Keywords

1 Introduction

The plasminogen-plasmin system is involved in not only the regulation of hemostatic balance but also a wide range of biologic processes. These include embryogenesis, development, wound healing, cell proliferation and migration. As such, the system plays an important functional role in both physiologic and pathologic conditions. When this system was first discovered, it was named the fibrinolytic system. However, with the realization that fibrin is not the only substrate for the active enzyme plasmin, the term plasminogen-plasmin system is more appropriate and will be used here. This chapter will review our current understanding of how components of this system affect tumor growth and metastasis.

2 Historical Background

The first association between the pathology of cancer and fibrin must be credited to Billroth [1], who first described the presence of tumor cells within a thrombus. Later, Iwasaki found that tumor cells within the vascular channels of recanalizing thrombi were viable [2]. The significance of fibrin within tumors was again noted by O’Meara and Jackson [3], who postulated that. if unresolved, the fibrin would induce further tumor growth and vascular proliferation. Subsequent observations by others led to experimental and clinical trials with anticoagulants [4], defibrinating agents [5], and thrombolytic agents [6]. Though the results were controversial, experimental tumor metastases were found to be enhanced by increasing the fibrin content in tumors through the use of antifibrinolytic agents, or by induction of hyperfibrinogenemia [7]. In the ensuing years, the role of the plasminogen-plasmin system gained attention not only in the pathogenesis of bleeding and thrombotic complications in the cancer patient, but also in tumor growth and metastasis [8].

3 Plasminogen-Plasmin System

The precursor of the active protease plasmin is plasminogen [9–17]. This is a single-chain glycoprotein of 92 kDa synthesized mainly in the liver. It is present in plasma and extracellular fluids at a concentration of 1–2 µM, with a biologic half-life of 2.2 days. In the native form, the amino-terminal is occupied by glutamic acid. On activation by a plasminogen activator or by plasmin, proteolytic cleavage at Lys-77 and Lys-78 results in the formation of Lys-plasminogen (Fig. 4.1). Further cleavage of the Arg-560 and Val-561 peptide bond results in the formation of a two-chain plasmin held together by two disulfide bonds.

Fig. 4.1
figure 4_1_152258_1_En

The plasminogen-plasmin system

Full size image

Its structure contains five triple-looped structures with three disulfide bonds known as kringles. These kringles are involved in binding of plasminogen to cell surfaces and to fibrin. Of interest is that the structure of a potent anti-angiogenic protein, angiostatin, is identical to the first four of the five kringles.

The lysine at the amino-terminal of plasminogen functions as the binding domain for many proteins, including fibrin, α2-antiplasmin, thrombospondin and the plasminogen receptor annexin II. It also enables plasminogen to bind to specific lysine binding sites on many cell surfaces. Furthermore, plasminogen activation is blocked by lysine binding to ε-aminocarboxylic acids (epsilon amino-caproic acid, tranexamic acid). These are clinically useful antifibrinolytic agents.

Plasminogen has no proteolytic activity prior to conversion to plasmin. On the other hand, plasmin is a serine protease with broad substrate specificity, which includes fibrin, fibrinogen and extracellular matrix (ECM) proteins such as laminin and fibronectin, either directly or indirectly through the activation of latent metalloproteinases [8, 17, 18]. Thus, it is an intermediary protease with a wide range of functions in health, such as tissue remodeling and wound healing, and involvement in pathologic processes including tumor growth and metastasis. Other substrates of importance include the pro-forms of growth factors, which can be cleaved and activated by plasmin. Plasmin can also proteolyse specific cleavage sites on plasminogen to generate angiostatin.

3.1 Plasminogen Activators

Though there are many proteases derived from bacteria, fungus, insects and other animals that can proteolyse plasminogen, only two activators are present in man.

The first one is tissue-type plasminogen activator (tPA) [19–21]. It is a 70-kDa glycoprotein which, under physiologic conditions, is synthesized mainly by endothelial cells. It is responsible for maintaining vascular patency in response to intravascular fibrin formation. However, it is also produced by neurons, keratinocytes, melanocytes and various tumor cells. Observations in transgenic mice with tPA null−/− have now shown that, in addition to maintaining vascular patency, tPA participates also in neuronal development and neurologic functions [22].

