Venous thromboembolism (VTE), encompassing deep-vein thrombosis (DVT) and pulmonary embolism (PE), affects approximately 1.3 to 1.5 per 1,000 adults annually and remains a significant cause of preventable morbidity and mortality worldwide (1). Standard diagnostic modalities, including compression ultrasonography and computed tomography pulmonary angiography, allow timely detection, and anticoagulant therapies, such as heparins, vitamin K antagonists (VKA), and direct oral anticoagulants (DOACs), achieve high rates of efficacy in the general population (1–3).

Among patients with malignancy, however, the incidence, severity, and clinical consequences of thrombosis are substantially amplified (4). Cancer-associated thrombosis (CAT) is now recognized as the second leading cause of death in oncology patients, surpassed only by cancer progression itself (5). Tumor-related mechanisms—including tissue factor overexpression, release of procoagulant microparticles, tumor–platelet interactions, neutrophil extracellular trap formation, and cytokine-mediated endothelial dysfunction—create a profoundly prothrombotic milieu while simultaneously increasing bleeding risk and complicating anticoagulant management (6).

A distinct and less common manifestation within the spectrum of CAT is tumor thrombus (TT), defined by direct intraluminal extension of malignant tissue into adjacent venous structures, such as renal vein or inferior vena cava involvement in renal cell carcinoma (RCC), or portal and hepatic vein invasion in hepatocellular carcinoma (HCC) (5). For the purposes of this review, TT is defined as an intravascular extension of malignant tissue, including contiguous macrovascular invasion (e.g., renal vein or inferior vena cava extension in RCC, portal or hepatic vein invasion in HCC), intracardiac tumor extension, and intravascular tumor emboli with confirmed malignant histology (7, 8). Unlike bland VTE, TT consists predominantly of viable malignant tissue interwoven with fibrin and platelets and is therefore biologically distinct from fibrin-rich thrombus (7, 9). Although these entities differ in anatomical location and clinical presentation, they share a common mechanism of malignant intravascular invasion and represent a unique pathophysiologic process within the spectrum of CAT.

In clinical practice, anticoagulant therapy is frequently prescribed in patients with TT based on extrapolation from CAT management paradigms. However, whether anticoagulation provides meaningful clinical benefit in TT, such as thrombus regression, prevention of embolic events, or survival improvement, remains uncertain. The available evidence is sparse, heterogeneous, and confounded by advanced malignancy, vascular invasion, and competing bleeding risks (10). As a result, major society guidelines offer only limited and sometimes conflicting recommendations regarding anticoagulation in TT, reflecting variability in tumor biology, thrombus extent, and available data (11, 12). Although recent observational studies and narrative reviews have highlighted substantial practice variability and the increasing use of DOACs, the comparative effectiveness and safety of anticoagulation in this population have not been systematically evaluated (9, 10, 13). Importantly, patients with TT have been systematically excluded from major randomized trials of CAT, including studies evaluating low-molecular-weight heparins and direct oral anticoagulants. As a result, the applicability of CAT trial evidence to patients with malignant intravascular invasion remains uncertain, and current anticoagulation practice for tumor thrombus is largely extrapolated from CAT paradigms, with limited direct supporting data. This evidence gap creates substantial clinical uncertainty regarding the effectiveness and safety of anticoagulation in this distinct population (14–16).

To date, no comprehensive systematic review has synthesized the available evidence addressing anticoagulant use specifically in solid tumor–associated tumor thrombus. Addressing this gap is critical to clarify the potential benefits and harms of anticoagulation in TT, inform multidisciplinary clinical decision-making, and guide future research and guideline development. Therefore, this systematic review aims to critically appraise and synthesize available studies evaluating the impact of anticoagulant therapy on thrombus resolution, thromboembolic recurrence, survival, and bleeding outcomes in adults with solid tumor–associated tumor thrombus.

