Tumor cell procoagulant and urokinase expression in carcinoma of the ovary.

BACKGROUND: An association between cancer and increased blood coagulation has been observed for many years. Generally, there is an equilibrium between the coagulation system (fibrin deposition) and the fibrinolytic system (degradation of fibrin by enzymes). However, in malignant disease such as ovarian carcinoma, this equilibrium is disrupted, resulting in the abnormal activation of coagulation or hypercoagulability. Also, evidence indicates that various components of these pathways may contribute to the disorderly characteristics of malignancy, such as proliferation, invasion, and metastasis. PURPOSE: Our purpose was to define the mode of interaction of tumor cells in ovarian carcinoma with both the coagulation (procoagulant-initiated) and fibrinolysis (urokinase-type plasminogen activator-initiated) (u-PA) pathways. METHODS: Studies were performed on acetone-methylbenzoate-xylene-fixed tissue prepared from fresh resected primary tumor specimens from 15 patients with cystic epithelial ovarian carcinoma. None of the patients had received prior treatment. Antibodies were tested on control and tumor tissues in concentrations that provided maximum staining intensity with minimum background staining. Laboratory immunohistochemical techniques used purified, monospecific antibodies to detect coagulant antigens. Tests were performed utilizing antibodies to recombinant human tissue factor; factor VII; factor X; factor XIIIA; high-molecular-weight and low-molecular-weight forms of u-PA; tissue-type plasminogen activator; plasminogen; and the plasminogen activator inhibitors 1, 2, and 3. Monoclonal antibodies used for specific antigen detection included 1-8C6 (fibrinogen), T2G1 (fibrin), and EBM-11 (macrophage-specific). RESULTS: The ovarian tumor cells expressed urokinase-type plasminogen activator in a pattern that was variable in intensity and distribution. Tumor cell plasminogen was not detected. Tumor cells also expressed tissue factor and coagulation pathway intermediates that resulted in local thrombin generation as evidenced by the conversion of fibrinogen (present in tumor connective tissue) to fibrin that was found to hug the surfaces of tumor nodules and individual tumor cells. Detected fibrin could not be accounted for on the basis of necrosis or a local inflammatory cell infiltrate. CONCLUSIONS: These results are consistent with the existence of a dominant tumor cell-associated procoagulant pathway that leads to thrombin generation and hypercoagulability in carcinoma of the ovary. IMPLICATIONS: In ovarian carcinoma the procoagulant pathway may contribute to tumor progression. Clinical trials of therapeutic drugs capable of limiting local coagulability (anticoagulants, protease inhibitors) are indicated in this tumor type.

Occurrence of components of fibrinolysis pathways in situ in neoplastic and nonneoplastic human breast tissue.

The occurrence and distribution of components of fibrinolysis pathways were determined using immunohistochemical techniques applied to 10 cases of primary carcinoma of the breast, normal breast tissue obtained from two patients undergoing reductive mammoplasty, and three cases of benign breast tumors. Tumor cells stained for urokinase- and tissue-type plasminogen activators, plasminogen activation inhibitor-1, plasminogen, and plasmin-antiplasmin complex neoantigen. The tumor connective tissue stained for fibrinogen and its D fragment plasmin digestion product. By contrast, only occasional nonneoplastic duct epithelial cells stained for urokinase- and tissue-type plasminogen activators and there was little or no staining for the other antigens tested. These results are consistent with the existence of local amplification of expression of enzymatically active plasminogen activators, and particularly of urokinase-type plasminogen activator, in situ in primary breast cancer tissue. These features distinguish malignant from benign breast tissue and may modulate neoplastic progression through an effect on tumor cell proliferation, invasion, and metastatic dissemination.

Fibrinogen deposition without thrombin generation in primary human breast cancer tissue.

