophilia included protein C deficiency, protein S deficiency, and antithrombin III deficiency. The detailed definitions of the patients’ characteristics are provided in Supplementary Appendix 2 LLE limited life expectancy, BMI body mass index, VTE venous thromboembolism, PE pulmonary embolism, DVT deep vein thrombosis, eGFR estimated glomerular filtration rate aRisk-adjusting variables for the multivariable Cox regression model to estimate the risk for recurrent VTE bRisk-adjusting variables for the multivariable Cox regression model to estimate the risk for major bleeding cRisk-adjusting variables for the multivariable Cox regression model to estimate the risk for all-cause death Transient risk factors for VTE included recent surgery (within 2 months), immobilization (bed rest ≥ 4 days), long-distance travel (≥ 6 h), central venous catheterization, pregnancy or puerperium, severe infection, and estrogen use. Patients with active cancer were defined as those being treated for cancer, such as with chemotherapy or radiotherapy, those scheduled to undergo surgery for cancer, or those with metastasis to other organs. Cancer patients with LLE were defined as a treating-physician–estimated life expectancy of ≤ 6 months at the time of the index VTE event. We selected the 6-month threshold because it aligns with the Medicare Hospice Benefit eligibility standard, which requires physician certification of a ≤ 6-month prognosis, and with international oncology guidelines recommending at least 6 months of anticoagulation for cancer-associated VTE [7, 10, 11]. In our participating institutions, prognostic estimation is performed clinically and globally, rather than through a single prognostic score. Physicians generally considered overall functional decline, tolerance to anticancer therapy, nutritional status, and disease progression when assessing life expectancy. While we did not systematically collect performance or prognostic scales such as PPS or PPI, the clinical indicators commonly informing physician judgment (e.g., ECOG ≥ 3, transition to best supportive care, reduced oral intake, involuntary weight loss, or hypoalbuminemia) are consistent with criteria used in palliative and hospice care frameworks [10, 11]. We have clarified this point to improve transparency in our operational definition. Initial anticoagulation therapy was defined as parenteral anticoagulation therapy in the acute phase (heparin or fondaparinux) for ≤ 10 days after the diagnosis, whereas anticoagulation therapy beyond the acute phase was defined as anticoagulation therapy (vitamin K antagonists, DOACs, or heparin) for > 10 days after the diagnosis [12]. Detailed definitions of other patients’ characteristics are given in Supplementary Appendix 2. Clinical follow-up and outcomes Collection of follow-up information was mainly conducted through review of hospital charts, and additional follow-up information was collected through contact with patients, relatives, and/or referring physicians by phone and/or mail, with questions regarding vital status, clinical events, invasive procedure, and status of anticoagulation therapy. In this cohort study, final data collection for follow-up events was performed between September 2021 and April 2022. The primary outcome measure in this study was VTE recurrence. VTE recurrence was defined as PE and/or DVT with symptoms accompanied by confirmation of new thrombus or exacerbation of the thrombus by objective imaging examinations and/or autopsy [13]. The secondary outcomes were major bleeding and all-cause death. Major bleeding was defined as ISTH major bleeding, which consisted of reduction in the hemoglobin level by at least 2 g/dL, transfusion of at least 2 units of blood, or symptomatic bleeding in a critical area or organ [14]. The independent clinical event committee (Supplementary Appendix 3), unaware of patients’ characteristics, reviewed all of the detailed clinical courses and adjudicated the clinical events. If there was inconsistency, final adjudication for clinical events was made on the basis of the full consensus of the independent clinical event committee. Final classifications for causes of deaths were also based on the full consensus of the independent clinical event committee. The adjudication procedure was conducted between June 2022 and August 2022. Discontinuation of anticoagulation was