The resting plasma level of tPA is low, around 5 ng/mL, but large amounts can be released from endothelial cells under a variety of circumstances. It is then quickly bound to the circulating inhibitor PAI-1. A lesser amount is also bound to α2-macroglobulin. It is then rapidly removed from the circulation by the liver with a plasma half-life of around 5 min.

The source of the circulating tPA is believed to be the vascular endothelium. Earlier observations from our laboratory indicated that fibrinolytic activity can be released from the vascular wall by ischemia [23, 24], serotonin [25] and other vasoactive stimuli [23]. This fibrinolytic activity was later shown to be derived mostly from tPA and, to a lesser extent, from uPA. However, recently, there is evidence that tPA present in the neuronal terminals in the autonomic nervous system can also be released through the vascular wall [26]. This new finding explains our earlier observation that stimulation of a vessel wall can release fibrinolytic activity from a vessel located distally to the site of stimulation, indicating that the stimuli is transmitted via perivascular sympathetic nerves [23].

The “finger” domain present at the amino-terminal enables tPA to have a high affinity for fibrin, thus making tPA a more efficient thrombolytic agent than uPA. In addition to its high affinity for fibrin, tPA also binds to extracellular matrix (ECM) proteins, including laminin and fibronectin, and the mannose-6-phosphate/insulin-like growth factor. Recently, a cellular surface receptor for both tPA and plasminogen, termed annexin II, was found in endothelial cells, macrophages and certain tumor cells [27, 28]. The close proximity of these two ligands on this dual receptor enhances plasminogen activation. The high expression of annexin II in acute promyelocytic leukemia and other malignant conditions may explain the high bleeding risk in these disorders [29].

The second plasminogen activator in man is urokinase-type plasminogen activator (uPA). This fibrinolytic enzyme was first discovered in urine, hence its name urokinase. It originates from kidney cells [30, 31]. When first released, it is in a single-chain form, pro-urokinase, which is then rapidly converted by plasmin or kallikrein to the two-chain form connected by a disulfide bond. The single-chain glycoprotein has a molecular mass of 53 kDa. Both forms are fibrinolytic and are used as thrombolytic agents, but the single-chained uPA has a higher affinity for fibrin. The tertiary structure of uPA is composed of the amino-terminal fragment (ATF) that contains a growth factor domain, as well as a kringle domain. Both the single-chain and the two-chain uPA binds to the uPA receptor (uPAR) via the ATF, forming a uPA-uPAR complex [32]. The complex form facilitates plasminogen activation by uPA.

In addition to being a potent activator of plasminogen, uPA also directly activates procollagenase. This allows it to exert a regulatory effect on cell migration as well as tumor growth and metastasis.

uPA is secreted by many other cell types including endothelial cells and tumor cells in addition to kidney cells. The role of tumor-derived uPA in tumor growth and metastasis has been the subject of many studies and will be reviewed in a later section of this chapter.

uPAR is anchored to the cell surface by its glycosyl-phosphatidylinositol (GPI) domain at the C-terminal, and released by phosphatilyl-inositol specific phospholipase C [33]. On the cell surface, uPAR has a high affinity for uPA, and can also bind uPA which had been inactivated by PAI-1, forming a uPA-uPAR-PAI complex. This complex is rapidly internalized [34], a process facilitated by several members of the low density lipoprotein receptor (LDLR) family [35]. Following internalization, the uPA-PAI-1 portion of the complex is degraded while uPAR is recycled and emerges at a different site on the cell surface [34]. This process is believed to be directional and important in cell migration.

3.2 Inhibitors of the Plasminogen-Plasmin System

Inhibitors of plasmin are α2-antiplasmin (α2-AP), and α2-macroglobulin. α2-AP belongs to the family of serpins (serine protease inhibitors) [36]. Plasmin generated in circulating blood binds to a lysine binding site on α2-AP and is rapidly inhibited. The level of the plasmin-α2-AP complex in blood is often used as an indicator of the intensity of fibrinolytic activity. α2-AP also cross-links with the α-chain of fibrin, preventing the latter’s proteolysis by plasmin. When excess circulating plasmin in blood has saturated all the available α2-AP, a slower acting inhibitor α2-macroglobulin acts as a second line of defense. In addition, there are other serine protease inhibitors including antithrombin, α1-antitrypsin, α1-antichymotrypsin, inter-α-trypsin inhibitor and C-1 inactivator. The recently discovered thrombin activatable fibrinolytic inhibitor (TAFI) inhibits plasmin as well as both plasminogen activators, tPA and uPA.

