Homocysteine and oxidized low density lipoprotein enhanced platelet adhesion to endothelial cells under flow conditions: distinct mechanisms of thrombogenic modulation

We investigated the effects of two well established risk factors for cardiovascular disease, homocysteine and oxidized low density lipoprotein (ox-LDL), on endothelial cell thrombogenicity. For this purpose we studied platelet adhesion to human endothelial cells (EC) under flow conditions at a shear rate of 350 s(-1) following EC treatment with either homocysteine or ox-LDL. Treatment of EC with either homocysteine (1 or 10 mmol/L for 16 h) or ox-LDL (100 microg/ml for 16 h) resulted in a 2-3 fold enhancement in platelet adhesion. The enhancement in platelet adhesion induced by 1 mmol/L homocysteine, but not that induced by 10 mmol/L homocysteine, was absolutely dependent on fibrin formation. Homocysteine treatment has significantly increased the cell surface tissue factor (TF) activity and slightly reduced the expression of the intercellular adhesion molecule I (ICAM-1). In contrast, ox-LDL treatment upregulated ICAM-1 expression and had no significant effect on endothelial TF activity. Neither homocysteine nor Ox-LDL affected surface expression of the alpha(v)beta3 integrin. The homocysteine-induced enhancement in platelet adhesion was almost completely abolished by blockade of the EC TF activity by a polyclonal antibody. The enhancing effect of homocysteine was also greatly reduced by inhibition of the EC alpha(v)beta3 integrin, but was not affected by blockade of EC ICAM-1. On the other hand, ox-LDL-induced enhancement in platelet - EC adhesion was greatly inhibited by blocking ICAM-1 or alpha(v)beta3, but remained unaffected by inhibition of TF activity. Preincubation of platelets with the glycoprotein IIb-IIIa (GPIIb-IIIa) antagonist Reo-Pro has virtually abolished the enhancing effect of both homocysteine and ox-LDL. Our results suggest that homocysteine and ox-LDL might increase endothelial thrombogenicity by distinct mechanisms: homocysteine - by inducing TF activity, and ox-LDL - by upregulating ICAM-1, both of which enhance GPIIb-IIIa/fibrinogen dependent platelet adhesion to EC. The alpha(v)beta3 integrin, although not affected by EC stimulation, seems to play a crucial role in platelet-EC interaction regardless of the mechanism of EC perturbation.     

Soluble tissue actor interferes with angiostatin-mediated inhibition of endothelial cell proliferation by lysine-specific interaction with plasminogen kringle domains

Experimental and clinical data suggest that tissue factor (TF), the major initiator of blood coagulation cascade, as well as proteases and components of the fibrinolytic system are involved in tumor growth at least in some solid tumors via effects on angiogenesis. Whereas the pro- and anti-angiogenic effects of the plasminogen/plasmin system and plasminogen kringle domains, respectively, are well characterized, the pathways responsible for the pro-angiogenic properties of TF remain poorly understood. To learn more about the biological significance of the recently described binding of plasminogen to the extracellular domain of TF, we examined the effects of soluble TF (sTF) on angiostatin-inhibited proliferation of endothelial cells. In solid phase binding assays, we found that sTF binds specifically to plasminogen, to the plasminogen kringle domains K1-3, K1-5, K4, as well as to mini-plasminogen. Inhibition of binding of plasminogen and its kringle domains to sTF by the lysine analog 6-aminohexanoic acid (AHA) suggests that lysine-binding sites are involved in plasminogen interaction with TF. Moreover, in the presence of sTF, the inhibitory effect of K1-5 on bFGF-mediated HUVEC proliferation was dose-dependently and saturably abolished. This suggests that TF can interfere with the antagonistic effect of K1-5 on endothelial cell proliferation. In contrast, sTF by itself had no effect on the endothelial cell proliferation. Whereas the interference of TF with K1-5-mediated effect was prevented by AHA, this lysine analog did not abolish the proliferation inhibition of K1-5. In conclusion, the binding of sTF to the plasminogen fragment K1-5 seems to antagonize the anti-angiogenic effects of this plasminogen fragment.     

