This article is mentioned in:

Introduction

Hepatocellular carcinoma (HCC) is becoming a critical malignancy affecting human health worldwide (1,2). The 5-year survival rate of patients with HCC is <12% (3,4). Although surgical resection is the common therapeutic approach for patients with HCC at an early stage, recurrence is observed in >70% of patients within 5 years (5). Local metastasis and chemoresistance remain the key causes of the poor prognosis of HCC (6).

It has been established that the coagulation and fibrinolytic systems serve important roles in the human body by maintaining the balance of the physiological coagulation process, which is closely associated with liver functions, including generation of clotting factors, and synthesis of fibrinolytic and antifibrinolytic substances (7). However, recent studies have indicated that both systems can also influence the development and prognosis of HCC (8,9). It has been demonstrated that elevated levels of serum urokinase plasminogen activator (uPA) are associated with higher mortality of patients with HCC after resection (8). The clinical course of certain tumor types can be slowed down by protease inhibitors, which inhibit uPA expression (9). Similarly, the majority of patients with cancer exhibit some changes in the coagulation and fibrinolytic systems (10,11). However, it is unclear whether these changes have any effects on the biological behavior of cancers and how they interact with each other.

Coagulation characteristics of cancer

The blood coagulation and fibrinolytic systems are mainly involved in maintaining a balance between thrombogenesis and bleeding, and in keeping blood vessels unobstructed. However, systemic activation of blood coagulation is also a prominent characteristic of cancer, and various coagulation components have been identified to be associated with malignant phenotypes (10). For example, high fibrinogen expression has been suggested to contribute to the poor prognosis of patients with advanced cancer (11).

Coagulation activation in cancer can be classified into two types: Non-specific and tumor cell-specific (Fig. 1). The causes of non-specific activation include damaged tissues, inflammation, surgery, chemotherapy drugs and radiation. The causes of tumor cell-specific activation can be ‘direct’ or ‘indirect’. For example, tissue factor (TF), which is expressed by tumor cells, can trigger coagulation directly. Furthermore, certain cytokines, which are expressed by tumor cells, may activate procoagulant expression of endothelial cells, which can trigger coagulation indirectly. Patients with tumors can be divided into three groups according to specific coagulation activation. Certain patients exhibit generation of thrombin and fibrin by tumor cells but minor systemic coagulation activation, and these patients belong to the direct activation group (12). Other patients exhibit marked and deadly coagulation activation but small tumor tissues, and these patients belong to the indirect activation group (13). The remaining patients who do not belong to either of these groups may exhibit only limited activation of coagulation (14).

Figure 1. Coagulation characteristics of cancer. Usually, the blood coagulation system and the fibrinolytic system are important physiological activities maintaining the balance between thrombogenesis and bleeding. However, systemic activation of the blood coagulation mechanism is also a prominent characteristic of cancer, which may be non-specific or tumor cell-specific.

Animal models suggest that numerous drugs that regulate coagulation activities also have antitumor effects (15,16). These drugs can attenuate tumor growth, decrease metastasis and prolong survival time. Furthermore, the majority of the components associated with coagulation and fibrinolysis contribute to HCC in some way (Table I).

Table I. A list of studies on the relationship between the coagulation system and HCC.

Table I.

A list of studies on the relationship between the coagulation system and HCC.