4 Plasminogen Activator Inhibitors (PAIs)

4.1 Plasminogen Activator Inhibitor Type 1 (PAI-1)

PAI-1 is a member of the serine protease inhibitor (SERPIN) family (Table 4.1). It is perhaps the most important component of the plasminogen-plasmin system in the regulation of many physiologic processes and in the pathogenesis of many disorders including cancer [37–40]. PAI-1 is synthesized by the endothelial cells, liver, adipose tissues, vascular smooth muscle cells, and a large number of tumor cells. In addition to being a potent inhibitor of both tPA and uPA, PAI-1 inhibits plasmin directly.

Table 4.1 Characteristics of plasminogen activator inhibitor type 1
Full size table

PAI-1 is stored in the α-granules of platelets. Most of the PAI-1 in platelets is in a latent form, but when the active portion is released into a thrombus or into the ECM, it can exert their effects on fibrinolysis as well as ECM functions.

The plasma level of PAI-1 in resting healthy individuals is around 1 nM, an amount which is two to three times more than needed to inhibit the circulating plasminogen activators. As discussed below, multiple stimuli, including inflammatory and tumor derived cytokines, can further increase this level. The plasma level of PAI-1 is elevated in obesity, metabolic syndrome, type II diabetes, inflammatory states, and cancer [37]. PAI-1 levels are often used as a prognostic marker for thromboembolic complications in patients with cancer, type II diabetes and veno-occlusive disease in the post-bone marrow transplantation setting. PAI-1 is present in many tumor cell types and in the stromal fibroblasts of the tumor microenvironment as well as in tumor associated endothelial cells. In these settings, it may modulate tumor growth, invasion and angiogenesis.

PAI-1 is the principal inhibitor of both uPA and tPA. In addition, it has a high affinity for ECM proteins, especially vitronectin. The interactions between vitronectin, uPA, uPAR and PAI-1 modulate multiple functions of uPA. The binding of PAI-1 to vitronectin stabilizes the adhesion of cells to the ECM. Both of these processes are essential for cell migration. PAI-1 bound to the uPA-uPAR complex is internalized by endocytosis. While PAI-1 and uPA are degraded intracellularly, uPAR is secreted by the cell and recycled. This process facilitates the propelling action observed in cell migration and tissue remodelling. In addition, PAI-1 binds to glycoaminoglycans, and to low density lipoprotein receptors (LDL-R) [38]. LDL-R facilitates the internalization of the PAI-1-uPA-uPAR complex.

PAI-1 added to cell cultures inhibits apoptosis of both normal vascular smooth muscle cells and tumor cells, and may thus contribute to tumor proliferation and to angiogenesis [39, 40].

4.2 Regulation of PAI-1

As the PAI-1 gene is expressed in almost every cell type in the body, transcription of this gene is regulated by numerous signals generated by cellular responses to various stimuli [41, 42] (Table 4.1). These responses include inflammatory cytokines such as TNFα and IL-1, growth factors, including TGFβ1, EGF, and FGF, hormones such as insulin and glucocorticoids, angiotensin II and angiotensin IV [43] and hypoxia-inducing factor and reactive oxygen species. Through these pathways, PAI-1 is up-regulated in obesity and metabolic syndrome, type II diabetes, hypertension and many types of cancer. These observations led to the concept that PAI-1 not only plays a major role in thrombogenesis by inhibiting fibrinolysis, but is also involved in the pathogenesis of many other disorders by its modulation of cellular interactions.

Several anti-PAI-1 agents are being developed for possible therapeutic use in cancer and other disorders affected by PAI-1 [3, 37].

4.3 Plasminogen Activator Inhibitor Type 2 (PAI-2)

PAI-2 is synthesized by the placenta, monocytes [44, 45], eosinophils and keratinocytes as well as by ovarian tumors and myeloid leukemic cells. PAI-2, as in the case of PAI-1, is an inhibitor of both uPA and tPA [46]. Its expression by the placenta is believed to contribute to the increased prothrombotic risk in late pregnancy. Also, like PAI-1, it binds the uPA-uPAR complex on the cell surface and is then internalized. However, unlike PAI-1, most PAI-2 exists within the cell in the cytosol and only a small fraction is secreted. The plasma level is barely detectable. While the secreted portion takes part with PAI-1 in the inhibition of the plasminogen activators, the main functions of PAI-2 within the cell are not clear. One known intracellular function is the protection of cells against TNFα-mediated apoptosis [47]. It is notable that the PAI-2 gene is located in chromosome 8, less than 300 mbp from the apoptosis-inhibiting BCL-2 gene. In follicular lymphoma with the t(14;18) translocation, BCL-2 is over-expressed, resulting in inhibited apoptosis of the lymphoma cells [48] . However, it is not known whether PAI-2 has a similar effect on lymphoma.

