The Biology of Hemodialysis Vascular Access Failure

Abstract

Arteriovenous fistulas (AVFs) are essential for patients and clinicians faced with end-stage renal disease (ESRD). While this method of vascular access for hemodialysis is preferred to others due to its reduced rate of infection and complications, they are plagued by intimal hyperplasia. The pathogenesis of intimal hyperplasia and subsequent thrombosis is brought on by uremia, hypoxia, and shear stress. These forces upregulate inflammatory and proliferative cytokines acting on leukocytes, fibroblasts, smooth muscle cells, and platelets. This activation begins initially with the progression of uremia, which induces platelet dysfunction and primes the body for an inflammatory response. The vasculature subsequently undergoes changes in oxygenation and shear stress during AVF creation. This propagates a strong inflammatory response in the vessel leading to cellular proliferation. This combined response is then further subjected to the stressors of cannulation and dialysis, eventually leading to stenosis and thrombosis. This review aims to help interventional radiologists understand the biological changes and pathogenesis of access failure.

Objectives: Upon completion of this article, the reader will be able to describe the biological forces that drive intimal hyperplasia leading to arteriovenous fistula failure.

Accreditation: This activity has been planned and implemented in accordance with the Essential Areas and Policies of the Accreditation Council for Continuing Medical Education (ACCME) through the joint providership of Tufts University School of Medicine (TUSM) and Thieme Medical Publishers, New York. TUSM is accredited by the ACCME to provide continuing medical education for physicians.

Credit: Tufts University School of Medicine designates this journal-based CME activity for a maximum of 1 AMA PRA Category 1 Credit™. Physicians should claim only the credit commensurate with the extent of their participation in the activity.

For the over 600,000 patients with end-stage renal disease (ESRD) in the United States alone, hemodialysis has been an indispensable lifeline. While there are several modes of vascular access including grafts and catheters, native fistulas are the most preferred, as they have lower rates of infection and complication compared with other alternatives.^{1 2 3 4} These outcomes and the Fistula First Initiative have helped to increase the numbers of arteriovenous fistulas (AVFs) used throughout the world and in diverse patient populations.^{4 5 6 7} However, AVFs can be affected by several problems that compromise venous access. Stenosis, thrombosis, infection, and aneurysm formation are the most common complications, and stenosis and thrombosis are the most relevant to AVF access failure.

Stenosis in the setting of hemodialysis (HD) usually occurs on the venous side, and is defined by the proliferation of several cell types leading to intimal hyperplasia (IH). These include inflammatory cells (mainly macrophages) along with vascular smooth muscle cells (SMCs), myofibroblasts, and fibroblasts. This rapid proliferation occurs due to uremic changes in ESRD patients along with stressors secondary to surgical trauma. To better understand the complications associated with AVFs and arteriovenous grafts (AVGs), it is helpful to study the chronology of these pathological changes. AVF failure can be thought of as occurring in three overly simplified steps: inflammation, proliferation, and thrombosis.

Initially, patients with ESRD have high baseline levels of inflammatory and platelet cell dysfunction secondary to uremic toxins and reactive species. During the creation of the AVF, there are changes in the vessel wall secondary to hypoxia and shear stress. These factors work together in a positive feedback loop propagating inflammation and cellular proliferation, which progresses to the point of stenosis and thrombosis.^{8 9 10 11 12 13} By exploring the nature of these biological mechanisms, HD access failure can be better understood, which can help guide future research aimed to ameliorate it.

Uremia

As stated earlier, even before fistula or graft creation several important systemic and vascular changes take place.^{14 15 16 17} The inherent uremia of ESRD increases inflammation and oxidative stress.^{18 19} These changes are evidenced by increases in many inflammatory cytokines, namely, interleukin-6 (IL-6) and tumor necrosis factor-α (TNF-α), and proliferate cytokines, such as transformative growth factor - β (TGF-β).