Platelet activation, coagulation and angiogenesis in breast and prostate carcinoma

In health, haemostasis and angiogenesis are tightly regulated processes, but may become deregulated in cancer. Recent evidence suggests that platelet activation may link these processes as platelets can release angiogenic factors such as vascular endothelial growth factor (VEGF). Furthermore, inflammation has also been implicated in regulating both coagulation and angiogenesis, possibly by activating platelets directly and increasing, for example, plasma fibrinogen. We hypothesized relationships between plasma markers of the processes in two common forms of cancer. Plasma levels of VEGF (reflecting angiogenesis), soluble P-selectin, (marking platelet activation), tissue factor [TF], fibrinogen and fibrin D-dimer (coagulation markers), and serum levels of IL-6 (inflammation) were measured by ELISA in 30 patients with biopsy-proven breast cancer, 30 patients with biopsy-proven prostate cancer, and 30 age- and sex-matched controls for each group. Prostate specific antigen was also measured in the men. Release of VEGF from IL-6 stimulated platelets was assessed by ELISA. Plasma levels of IL-6 (P <0.02), VEGF, soluble P-selectin, fibrinogen, and fibrin D-dimer (all p <0.01) were significantly raised in breast cancer, whereas VEGF, soluble P-selectin, fibrin D-dimer (all p <0.01) and fibrinogen (p <0.05) were significantly raised in prostate cancer. Significant correlations were found between IL-6 and VEGF (p <0.01), and IL-6 and soluble P-selectin (p = 0.038) in breast cancer. Further experiments demonstrated an in vitro IL-6 induced dose-dependent release of VEGF from platelets. In conclusion, strong relationships between IL6 and VEGF, but not with coagulation or platelet markers, and release of VEGF from IL-6 stimulated platelets, suggest a role for inflammation and platelets in angiogenesis.     

Inflammatory response and the endothelium

Antiphospholipid-mediated endothelium perturbation plays a role in antiphospholipid syndrome (APS)-associated vasculopathy. Antiphospholipid antibodies activate endothelium both in vitro and in vivo experimental models by inducing a pro-inflammatory/-coagulant phenotype; the antibodies recognize β2 glycoprotein I (β2GPI) on human endothelial cells (EC) from different parts of the vasculature.

In spite of such large in vitro evidence, few studies have addressed the issue whether or not a comparable endothelial perturbation might be detectable in vivo.

We investigated several indirect ex vivo parameters of endothelial dysfunction: plasma levels of soluble adhesion molecules (sADM), soluble thrombomodulin (sTM), von Willebrand factor (vWF) and tissue plasminogen activator (t-PA) by solid-phase assays. The study included: patients with primary antiphospholipid syndrome (n=32), with the syndrome secondary to non-active systemic lupus erythematosus (SLE, n=10), six patients with persistent antiphospholipid positivity at medium/high titre without any clinical manifestation of the syndrome. Fifty-two age and sex matched healthy subjects have been enrolled as controls. In addition, circulating endothelial cells identified by flow cytometry and the brachial artery flow-mediated vasodilation (FMV) were evaluated in 26 patients (20 primary and 6 lupus syndromes) and 30 healthy controls.

Plasma levels of soluble adhesion molecules did not differ from controls, while a significant increase in von Willebrand factor titres (P<0.05) was found. No significant difference was found regarding the number of circulating endothelial cells and flow-mediated vasodilation.

As a whole, these findings do suggest that antiphospholipid antibodies per se are not able to support a full-blown endothelial perturbation in vivo. As shown in antiphospholipid syndrome experimental animal models, a two-hit hypothesis is suggested.     

Effect of intradermal tumor necrosis factor-alpha-induced inflammation on coagulation factors in dermal vessel endothelium. An in vivo study of human skin biopsies

Inflammatory mediators were shown to exert procoagulant effects on cultured human endothelial cells (EC). In the present study the effect of intradermal application of tumor necrosis factor-alpha (TNF-alpha) on the expression of factors involved in regulation of coagulation at the EC surface, i.e. tissue factor (TF), thrombomodulin (TM) and tissue factor pathway inhibitor (TFPI) was studied in humans in vivo. The endothelial expression of these factors was evaluated immunohistochemically in biopsies taken after intradermal application of 5000 U TNF-alpha in 8 healthy volunteers. After 6 and 22 h biopsies were taken from the injection sites. At TNF-alpha injected sites typical inflammatory changes. e.g. EC upregulation of adhesion molecules and accumulation of leukocytes were detected. In parallel we could document EC expression of TF, downregulation of TM and depletion of tissue factor pathway inhibitor (TFPI) in inflamed areas. Early depletion of endothelial IkappaB alpha at the site of inflammation after application of TNF-alpha points to an activation of the NF-kappaB pathway. Our data suggest that, as shown in in vitro experiments, TNF-alpha activates the NF-kappaB pathway and induces specific procoagulant changes of EC due to expression of TF, down-regulation of TM and depletion of TFPI in vivo in humans. This procoagulant shift in the haemostatic balance on the cell surface, caused by TNF-alpha-induced inflammation, is likely to contribute to thrombosis associated with tissue inflammation in humans.     