There are a number of observations of a tumorogenic effect of PAI-2. Studies with oligonucleotide microarrays indicate there is at least a 12-fold increase in PAI-2 genes in ovarian serous papillary carcinoma compared to normal ovarian tissues [49]. The level of PAI-2 has been correlated with a poor prognosis for ovarian, and colorectal carcinomas, while low expression in epithelial carcinoma such as head and neck squamous cell carcinoma signifies invasion [50]. In contrast, a high PAI-2 level indicates a favorable prognosis in breast carcinoma. Recently, high expression of PAI-2 was found in the gastric mucosa of patients with Helocobacter pylori infection, suggesting another possible link between the proliferative and apoptotic inhibitory actions of PAI-2 and the ultimate gastric cancer formation [51].

4.4 Thrombin-Activatable Fibrinolytic Inhibitor (TAFI)

Thrombin-activatable fibrinolytic inhibitor, also known as procarboxypeptidase U, is a 55-kDa carboxypeptidase synthesized in the liver that has been shown to play an important role in fibrinolysis. It is initially produced as a proenzyme and is converted to an active, zinc ion dependent carboxypeptidase B-like enzyme (TAFIa) through cleavage at the Arg92-Ala93 bond. This process can be mediated by trypsin or plasmin. However, these activators are much less efficient than thrombin. Thrombin conversion of TAFI to TAFIa is accelerated 1250-fold in the presence of thrombomodulin [52], and the activation is calcium-ion dependent. TAFIa production occurs in two peaks, with the first peak occurring shortly after initiation of clot formation and the second stimulated by thrombin generation. The half-life of the enzyme is approximately 10 min and therefore the first increase in concentration is transient. The second peak is prompted by plasmin formation and unlike the first, has little effect on fibrinolysis [53].

TAFIa is able to attenuate fibrinolysis by preventing plasmin formation through removal of carboxyterminal arginine and lysine residues from fibrin and fibrin cleavage products. This blocks the binding of tPA and plasminogen, thereby interfering with the positive feedback of plasmin formation [54]. Removal of the arginine and lysine residues from fibrin also removes the inhibitory effect fibrin has on antiplasmin. Fibrinolysis is then prevented by the greater resulting influence of antiplasmin on free plasmin [55]. Plasmin may also be directly inhibited by TAFI, to impair fibrinolysis further. The increased amounts of TAFIa generated by activation of the intrinsic system continue to propagate the downregulation of fibrinolysis. There have been no reported inhibitors of TAFIa to date, and the activity appears to decay spontaneously.

Due to its effect on fibrinolysis, levels of TAFIa may correspond to an increased risk of thrombosis or bleeding. Increased levels have been shown to be a weak risk factor in incident (twofold increased risk) [56] and recurrent (twofold increased risk) [57] venous thromboembolism. Concurrent elevations in factor VIII may enhance this thrombogenic risk. On the other hand, low TAFI activity is seen in patients with acute promyelocytic leukemia, and this finding may in part explain the hyperfibrinolysis and propensity for hemorrhage seen in this disease [58]. TAFI levels have also been investigated in a number of other disease states where perturbations of clotting may play a central role. Examples include demonstration of increased TAFI levels in ischemic stroke, in patients with angina, and in men requiring coronary artery bypass grafting [59–61].

It is unclear what role, if any, TAFI has in malignant disease states. A mouse model failed to show any affect of TAFI deficiency on growth or metastasis of different tumor cell types [62]. Malignancy may increase TAFI expression through a cytokine-mediated process. Theoretically, increased levels of TAFI may promote growth and spread of tumor cells through intra-tumoral fibrin deposition, and may accentuate the several prothrombotic features of various malignant states. Clinical data in humans supporting these possibilities is lacking.