There are many other effects of uremia that involve the vascular system. Several studies have shown increased vessel thickness and calcification in patients with ESRD. This increased wall thickness is likely secondary to the proliferation of vascular smooth muscle cells in response to inflammation secondary to advanced glycation products and reactive oxygen species (ROS), both of which are increased in ESRD patients.^{20 21 22} The increased thickness of the vessel wall has implications in cardiovascular risk and also in access failure as it, along with inflammation, predisposes the vessel to stenosis.^{23 24 25 26} Increased vessel fibrosis, calcium phosphate deposition, and cellular calcium extrusion have been noted in ESRD patients. These changes are likely secondary to oxidative stress, dysfunctional cell signaling, and changes in mineral metabolism.^{27 28} These effects are compounded by platelet and endothelial wall dysfunction leading to rapid atherosclerotic progression. These combined changes predispose the vessel wall to negative, stenotic, inward remodeling after access placement.

Hypoxia

It is in this background of widespread cellular dysfunction that patients usually undergo placement of an AVG or AVF creation. The trauma induced by surgery induces hypoxia secondary to ischemia, both from cross-clamping and disruption of the vasa vasorum. This hypoxic injury is thought to also occur due to repeat needle sticks during cannulation for dialysis. This insult induces a series of cytokine cascades that promote inflammation, angiogenesis, and proliferation, which ultimately results in IH.

Ischemia and the release of ROS activate early response elements such as hypoxia-inducible factor 1-alpha (HIF-1α) and immediate-early responsive gene (IEX-1).^{29 30 31} These in turn activate endothelial cells (ECs), inflammatory cells, and vascular SMCs. HIF-1α, ROS, and hypoxia have been shown to upregulate vascular endothelial growth factors (VEGFs), which increase inflammation and promote EC proliferation. In addition, VEGF-A subsequently activates metalloproteinase (MMPs), namely, MMP-2, MMP-9, and ADAMTS-1, which allows for remodeling of the extracellular matrix (ECM) allowing for increased cellular proliferation.^{25 30 32 33 34 35 36} Hypoxia also activates platelet-derived growth factor, which has been shown to increase myofibroblasts.^{13 31 35 37}

These are the main proliferative changes that occur. However, cytokines such as VEGF-A, IEX-1, HIF-1α have also been shown to propagate inflammation, which can subsequently induce proliferation resulting in a dangerous positive feedback loop. VEGF-a can serve as a leukocyte attractant. HIF-1α can induce activation of inflammatory cells.³⁸ These and other early cytokines have been linked to activation of IL-8 and MCP-1. MCP-1 serves a leukocyte attractant and activator via CXCR motif receptors drawing monocytes, memory T lymphocytes, and natural killer cells toward the AVF.³⁹ IL-8 also recruits monocytes via the same receptors. Macrophage migration inhibitory factor also becomes elevated, and is thought to direct inflammatory cells toward the neointima, inducing negative remodeling through the proliferation of vascular SMCs and ECs.^{40 41} Another set of molecules implicated in AVF failure are Heme oxygenase-1 (HO-1) and Heme oxygenase-2 (HO-2). HO-1 is a stress protein that acts in a vasoprotective capacity by reducing inflammation, proliferation, fibrosis, and oxidant stress. Based on knockout (KO) gene models, it is likely that it acts on plasminogen activator inhibitor-1 (PAI-1), MCP-1, MMP2, and MMP9.^{42 43} HO-2 is thought to act in a similar manner.⁴⁴

Along with these agents, TGF-β and TNF-α have also been implicated in IH. TGF-β1 has been shown to lead to changes in the ECM, cellular proliferation, and thrombus formation. TNF-α is an inflammatory cytokine that has been shown to be activated by local inflammation and to hypoxia.^{45 46} TNF-α is released by leukocytes and also by SMCs and ECs.⁴ TNF-α can act via the receptor for advanced glycation end products (RAGE). This receptor has been linked to upregulation of VCAM-1. TNF-α can also act through more specific receptors upregulating inflammatory cytokines such as IL-1β; it also promotes cellular proliferation in fibroblasts. The ability of TNF-α to be released from and act on cells enables it to serve as a strong mediator of cellular proliferation and inflammation in the setting of AVF failure.

