Breaking Down Tumor Thrombus: Current Strategies for Medical Management

Abstract

Tumor thrombus, the intravascular extension of tumor into adjacent blood vessels, is frequently encountered in patients with renal cell carcinoma and hepatocellular carcinoma, and often involves the abdominal vasculature including the renal vein, portal vein, and the inferior vena cava. While a bland thrombus is composed of platelets and fibrin, in contrast, a tumor thrombus refers to an organized collection of tumor cells. Though oftentimes detected incidentally on imaging, tumor thrombus may have significant clinical implications and can be challenging to differentiate from bland thrombus. Additionally, the optimal management of tumor thrombus, including the use of anticoagulation, remains poorly described. This review summarizes common causes of tumor thrombus, as well as its impact on staging, prognosis, and treatment.

I. Introduction

Cancer-associated thrombosis (CAT) is one of the leading causes of death in patients with cancer.¹ A complex interplay of factors, including tumor cell secretion of procoagulant substances and the nonspecific generation of inflammatory cytokines, acute phase reactants, abnormal protein metabolism, and blood stasis around the tumor site ultimately culminate in a hypercoagulable state and subsequent clot formation. Additional treatment-specific factors, such as chemotherapy, immunotherapy, hormone therapy, and surgery can also increase risk of thromboembolic events through similar mechanisms, such as stimulating tissue factor production by host cells.^2–4

Venous thromboembolism (VTE), which refers most commonly to deep venous thrombosis (DVT) and pulmonary embolism (PE), is typically thought of as a bland thrombus (BT), or one composed of platelets, macrophages, and fibrin.⁵ In contrast, tumor thrombus (TT) is a less commonly described phenomenon referring to the intravascular extension of a tumor into adjacent blood vessels. TT can occur in various tumor types, but it is most commonly described in patients with renal cell carcinoma (RCC), hepatocellular carcinoma (HCC), pancreatic neuroendocrine tumors (PNETs), and lung cancer. Other tumors associated with TT include gastric adenocarcinoma, adrenocortical carcinoma (ACC), osteosarcoma, Wilms tumor (WT), and uterine leiomyosarcoma.^6–8 Though TT is often asymptomatic and incidentally detected on routine surveillance imaging, the presence of TT worsens prognosis, affects cancer staging, and can impact treatment options for patients. Strong observational data stratifying overall survival based on the presence or absence of TT is lacking in most tumor types, and as such the true impact of TT remains unknown. Common sites of TT by specific tumor type, including a summary of the incidence, suggested treatment modalities, and prognosis is summarized in Table 1.

Table 1:

Summary of different tumor types, possible sites of tumor thrombus, treatment modalities and prognosis