5 Role of the Plasminogen – Plasmin System in Tumor Growth and Metastasis

The relationship between the plasminogen-plasmin system and tumor biology is complex [8, 42, 63, 64] (Fig. 4.2). It involves several important steps as shown in Fig. 4.2. The major ones are cell proliferation, apoptosis, cell migration and invasion and angiogenesis. In all these steps, one or more of the components of the plasminogen-plasmin system participate in the process.

Fig. 4.2
figure 4_2_152258_1_En

Plasminogen-plasmin system in tumor growth and metastasis

Full size image

5.1 Tumor Cell Proliferation and Apoptosis

uPA bound to uPAR is mitogenic. On the surface of tumor cells, initiation of intracellular signaling follows the assembly of the amino-terminal fragment of uPA (containing an epidermal growth factor-like domain), and uPAR, along with an ECM protein (such as vitronectin or fibronectin) and the epidermal growth factor receptor, leads to growth stimulation [65, 66]. The same events occur with stromal cells, vascular smooth cells and endothelial cells in the tumor microenvironment. These characteristics of uPA form the basis for a high uPA expression in tumors signifying a poor prognosis. On the other hand, we had shown previously that, in vitro and in animals, high expression of PAI-1, by inhibiting uPA, impairs tumor growth, angiogenesis and metastasis [67]. When PAI-1 was transfected into an aggressive human prostate cancer cell, PC-3, tumors with transfected cells that over-expressed PAI-1 had a slower growth rate in vitro. Tumors in athymic mice given PC-3 clones with high PAI-1 expression were smaller, less metastatic and contained less vasculature. This anti-tumorogenic effect is attributed to the inhibitory action of uPA on tumor invasion, and on uPA activation of plasmin. However, in a different setting, PAI-1 has an opposite effect by being inhibitory to apoptosis. When the stable form of PAI-1 is added to tumor cell cultures, or endothelial cell culture, both spontaneous and induced apoptosis in tumor cells are inhibited [39]. In this respect, PAI-1 enhances tumor growth. PAI-2 added to tumor cell cultures also has an anti-apoptotic action. As tumor growth involves both cell proliferation and apoptosis, PAI-1 may favor tumor growth by inhibiting apoptosis. In certain tumors, a high PAI-1 content signifies a poor prognosis. These observations of stimulated and inhibited tumor growth were confirmed by others [38]. Which of these opposing effects of PAI-1 are acting on the tumor must be dependent on a number of factors, many of which are still not fully understood. PAI-1 injected into animals or added to tumor cell cultures likely inhibits uPA on the tumor cell surface, where the assembly of PAI-1 and the uPA-uPAR complex are internalized, leading to the degradation of uPA. Also, it is likely that the effect may be dose-dependent. One determinant may also be the tumor type. For example, in carcinoma of the breast, the known tumorogenic effect of estradiol may in part be mediated through the down-regulation of tPA, uPA, uPAR and PAI-1 [68].

The activation of pro-forms of growth factors is another way in which plasmin, tPA or uPA can participate in tumorogenesis. One example is that plasmin can activate latent β-FGF. Plasmin has also been shown to activate the propeptide of VEGF-C and VEGF-D, both angiogenic and lymphangiogenic factors [69]. Of interest, tPA can directly activate the latent form of PDGF-CC by proteolytic removal of the CUB domains [70].

5.2 Tumor Cell Migration and Invasion

The regulation of cell migration by uPA, uPAR and PAI-1 in cell migration is important in wound healing as well as in cancer cell invasion, and metastasis. uPA bound to uPAR is present at focal adhesion sites on the cell surface [71, 72]. Locally generated plasmin from the activation of plasminogen at these sites in turn activates latent metalloproteinases and latent growth factors [5]. It is believed that the resulting proteolysis of the extracellular matrix (ECM) frees the cell from its adhesion site allowing cell migration. Since uPA bound to uPAR is a much more potent activator of plasminogen, the location of uPAR determines the direction of the cell migration. There is also another major cellular process, involving vitronectin in the ECM, uPA-uPAR complex, PAI-1 and the recycling of uPAR [34]. Both monomeric and dimeric forms of uPAR are present on the cell surface [33, 73]. Vitronectin in the ECM binds to uPAR, preferentially the dimeric form [74], and affects the distribution of uPAR on the cell surface. The complex of uPA, uPAR and PAI-l binds to vitronectin on the cell surface, resulting in changes in the cell shape and cell adhesiveness to the ECM. With the endocytosis of the complex assembly, de-adhesion occurs. Following the recycling of uPAR, readhesion takes place [33]. These steps regulate cell migration and propulsion.