Factors	Author (year of publication)	Functions in hepatocellular carcinoma (HCC) (Ref.)
TF/FVII/PAR2	Dupuy et al (2003);	TF promoted tumoral angiogenesis (20);
	Poon et al (2003);	TF promoted angiogenesis and invasiveness (21);
	Panasiuk et al (2007);	TF contributed to degree of tumor differentiation, tumor size (22);
	Zhou et al (2011);	TF promoted genesis, invasion and metastasis (18);
	Tian et al (2011);	TF promoted hepatic metastasis of CRC (23,24);
	Li et al (2012);
	Chen et al (2014);	TF/FVII/PAR2 inhibited autophagy (28);
	Tsai et al (2015);	FVII/PAR2 promoted migration (27);
	Huang et al (2019);	TF promoted growth (19);
FVIII	Zhuang et al (2019);	Promoted lung metastasis (31);
FXII	Srivastava et al (2012);	Promoted angiogenesis and progression (33);
uPA/uPAR	Sun et al (2008);	Enhanced the invasion and migration (36,37);
	Cao et al (2013)
	Tsai et al (2019);	Predicted poor prognosis (8);
	Lee et al (2008);	Promoted invasion (40,41);
	Sun et al (2009);
	Nakagawa et al (2014);	Promoted progression (83);
	Wei et al (2019);	Promoted intrahepatic metastasis (9)
PAI-1	Weng et al (2010);	Leaded to higher susceptibility and pathological development of HCC (45);
	Chen et al (2015);	Promoted hepatic metastasis of CRC (46);
	Divella et al (2015);	Predicted poor prognosis of HCC patients undergone TACE (47);
	Wang et al (2016);	Suppressed invasiveness and motility (48);
PAI-2	Zhou et al (2013);	Low expression predicted portal vein thrombosis and poor prognosis (44);
	Jin et al (2019);	Inhibited invasiveness (49);
DCP	Tang et al (2003);	Correlated with intrahepatic metastasis and recurrence (54);
	Suzuki et al (2005);	Promoted growth (55);
	Gao et al (2008);	Promoted angiogenesis (56);
	Pang et al (2008);	Elevated in serum of HCC patients (50);
	Yue et al (2011);	Promoted growth and metastasis (57);
	Kudo et al (2020);	Predicted poor prognosis (51,52)
	Li et al (2019);
TM	Suehiro et al (1985);	Inhibited portal vein thrombosis and intrahepatic metastasis (63);
	Huang et al (2010);	Inhibited migration (64);
Thrombin	Xue et al (2010);	Predicted poor prognosis (66);
	Zacharski (2002)	Inhibitors of thrombin suppressed metastasis of HCC (10)

[i] TF, tissue factor; FVII, coagulation factor VII; FVIII, coagulation factor VIII; FXII, coagulation factor XII; uPA, urokinase plasminogen activator; uPAR, urokinase plasminogen activator receptor; PAR2, proteinase-activated receptor 2; CRC, colorectal cancer; PAI, plasminogen activator inhibitor; TACE, transcatheter arterial chemoembolization; TM, thrombomodulin; DCP, Des-γ-carboxy prothrombin.

Regulation of the coagulation and fibrinolytic systems in HCC

Clotting factors and HCC

Coagulation factor VII (FVII) combines with TF to initiate the exogenous coagulation signaling pathway, which activates the coagulation cascade and platelets (17). Upregulation of TF has been identified in numerous types of cancer, including HCC. Zhou et al (18) demonstrated that both plasma and tissue samples from patients with HCC exhibited high levels of TF, and this promoted invasion and metastasis via regulation of the AKT/ERK-EGFR signaling pathway (19). Mouse models and clinical follow-up revealed that TF expression is associated with angiogenesis and invasiveness of HCC (20,21). The plasma levels of TF in patients with HCC are associated with the degree of tumor differentiation, tumor size and occurrence of hepatic cirrhosis (22).

Hepatic metastasis of colorectal cancer (CRC) is closely associated with TF expression, and TF knockdown inhibits hepatic metastasis in vitro and in vivo via activation of apoptosis and autophagy (23,24). Furthermore, a TF-FVIIa inhibitor (FFRFVIIa) was found to be able to attenuate hepatic metastasis of CRC (25). Neaud et al (26) revealed that TF pathway inhibitor-2 could induce invasion of HCC cells following binding to the TF-FVIIa complex, and this could be inhibited in a dose-dependent manner by an antibody against human factor VII.

FVII/proteinase-activated receptor 2 (PAR2) promotes migration of HCC cells via activation of the MEK/ERK, TSC complex subunit 2 and mTOR signaling pathways (27). Additionally, Chen et al (28) demonstrated that TF, FVII and PAR2 increased the invasion and migration of HCC cells via mTOR signaling to inhibit autophagy. Furthermore, Lin et al indicated that the four common polymorphisms in the promoter region of the FVII gene (−122T/C, −323ins10-bp, −401G/T and −402G/A) had no effects on the incidence, survival and recurrence in patients with HCC (29).

Coagulation factors VIII (FVIII) and XII (FXII) may also contribute to HCC to some extent. Dihydrodiosgenin was found to inhibit the lung metastasis of mouse HCC cells by decreasing FVIII secretion via the PI3K, MAPK and NF-κB signaling pathways, and this led to anoikis of HCC cells (30,31). Astrocyte elevated gene-1 induced marked upregulation of FXII levels, and increased angiogenesis and progression of HCC (32,33).

uPA system and HCC

uPA, uPA receptor (uPAR), and plasminogen activator inhibitor (PAI)-1 and PAI-2, are four components of the uPA system. All of them may contribute to the metastasis of HCC.