Tumor Type	Possible Sites & Frequency of TT	Treatment Modalities	Prognostic Data	References
Renal cell carcinoma (RCC)	Renal vein: 10–18% Inferior vena cava (IVC): 4–10% Right atrium/right ventricle: 1% Pulmonary artery: 1.5–3.5%	Nephrectomy Tumor thrombectomy	RCC with IVC TT: Median OS and 1- year DSS with nephrectomy & tumor thrombectomy vs untreated:21.5 months and 75.5% vs 5 months and 30%	10,23,24
Hepatocellular carcinoma (HCC)	Portal vein: 44–66% Hepatic vein: 2–5% IVC: 3–4% Right atrium Pulmonary artery	Systemic therapy Radiation therapy Liver-directed therapies Hepatectomy with tumor thrombectomy	Median OS of untreated HCC with PVTT: 2–4 months Median OS of untreated HCC with HVTT: 5–7 months	46,50,52,61–63
Intrahepatic cholangiocarcinoma (ICC)	Portal vein: 15–20% Hepatic vein	Liver resection Tumor thrombectomy	5-year OS after resection with PVTT vs without PVTT: 0% vs 23%	69
Adrenocortical carcinoma (ACC)	Adrenal vein Renal vein IVC:25%	Surgical resection with tumor thrombectomy Systemic chemotherapy	Median OS: Complete resection with IVCTT vs without IVCTT: 14.8 months vs 43 months	49
Wilms Tumor (WT)	Renal vein IVC: 4–10% Right atrium Pulmonary artery	Neoadjuvant chemotherapy Radical nephrectomy	OS in children vs adults: 90% vs 70% OS data specific to WT with TT is lacking	34–38
Pancreatic neuroendocrine tumor (PNET)	Splanchnic veins: 33%	Pancreatectom y (extent depending on location) Splenectomy if SV involved tumor thrombectomy	OS and DFS post resection of PNET: 60% and 30% respectively	70,111
Colorectal cancer (CRC)	Splanchnic veins: 1.7–2.8%	Complete tumor resection Adjuvant chemotherapy	OS data specific to CRC with TT is lacking. Long survival expected with complete resection and adjuvant chemotherapy	71,72,75
Pelvic Osteosarcoma	Overall:45% Pelvic veins Common femoral vein Common iliac vein Internal iliac vein Heart	Complete thrombectomy (limb salvage) Resection (amputation)	Median OS of surgically treated pelvic osteosarcoma with TT vs without TT : 54 months Vs 21.5 months	7,88
Intravenous Leiomyomatosis (IVL)	Pelvic veins IVC Heart	Complete resection with hysterectomy Bilateral salpingo-oophorectomy Tumor thrombectomy	Good prognosis with complete resection.	89,112
Testicular cancer	Overall: 3–11% Gonadal veins IVC	Chemotherapy Orchiectomy Tumor thrombectomy	Prognostic data specific to TT is lacking	91,113
Thyroid: Follicular carcinoma	Neck veins: 0.2–3.8% Jugular brachiocephalic Superior vena cava (SVC) Right atrium	Total thyroidectomy Tumor thrombectomy Radioactive iodine (RAI)	OS with TT (treated): <24 months to 96 months	94,95,114

Abbreviations: TT, tumor thrombus; OS, Overall Survival; DSS, Disease specific survival; DFS, Disease free survival; PVTT, portal vein tumor thrombus; HVTT, hepatic vein tumor thrombus;

Additionally, BT may sometimes co-exist with TT. It is postulated that the presence of TT disrupts vascular integrity, activates vascular endothelium, and disturbs the venous blood flow leading to a local procoagulant state and increased risk of local BT development.⁹ Due to its composition, BT is prone to propagation and embolization if left untreated.^5,9,10 In contrast, TT contains organized tumor cells and is believed to be more stable and thus less prone to embolization.^9,11 Though biopsy is the gold standard to differentiate TT from BT, this is rarely performed in clinical practice. Instead, multimodality imaging plays a critical role in distinguishing TT from BT.¹²

Finally, though anticoagulation is the standard of care for the treatment of BT, the optimal management of TT has not been fully delineated. Anticoagulant use in patients with cancer can present challenges, particularly in those with chemotherapy-induced thrombocytopenia or those with other risk factors for bleeding. High-quality data is sparse regarding the natural history of TT as are studies informing clinical outcomes of those treated with or without anticoagulation. Published guidelines from the major hematologic and oncologic societies do not include recommendations for the management of TT, likely due a paucity of evidence.^13–16 This review aims to summarize existing literature regarding the role of anticoagulation in the management of TT and to serve as a practical guide for healthcare professionals faced with this challenging clinical scenario.

