Increased Risk of Thrombosis
The risk of venous thromboembolism is increased in patients with renal failure. The mortality related to pulmonary embolism (PE) is greater in patients with renal failure when compared with those without renal disease. Thromboembolism itself has a 2-fold increased risk in patients with advanced kidney disease, while a higher risk has been shown in hospitalized patients with renal impairment. The risk begins to rise when the estimated glomerular filtration rate (eGFR) falls <75 mL/min/1.73 m. During early stages of chronic kidney disease (CKD), the risk of thrombosis seems to be related to albuminuria.
Clinical relevant thrombosis in patients with renal failure may present as deep venous thrombosis with/without PE, HD vascular access-associated thrombosis including arteriovenous graft thrombosis as well as native AV fistula thrombosis, central venous catheter thrombosis with/without central vein thrombosis, right atrial thrombus. Furthermore, thrombus formation may also occur within arteries that are often atherosclerosis-associated, and could present as acute coronary syndrome, cerebrovascular event or peripheral artery occlusion (Table 1).
What are the pathophysiological pathways associated with an increased risk of thrombosis?
Coagulation Cascade
Patients with CKD have increased levels of fibrinogen that directly contribute to a hypercoagulable state. This is associated with increased levels of pro-inflammatory markers such as C-reactive protein and interleukin-6. Furthermore, increased levels of plasma tissue factor (TF) have been observed in patients with renal failure. Apart from coagulation it can contribute also to inflammation as it can induce the proinflammatory transcription factor Nf-κB as well as protease-activated receptor-1. It has also been shown that the concentrations of the coagulation factors XIIa and VIIa as well as activated protein C complex and thrombin-anti thrombin complexes are increased in patients with renal failure. On the other hand, the activity of antithrombin is reduced.
A clinically important system that may be involved in the hypercoagulable state of patients with renal failure can be the renin–angiotensin–aldosterone system as its activation has been associated with increased levels of plasma fibrinogen, d-dimer and plasminogen activator inhibitor (PAI)-1. PAI-1 has been associated with an inhibition of extracellular matrix turnover, stimulation of macrophage and myofibroblast infiltration as well as the regulation of TGF-β, thus promoting tissue fibrosis with progression of CKD. Furthermore, PAI-1 inhibits the activation of the fibrinolytic system through inhibition of the tissue plasminogen activator (t-PA) and urokinase.
Platelets
In patients performing peritoneal dialysis, platelets can be activated which is thought to be related to hypoalbuminaemia. It has been demonstrated that in catabolic uraemic patients, reduced plasma levels of l-arginine and NO were associated with an increased platelet aggregability. This can be supported through an accumulation of phenyl acetic acid, a uraemic toxin, inhibiting inducible NO synthase (iNOS) resulting in a reduced production of NO.
In patients with renal failure, increased levels of phosphatidylserine can be observed at the surface of platelets that is related to caspase-3 activation. Phosphatidylserine binds to activated factor V that promotes binding of factor X leading to the formation of thrombin with thrombus formation. Platelets of uraemic patients contain increased levels of p-selectin as well as the fibrinogen receptor PAC-1 resulting in platelet/leucocyte aggregates, followed by an increased reactivity of platelets. In addition, this mediates the formation of platelet/leucocyte aggregates, related to the formation of free oxygen radicals by neutrophil granulocytes leading to thrombus formation in patients with renal failure.
Endothelium
The endothelium is of crucial importance for haemostasis. It is responsible for the secretion of factors modulating the coagulation cascade such as PAI-1 and vWF, participates in the regulation of the vascular tone, regulates oxidant stress and thus also inflammatory responses and produces endothelial microparticles (MPs). Furthermore, it influences haemostasis through proliferation/repair processes that also include endothelial progenitor cells (EPCs). The endothelium may lose its anti-thrombogenic properties if it is stimulated by thrombin, hypoxia, shear stress, oxidants, interleukin-1, tumour necrosis factor, γ-interferon, desmopressin acetate and endotoxin. In patients with end-stage renal disease (ESRD), endothelial cell damage can lead to coagulation disorders together with thrombophilia. Homocysteine can play a role as a mediator between renal dysfunction and endothelial cell damage. It can inhibit the thrombomodulin-dependent activated protein c system that results in permanent activation of thrombin with subsequent formation of fibrin. It also interferes with endothelial release of t-PA predisposing to hypofibrinolysis. This can also be due to an impaired release of t-PA from the endothelium with an intact endothelium-dependent vasodilation.
Hyperhomocystinaemia also interferes with subendothelial cell proliferation through metalloproteinase-inducible genes as it leads to an activation of matrix metalloproteinase-9. Again, also in this setting, increased levels of PAI-1 have been suggested as markers of endothelial cell activation. However, high plasma concentration of fibrinogen, d-dimer, thrombin–antithrombin complex, coagulation factor VII, vWF, thrombomodulin and PAI-1 can all indicate endothelial cell damage and a thrombophilic state in uraemic patients.