Platelet activation rather than endothelial injury identifies risk of thrombosis in subjects positive for antiphospholipid antibodies

BACKGROUND: Anti-phospholipid antibodies (APLA) are often associated with thrombosis, defining the antiphospholipid syndrome (APS) but it remains unclear why many subjects who are positive for APLA chiefly anti-cardiolipin (aCL) or anti-beta2GPI (abeta2GPI) do not develop thrombosis. A related question addressed in this study is whether the target of cellular injury in APS is predominately platelets or endothelial cells (EC). METHODS: aCL and abeta2GPI were determined by ELISA in 88 patients, 60 of whom were thrombotic and 28 non-thrombotic. Platelet activation was measured by CD62P and by concentration of platelet microparticles (PMP) and EC activation was assessed by endothelial microparticles (EMP), both by flow cytometry. Lupus anticoagulant (LAC) was measured in the hospital laboratory. RESULTS: There was no difference in frequency of aCL or abeta2GPI, neither IgG or IgM, between the thrombotic and non-thrombotic groups. Both groups showed elevated EMP compared to controls but this did not differ between thrombotic and non-thrombotic groups. In contrast, PMP were not significantly elevated in non-thrombotic but were elevated in thrombotic compared to non-thrombotic (p=0.03) and controls. CD62P, an independent marker of platelet activation, was also elevated in thrombotic vs. non-thrombotic. There was a trend for increased LAC in the thrombotic group but not significant. CONCLUSION: Although all subjects had evidence of endothelial activation, only platelet activation differed between thrombotic and non-thrombotic. This supports the hypothesis that platelet activation predisposes to thrombosis in the presence of chronic EC activation. These data also raise the possibility of distinguishing risk-prone APLA-positive individuals.     

Measures of endothelial dysfunction in plasma of patients with posttraumatic stress disorder

Posttraumatic stress disorder (PTSD) confers an increased cardiovascular risk. In 14 otherwise healthy patients with PTSD and in 14 age- and gender-matched non-PTSD controls, we investigated whether the categorical diagnosis of PTSD and severity of PTSD symptom clusters (i.e. re-experiencing, avoidance, arousal, and overall score) would be associated with plasma concentrations of three markers of endothelial dysfunction [soluble tissue factor (sTF), von Willebrand factor (VWF), and soluble intercellular adhesion molecule (sICAM)-1]. Compared with controls, patients had significantly higher sTF; this difference became nonsignificant when controlling for psychological distress. VWF and sICAM-1 levels were not significantly different between patients and controls. In the entire sample virtually all PTSD symptom clusters correlated significantly and positively with sTF and VWF but not with sICAM-1. The correlation between symptoms of re-experiencing and sTF was significantly different between patients and controls. Controlling for symptoms of anxiety and depression (i.e. psychological distress) rendered most associations between PTSD symptom clusters and sTF nonsignificant, whereas controlling for age retained significance of associations with VWF. Posttraumatic stress showed a continuous relationship with sTF and VWF, with the former relationship being partly affected by psychological distress. This suggests one mechanism by which posttraumatic stress could contribute to atherosclerosis.     