II. Common Sites of TT

TT in the Inferior Vena Cava (IVCTT)

The most common tumors involving the inferior vena cava (IVC) are RCC, HCC, and ACC. RCC has a known predilection for vascular invasion and TT formation with a cited prevalence of approximately 4–10%.⁶ The presence of TT in the setting of a new RCC diagnosis complicates the staging, prognosis, and surgical approach.^17–21 For example, the presence of TT upstages a tumor from T2 to T3a for segmental or central renal vein involvement, and to T3b for the involvement of the IVC.²² Clinical upstaging can significantly alter the surgical approach in locally advanced RCC and potentially preclude nephron-sparing surgery as an option for patients. The median survival of patients with untreated RCC and IVCTT is about 5 months, with an estimated 1-year disease-specific survival of 29%.²³ In comparison, cytoreductive nephrectomy with tumor thrombectomy improves the median survival and disease-specific survival to 21.5 months and 75.5% respectively.²⁴ Complete nephrectomy and tumor thrombectomy improve 5-year survival to greater than 50% compared to 10% with partial nephrectomy.^25–29

The Mayo Clinic thrombus classification is widely used to describe levels of IVCTT involvement in patients with RCC to guide surgical planning.³⁰ Level III TT, defined as TT extending above the hepatic veins but below the diaphragm, and Level IV TT, defined as TT extending above the diaphragm, require complex surgical planning with concurrent cardiopulmonary bypass. IVC wall invasion of TT is equally important in preoperative assessment, as this requires IVC resection and reconstruction. Anticoagulation is generally indicated if BT is identified in addition to IVCTT, and the coexistence of both is independently associated with poor survival, adverse pathological features, advanced tumor stage, and increased risk of nodal metastases.^24,31 Perioperative PE occurs in less than 6% of patients with IVCTT but carries a mortality rate as high as 60–75% due to the requirement of cardiopulmonary bypass for embolectomy.^29,32,33

WT, also known as nephroblastoma, is rarely diagnosed in adults and accounts for 0.5% of renal cancers. Like RCC, WT tends to invade the renal vein and extend into the IVC in about 4–10% of the cases.^34–37 Adults with WT tend to present at a later stage and have worse prognosis compared to pediatric cohorts, which is largely attributed to delay in diagnosis and management.^38–40 Neoadjuvant chemotherapy is recommended for WT with TT invading the IVC or right atrium in order to shrink the size of TT, thus decreasing surgical complexity and perioperative complication rates.^34,41,42

Another cancer type that increases risk of IVCTT is HCC, which is estimated to occur in about 3–4% cases and is associated with a dismal prognosis of 1–4 months for untreated cases.^43–47 Hepatectomy with thrombectomy, although technically challenging, improves overall survival, though is limited to patients with a resectable primary tumor and preserved liver function.⁴⁵ TT involving the portal and hepatic veins is more frequently encountered in patients with HCC and is discussed in greater detail below.

Finally, ACC, a rare and aggressive neoplasm, tends to invade locally into the adrenal veins and extend into the renal vasculature and the IVC in approximately 25% of the cases.⁴⁸ Patients with ACC and IVCTT have a significantly poorer prognosis compared to those without vascular invasion. In fact, a retrospective analysis of 65 patients with locally advanced ACC who underwent complete resection found a reduced median overall survival of 14.8 vs 43 months in those with and without IVCTT.⁴⁹ Given the morbidity of complete surgical resection and thrombectomy, patients with ACC must be carefully selected.

TT in the Portal Vein (PVTT) and Hepatic Vein (HVTT)

HCC has a distinct ability to extend into the hepatic vasculature, most frequently involving the portal vein. Macrovascular invasion (MVI) of the portal vein, or PVTT, occurs in 44–66% of patients with HCC and is present at the time of diagnosis in approximately 20% of the patients.^50,51 MVI has significant clinical consequences, including the development of portal hypertension and associated liver decompensation. The presence of PVTT in HCC portends a poor prognosis, with an estimated median survival of only 2–4 months without treatment.^46,52 The extent of PVTT also has prognostic implications among those receiving treatment, with an estimated median survival of 9 months for those with segmental PVTT compared to 4.6 months with main portal vein involvement.⁵³

The presence of PVTT in patients with HCC, classified as Barcelona Clinic Liver Stage (BCLC) Stage C, carries significant clinical consequence as surgical resection is no longer a treatment option for patients.^54,55 Surgical hepatectomy with thrombectomy, though rarely performed in western countries, is increasingly performed in the Asian countries depending on the extent of PVTT.^6,56 Additional treatment modalities are also limited, including liver transplantation due to high rates of PVTT recurrence and transarterial chemoembolization due to risk of hepatic necrosis and worsening liver function.^57–60 Systemic therapy with vascular endothelial growth factor inhibitors, immunotherapy, and tyrosine kinase inhibitors remains the backbone of treatment for patients with HCC who are not candidates for localized therapy.