Atherosclerosis itself seems to be associated with an increased risk of the development of venous thrombosis in patients with renal failure. The reason for this phenomenon could be an overlap of the respective risk factors such as obesity, hypertension, smoking, diabetes and dyslipidaemia. Furthermore, in patients with renal failure, platelets and the coagulation system could be activated in atherosclerotic vessels contributing to the formation of venous thrombosis at different vessel sites. In a recent population-based study, 26% of patients with venous thrombosis also had a history of symptomatic atherosclerosis. Interestingly, microalbuminuria is also associated with the development of venous thrombosis. This could be related to the fact that microalbuminuria reflects the severity of endothelial damage which in turn can promote thrombosis.
Microparticles
MPs have recently been discovered to have potent procoagulatory capacity and thus could play an important role in coagulation. MPs are formed from plasma membranes of many cells including endothelial cells, platelets as well as monocytes/macrophages. They are the result of cell activation during inflammatory processes but also occur during physiological processes such as cell differentiation and senescence. Increased levels of MPs have been described in diseases with an increased procoagulant state such as chronic kidney insufficiency as well as cancer. Their procoagulatory effects are derived from the presentation of phosphatidylserine facilitating the conversion from prothrombin to thrombin as well as the presence of TF on their surface. Apart from membrane bound TF, the MPs also release soluble TFs further promoting coagulation resulting in excessive thrombus formation. Furthermore, MPs could influence coagulation by another mechanism, which is through the recently discovered microRNAs (miRNAs). miRNAs are small non-coding single-strand RNAs that modulate target gene expression by post-transcriptional modulation and are expressed in the majority of cells. The connection between miRNAs and the coagulation system is not clear so far. However, some data exist, linking miRNAs to the function of platelets through regulation of platelet mRNA translation. Here, the expression of the P2Y12 receptor, that is important for the ADP-stimulated activation of the GP IIb/IIIa receptor resulting in prolonged platelet aggregation, is regulated through miRNAs. Vesicle-associated membrane protein 8 (VAMP8) is important for the secretion process of platelets with hyperreactive platelets demonstrating increased VAMP8 levels, while hyporeactive platelets show decreased levels. Thus, it could be shown that the concentration of miRNA96 was 2.6 times higher in hyporeactive platelets. How such regulation processes are influenced in renal failure is not known so far but it is tempting to speculate that an important influence exists.
As MPs and miRNAs could regulate platelet activity, this also brings the platelet proteome/transcriptome into focus. Contact of platelets to uraemic toxins and artificial surfaces during HD could lead to a change in the platelet proteome, which could be a result of an altered platelet transcriptome. Interestingly, analysis of the platelet mRNA and miRNA transcriptome revealed alterations when compared with healthy subjects. Dialysis seemed to correct the levels of some mRNAs and most miRNAs. Here, analysis of hsa-miR-19b, a miRNA involved in the regulation of platelet reactivity through phosphatidylcholine transfer protein and WD repeat-containing protein 1, was increased in platelets of uraemic patients. This suggests that altered miRNA-based mRNA regulatory mechanisms could influence the platelet response to uraemia leading to platelet-related complications in CKD.
Antiphospholipid Antibodies
Antiphospholipid antibodies (lupus anticoagulant, cardiolipin antibodies) can be detected in many patients on HD. Their significance in patients with renal failure with or without HD is not clear so far. They could represent a risk factor for thrombosis, particularly vascular access thrombosis in this group of patients. One group detected an increased prevalence of IgG anticardiolipin antibodies in patients with recurrent vascular access thrombosis. On the other hand, the presence of antiphospholipid antibodies could be an epiphenomenon of aspirin use or simply HD in these patients. Even vascular access stenosis but not thrombosis has been associated with the presence of such antibodies. Furthermore, an increased prevalence of anti-protein C and anti-protein S has been observed in HD patients with vascular access thrombosis. Again, the pathogenic significance of this observation is not clear so far. Anti-protein C antibodies would result in an increased activity of factor VII and potentially also factor Va resulting in an increased formation of thrombin leading to thrombus formation.
In conclusion, the increased risk of thrombosis in patients with renal failure is related to changes in the coagulation cascade with increased levels of fibrinogen and plasma TF, factor XIIa and VIIa, F VIII, activated protein C complex, thrombin–antithrombin complexes, d-dimers, prothrombin fragments 1 + 2 (F1 + 2) and a reduced activity of antithrombin (Figure 2). Furthermore, the activity of platelets is increased by a reduced amount of vasoactive substances (i.e. NO), increased levels of phosphatidylserine, p-selectin, and fibrinogen receptor PAC-1. Changes of the endothelium also contribute to the risk of thrombosis by increased thrombin, hypoxia, shear stress, oxidants and cytokines. A new mechanism is related to MPs that promote the presentation of phosphatidylserine and TF, and also contain miRNAs potentially regulating the function of platelets. In addition, antiphospholipid antibodies can also promote thrombosis in patients with renal insufficiency (Figure 2).
(Enlarge Image)
Figure 2.
Factors involved in the increased risk of thrombosis in patients with renal failure (for details see the text).