Thrombin activation of endometrial endothelial cells: a possible role in intrauterine growth restriction

Preeclampsia (PE), intrauterine growth restriction (IUGR) and abruption with or without fetal loss are associated with reduced uteroplacental blood flow, decidual vasculopathy, endothelial cell dysfunction, thrombosis, inflammation and hemorrhage. Our hypothesis is that reduced uteroplacental blood flow causes focal decidual hypoxia that generates vascular endothelial growth factor (VEGF). The latter acts directly on decidual endothelial cells to induce aberrant expression of tissue factor (TF), the primary initiator of coagulation. This in turn generates thrombin that induces: i) further TF expression; and ii) inflammatory cytokines. Both VEGF and TF induce aberrant angiogenesis-vessel maintenance reflected by endothelial cell fenestrations and induction of a prothrombotic surface causing both the decidual hemorrhage (i.e. abruption) and thrombosis (i.e. uteroplacental vascular insufficiency) observed in these adverse pregnancy outcomes. This novel hypothesis is supported by our finding of TF expression in decidual endothelium of pregnancies complicated by IUGR and/or fetal loss. Moreover, treatment of cultured endometrial endothelial cells with VEGF or thrombin induces TF protein and mRNA expression. Quantitative RT-PCR analysis indicates that thrombin enhances (>10-fold) the output of diverse inflammatory cytokines in these cultures. The greatest effect (>2-log) was seen on macrophage inflammatory protein 3alpha (MIP3alpha). In vitro, thrombin results in endometrial endothelial cell aggregations and changes in the apoptotic pathway. Thus, we postulate that reductions in uteroplacental flow initiate a cascade of molecular effects leading to hypoxia, thrombosis, inflammation, and endothelial cell dysfunction resulting in untoward pregnancy outcomes.     

Endothelial Cell Markers in Chronic Uremia: Relationship with Hemostatic Defects and Severity of Renal Failure

Plasma von Willebrand factor antigen, soluble thrombomodulin, and tissue factor were increased in 31 patients with severe chronic renal failure (creatinine clearance <20 ml/min) under conservative treatment, whereas plasminogen activator inhibitor antigen did not differ significantly from healthy controls. No correlation among plasma levels of these proteins was found. Three patterns of relationship between endothelial cell markers and hemostatic defects were identified: 1) Plasma thrombomodulin, a marker of endothelium damage, was found an independent predictor of bleeding time and platelet aggregation, and secretion defects, and was also related to the severity of renal failure; 2) von Willebrand factor antigen, an index of endothelial cell activation and secretion, was significantly correlated with intravascular markers of thombin and plasmin generation and with platelet adenosine triphosphate content, but not with plasma creatinine levels; and 3) tissue factor and plasminogen activator inhibitor antigen levels were not statistically correlated with the diverse hemostatic defects. Activation of coagulation and fibrinolysis, secondary to endothelial cell activation, appearing early during the evolution of chronic renal failure, is pathogenically related to the platelet dysfunction, and probably to development of atherosclerosis and thrombotic events in this disease. The progression of chronic renal failure, through endothelial cell damage, would lead to aggravation of the platelet functional defect potentiating the hemorrhagic risk.     

Predictors of inflammation in response to anthracycline-based chemotherapy for breast cancer

Although chemotherapy for breast cancer can increase inflammation, few studies have examined predictors of this phenomenon. This study examined potential contributions of demographics, disease characteristics, and treatment regimens to markers of inflammation in response to chemotherapy for breast cancer. Thirty-five women with stage I-III-A breast cancer (mean age 50 years) were studied prior to cycle 1 and prior to cycle 4 of anthracycline-based chemotherapy. Circulating levels of inflammatory markers with high relevance to breast cancer were examined, including C-reactive protein (CRP), tumor necrosis factor-alpha (TNF-alpha), Interleukin-1 receptor antagonist (IL1-RA), vascular endothelial growth factor (VEGF), soluble intercellular adhesion molecule-1 (sICAM-1), Interleukin- (IL-6), soluble P-selectin (sP-selectin), and von Willebrand factor (vWf). Chemotherapy was associated with elevations in VEGF (p < or = 0.01), sICAM-1 (p < or = 0.01), sP-selectin (p < or = 0.02) and vWf (p < or = 0.05). Multiple regression analysis controlling for age and body mass index (BMI) showed that higher post-chemotherapy levels of inflammation were consistently related to higher pre-chemotherapy levels of inflammation (ps < or =0.05) as well as to certain disease characteristics. Post-chemotherapy IL-6 levels were higher in patients who had larger tumors (p < or = 0.05) while post-chemotherapy VEGF levels were higher in patients who had smaller tumors (p < or = 0.05). Post-chemotherapy sP-selectin levels were highest in women who had received epirubicin, cytoxan, 5-fluorouracil chemotherapy (p < or = 0.01). These findings indicate that chemotherapy treatment can be associated with elevations in certain markers of inflammation, particularly markers of endothelial and platelet activation. Inflammation in response to chemotherapy is most significantly related to inflammation that existed prior to chemotherapy but also potentially to treatment regimen and to certain disease characteristics.     