This systematic review represents the first comprehensive synthesis of available evidence evaluating the effectiveness and safety of anticoagulant therapy in adults with solid tumor–associated TT. Several large surgical and oncologic series describing TT were identified but were excluded because they did not evaluate anticoagulation exposure or report anticoagulant-related outcomes. By systematically appraising eight studies encompassing multiple malignancies and anticoagulant classes, our findings provide a unified assessment of a clinical practice that remains widespread despite minimal empirical support.

Collectively, the included studies demonstrate that anticoagulation alone does not result in radiographic regression of TT, nor does it consistently protect against thromboembolic recurrence or improve survival outcomes. In contrast, a clinically meaningful risk of bleeding, particularly in patients with RCC, was repeatedly observed (14–21). Together, these findings suggest a limited therapeutic role for anticoagulation in modifying TT biology while highlighting a non-trivial safety burden.

A central and consistent observation of this review is the absence of TT regression attributable to anticoagulant therapy. All documented cases of complete thrombus resolution occurred exclusively following tumor-directed or mechanical interventions, including surgical thrombectomy, percutaneous aspiration (e.g., AngioVac), cryoablation, embolization, or systemic oncologic therapy (16–19). These observations reinforce the biological distinction between TT and bland venous thromboembolism: TT consists predominantly of viable malignant tissue rather than fibrin-rich clot, rendering it intrinsically resistant to anticoagulant-mediated dissolution (8, 10). Our synthesis therefore provides empirical confirmation, across tumor types and anticoagulant classes, that anticoagulation should not be expected to modify TT biology.

The evidence regarding thromboembolic recurrence was limited and heterogeneous. While TT was consistently associated with a higher baseline risk of VTE, most notably in RCC, anticoagulation did not reliably mitigate this risk. In the largest cohort study, anticoagulated patients with RCC-related TT continued to experience high rates of VTE despite treatment (15), suggesting that malignant embolic risk may be driven more by tumor burden and vascular invasion than by hypercoagulability alone. Data from HCC were similarly inconclusive, with numerical trends toward fewer non-malignant thrombotic events failing to reach statistical significance and remaining confounded by concurrent interventions (14, 21).

An important distinction in the management of patients with TT is between prophylactic-dose anticoagulation administered for routine cancer-associated VTE prevention and therapeutic-dose anticoagulation prescribed for established thrombosis or empirically for TT itself. These strategies serve different clinical objectives and should not be conflated. Most included studies evaluated therapeutic anticoagulation, and data on thromboprophylaxis in patients with TT were sparse and not systematically reported. As a result, current evidence does not permit conclusions regarding whether prophylactic anticoagulation prevents progression from localized tumor thrombus to systemic VTE or prevents the formation of bland thrombus on TT surfaces. These questions remain clinically important but are not addressed by the available literature. Future prospective studies should explicitly distinguish prophylactic and therapeutic anticoagulation strategies and evaluate their differential impact on thromboembolic risk and bleeding outcomes in patients with TT.

Survival outcomes across all included studies were predominantly determined by tumor-related factors rather than by anticoagulation exposure. Neither cohort studies nor case reports demonstrated a survival advantage associated with anticoagulant therapy (14–21). In RCC, anticoagulated patients exhibited higher mortality; however, this association was confounded by more advanced disease and greater thrombus burden (15). In HCC and other malignancies, survival correlated with tumor stage, extent of vascular invasion, and feasibility of definitive tumor-directed treatment rather than with anticoagulant use (8, 14). Based on available evidence, anticoagulation does not appear to alter the natural history of TT or the oncologic trajectory of affected patients.

In contrast, bleeding risk emerged as a consistent and clinically important safety signal. Anticoagulation was associated with a significantly increased risk of major bleeding in RCC cohorts (15), and clinically relevant bleeding events were reported across tumor types (21). Although HCC cohorts appeared to demonstrate lower rates of variceal bleeding among anticoagulated patients, this finding was likely attributable to aggressive prophylactic variceal management rather than a protective effect of anticoagulation itself (14). The absence of major bleeding in case reports must be interpreted cautiously, given limited follow-up and strong selection bias. Overall, the balance of evidence suggests that anticoagulation exposes patients with TT to substantial harm without demonstrable disease-modifying benefit.