HVTT is a rarer complication compared to PVTT in patients with HCC, with an estimated prevalence of 2–5%.^44,61–63 HVTT also portends a poorer prognosis with an estimated median overall survival of about 5–7 months.⁶³ While consensus guidelines from the European Association for the Study of the Liver (EASL) and the American Association for the Study of Liver Diseases (AASLD) do not specifically address HVTT, surgical resection with hepatectomy with thrombectomy is often required.^44,64,65

A disorder worth mentioning is Budd Chiari syndrome (BCS), which is characterized by obstruction of the hepatic vein outflow tract at any level ranging from the hepatic venules, large hepatic veins to the hepatic IVC and right atrium.^66,67 TT-associated BCS is a very rare condition and not commonly encountered. Clinical features range from no symptoms to the classical triad of abdominal pain, hepatomegaly, and ascites. Symptomatic and untreated BCS carries a mortality rate of 90% in three years.⁶⁸

Other tumors with a predilection for PVTT include intrahepatic cholangiocarcinoma (ICC), an aggressive malignancy with an overall poor prognosis. Macrovascular PVTT occurs in 15–20% of cases of ICC and is an independent negative prognostic factor associated with worse overall survival and disease-free survival. In fact, estimated 5-year overall survival for patients with ICC and PVTT after curative resection is 0% compared to 23% without PVTT.⁶⁹

TT in the Non-portal Splanchnic and Other Abdominal Veins

TT involving non-portal splanchnic veins, including the splenic and superior and inferior mesenteric veins, is often seen in PNETs, which account for less than 2% of pancreatic tumors. While BT is prevalent among patients with pancreatic adenocarcinoma, local venous invasion/TT is more common in PNETs and can be seen in up to 33% of cases.⁷⁰ Oftentimes detected incidentally on imaging, TT impacts surgical planning and can lead to clinical complications.⁷⁰ One such complication of splenic vein TT is an entity known as sinistral portal hypertension, which can lead to the development of isolated gastric varices and gastrointestinal bleeding. Surgical resection with tumor thrombectomy is the recommended treatment modality in patients with PNET and TT when feasible.

Another rarer cause of TT involving the abdominal vasculature is colorectal cancer (CRC), reported to occur in 1.7–2.8% of cases.^71,72 The specific site of TT depends on the location of the primary tumor, with tumors of the ascending and transverse colon tending to invade the superior mesenteric vein (SMV) while descending and sigmoid favor the inferior mesenteric vein (IMV). TT in the setting of rectal cancer tends to involve the internal iliac vein. In rare circumstances, however, involvement of the portal vein has also been described in the setting of liver metastases due to the dual venous drainage system.^71,73,74 Complete tumor resection followed by adjuvant chemotherapy improves prognosis.⁷⁵