The occurrence and distribution of components of coagulation pathways in situ were determined using immunohistochemical techniques applied to 10 cases of primary carcinoma of the breast, normal breast tissue obtained from two patients undergoing reductive mammoplasty, and three patients with benign breast tumors. Tumor cells stained for factor X and thrombomodulin but not for tissue factor, factor V, factor VII, or factor XIII. Rare nonneoplastic duct epithelial cells stained for thrombomodulin, but these tissues did not otherwise stain for any of these antigens. Macrophages within the tumor stroma stained for tissue factor, factor VII, and factor XIII but not for factor V or factor X. These features of macrophages were the same in malignant and nonmalignant breast tissue. Fibrinogen was present in abundance throughout the connective tissue in breast cancer but not in nonmalignant tissues. By contrast, no staining was observed using fibrin-specific antibodies. These results suggest that an intact coagulation pathway does not exist in breast cancer tissue and that thrombin capable of transforming fibrinogen to fibrin is not generated in significant amounts in this tumor type. While fibrin is not a feature of the connective tissue stroma in breast cancer, it is conceivable that the abundant fibrinogen present in the tumor connective tissue (and factor XIII present in connective tissue macrophages) might contribute to the structural integrity of breast tumor tissues.

Thrombus imaging: a comparison of radiolabeled GC4 and T2G1s fibrin-specific monoclonal antibodies.

Radioimmunoimaging of experimentally-induced canine thrombi has previously been achieved with iodine-131- and indium-111-labeled (131I and 111In) anti-fibrin T2G1s monoclonal antibody (MAb). We now compare T2G1s to another anti-fibrin MAb, designated GC4, for imaging fresh and aged canine thrombi. GC4 is specific for a neoepitope exposed on fibrin later in the thrombolytic process after plasmin digestion. Femoral venous thrombi were induced in six groups of dogs, each containing three dogs. In two groups, the MAbs were compared when the thrombi were 3-hr or 3-days old at the time of injection, and the dogs were killed at 48 hr. In thrombi 3-hr-old, the GC4/T2G1s concentration ratio averaged 0.53 compared to 1.9 in 3-day-old thrombi. Two groups of dogs with thrombi 1- or 3-days-old were heparinized before MAb injection and were killed at 24 hr. The heparinized dogs with thrombi 1- or 3-days-old had GC4/T2G1s mean ratios of 2.3 and 2.9, respectively. In the unheparinized groups, the corresponding ratios were 1.1 and 1.9. GC4 may be more useful for clinical thrombus imaging than T2G1s because spontaneous venous thrombi are usually several days old at the time of presentation and patients are often heparinized immediately.

Thrombus imaging with indium-111 and iodine-131-labeled fibrin-specific monoclonal antibody and its F(ab')2 and Fab fragments.

We have previously reported successful imaging of fresh (2-4 hr old) and aged (1-5 days old) canine thrombi with 131I-labeled intact monoclonal antibody (MAb) specific for fibrin. We now report thrombus imaging with 131I-labeled F(ab')2 and Fab and 111In-labeled intact MAb, F(ab')2, and Fab. Indium-111-labeled F(ab')2 proved to be the best imaging agent due to less nonspecific binding in the liver than whole IgG. Image quality was improved by the higher administered dose permissible with 111In and its better physical characteristics for imaging, compared to 131I. Immunofluorescence of fresh human histologic sections showed intact MAb and F(ab')2 binding to thrombi, pulmonary emboli, and atherosclerotic plaques, strengthening the feasibility of clinical thrombus imaging.

Localization of blood coagulation factors in situ in pancreatic carcinoma.

Blood coagulation is activated commonly in pancreatic carcinoma but the role of the tumor cell in this activation is undefined. Immunohistochemical procedures were applied to fixed sections of 22 cases of resected adenocarcinoma of the pancreas to determine the presence of components of coagulation and fibrinolysis pathways in situ. Tumor cell bodies stained for tissue factor: prothrombin: and factors VII, VIIIc, IX, X, XII, and subunit "a" of factor XIII. Fibrinogen existed throughout the tumor stroma, and tumor cells were surrounded by fibrin. Staining for tissue factor pathway inhibitor, and plasminogen activators was minimal and inconsistent. Plasminogen activator inhibitors -1, -2, and -3 were present in the tumor stroma, and on tumor cells and vascular endothelium. Extravascular coagulation activation exists associated with pancreatic carcinoma cells in situ that is apparently unopposed by naturally occurring inhibitors or the plasminogen activator-plasmin system. We postulate that such local coagulation activation may regulate growth of this malignancy. These findings provide a rationale for testing agents that modulate the blood coagulation/fibrinolytic system (that inhibit tumor growth in other settings) in pancreatic carcinoma.