defined as withdrawal of anticoagulation therapy for more than 14 days for any reason, such as physician’s judgment in the absence of adverse events, bleeding events, drug side effects, and patient’s non-adherence. The definitions of other clinical events are provided in Supplementary Appendix 4. Follow-up information was obtained during hospitalization and at subsequent outpatient visits or rehospitalizations according to each participating center’s routine practice; therefore, the frequency of follow-up was not standardized across the registry. Statistical analysis Categorical variables are presented as numbers and percentages. Continuous variables are presented as mean and standard deviation or median and interquartile range values based on their distributions. Categorical variables were compared with the chi-squared test when appropriate; otherwise, Fisher’s exact test was used. Continuous variables were compared using Student’s t-test or Wilcoxon’s rank-sum test based on their distributions. For the analyses of discontinuation of anticoagulation therapy, VTE recurrence, and major bleeding, the Aalen-Johansen method was used to estimate cumulative incidence, and differences were assessed with Gray’s test. This approach was chosen because death was a competing risk due to the higher mortality of cancer patients with LLE. The Kaplan–Meier method was used to estimate the cumulative incidence, and differences were assessed with the log-rank test for the analysis of all-cause death. To adjust clinically relevant confounders, Fine and Gray’s methods were used to estimate the subdistribution hazard ratio (SHR) and 95% confidence interval (CI) for the risk of clinical outcomes in patients with LLE compared with those without LLE. Based on the previous reports [15–17] and consideration of clinical relevance, covariate sets were prespecified. For the VTE recurrence model, known prothrombotic risk factors, such as obesity, varicose vein, a history of VTE, and hereditary thrombophilia, and transient provoking factors were included. For the major bleeding model, predictors of bleeding, such as hypertension, diabetes mellitus, chronic kidney disease, history of stroke, liver cirrhosis, history of major bleeding, anemia, and thrombocytopenia were included. The model for all-cause death included a comprehensive set of variables associated with both thrombotic and bleeding risks. Thus, 9 risk-adjusting variables for VTE recurrence, 13 risk-adjusting variables for major bleeding, and 17 risk-adjusting variables for all-cause death were selected (Table 1). Several sensitivity analyses were performed. First, to account for treatment changes over time, a Fine and Gray’s model with anticoagulation status (on or off treatment) as a time-varying covariate was used. Second, to further adjust for cancer-related confounders, another analysis that included variables for cancer status (on chemotherapy, scheduled for operation, distant metastasis, and receiving radiotherapy) was performed. Finally, a separate analysis was conducted with additional adjustment for cancer-site risk categories based on the Khorana risk model: very high-risk (stomach, pancreas), high-risk (lung, lymphoma, gynecologic, bladder, testicular, and kidney), and low-risk (all other sites)[18]. Stata/BE 18.0 (StataCorp) and R software version 4.4.0 (R Foundation for Statistical Computing, Vienna, Austria) were used for all analyses. All reported P-values are two-sided, and P < 0.05 was considered significant. Results Patients’ characteristics Of the 1,507 patients with active cancer enrolled in the registry, 222 patients (14.7%) were defined as having LLE, and 1,285 patients (85.3%) were defined as patients without LLE (Fig. 1). The baseline clinical characteristics were different in several aspects between the two groups. Compared with patients without LLE, those with LLE were slightly older and had significantly lower body weight and body mass index. Patients with LLE were more likely to have a history of major bleeding (9.0% vs. 4.8%, P = 0.01), whereas patients with LLE were less likely to have hypertension, dyslipidemia, and atrial fibrillation (Table 1). Patients with LLE were less likely to be on chemotherapy (20.3% vs. 56.0%, P < 0.001) and more likely to have distant metastasis (38.7% vs. 25.8%, P < 0.001) than patients without LLE. Patients with LLE