TT in the Pulmonary Vasculature

TT involving the pulmonary vasculature can occur in three different ways: direct extension, macroscopic tumor embolism, and tumor microembolism. First, direct extension of a cavoatrial thrombus into the proximal pulmonary vasculature has been reported to occur in patients with RCC, as well as cases of non-small cell lung cancer (NSCLC), in which TT invades the pulmonary artery and extends in retrograde fashion into the main pulmonary trunk.⁷⁶ Second, macroscopic tumor embolism can occur when cell clusters of tumors embolize into the pulmonary vasculature and mimic acute PE from BT.⁷⁷ Lastly, and quite rare, is involvement of pulmonary microvasculature by two entities known as pulmonary tumor microembolism (PTE) and pulmonary tumor microangiopathy (PTTM). PTE and PTTM are frequently associated with mucin-secreting adenocarcinomas of the stomach, lung, breast, and colon.^78,79 PTE refers to the occlusion of pulmonary microvasculature by tumor emboli, while PTTM represents the natural progression of PTE, characterized by extensive remodeling of pulmonary vessels with nests of tumor emboli.⁸⁰ PTE and PTTM are rarely diagnosed antemortem and instead are frequently discovered on autopsy studies, with an estimated prevalence of 3–26% for PTE and 1–3% for PTTM.^78,80–82 Both PTE and PTTM can rapidly progress to cause severe pulmonary hypertension, acute hypoxic respiratory failure, right heart failure, and sudden death.^82,83

Differentiating between BT and TT using traditional imaging techniques is challenging. On CT imaging, both bland and tumor pulmonary emboli appear as focal hypodense filling defects, while on V/Q scan, pulmonary microemboli appear as small peripheral subsegmental perfusion ventilation mismatched defects or have a checkerboard appearance, known as the “checkerboard sign”.⁸⁴ Though definitive diagnosis of PTE or PTTM can be made by cytological exam from a wedged pulmonary artery catheter or by lung biopsy, this is typically not pursued as PTE/PTTM is often a manifestation of end-stage cancer with a very limited prognosis.⁸⁵ Anticoagulation can be considered based on anecdotal reports.⁸⁰ Pulmonary thrombectomy or embolectomy can be attempted for macroscopic or proximal vasculature involvement if feasible.^9,86

III. Other Less Common Sites of TT Involvement

TT involving the pelvic veins can be seen in primary pelvic osteosarcoma, which is associated with venous invasion in approximately 45% of cases.⁸⁷ Median survival of pelvic osteosarcoma with TT is 21.5 months compared to 54 months without TT. Surgical resection and tumor thrombectomy is usually recommended and is associated with improved prognosis.⁸⁸

Intravenous leiomyomatosis (IVL) is a rare, atypical form of uterine leiomyomatosis with growth of tumor within the intrauterine veins extending to the extrauterine pelvic veins, IVC, and heart. The recommended treatment for IVL that extends into the IVC is complete resection with hysterectomy, bilateral salpingo-oophorectomy, and tumor thrombectomy with cardiopulmonary bypass depending on the level of the TT. As expected, early diagnosis and management is associated with improved prognosis.⁸⁹

Testicular cancer may also rarely cause TT through extension of the primary tumor into the gonadal veins and may also involve the IVC via metastatic lymphadenopathy. Autopsy reports suggest an incidence of approximately 11%, though only 1% of cases were detected by imaging.^90,91 Similarly, early diagnosis and tumor resection is critical to improve prognosis and prevent complications of TT.

TT in follicular thyroid carcinoma has also been reported to involve the great veins of the neck, in about 0.2–3.8% of cases including the jugular and brachiocephalic veins, as well as the mediastinum, including the superior vena cava and right atrium, and is independently associated with poor prognosis. ^92–95

As mentioned above, NSCLC can rarely cause TT that extends into the pulmonary vasculature and left atrium, leading to arterial embolism, stroke, bowel infarction, and acute limb ischemia.^96,97 In such cases, prognosis is poor, and embolectomy should be attempted if feasible.