Indirect activation of blood coagulation in colon cancer.

Systemic activation of the coagulation mechanism is known to exist in patients with colon cancer. The mechanism of such activation was investigated using immunohistochemical techniques applied to fresh frozen sections of resected primary colon cancer specimens. Tumor cells stained for tissue factor, factor V, and urokinase-type plasminogen activator. Perivascular and intercellular areas stained for fibrinogen and the "a" subunit of factor XIII. Staining was minimal or absent for protein C, protein S, plasminogen activator inhibitors 1-3, factor VII, factor X, and fibrin (the antigenic site on the amino-terminal portion of B beta chain that is exposed following thrombin cleavage of fibrinopeptide B was not detected). The lack of an intact thrombin-generating pathway in situ associated with viable colon cancer cells is consistent with the findings of others that coagulation activation in colon cancer may be triggered by a soluble tumor product that exerts its effect at sites distant from the tumor. These results may explain the absence of clinical responsiveness of colon cancer to antithrombotic drug therapy and may clarify therapeutic strategies for this common tumor.

Expression of tissue factor and tissue factor pathway inhibitor in situ in laryngeal carcinoma.

The blood coagulation mechanism may support tumor progression by several mechanisms including promotion of cell proliferation and angiogenesis. Immunohistochemical procedures were applied to AMeX-fixed sections of twelve cases of squamous cell carcinoma of the larynx obtained at surgical resection to determine the presence and distribution of tissue factor (TF), tissue factor pathway inhibitor (TFPI), other coagulation factors, fibrinogen, and fibrin in situ. TF antigen was present in normal squamous epithelial cells and tumor cells, predominantly in immature tumor cells in the vicinity of the host-tumor interface. Tumor cells stained also for factors VII and X. Staining for TFPI antigen was demonstrated in the connective tissue stroma adjacent to the tumor, in microvascular endothelial cells, and in normal squamous epithelial cells. Fibrinogen and factor XIIIa were distributed throughout the tumor connective tissue stroma. Fibrin (thrombin-cleaved fibrinogen) was detected at the host-tumor interface and along the margins of tumor nodules. Tumor cells in carcinoma of the larynx express a functional, TF-initiated pathway of blood coagulation. Interpretation of these findings together with the results of clinical trials of inhibitors of TF-induced coagulation activation versus effects of inhibitors of TF expression suggest novel approaches to the experimental therapy of laryngeal carcinoma.

Malignant melanoma. Interaction with coagulation and fibrinolysis pathways in situ.

Immunohistochemical techniques applied to fresh frozen sections of metastatic malignant melanoma tissue revealed abundant fibrinogen (or fibrin I) in perivascular areas throughout the tumor connective tissue stroma. Fibrin was readily detected in a focal distribution in the connective tissue around nodules of viable tumor. Staining for D-dimer of cross-linked fibrin (using an antibody that cross-reacted with fragment D of fibrinogen) coincided with staining for fibrin. Diffuse staining of tumor cell bodies was observed for Factor X, and Factor XIII ("a" subunit) was detected in scattered areas of connective tissue throughout the tumors. Factor VII was not detected, and only rare tumor cells stained for tissue factor. These results support the concept that a tumor cell-associated, thrombin-generating pathway exists in situ in malignant melanoma tissue that includes Factor X but neither tissue factor nor Factor VII. By contrast, tumor cell staining was observed rarely for urokinase and to a variable extent for tissue plasminogen activator.

Flow and antibody binding properties of hydrated fibrins prepared from plasma, platelet rich plasma and whole blood.