5.3 Tumor Angiogenesis

The plasminogen-plasmin system is involved in tumor angiogenesis in two aspects [75–79]. First, along with tumor derived cytokines and VEGF, uPA and PAI-1 modulate endothelial cell proliferation. Second, plasmin and plasminogen activators proteolyse plasminogen and, acting with several cofactors, release one or more of the kringle structures. These kringles possess inhibitor effects on tumor angiogenesis, best exemplified by angiostatin. Plasmin, derived from uPA activation of plasminogen, can be reduced by plasmin reductase (phosoglycerate kinase) in the presence of co-factors, including glutathione and cysteine. The disulfide bonds between kringles are further proteolysed to form kringle peptides with potent inhibitory activity against the proliferation of microvascular endothelial cells. Despite encouraging results of tumor reduction in animals, these results were not reproduced in early human clinical trials [79].

6 Role of the Plasminogen-Plasmin System in Thrombotic Complications in Cancer

Though procoagulants and their activation in cancer play a major role in thrombogenesis, changes in fibrinolytic components are also important contributory factors. Evidence for this is provided by in vitro studies as well as observations of inhibited fibrinolysis in vivo. The inhibited fibrinolysis is primarily due to increased in PAI-1 activity in the plasma of cancer patients. Both uPA and tPA are regulated by PAI-1. While uPA acts on cell proliferation and migration and thus plays an important role in cancer progress, tPA, on the other hand, is the body’s defense against extension of intravascular fibrin thrombi, and thus a high PAI-1 level is thrombogenic.

7 The Plasminogen-Plasmin System in Acute Leukemia

Both thrombotic and bleeding complications are linked to perturbation of the plasminogen-plasmin system in acute leukemia [80, 81]. In acute promyelocytic leukemia (FAB: M-3), annexin II, a dual receptor for tPA and for plasminogen [27, 28], is highly expressed in the leukemic cells [29, 82]. Its level is also increased in a small number of patients with AML (FAB: M4-5) or with ALL. Annexin II is a cell membrane surface protein found in endothelial cells, macrophages and several malignant cell lines. It is a protein with a molecular weight of 40 kDa. In addition to binding tPA, it is a co-receptor for plasminogen, with tPA binding at the amino-terminal of the core 1 domain, while plasminogen binding occurs at the lysine binding site in the core 4 domain. The close proximity of the two ligands on the cell surface facilitates their interaction in plasmin generation, with in vitro enhancement of 60-fold. Annexin II on endothelial cell surface has the highest expression on the microvascular endothelial cells in the brain [83]. In acute promyelocytic leukemia (APL), this location of a higher Annexin II expression is of clinical significance. APL cells in the bone marrow and peripheral blood as well as the APL cell line NB4, express both uPA and tPA [84]. After treatment with trans-retinoic acid, the uPA expression is further dysregulated in NB4 cells for 24 h before returning to normal levels as these cells undergo differentiation [85]. Thus, an increased expression of uPA by APL cells, in conjunction with the expression of tPA as well as the presence of annexin II contributes to the excessive fibrinolysis [29]. In addition, TAFI level is low in APL and thus there is less inhibitory control over fibrinolysis [59]. Larger amounts of plasmin were also found to be generated in vitro by the brain endothelial cells [84]. This may provide an explanation for the relatively higher incidence of intracranial hemorrhage in APL. In a recent analysis of the early deaths seen in patients with APL, hemorrhagic complications were found to be the major cause of early deaths, accounting for over 50% of patients with or without ATRA therapy [86–90]. Intracranial bleeding accounts for most of the fatal hemorrhages, with other sites including diffuse pulmonary alveolar and gastrointestinal hemorrhage. In addition to plasmin induced fibrinolysis, elastase and chymotrypsin released by leukemic blasts may also contribute to the impaired hemostasis by proteolysis of von Willebrand factor [91].

8 Clinical Observations of the Plasminogen-Plasmin System in Cancer

As discussed, the plasminogen-plasmin system has potential biological effects extending beyond thrombosis and hemorrhage. Tumor growth and metastasis may be promoted by uPA-mediated enhancement of cell proliferation, adhesion, migration, and degradation of the extracellular matrix. Alterations of this system in cancer could lead to a variety of altered clinical outcomes, which has been demonstrated in a number of cancer types.