IV. Multimodality Imaging to Differentiate TT from BT

As summarized above, the accurate identification and differentiation of TT from BT is critical for tumor staging, clinical management, and treatment planning. Overall, it is important for clinicians to maintain a high index of suspicion for TT, especially in cancers with a predilection for TT, such as RCC, HCC, PNET, and NSCLC. In general, when comparing imaging characteristics of TT compared to BT, TT is more expansile, demonstrates enhancement, appears contiguous with the primary mass, and has imaging features similar to that of primary malignancy.⁸ While grey-scale ultrasound (US) is sensitive for detecting venous thrombosis, it can be challenging to discern BT from TT. Instead, multimodality imaging techniques using contrast-enhanced ultrasound (CEUS), contrast-enhanced computed tomography (CECT), contrast-enhanced magnetic resonance imaging (CEMRI), and combined fluorodeoxyglucose/positron emission tomography/CT (FDG PET/CT) have demonstrated superiority in differentiating TT from BT compared to US alone.

With regards to imaging technique, CEUS is fast, inexpensive, and reliable with a high sensitivity (94%) and specificity (96%) in differentiating BT from TT of the portal vein.^98,99 The presence of enhancing soft tissue on CECT or CEMRI in a portal vein or hepatic vein is indicative of tumor thrombus. CECT and CEMRI are frequently used to evaluate HCC and have moderate sensitivity (64% for CECT vs. 62% for CEMRI) but are nearly 100% specific for detecting macroscopic PVTT or HVTT.^100,101 MRI T2-weighted and diffusion-weighted imaging (DWI) sequences can also be effective tools in evaluating TT. Classically, TT is associated with intermediate T2 signal intensity compared to BT, which shows low intensity.⁸ TT on FDG PET/CT typically has a higher maximum standardized uptake value (SUV) than BT, with a reported sensitivity, specificity, and accuracy of 94%, 80%, and 89%, respectively, when an SUV of greater than 3.35 is used for the diagnosis of PVTT.¹⁰² Ultimately, the optimal imaging technique has yet to be elucidated and multidisciplinary input from radiologists as well as center-specific resources play a key role in determining the recommended modality.

V. Risk of VTE in TT and Available Data to Guide Management

The presence of TT can increase the risk of BT, as seen in a retrospective study of 176 patients with RCC and TT undergoing nephrectomy. The analysis identified an increased incidence of preoperative VTE, defined as proximal lower-extremity DVT or PE, (hazard ratio (HR) 5.7, 95% Confidence Interval (CI) 1.4–16.8).²⁹ In a separate analysis of 170 patients with RCC , 97 of whom (57.1%) had tumor thrombus, patients with residual TT following surgery had a higher risk of VTE at 2-year follow-up compared to those with complete tumor thrombectomy (HR 8.7, 95% CI 1.7–43.4) and those with no tumor thrombus (HR 6.5, 95% CI 1.7–24.7).¹⁰³ Similar results were also observed in a separate retrospective study of 2762 patients with RCC and TT, which identified an estimated odds ratio for VTE of 8.16 (95% CI 1.48–45).¹⁰⁴ Lastly, a retrospective analysis of 183 patients who underwent nephrectomy and IVC tumor thrombectomy found post-operative VTE was significantly associated with an increase in all-cause mortality.(HR 1.53, 95% CI 1.04–2.23).¹⁰⁵

Few retrospective studies have investigated the natural history and clinical outcomes of patients with TT treated with or without anticoagulation (Table 2). One particular retrospective study of 50 pediatric patients with TT, of which 20% (n=10) were treated with anticoagulation, found no significant difference in the rate of BT formation, thromboembolization, or overall survival in those treated with or without anticoagulation.¹⁰⁶ The most common cancers in this pediatric cohort were WT, followed by osteosarcoma and Ewing sarcoma. Notably, 40% (4 of 10) of those treated with anticoagulation reported one major and three minor bleeding events. Another retrospective analysis explored the management and outcomes of TT in 68 patients with different tumor types, including renal, lung, pancreas, gastrointestinal and liver malignancies.¹⁰⁷ In this study, TT resolution occurred in 14% of those on therapeutic anticoagulation, 10% of those on prophylactic dosing, and 9% of those not anticoagulated. Bleeding events were more frequent in the anticoagulation group, with 7% experiencing a major bleeding event and 16% a minor bleeding event. Notably there was no significant difference in overall survival in those treated with or without anticoagulation.