Previous studies, using cross-linked fibrin prepared from purified fibrinogen, showed low binding of a fibrin-specific monoclonal antibody designated T2G1 (Procyk et al., Blood 77:1469-75, 1991). In this study we investigated the binding of T2G1 and one other antibody to clots prepared from platelet poor plasma (PPP), platelet rich plasma (PRP) and whole blood. In contrast to our previous study, we used unlabelled antibodies and quantitated the level bound by ELISA, measuring antibody concentration in the non-adsorbed fraction. Antibody T2G1 bound 1.35 +/- 0.10 pmol/pmol fibrin (n = 11) to whole blood columns, 1.64 +/- 0.18 (n = 10) to PRP columns and 1.58 +/- 0.13 (n = 8) to PPP columns. The binding of T2G1 to columns made from purified fibrinogen was 0.78 +/- 0.05 pmol/pmol fibrin (n = 15). An antibody to a conformation-dependent epitope on Fragment D (Fd4-7B3) bound in comparable amounts to the different fibrins. Flow data show that whole blood columns, and also, but to a lesser extent those made with plasma, had a higher flow rate, permeability and fiber mass-length ratio than columns prepared from fibrinogen indicating a more coarse fibrin network. These data show that the presence of other proteins and blood cells, similar to what might occur in vivo, not only lead to an increase in the permeability of gels but also allow for better exposure of some epitopes.

Expression of prothrombin fragment 1+2 in cancer tissue as an indicator of local activation of blood coagulation.

Immunohistochemistry was applied to AMeX-fixed tissue sections of 12 adenocarcinomas of the stomach (seven intestinal adenocarcinomas and five diffuse carcinomas), 12 adenocarcinomas of the pancreas (nine ductal adenocarcinomas and three signet ring carcinomas), and 12 squamous cell carcinomas of the larynx obtained at surgical resection to examine the possibility of extravascular activation of blood coagulation in cancer tissues by exploring the in loco patterns of distribution of fibrinogen, a final product of blood coagulation, fibrin, and a by-product of coagulation reactions (prothrombin fragment 1+2). Gastric, pancreatic, and laryngeal cancers exhibited fibrinogen antigen in abundance throughout the tumor stroma. Fibrin was detected along the edges of nests of carcinoma cells and at the host-tumor interface. Prothrombin fragment 1+2 was present in the blood vessels in areas of neoangiogenesis at the host-tumor interface (gastric and pancreatic cancer tissues) and on the tumor cell bodies (pancreatic and laryngeal cancer tissues). The presence of prothrombin fragment 1+2 in cancer tissues appears to be a good indicator of coagulation activation and thrombin generation at the tumor burden.

Presence of vitronectin and activated complement factor C9 on ventriculoperitoneal shunts and temporary ventricular drainage catheters.

OBJECT: The pathogenesis of cerebrospinal fluid (CSF) shunt infection is characterized by staphylococcal adhesion to the polymeric surface of the shunt catheter. Proteins from the CSF--fibronectin, vitronectin, and fibrinogen--are adsorbed to the surface of the catheter immediately after insertion. These proteins can interfere with the biological systems of the host and mediate staphylococcal adhesion to the surface of the catheter. In the present study, the presence of fibronectin, vitronectin, and fibrinogen on CSF shunts and temporary ventricular drainage catheters is shown. The presence of fragments of fibrinogen is also examined. METHODS: The authors used the following methods: binding radiolabeled antibodies to the catheter surface, immunoblotting of catheter eluates, and scanning force microscopy of immunogold bound to the catheter surface. The immunoblot showed that vitronectin was adsorbed in its native form and that fibronectin was degraded into small fragments. Furthermore, the study demonstrated that the level of vitronectin in CSF increased in patients with an impaired CSF-blood barrier. To study complement activation, an antibody that recognizes the neoepitope of activated complement factor C9 was used. The presence of activated complement factor C9 was shown on both temporary catheters and shunts. CONCLUSIONS: Activation of complement close to the surface of an inserted catheter could contribute to the pathogenesis of CSF shunt infection.

Fibrinogen-fibrin transformation in situ in renal cell carcinoma.