Importantly, the behavior of TT is increasingly influenced by modern systemic anticancer therapies, which may induce tumor regression—including intravascular components—independent of anticoagulation. Targeted agents such as tyrosine kinase inhibitors (e.g., sunitinib, cabozantinib) and immune checkpoint inhibitors have demonstrated activity against RCC and HCC and may contribute to regression of vascular tumor invasion in selected patients (22–25). In this context, apparent thrombus regression reported in isolated cases is difficult to attribute to anticoagulation when administered concurrently with effective systemic therapy. Conversely, anti-angiogenic agents, including bevacizumab and VEGF pathway inhibitors, are associated with increased bleeding risk and may further shift the anticoagulation risk–benefit balance in patients with TT (26–29). These considerations underscore the need for multidisciplinary evaluation and integration of oncologic treatment plans when anticoagulation is contemplated.

These findings reinforce the concept that TT represents a spectrum of malignant intravascular invasion with shared biological features but distinct anatomical and clinical manifestations. Although contiguous venous extension in RCC and HCC, intracardiac tumor extension, and pulmonary tumor embolism differ in presentation and prognosis, all are characterized by malignant tissue occupying the vascular lumen and demonstrating limited responsiveness to anticoagulation (8, 10, 14–20). Across tumor types and locations, thrombus regression attributable solely to anticoagulant therapy was not observed, and apparent thrombus resolution was consistently associated with tumor-directed or mechanical interventions (16–19). In this context, thrombus regression described in isolated reports is difficult to interpret when anticoagulation is administered concurrently with definitive tumor-directed therapy.

Importantly, the current literature does not exclude the possibility that anticoagulation may provide benefit in narrowly defined clinical contexts, such as in patients with mixed malignant and bland thrombotic components or those with concurrent VTE. However, these subgroups were not systematically evaluated, and available data are insufficient to draw definitive conclusions (14, 15, 20). The overall certainty of evidence remains low due to retrospective study designs, small sample sizes, heterogeneity in tumor types and thrombus extent, and confounding by tumor stage and treatment (14–20).

These findings highlight the importance of distinguishing isolated TT from clinical scenarios in which anticoagulation is routinely indicated regardless of tumor involvement, including confirmed PE, proximal DVT, atrial fibrillation with elevated thromboembolic risk, or catheter-associated thrombosis (1, 11, 12). In such cases, anticoagulation should be guided by established principles for cancer-associated thrombosis, with careful consideration of bleeding risk.

From a clinical perspective, tumor-directed strategies, including surgical thrombectomy, locoregional therapies, and systemic anticancer treatment, remain the cornerstone of TT management (8, 11). When used, anticoagulation should be considered an adjunctive, indication-driven therapy rather than a disease-modifying therapy. Multidisciplinary collaboration among oncology, surgery, cardiology, hepatology, and thrombosis specialists is essential to individualize care and avoid unnecessary harm.

In summary, available evidence suggests that anticoagulation does not appear to meaningfully modify tumor thrombus biology, induce radiographic regression, reduce thromboembolic recurrence, or improve survival in patients with solid tumor–associated tumor thrombus, while exposing patients to a clinically relevant risk of bleeding. Tumor-directed and mechanical interventions remain the primary determinants of clinical outcomes. Anticoagulation should therefore not be used routinely for isolated tumor thrombus, but may be appropriate when conventional indications are present, such as confirmed pulmonary embolism, proximal deep-vein thrombosis, atrial fibrillation with elevated thromboembolic risk, or catheter-associated thrombosis. Given the low certainty and heterogeneity of existing data, prospective multicenter studies are needed to better define patient subgroups that may derive benefit and to inform evidence-based clinical guidance.