Functional analyses of patient-derived IgG monoclonal anticardiolipin antibodies using in vivo thrombosis and in vivo microcirculation models

Antiphospholipid antibodies (aPL) have been associated with thrombosis and pregnancy losses in patients diagnosed with antiphospholipid syndrome (APS) and enhance thrombus formation in vivo in mice, but the mechanism of thrombosis by aPL is not completely understood. It has been proposed that aPL may affect endothelial cell (EC) function and/or induce their activation, transforming their anticoagulant surface into procoagulant, thus predisposing to thrombosis. It has been proposed that aPL may affect EC cell function and/or induce their activation, transforming their anticoagulant surface into procoagulant, thus predisposing to thrombosis. This study proposes to test the hypotheses that some IgG anticardiolipins (IgG aCL) with thrombogenic properties in mice, exert their effects through activation of endothelium. We studied seven patient-derived monoclonal aCL for their thrombogenic properties in an in vivo pinch-induced thrombosis model, and their functional activities in activating EC by analyzing in vivo leukocyte adhesion to endothelium in microcirculation in venules in exposed murine cremaster muscle and in vitro adhesion molecule expression in cultured EC. The binding of the monoclonal aCL to EC was also tested. In addition to the previous identified thrombogenic IS2, four of the five new more IgG monoclonal aCL (from two patients) were found to be thrombogenic. Of these five thrombogenic aCL, three caused more in vivo leukocyte adhesion to EC in microcirculation, as compared to that induced by the H2 control human monoclonal IgG, and enhanced expression of adhesion molecules (particularly VCAM-1) on cultured EC. These data show that about 2/3 patient-derived IgG monoclonal aCL are thrombogenic and suggest that some thrombogenic IgG aCL exert their effects through activating EC.     

Monocyte-derived microparticles and exosomes induce procoagulant and apoptotic effects on endothelial cells

Microvesicles (MVs) which include microparticles (MPs) and exosomes are found in blood circulation in normal physiologic conditions and are increased in a variety of diseases. This study evaluated the effects of MVs on human umbilical vein endothelial cells (HUVEC) by morphologic changes, apoptosis, and thrombogenicty, in vitro. Stimulation of monocyte cell line (THP-1) by starvation or by endotoxin and calcium ionophore A23187 resulted in the release of MVs which express exosome marker Tsg 101, negative phospholipids in their leaflets, monocyte markers (CD18, CD14) and active tissue factor (TF). MVs were found to disrupt EC integrity and rapidly induce membrane blebbing. Brief exposure (2-4 hours) to MVs resulted in EC membrane phospholipids "flip-flop" while longer stimulation (20 hours) led to two contradicting outcomes - tube formation as well as apoptosis, as assessed by nuclear fragmentation. Additionally, MVs exposure resulted in increased cell surface thrombogenicity and perturbation of the endothelial haemostatic balance, which were enhanced during longer exposure time. Activity, antigen level and mRNA expression of the coagulation initiator TF were elevated due to (i) adherence of MVs derived TF to the EC membrane, and (ii) an increase in endothelial TF expression. Furthermore, levels of the anticoagulant tissue factor pathway inhibitor (TFPI) and thrombomodulin (TM) were decreased. These findings demonstrate that monocyte MVs increase endothelial thrombogenicity and apoptosis. In addition, they induce tube formation which may indicate their angiogenic effect. These findings may clarify, in part, the role of MVs in EC dysfunction associated with inflammatory diseases and hypercoagulable states.     