The mechanism of coagulation activation in renal cell carcinoma was investigated using immunohistochemical techniques applied to fresh frozen sections of resected primary tumors. Tissue factor antigen was detected in the endothelium of vascular channels within the tumors. Fibrinogen and factor V were distributed diffusely in the perivascular tumor connective tissue. Fibrin was readily detected in a linear pattern along the edges of nodules of viable tumor indicating that thrombin had formed from the interaction of coagulation factors demonstrated previously in renal cell carcinoma tissue. Tissue plasminogen activator was detected in the endothelium of blood vessels in the vicinity of the tumor and urokinase in areas of necrosis but neither were associated with viable tumor cells. These results indicate that thrombin is formed locally in renal cell carcinoma tissue that transforms fibrinogen to fibrin. There also appears to be a net deficit in fibrinolysis in situ in this tumor. We postulate that these conditions might contribute to stabilization and progression of renal cell carcinoma and that clinical trials of antithrombotic agents are justified in this tumor type.

Association between decreased pulmonary endothelial cell thrombomodulin and local fibrin deposition in pneumonia.

Thrombomodulin (TM) plays an important role in anticoagulation by forming a complex with thrombin, which subsequently activates protein C. TM is inactivated and downregulated by inflammatory cell mediators. This study examined whether bronchopneumonia is associated with changes in TM immunoreactivity, and whether a decrease in TM is accompanied by evidence of hypercoagulability, i.e. local deposition of fibrin. Double antibody staining for TM and fibrin was performed on lung tissue sections from patients who had died of pneumonia and from patients who had died rapidly, secondary to trauma. Inflammatory changes were assessed histologically and immunohistochemically using antibodies against interleukin-1alpha, tumor necrosis factor-alpha, and myeloperoxidase. Areas with bronchopneumonia exhibited markedly decreased endothelial TM staining of alveolar walls and small vessels. These changes were associated with prominent fibrin immunoreactivity. Some areas exhibited mild to moderate inflammation with little fibrin deposition and variable amounts of TM in adjacent vessels. This study is the first to relate changes of TM immunoreactivity levels to fibrin deposition in a human disease process. These data may have implications for pulmonary pathophysiology in patients with bronchopneumonia.

Evaluation of fibrinogen self-assembly: role of its αC region.

BACKGROUND: Exposure of cryptic, functional sites on fibrinogen upon its adsorption to hydrophobic surfaces of biomaterials has been linked to an inflammatory response and fibrosis. Such adsorption also induces ordered fibrinogen aggregation which is poorly understood. OBJECTIVE: To investigate hydrophobic surface-induced fibrinogen aggregation. METHODS: Contact and lateral force scanning probe microscopy, yielding topography, image dimensions and fiber elastic modulus measurements were used along with transmission and scanning electron microscopy. Fibrinogen aggregation was induced under non-enzymatic conditions by adsorption on a trioctyl-surface monolayer (trioctylmethylamine) grafted onto silica clay plates. RESULTS: A more than one molecule thick coating was generated by adsorption on the plate from 100 to 200 µg mL⁻¹ fibrinogen solutions, and three-dimensional networks formed from 4 mg mL⁻¹ fibrinogen incubated with uncoated or fibrinogen-coated plates. Fibrils appeared laterally assembled into branching and overlapping fibers whose heights from the surface ranged from approximately 3 to 740 nm. The elastic modulus of fibrinogen fibers was 1.55 MPa. No fibrils formed when fibrinogen lacking αC-domains was used as a coating or was incubated with intact fibrinogen-coated plates, or when the latter plates were sequentially incubated with anti-Aα529-539 mAb and intact fibrinogen. When an anti-Aα241-476 mAb was used instead, fine, long fibers formed. Similarly, sequential incubations of fibrinogen-coated plates with recombinant αC-domain (Aα392-610 fragment) or αC-connector (Aα221-372 fragment) and fibrinogen resulted in distinctly fine fiber networks. CONCLUSIONS: Adsorption-induced fibrinogen self-assembly is initiated by a more than one molecule-thick surface layer and eventuates in three-dimensional networks whose formation requires fibrinogen with intact αC-domains.