were more likely to have anemia (84.2% vs. 76.0%, P = 0.007). There was no significant difference in the prevalence of PE between patients with LLE and patients without LLE (49.1% vs. 51.5%, P = 0.51), although patients with LLE had a higher proportion of hypoxemia (24.3% vs. 20.0%, P = 0.04) (Table 1). Laboratory values showed higher median D-dimer levels in patients with LLE (16.9 µg/mL vs. 9.8 µg/mL, P < 0.001). A subgroup analysis comparing patients’ characteristics based on anticoagulant use in cancer patients with LLE is shown in Supplementary Table 1. Study flow chart. VTE includes pulmonary embolism and/or deep vein thrombosis. VTE, venous thromboembolism; LLE, limited life expectancy As for cancer sites, lung (14.6%), blood (11.6%), and colon (10.2%) were the most frequent in patients with active cancer. The details are shown in Supplementary Table 2. Treatment strategies There was no significant difference in the use of initial parenteral anticoagulation therapy between the two groups of patients with and without LLE (50.9% vs. 47.4%, P = 0.33) (Table 2). Similarly, there were no significant differences in the proportions of heparin use (49.5% vs. 46.2%, P = 0.36), inferior vena cava filter use (9.5% vs. 9.0%, P = 0.84), ventilator support use (0.5% vs. 1.4%, P = 0.24), and cardiopulmonary support use (0.0% vs. 0.5%, P = 0.31) between the two groups. However, the proportion of patients who did not receive any oral anticoagulant was significantly higher in patients with LLE than in patients without LLE (19.4% vs. 7.9%, P < 0.001). Overall dosing patterns of DOACs did not differ substantially between patients with and without LLE, although there were significant differences in the proportions of patients receiving maintenance-dose rivaroxaban (P < 0.001) and edoxaban (P < 0.001) (Supplementary Table 3). The cumulative one-year incidence of discontinuation of anticoagulation therapy was not different between the two groups in the competing risk analysis (37.0% vs. 38.3%, Gray’s test P = 0.42) (Fig. 2A). A subgroup analysis comparing treatment strategies by anticoagulant use in cancer patients with LLE is shown in Supplementary Table 4. Within the LLE subgroup, the comparison was made with and without oral anticoagulants, but there were no significant differences between the two groups in VTE treatment other than oral anticoagulants (heparin, IVC filters, etc.) and concomitant medication use. Table 2Treatment strategiesWith LLE(N = 222)Without LLE(N = 1285)P-valueInitial parenteral therapy113 (50.9%)609 (47.4%)0.33Heparin110 (49.5%)594 (46.2%)0.36Single injection at diagnosis6 (2.7%)28 (2.2%)0.74Continuous injection104 (46.8%)566 (44.0%)0.74Thrombolysis3 (1.4%)40 (3.1%)0.15Inferior vena cava filter use21 (9.5%)116 (9.0%)0.84Ventilator support1 (0.5%)18 (1.4%)0.24Cardiopulmonary support0 (0.0%)6 (0.5%)0.31Oral anticoagulation < 0.001None43 (19.4%)101 (7.9%)DOACs150 (67.6%)1049 (81.6%)VKA29 (13.1%)135 (10.5%)Type of DOAC < 0.001Dabigatran0 (0.0%)2 (0.2%)Rivaroxaban21 (9.5%)276 (21.5%)Apixaban25 (11.3%)232 (18.1%)Edoxaban104 (46.8%)539 (41.9%)Concomitant medicationsCorticosteroids42 (18.9%)168 (13.1%)0.02NSAIDs35 (15.8%)155 (12.1%)0.12PPIs/H2-blockers115 (51.8%)593 (46.1%)0.12Statins14 (6.3%)176 (13.7%)0.002Antiplatelet agents9 (4.1%)76 (5.9%)0.27Categorical variables are presented as number and percentage values. Categorical variables were compared using the χ2 testInitial parenteral anticoagulation therapy included heparin (single or continuous injection) and fondaparinux within 10 days after the diagnosisOral anticoagulation therapy included a vitamin K antagonist (warfarin) and DOACs. Antiplatelet drugs included aspirin, ticlopidine, clopidogrel, prasugrel, ticagrelor, and cilostazolLLE limited life expectancy, DOAC direct oral anticoagulant, VKA vitamin K antagonist, NSAID nonsteroidal anti-inflammatory drug, PPIs proton pump inhibitors, H2-blockers, histamine type 2 receptor blockers Treatment strategies Categorical variables are presented as number and percentage values. Categorical variables were compared using the χ2 test Initial parenteral anticoagulation therapy included heparin (single or continuous injection) and fondaparinux within 10 days after the diagnosis Oral anticoagulation therapy included a vitamin K antagonist (warfarin) and DOACs. Antiplatelet drugs included