Plasminogen can be converted to plasmin by uPA, and promotes angiogenesis and tumor growth, degrading the extracellular matrix (ECM) and activating pro-MMPs (34). Matrix metalloproteinases (MMPs), particularly MMP-2 and MMP-9, are able to degrade type IV collagen (an ECM component) to promote invasion and metastasis of cancer (35). Previous studies have demonstrated that uPA could activate MMP-2/MMP-9 to enhance the invasion and migration of HCC (36,37).

Higher expression levels of uPA have been observed in aggressive HCC cells compared with levels in HCC cells with low levels of invasiveness (9). Patients with HCC, particularly those with a portal tumor embolus and metastasis, also exhibit high levels of uPA and uPAR. The positive rates of uPA and uPAR in patients with HCC were found to be 72 and 86%, respectively, and a positive rate of uPAR of 100% was observed in patients who died within 2 years after surgery (8). Phosphorylation of ERK1/2 and PI3K/Akt, and activation of the protein tyrosine kinase 2 (FAK), NF-κB and STAT3 signaling pathways may contribute to these processes (38,39). Hepatocyte growth factor (HGF) was demonstrated to promote the invasion of HCC cells by enhancing the levels of uPA and uPAR. Monoclonal antibodies against uPAR were found to inhibit tumor cell invasion mediated by HGF in a dose-dependent manner (40). Bicyclol inhibited the invasiveness of HCC cells by decreasing the expression levels of uPAR (41). A novel inhibitor of uPA, serine peptidase inhibitor Kazal type 13, also inhibited intrahepatic metastasis of HCC in vivo (9).

PAI-1 and PAI-2 are regulators of uPA and uPAR involved in activating plasminogen, initiating signal transduction and inducting cell chemotaxis (42). High PAI-1 levels in tumors are associated with poor prognosis; however, the PAI-2 level may be associated with a beneficial prognosis (43,44). In a Taiwanese population, PAI-1 genotypes were considered as a critical factor which led to higher susceptibility and pathological development in patients with HCC (45). Furthermore, patients with CRC with liver metastasis were demonstrated to have higher plasma levels of PAI-1, and the levels were associated with tumor differentiation, tumor size, Duke's stage and lymphatic metastasis (46). Elevated plasma levels of PAI-1 have been suggested to be a predictor of unfavorable prognosis in patients with HCC with serpin family E member 1 4G/4G polymorphism undergoing transarterial chemoembolization (47).

Other studies have made opposite observations regarding the effects of PAI-1 on HCC. Wang et al (48) revealed that berberine could trigger cell apoptosis by generating reactive oxygen species, and could inhibit the migration and invasion of HCC cells. This may be associated with the upregulation of PAI-1, which could decrease the expression levels of cyclooxygenase-2 (Cox-2), NF-κB, uPA and uPAR via inactivation of the p38 and ERK1/2 signaling pathways (48).

There is limited research on PAI-2 in HCC. PAI-2 was found to inhibit invasion of HCC MHCC97-H cells via uPA- and retinoblastoma/transcription factor E2F1-associated mechanisms (49).

Des-γ-carboxy prothrombin (DCP) and HCC

DCP is produced in HCC cells in the absence of vitamin K or the presence of vitamin K antagonists, and may stimulate growth, invasion and metastasis of HCC. Previous research has indicated that 44–81% of patients with HCC exhibit elevated serum levels of DCP (50). DCP is widely used as a serologic tumor marker for the diagnosis and follow-up of HCC (51). Although its sensitivity is lower than that of α-fetoprotein (AFP), it has a higher specificity than AFP. A number of studies have suggested that higher levels of DCP may be associated with poor tumor behavior and prognosis of patients with HCC (52,53). Additionally, positive serum DCP expression is closely associated with vascular invasion, intrahepatic metastasis, TNM stage and recurrence in patients with HCC (54).

DCP induces the proliferation of HCC cells via activation of the MET proto-oncogene receptor tyrosine kinase (Met)-Jak family tyrosine kinases-STAT signaling pathway (55). Furthermore, binding of DCP to c-Met can activate various downstream effectors, such as autophosphorylation of EGFR, to increase the proliferation and angiogenesis of HCC (56). Yue et al (57) suggested that DCP could increase the activity of MMP-2/MMP-29 via activation of the ERK1/2-MAPK signaling pathway to promote the growth and metastasis of HCC.