Table 2:

Studies evaluating the role of anticoagulation in TT

Study	Patient population	Tumor Types	Primary Outcomes	Safety outcomes
Marcoux et al 2019	N=153 41 (26.8%) on AC 112 (73.2%) not on AC	RCC HCC	BT: 18 (11.8%) within 6 months - 11 (61%) on AC - 7 (39%) not on AC OS: 42.5% at 6 months, p=0.42 No difference between groups	Major Bleeding: 45% in AC group
Agarwal et al 2021	N=50 pediatric patients 10 (20%) on AC -6 therapeutic AC -4 prophylactic AC 40 (80%) not on AC	WT Hepatoblastoma Neuroblastoma Ewing’s sarcoma Osteosarcoma	BT: 0 on both groups All-cause mortality: AC group- 3 (30%) No AC group-16 (40%) p=0.722	Bleeding: 40% in AC group, p=0.01 Major-1 Minor- 3
Faruqi et al 2022	N=68 68% on AC (therapeutic or prophylactic) 32 % not on AC	Renal Lung Pancreas Gastrointestinal Liver Other	Tumor resolution: -14% on therapeutic AC -10% on prophylaxis -9%not on AC OS: -39% on AC vs 41% not on AC at 6 months	Bleeding: Major – 7% – AC group Minor – 16%- AC group
Kaptein et al 2022	N=86 86 with TT in 647 patients 24 (28%) on AC 62 (72%) not on AC	RCC	BT: 23.9 % on AC 17.6% not on AC aHR 0.56 (95% CI 0.13–2.5)	Major bleeding: 32.5% on AC 11.9% not on AC HR 3.4 (95% CI 0.95–12)

Abbreviations: TT, tumor thrombus; AC, anticoagulation; BT, bland thrombus; HR, hazard ratio; aHR, adjusted hazard ratio; CI, confidence interval; OS, overall survival; IVC, inferior vena cava; RCC, renal cell carcinoma; HCC, Hepatocellular carcinoma

Prevention of BT in patients with TT has not been well studied. A single-center retrospective analysis of 153 adult patients with RCC and HCC-associated TT, found a 12% incidence of VTE in the total study population within six months of TT diagnosis.¹⁰⁸ Interestingly, the majority of patients who developed VTE (61%) were on anticoagulation for the management of TT, compared to 39% of those who were not initially treated with anticoagulation. Of note, there was no difference in overall survival in those who did or did not receive anticoagulation, but the anticoagulation group had a higher rate of major bleeding (45%). Similarly, a recent retrospective study of 86 patients with RCC and TT found that patients with TT were at higher risk of developing VTE (adjusted HR 6.61, 95% CI 2.18–13.73).¹⁰ The anticoagulation group (n=24) was associated with a 44% decreased risk of VTE (HR 0.56, 95% CI: 0.13–2.48) but had a higher risk of major bleeding (HR 3.4, 95% CI 0.95–12). The anticoagulation group had 32.5% major bleeding rate compared to 12% in the group not on anticoagulation. This study also showed that the incidence of VTE was lowest when TT involved the renal vein only (7.4%) compared to IVC below (22%) and above the diaphragm (55.3%), suggesting that the level of TT also carries important clinical implications.

Overall, the cumulative findings from these retrospective studies demonstrate an increased incidence of VTE in patients with TT with significant clinical implications. It is important to note that these studies were unable to distinguish BT from macroscopic TT in the pulmonary vasculature, as in the case of PTE. Furthermore, the few studies investigating the role of anticoagulation in the management of TT have not shown significant difference in preventing VTE or improving survival. Overall, the paucity of studies, many of which were limited to patients with RCC, limit robust conclusions. It is also important to recognize that significant heterogeneity among available studies, specifically regarding differences in tumor type, diagnostic techniques, outcome measures, and the type and dosing of anticoagulation, preclude broader interpretation.