aspirin, ticlopidine, clopidogrel, prasugrel, ticagrelor, and cilostazol LLE limited life expectancy, DOAC direct oral anticoagulant, VKA vitamin K antagonist, NSAID nonsteroidal anti-inflammatory drug, PPIs proton pump inhibitors, H2-blockers, histamine type 2 receptor blockers Clinical outcome curves. Cumulative incidence curves for discontinuation of anticoagulation therapy (A), VTE recurrence (B), and major bleeding (C), accounting for death as a competing risk. Kaplan–Meier curve for all-cause death (D). VTE includes pulmonary embolism and/or deep vein thrombosis. VTE, venous thromboembolism; LLE, limited life expectancy Clinical outcomes The cumulative one-year incidence of VTE recurrence was not significantly different between the two groups of patients with and without LLE (2.9% vs. 4.3%, Gray’s test P = 0.059) (Fig. 2B). Even after adjusting for potential confounders, the risk of patients with LLE relative to those without LLE was not significant for VTE recurrence (SHR: 0.47, 95% CI: 0.20–1.10, P = 0.08) (Table 3). In a sensitivity analysis using a time-varying covariate for anticoagulation status, the risks for patients with LLE were lower than for patients without LLE (SHR: 0.39, 95% CI: 0.16–0.92, P = 0.03) (Supplementary Table 5), whereas these were not significant in the sensitivity analysis including cancer status variables as adjusted covariates (SHR: 0.47, 95% CI: 0.19–1.17, P = 0.10) (Supplementary Table 6). The results were similar in a sensitivity analysis with adjustment for cancer-site risk categories according to the Khorana risk model (SHR: 0.43, 95% CI: 0.18–1.00, P = 0.05) (Supplementary Table 7). Table 3Crude and adjusted clinical outcomes between patients with LLE and without LLECumulative one-year incidence(Number of patients with event)With LLEN = 222Without LLEN = 1,285Crude SHR(95% CI)P-valueAdjusted SHR(95% CI)P-valueVTE recurrence2.9% (6)4.3% (53)0.48 (0.21–1.12)0.090.47 (0.20–1.10)0.08Major bleeding11.0% (24)10.5% (118)1.07 (0.70–1.64)0.761.10 (0.72–1.70)0.66All-cause death94.1% (192)35.4% (433)N/AN/AN/AN/AFor the analyses of VTE recurrence and major bleeding, the Aalen-Johansen method was used to estimate cumulative incidence, and differences were assessed with Gray’s test. For the analyses of all-cause death, the Kaplan–Meier method was used to estimate cumulative incidence, and differences were assessed with the log-rank test. Crude and adjusted HRs and 95%CIs of the terminal group relative to the non-terminal group were estimated using Fine and Gray’s subdistribution hazard modelsAdjusted covariates were as follows:VTE recurrence: age, sex, BMI, varicose veins, history of VTE, autoimmune disease, transient risk factor, thrombus site (PE, proximal DVT, or distal DVT), and hereditary thrombophiliaMajor bleeding: age, sex, hypertension, diabetes mellitus, history of stroke, liver cirrhosis, history of major bleeding, autoimmune disease, transient risk factor, thrombus site (PE, proximal DVT, or distal DVT), anemia, and thrombocytopeniaAll-cause death: age, sex, BMI, hypertension, diabetes mellitus, chronic kidney disease, history of stroke, varicose veins, liver cirrhosis, history of VTE, history of major bleeding, autoimmune disease, transient risk factor, thrombus site (PE, proximal DVT, or distal DVT), anemia, thrombocytopenia, and hereditary thrombophiliaLLE limited life expectancy, SHR subdistribution hazard ratio, CI confidence interval, VTE venous thromboembolism, BMI body mass index, PE pulmonary embolism, DVT deep vein thrombosis, N/A not applicable Crude and adjusted clinical outcomes between patients with LLE and without LLE For the analyses of VTE recurrence and major bleeding, the Aalen-Johansen method was used to estimate cumulative incidence, and differences were assessed with Gray’s test. For the analyses of all-cause death, the Kaplan–Meier method was used to estimate cumulative incidence, and differences were assessed with the log-rank test. Crude and adjusted HRs and 95%CIs of the terminal group relative to the non-terminal group were estimated using Fine and Gray’s subdistribution hazard models Adjusted covariates were as follows: VTE recurrence: age, sex, BMI, varicose veins, history of VTE, autoimmune disease, transient risk factor, thrombus site (PE, proximal DVT, or distal DVT), and hereditary thrombophilia Major bleeding: age, sex, hypertension, diabetes mellitus, history