It is crucial for DCP generation to decrease vitamin K uptake through cytoskeletal changes in the process of epithelial-to-fibroblastoid conversion induction by chemical methods (58). Hypoxia can induce the production of DCP in HCC cells in this way (59). Vitamin K2 can inhibit the generation of DCP, and DCP levels can also be decreased by vitamin K2 analogues in patients with HCC (60,61).

Other factors and HCC

In addition to the aforementioned components of the coagulation and fibrinolytic systems, there are other factors that also contribute to HCC to a certain extent.

Thrombomodulin (TM) acts as a natural anticoagulant factor and is present at high levels in endothelial cells to maintain unobstructed circulation. TM has been demonstrated to serve roles in inflammation, thrombosis and carcinogenesis (62). A study of TM in 141 patients with HCC who underwent surgery suggested that TM may inhibit tumor cell invasion to the portal vein and prevent intrahepatic metastasis (63). Furthermore, silencing of TM expression markedly increased the migration of HCC cells by decreasing E-cadherin levels and increasing zinc finger E-box binding homeobox 1 (ZEB1) levels (64).

Thrombin is a serine protease and serves multiple roles in coagulation. A previous study revealed that thrombin could induce changes in osteopontin (OPN) activity (a potential therapeutic target for inhibiting HCC metastasis) (65). Furthermore, another study demonstrated that these changes served a key role in the aggressive phenotype of HCC mediated by OPN via activation of integrin β1-FAK, and thrombin was associated with poor prognosis in HCC (66). Several thrombin inhibitors available for clinical treatment, including non-specific anticoagulants and thrombin-specific inhibitors, have been demonstrated to inhibit metastasis in experimental models (10).

Role of the coagulation system in resistance of HCC cells to anoikis

Anoikis is critical to ensure that cells contact the ECM appropriately and to limit the invasiveness of cancer. However, numerous cancer cells can survive by suppressing anoikis to stimulate tissue invasion and to confer metastatic abilities to cancer cells (67). Resistance to anoikis is a prerequisite for the metastasis of epithelial cancer cells (68,69). As a typical malignant tumor derived from epithelial cells, HCC obtains the ability of anoikis resistance to survive in the bloodstream, and to subsequently metastasize and become more resistant to anticancer agents (70).

Various coagulation components may be involved in the anoikis resistance of HCC. Cellular communication network factor 3 downstream genes TF and thrombin, which are positively associated with the malignant phenotype of HCC and vascular thrombosis, prolong survival time and improve pulmonary metastasis in vivo. Mechanistically, ERK and NF-κB signaling are activated in order to maintain cell survival by inhibiting anoikis (71).

Dihydrodiosgenin inhibits lung metastasis of mouse HCC cells by attenuating the adhesion of cancer cells and platelets to endothelial cells by decreasing FVIII production. The direct antitumor activity may involve the PI3K, MAPK and NF-κB signaling pathways, which lead to anoikis of HCC cells (30,31).

Additionally, various coagulation-associated factors contribute to anoikis or have anti-anoikis effects in other types of cancer. Coagulation factor IXa was found to attenuate cell adhesion to the matrix and induces anoikis of epidermal cancer cells by increasing MAPK levels (72). Versteeg et al (73) observed that the TF-FVIIa complex enhanced invasion and migration of baby hamster kidney cells via inhibition of anoikis, which was mediated via the PI3K and MAPK signaling pathways.

The liver, unlike other organs, is the main site of coagulation and fibrinolytic factor production. Acquisition of anoikis resistance is a prerequisite for cancer metastasis. In the present review, it was speculated that coagulation- and fibrinolysis-related factors may contribute to intrahepatic metastasis of HCC via resistance to anoikis (Fig. 2).

Figure 2. Signaling pathways of the coagulation and fibrinolytic systems contribute to HCC. The majority of the components associated with the coagulation and fibrinolytic systems contribute to HCC via certain signaling pathways. These signaling pathways are not biologically divided. Indeed, they are connected to each other, and are associated with angiogenesis, anoikis resistance and metastasis of HCC cells. HCC, hepatocellular carcinoma; TF, tissue factor; FVIIa, coagulations factor VIIa; FXII, coagulation factor XII; FVIII, coagulation factor VIII; FIXa, coagulation factor IXa; PAR2, proteinase-activated receptor 2; EGFR, epithelial growth factor receptor; uPAR, urokinase plasminogen activator receptor; DCP, Des-γ-carboxy prothrombin; COX-2, cyclooxygenase-2; ROS, reactive oxygen species; PAI, plasminogen activator inhibitor; MMP, matrix metalloproteinase; STAT3, signal transducer and activator of transcription 3; VEGF, vascular endothelial growth factor; mTOR, mammalian target of rapamycin; TGF, transforming growth factor; PI3K, phosphatidylinositol 3-kinase; MAPKs, mitogen-activated protein kinases; PLCγ, Phospholipase C-γ; KDR, kinase domain region receptor; NF-κB, nuclear factor-kappaB.