The majority of the clinical literature evaluating upregulation of the plasminogen-plasmin system has focused on prognosis through the enhancement of tumor growth and progression. As mentioned in detail previously, this may come about through inhibition of apoptosis and through increased release of TGFβ, FGF2, ILGF-1, and hepatocyte growth factor. Degradation of the extracellular matrix and promotion of cellular adhesion may assist in the development of metastatic disease. These features of the plasminogen-plasmin system have been correlated with its over-expression, leading to adverse outcomes in a number of malignancies (Table 4.2). High uPA and/or PAI-1 levels have been shown to be adverse prognostic markers in breast, colorectal, esophageal, gastric, ovarian, prostate, renal, and endometrial cancers. These findings may prove to be a vital addition to previously known prognostic markers and potentially assist in individualizing cancer treatments.

Table 4.2 Plasminogen-plasmin system: clinical and laboratory features in malignancy
Full size table

The prognostic significance of the plasminogen-plasmin system has best been demonstrated in breast cancer. Levels of uPA and PAI-1 were more predictive of both disease-free and overall survival than ER status and tumor size in a pooled analysis of over 8,000 patients with breast cancer [92]. Increased uPA and PAI-1 were associated with a worse prognosis. Application of this finding to clinical practice was demonstrated in a prospective study of 761 patients [93]. Based on the hypothesis that poor outcome is associated with high PAI-1 and uPA, patients with lymph node negative disease, but with high PAI-1 and uPA, were given adjuvant cyclophosphamide-methotrexate-5-fluorouracil (CMF) chemotherapy. Another large trial of over 3,000 women showed increased levels of PAI-1 and uPA correlated with greater benefit from adjuvant chemotherapy vs those with lower levels [94]. Similar findings are present with regard to adjuvant hormonal treatment, with uPA and PAI-1 negative breast tumors responding better to intervention with tamoxifen than those with high expression, independent of ER/PR status [95].

Application of the prognostic information given by the plasminogen-plasmin system to clinical care has also been extended to other malignant diseases. For example, it has already been shown that higher TAFI levels in lung cancer directly correlate with a more favorable response to chemotherapy [96]. Such observations allow better individualization of cancer care, with administration of more aggressive treatment to those patients who are likely to benefit most. In addition, those who are unlikely to benefit could be spared the toxicities associated with many therapeutic interventions. Additional data is needed in this area to justify its wider application.

Available evidence indicates a pathophysiologic role of the plasminogen-plasmin system in the prothrombotic nature of malignant disease, with high tumor expression of PAI-1 and resultant inhibition of fibrinolysis potentially exacerbating the hypercoagulability associated with malignancy. To date, clinical studies demonstrating this correlation have largely been lacking. Whether components of this system can be used successfully in the treatment or prevention of thrombosis in cancer remains to be established. It is certainly plausible that alterations in the expression of the various components of the uPA system may predict risk for thrombosis in the same way as with disease outcome. This type of information may prove useful in targeting those at highest risk, with greater surveillance and possibly prophylactic treatment for those patients whose thrombotic potential is greatest.

9 Conclusion

Though fibrin was found in cancer tissues as early as the late nineteenth century, the active investigation of the role of the plasminogen-plasmin system in cancer has accelerated only in the past two decades. This has greatly increased our understanding of how the components of this system, especially uPA, uPAR and PAI-1, affect tumor growth, invasion and angiogenesis. Undoubtedly these findings have contributed to the elucidation of the pathogenesis of many forms of malignant disorders. However, little progress has been made in translating the findings from in vitro studies and animal experiments into innovative therapeutic approaches. In experimental tumors in animals, perturbation of uPA and of PAI-1 has been found to impair tumor growth and metastasis, while only a few anecdotal results have been reported in humans. As the pathogenesis of cancer is complex, one would expect that the influence of uPA and PAI-1 is only one part of this process. New agents are being designed to interdict these effects, especially those of PAI-1. Whether they will be effective remains to be determined by clinical trials. Questions to be addressed in future clinical trials will concern the effect of anti-PAI-1 or anti-uPA agents by themselves, or whether effectiveness will require a combination with cyto-reductive measures including chemotherapy or radiation, and in addition, the combination with anti-angiogenic agents or with hormonal therapy wherever applicable.