of stroke, liver cirrhosis, history of major bleeding, autoimmune disease, transient risk factor, thrombus site (PE, proximal DVT, or distal DVT), anemia, and thrombocytopenia All-cause death: age, sex, BMI, hypertension, diabetes mellitus, chronic kidney disease, history of stroke, varicose veins, liver cirrhosis, history of VTE, history of major bleeding, autoimmune disease, transient risk factor, thrombus site (PE, proximal DVT, or distal DVT), anemia, thrombocytopenia, and hereditary thrombophilia LLE limited life expectancy, SHR subdistribution hazard ratio, CI confidence interval, VTE venous thromboembolism, BMI body mass index, PE pulmonary embolism, DVT deep vein thrombosis, N/A not applicable The cumulative one-year incidence of major bleeding was not significantly different between the two groups in the competing risk analysis (11.0% vs. 10.5%, Gray’s test P = 0.91) (Fig. 2C). Even after adjusting for potential confounders, the risk of patients with LLE relative to those with non-LLE remained non-significant for major bleeding (SHR: 1.10, 95% CI: 0.72–1.10, P = 0.66) (Table 3). In a sensitivity analysis using a time-varying covariate for anticoagulation status, the risks for patients with LLE were not significantly different from those of patients without LLE (SHR: 1.03, 95% CI: 0.67–1.60, P = 0.87) (Supplementary Table 5). In a sensitivity analysis using a competing risk analysis including cancer status variables as adjusted covariates, there was no significant difference (SHR: 0.91, 95% CI: 0.57–1.46, P = 0.70) (Supplementary Table 6). Furthermore, in a sensitivity analysis with additional adjustment for cancer-site risk categories according to the Khorana risk model, the association between LLE and major bleeding remained not significant (SHR: 0.89, 95% CI: 0.57–1.40, P = 0.62) (Supplementary Table 7). The cumulative one-year incidence of all-cause death was significantly higher in patients with LLE than in those without LLE (94.1% vs. 35.4%, log-rank P < 0.001) (Fig. 2D). A subgroup analysis comparing clinical outcomes based on anticoagulant use in cancer patients with LLE is shown in Supplementary Table 8. The prognosis in the LLE group was very poor with or without oral anticoagulants, but mortality was even higher in the group that did not receive oral anticoagulants. The distribution of recurrent VTE sites between patients with and without LLE is shown in Supplementary Table 9. Patients without LLE had a higher incidence of both PE and DVT compared to patients with LLE, and no clear differences were observed in recurrence patterns by VTE subtype. Discussion The main findings of the present study were as follows. (1) In this large real-world registry, cancer patients with LLE had a significantly higher one-year all-cause mortality. (2) A significantly greater proportion of patients with LLE did not receive oral anticoagulation. (3) There were no significant differences in the cumulative incidence of VTE recurrence or major bleeding between cancer patients with LLE and without LLE after accounting for the competing risk of death, whereas a sensitivity analysis that accounted for anticoagulation status over time suggested a lower risk in the LLE group. Current guidelines recommend extended anticoagulation therapy for cancer-associated VTE. However, these recommendations are based on trials that often excluded patients with poor prognoses or LLE [7, 19]. The present study addressed this evidence gap by providing real-world data on this vulnerable and underrepresented population. A recent European, multicenter study reported significant institutional variations in the discontinuation of anticoagulation for cancer patients with LLE [20]. Hemato-oncologists and cardiologists tended to discontinue anticoagulation [20]. The present findings align with these observations, since the present cohort was derived from cardiology-focused institutions and reflect a similar trend of risk-based treatment de-escalation in cancer patients with LLE. In the present cohort, patients with LLE were significantly older, had a lower BMI, and a higher prevalence of a history of major bleeding. These factors are often interpreted as a heightened bleeding risk [21]. Therefore, it is plausible that clinicians weighed the benefits of VTE prevention against these risks and opted for more conservative treatment