uPA system and immunotherapy for HCC

Immunotherapy, particularly immune checkpoint therapy targeting programmed death-ligand 1 (PD-L1), cytotoxic T-lymphocyte associated protein 4 (CTLA-4), T-cell immunoglobulin mucin family member 3 (TIM3) and lymphocyte activating 3 (LAG3), serves an important role in tumor treatment. However, most solid tumors only have a 15–30% effective response rate to immune checkpoint inhibitors (74). Additionally, anti-programmed cell death protein 1 (PD1) therapy has been demonstrated to only have a 14.3% overall response rate for the treatment of patients with advanced HCC who have been previously treated with sorafenib (75). Studies have suggested that tumor PD-L1 levels may be associated with the response to PD-L1/PD1 blockade (76,77). Previous studies have revealed that low expression levels of PD-L1 in numerous types of HCC could be induced by stimulation of interferon (INF)-γ, targeted therapies and MYC activation, which decreased antitumor immunity and enhanced HCC progression (78–81).

Nakagawa et al (82) found that major urinary protein (MUP)-uPA mice with high expression levels of uPA specifically in the hepatocytes could develop nonalcoholic steatohepatitis-like disease that spontaneously progressed to HCC (82). They suggested that HCC progression in MUP-uPA mice depended on the numbers of immunosuppressive IgA+ plasma cells. Notably, IgA+ plasma cell levels most likely depend on PD-L1 and interleukin (IL)-10 expression (83). This indicated that uPA may be associated with PD-L1 levels in HCC. It would be beneficial to further increase the efficacy of immune checkpoint inhibitors in patients with HCC if coagulation and fibrinolytic components contribute to the expression of checkpoint molecules.

Summary and outlook

The majority of patients with tumors present with hypercoagulability, and increasing evidence has demonstrated that the coagulation and fibrinolytic systems are associated with tumor progression. A common characteristic of HCC is intrahepatic metastasis, which leads to a poor prognosis of patients with HCC. Studies have demonstrated that coagulation and fibrinolytic factors may regulate the metastasis of HCC in two ways. First, by conferring anoikis resistance to HCC cells; second, by promoting HCC cell escape from immune attacks. In conclusion, it not only protects HCC cells themselves in terms of survival, but also changes the tumor immune microenvironment, both of which eventually lead to intrahepatic metastasis of HCC.

Various treatment methods currently exist for advanced HCC. Molecular targeted therapy and immunotherapy have become popular research topics. However, targeted therapy or immunotherapy only show efficacy in certain patients with HCC, while they have poor efficacy or are even ineffective in the majority of patients. The authors of the present review are currently registering a relevant clinical trial and establishing related animal models, aiming to improve the efficacy of immunotherapy by changing the coagulation and fibrinolysis status of patients with HCC and exploring a specific molecular target, in order to eventually improve the prognosis of patients with HCC.

Acknowledgements

Not applicable.

Funding

The present study was supported by the National Natural Science Foundation of China (81670594, 81470791, 81376597); the Basic Research Innovation Group Program of Gansu Province (1606RJIA328); the Scientific Research of Health Services Program of Gansu Province (GSWSKY2017-09); the Talent Staff Fund of the Second Hospital of Lanzhou University (ynyjrckyzx2015-1-01); Talents Innovation and Entrepreneurship Program of Lanzhou City (2017-RC-62); the Cuiying Scientific and Technological Innovation Program of Lanzhou University Second Hospital (CY2017-ZD01); Fundamental Research Funds for the Central Universities (lzujbky-2017-79); Key Project of Science and Technology in Gansu Province (19ZD2WA001).

Availability of data and materials

All information included in this Review is documented by relevant and current references.

Authors' contributions

XL, BG and BW wrote the manuscript; XL, AL and HC designed the paper; ZF, YM and HL revised the manuscript. All authors read and approved the final manuscript and agreed to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved.

Ethics approval and consent to participate

Not applicable.

Patient consent for publication

Not applicable.

Competing interests

The authors declare that they have no competing interests.

References