A novel anti-tissue factor monoclonal antibody with anticoagulant potency derived from synthesized multiple antigenic peptide through blocking FX combination with TF

Tissue factor (TF) has been implicated in the pathogenesis of various thrombotic disorders. Monoclonal antibodies (mAb) that specifically target TF may have potential as antithrombotic therapy. We designed a unique TF peptide (TFP) that was specific for the binding site to factor X (FX). This peptide was used to develop TF mAb that block the coagulation cascade by interfering with the combination of FX with the TF/FVIIa complex. Chemically synthesized TFP coupled to polylysine matrix was used as multiple antigenic peptide (TF-MAP) and this was used to immunize Balb/c mice for the preparation of hybridomas. One hybridoma cell line released an antibody, named TF4A12, which had high anticoagulant potency (by dilute prothrombin time assay). Western blotting showed that TF4A12 could bind TF-MAP and the soluble TF extracellular domain (sTF(1-219)). Results of FX activation assay and amidolytic activity assay showed that the anticoagulant ability of TF4A12 is due to blocking FX, but not FVII, binding to TF. Our study identified an efficient method of developing TF mAb that could block the coagulation cascade.     

Soluble adhesion molecules and angiotensin-converting enzyme in dementia

We aimed to determine plasma and cerebrospinal fluid (CSF) levels of angiotensin-converting enzyme (ACE) and the soluble forms of intercellular adhesion molecule-1 (sICAM-1), vascular cell adhesion molecule-1 (sVCAM-1) and platelet endothelial cell adhesion molecule-1 (sPECAM-1) as surrogate markers for endothelial cell activation in clinically diagnosed patients with Alzheimer's disease (AD, n=260), dementia with Lewy bodies (DLB, n=39) and non-demented controls (n=34). Plasma sICAM-1 and sPECAM-1 were higher and CSF sVCAM-1 were lower in AD and DLB patients than in controls (p<0.001). DLB patients had higher CSF sICAM-1, but lower CSF sVCAM-1 (p<0.001). No difference in ACE levels was found between the dementia groups and controls. In controls and AD patients CSF sICAM and sVCAM-1 strongly correlated with each other and with blood barrier permeability whereas in DLB group these correlations were weaker. The observed patterns in adhesion molecules may reflect distinctions in the pathophysiological basis of their generation in dementia patients.     

Tissue factor--a receptor involved in the control of cellular properties, including angiogenesis

Tissue factor (TF), the major initiator of blood coagulation, serves as a regulator of angiogenesis, tumor growth and metastasis. In several models, TF expression mediates upregulation of the proangiogenic vasular endothelial growth factor (VEGF) that can directly act on endothelial cells to promote vessel formation. This occurs through ligand binding, activation of signaling cascades, signal transduction and alteration of growth factor expression and is mediated by both, coagulation-dependent and -independent pathways. Depending on the cell type and the biological settings, TF seems to affect cellular properties through (i) factor VIIa (FVIIa)-dependent proteolysis of factor Xa (FXa) and thrombin and subsequent activation of proteinase activated receptor (PAR) -1 and PAR-2, (ii) through direct FVIIa signaling and mitogen activated protein (MAP) kinase activation, that is conferred by a not yet identified receptor, (iii) through interaction of FVII(a) proteolytic activity and signaling of the cytoplasmic domain and (iv) through cytoplasmic signaling independent of ligand binding. The role of phosphorylation of the cytoplasmic domain and the pathways controlling phosphorylation of TF remain poorly understood.     

Elevated platelet angiostatin and circulating endothelial microfragments in idiopathic pulmonary arterial hypertension: A preliminary study

Introduction

Idiopathic pulmonary arterial hypertension (IPAH) is characterized by pulmonary arteriolar narrowing and degeneration associated with in situ thrombosis. We hypothesized that microvascular endothelial injury and apoptosis may be an initiating mechanism in IPAH. Endothelial apoptosis generates endothelial microfragments (EMF), which can activate platelets. Platelets release both VEGF and angiostatin, which depending the balance can inhibit or induce endothelial apoptosis, respectively.

Materials and Methods

We measured EMFs from blood of IPAH patients as index of endothelial cell apoptosis/injury and levels of pro- and anti- EC apoptotic factors found in platelets. EMFs and platelets in blood samples from control subjects and patients with IPAH were measured using a 4-color flow cytometry protocol, and platelet levels of VEGF and angiostatin were determined by ELISAs and immunoblotting.

Results

Compared to controls, IPAH patients exhibited higher numbers of circulating EMFs and more activated/apoptotic platelets. IPAH patients also exhibited higher levels of platelet angiostatin; however, no significant difference was detected in platelet VEGF levels between the two groups.

Conclusions

These results are consistent with an increase in EC dysfunction in patients with IPAH, possibly contributing to the progression of IPAH and its associated thrombosis.