strategies, such as withholding oral anticoagulants or using reduced doses [7, 21, 22]. Indeed, the present data support this interpretation, since the reduced 30 mg/day dose of edoxaban (both on- and off-label) was used more frequently in patients with LLE than in those without. This aligns with the principles of individualized care in palliative settings, where the treatment burden is a critical consideration [20, 21, 23, 24]. The lack of significant differences in VTE recurrence and major bleeding may be explained by the overwhelming impact of all-cause death as a competing event. A large proportion of patients in the LLE group likely died before they might have had either a recurrent VTE or a major bleeding event during a year. This reality underscores a critical clinical insight: in patients with terminal cancer, outcomes are predominantly driven by the underlying malignancy, rather than thrombotic or hemorrhagic events. The findings for VTE recurrence warrant a more nuanced interpretation. In the primary competing risk analysis, the difference in cumulative incidence was not significant, although there was a trend toward a lower risk in the LLE group. This suggests that the present study may have been underpowered to detect a clinically relevant difference. Furthermore, a sensitivity analysis incorporating anticoagulation status as a time-varying covariate showed a significantly lower risk of VTE recurrence in the LLE group, whereas this difference was not significant after adjustment for cancer status. This complexity highlights the difficulty in concluding that LLE itself has a protective effect. Instead, this observational finding likely results from a combination of overwhelming competing mortality and surveillance bias. We acknowledge that reduced clinical encounters in patients with LLE could lead to under-ascertainment of some VTE recurrences. However, in our dataset, pulmonary embolism represented a considerable share of recurrent events among patients without LLE, and isolated DVT was not over-represented. Although differences in imaging intensity near the end of life may play some role, this pattern suggests that surveillance bias alone is unlikely to fully explain the observed trend. Furthermore, results remained consistent in a sensitivity analysis adjusting for cancer-site risk categories, supporting our conclusions despite differences in cancer type distribution. These findings carry important implications for the management of VTE in patients with LLE. The absolute benefit of anticoagulation for VTE prevention may be marginal in this population, emphasizing the importance of individualized treatment decisions that weigh the potential benefits against the burdens of therapy. The present data provide objective evidence to support a shift in the primary goal of care from long-term VTE prevention to the palliation of symptoms and maximization of quality of life in cancer patients with LLE. In this context, the decision to withhold or discontinue anticoagulation should not be viewed as a failure of care, but rather as a rational, patient-centered therapeutic choice based on a shared decision-making process. This study had strengths and limitations. The strengths of this study included its large sample size derived from a multicenter registry, providing real-world insights into VTE management in the DOAC era [9]. Several limitations of this study should be acknowledged. First, as an observational, registry-based analysis, the possibility of selection bias and residual confounding such as performance status or specific chemotherapy regimens cannot be completely ruled out. Second, the definition of LLE was based on treating-physician judgment and was not determined using a standardized prognostic tool. Although such clinical prognostication reflects real-world practice and was supported by the high one-year mortality observed in the LLE group, it may vary among physicians. While indicators such as functional decline, inability to continue anticancer therapy, nutritional deterioration, and progressive disease commonly inform prognostic estimation, formal prognostic scores (e.g., PPS or PPI) were not collected in this registry. Future studies incorporating objective prognostic indices may help refine and standardize the identification of patients with limited life expectancy. Third, the incidence of VTE