VI. Patient Selection and Choice of Anticoagulation

Although limited, the available data suggests that patients with TT are at high risk of BT, yet the use of anticoagulation has not been shown to improve overall survival in any study to date and is associated with an increased risk of bleeding in this population. We can conclude based on the available data that presence of TT is a risk factor for development of BT, which is less stable and prone to embolization. Surgical thrombectomy in addition to anti-neoplastic directed medical therapy remain the mainstay of treatment for malignancy with accompanying TT. Therefore, the goal of anticoagulation in the management of TT should be to decrease the occurrence of BT and thromboembolism, rather than attempting to resolve or decrease the progression of the TT itself. This must be weighed against the competing risks of bleeding and cancer-related mortality. Specifically, patients with high bleeding risk and a short-anticipated survival are unlikely to benefit from anticoagulation and may be harmed if severe bleeding occurs.

Guidelines from major societies including the American Society of Hematology (ASH), the International Society on Thrombosis and Hemostasis (ISTH), the American society of clinical oncology (ASCO) and the American College of Chest Physicians (ACCP) do not address the use of anticoagulation in the setting of TT, at least in part due to lack of high quality prospective or randomized controlled studies.^13–16 Although no firm conclusions can be drawn from the currently available retrospective studies, our practice has been to consider the use of anticoagulation in select patients on a case-by-case basis.

Low molecular weight heparin (LMWH) was the most common anticoagulant used in retrospective studies evaluating anticoagulation in the setting of TT. Direct oral anticoagulants (DOACs) have changed the landscape of cancer associated thrombosis and are now incorporated as first line treatment by major society guidelines.^13,109,110 As such, our practice has been to generally use DOACs, although it should be noted that select DOACs have been found to carry increased rates of bleeding in patients with luminal GI cancers.¹⁶

Acknowledging a lack of data to guide recommendations, our approach is generally as follows,

• In patients with isolated TT (no associated BT), it is reasonable to monitor select patients without anticoagulation. Generally, the use of routine CT imaging as part of their cancer care allows the opportunity for serial monitoring of thrombosis extension or embolization.

• In patients with TT who undergo complete cancer resection/thrombectomy (major abdominopelvic, gynecological, or colorectal surgery), consideration of extended thromboprophylaxis up to 4 weeks post-operatively to prevent BT is reasonable

• In patients with TT who had partial resection/thrombectomy, continuing extended thromboprophylaxis past 4 weeks could be considered in select patients with favorable risk-benefit profile.

• In patients with TT and associated BT or more distant VTE, therapeutic anticoagulation should be considered as per cancer associated thrombosis guidelines. Careful decision making is often needed to account for the potential competing risks of bleeding and cancer associated mortality.

Conclusion

TT is commonly associated with certain cancers, including RCC and HCC. Available data suggests that patients with TT have a high incidence of developing BT, and the use of anticoagulation can help decrease the occurrence of bland VTE and thromboembolism. However, it has not been shown to improve overall survival in any study to date and is associated with a high rate of bleeding in this population. Therefore, the goal of anticoagulation in the management of TT should be to prevent bland VTE/thromboembolism rather than attempting to resolve or decrease the progression of the TT itself. The decision to use anticoagulation should be made on a case-by-case basis, taking into account the potential competing risks of bleeding and cancer-associated mortality. Future prospective studies evaluating the role of anticoagulation across different cancer types with TT are necessary to identify which patient subgroups benefit from prophylactic or therapeutic anticoagulation and the optimal choice of anticoagulant.

Highlights.

• Tumor thrombus (TT) refers to the intravascular extension of a tumor into adjacent blood vessels

• Differentiating TT from bland thrombus (BT) on multimodality imaging has significant staging, prognostic, and treatment implications

• The role of anticoagulation in the treatment of TT has not been well delineated

• Because the presence of TT increases risk of BT, anticoagulation should be considered for BT prevention in select patients with a favorable risk-benefit profile