recurrence and major bleeding may have been underestimated in the LLE group, since these events are often not aggressively investigated in patients approaching the end of life. Finally, external validation in multiethnic populations is warranted to enhance the generalizability of the present findings. Conclusion In this nationwide, real-world cohort, cancer patients with LLE were less likely to receive oral anticoagulants. After accounting for the competing risk of death, there were no significant differences in the cumulative incidence of VTE recurrence or major bleeding between cancer patients with and without LLE in the DOAC era. These findings support a more individualized, palliative-focused approach to VTE management in patients with advanced cancer. Supplementary Information Supplementary Material 1. Supplementary Material 2. Acknowledgements The authors appreciate the support and collaboration of all co-investigators and participating centers in the COMMAND VTE Registry-2. The authors thank FORTE Science Communications (https://www.forte-science.co.jp/) for English language editing. Abbreviations Confidence interval Direct oral anticoagulant Deep vein thrombosis International Society of Thrombosis and Hemostasis Limited life expectancy Pulmonary embolism Subdistribution hazard ratio Strengthening the Reporting of Observational Studies in Epidemiology Venous thromboembolism Authors’ contributions TK conceived and designed the study, performed the statistical analysis, and drafted the manuscript. NK conceived and designed the study and critically reviewed the manuscript for important intellectual content. YY, TM, YO, NT, MO, RC, KK, YN, NI, YK, SI, KK, MI, TT, ST, MO, TT, KO, JS, YO, TI, SU, PMC, KT, SH, KD, HM, YT, KM, KT, HN, DS, WS, TD, RN, HI, and TK contributed to data acquisition and interpretation. All authors critically reviewed and revised the manuscript for important intellectual content and approved the final version to be published. Funding This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Data availability The datasets generated and/or analyzed during the current study are not publicly available due to patient privacy regulations but are available from the corresponding author (Norimichi Koitabashi) on reasonable request. Declarations Ethics approval and consent to participate The study protocol was approved by the institutional review boards or ethics committees of all 31 participating centers, as listed in Supplementary Appendix 1. In accordance with the guidelines for epidemiological studies issued by the Ministry of Health, Labour and Welfare in Japan, the requirement for written, informed consent from each patient was waived because the study used de-identified clinical information obtained from routine clinical practice. Consent for publication Not applicable. Competing interests Dr. Koitabashi received lecture fees from Bayer Healthcare and grant support from Pfizer. Dr. Yamashita received lecture fees from Bayer Healthcare, Bristol-Myers Squibb, Pfizer, and Daiichi-Sankyo, and grant support from Bayer Healthcare and Daiichi-Sankyo. Dr. Kaneda received lecture fees from Bristol-Myers Squibb, Pfizer, and Daiichi-Sankyo. Dr. Morimoto reports lecturer's fees from Bristol-Myers Squibb, Daiichi Sankyo, Japan Lifeline, Kowa, Kyocera, Novartis, and Toray; manuscript fees from Bristol-Myers Squibb and Kowa; advisory board for Sanofi. Dr. Nishimoto received lecture fees from Bayer Healthcare, Bristol-Myers Squibb, Pfizer, and Daiichi-Sankyo. Dr. Ikeda N. received lecture fees from Bayer Healthcare, Bristol-Myers Squibb, and Daiichi-Sankyo. Dr. Ikeda S. received lecture fees from Bayer Healthcare, Bristol-Myers Squibb and Daiichi-Sankyo. Dr. Ogihara received lecture fees from Bayer Healthcare, Bristol-Myers Squibb, Pfizer, and Daiichi-Sankyo, and research funds from Bayer Healthcare and Daiichi-Sankyo. All other authors have reported that they have no relationships relevant to the contents of this paper to disclose. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Supplementary Materials Supplementary Material 1. Supplementary Material 2. Data Availability Statement The datasets generated and/or analyzed during the current study are not publicly available due to patient privacy regulations but are available from the corresponding author (Norimichi Koitabashi) on reasonable request.