Shear Stress

In addition to hypoxia and inflammation, shear stress also plays a role in negative remodeling. After AVF creation, the venous ECs undergo increases in shear stress due to greater flow. Over the course of maturation, levels of shear stress decrease as the vessel dilates and laminar flow is reestablished. While there is normalization of shear stress distal to anastomoses, ECs near the anastomosis experience continued wall shear stress (WSS). ECs respond in three interrelated mechanisms: by rearranging the cytoskeleton, expressing different signaling cytokines, and affecting vasodilation.^{47 48 49 50}

Increased WSS increases inflammatory cytokines such as Il-1β, TNF-α, and interferon (IF-γ).^{51 52 53} Additionally, WSS has been shown to increase VEGF-A, MMP2, MM9, intercellular adhesion molecule-1 (ICAM-1), and other cytokines implicated in the proliferative aspect of negative remodeling.⁵⁴ Many of these changes are likely tied to nuclear factors such as decreased Krüppel-like factor 2 (KLF-2) and increased AKT signaling. KLF-2 has been shown to maintain a resting cell state and suppress inflammatory cytokines such as MCP-1 and IL-8. AKT is linked to the ERK/MAPK pathways, which propagate cell survival and mitotic changes.

WSS can induce mechanical tension on vascular endothelial (VE)-cadherin and platelet endothelial cell adhesion molecule (PECAM)-1.^{55 56} This mechanical stress serves to remodel the actin fibers and intracellular junctions. This allows the cell to change shape, and also induces cellular mobility.^{57 58} Beyond structural changes within cells, WSS also affects the vessel structure by affecting heme oxygenase and nitric oxide (NO) synthesis. NO acts to dilate the vessel, but this response can become blunted in cases of persistent activation such as hypertension.⁵⁹

Thrombosis

Thrombosis of HD access is a cumulative step and often signals the end of access viability. The same factors that drive IH lead to thrombosis. It is important to note that in the setting of ESRD, patients often experience increased bleeding time. The increased bleeding is secondary to dysfunctional platelets and also due to impaired platelet endothelial interactions. It is likely that much of the platelet dysfunction seen can be attributed to metabolic changes, and activation of coagulative and fibrinolytic cascades.⁶⁰ Platelets in uremia have accelerated apoptosis. Levels of ADP and serotonin are lower in ESRD patients. These reduced levels limit platelet aggregation and activation.⁶⁰ Thromboxane A2 is also systemically decreased in ESRD patients, which limits vasoconstriction and platelet aggregation. Additionally, uremia appears to have an impact on the cytoskeletal structure of platelets, which are found to have lower levels of actin, inhibiting contractile mechanisms and mobility.⁶¹ These and other changes account for the bleeding aspect caused by uremia.

On a systemic level, these factors lead to increased bleeding, but at the site of access, venous stenosis, shear stress, and endothelial damage can precipitate clot formation.^{62 63 64 65 66 67 68 69} The process of dialysis has also been shown to activate platelets, depending mainly on the membrane used in the dialysis machine. This can result in thrombocytopenia due to platelet destruction, and can also yield cell aggregates. Platelet–leukocyte aggregates can induce atherosclerosis and thrombosis.⁷⁰ Additionally, dialysis has been shown to increase circulating levels of p-selectin, CD40L, von Willebrand factor (WF), and D-dimer. This suggests that platelet activation in dialysis may lead to temporarily higher coagulability in the postdialysis period.^{71 72 73}

Another major factor specific to the AVF is inflammation due not just to AVF creation but also to repeat needle stick injury. Platelets can adhere to the already inflamed endothelial tissue and potentiate the process by releasing MCP-1, VCAM-1, ICAM-1, IL-1 β, and TNF-α, among others. These cytokines further the activation cascade leading to increased inflammation, adhesion, and eventually plaque or thrombus formation. Additionally, platelets can also directly induce these cytokines by direct interaction via P-selectin and with the monocyte receptor Mac-1 (α_Mβ₂) via α_IIbβ₃ or GPIbα.⁷⁴ Uremia along with shear stress and indoxyl sulfate have also been linked with increased endothelial micro particles. These by-products of EC activation or apoptosis can inhibit endothelial NO, further reducing the dilatory capacity of the vessel.^{75 76}

Genetics

There is a growing body of evidence examining the genetics of fistula failure. Many of the driving forces implicated in the activation of genes leading to AVF failure can also lead to epigenetic changes. Epigenetic changes affect the expression of genes through changes in histone modifications without changes in the DNA sequence. For example, HIF-1α has been shown to alter a set of histone-modifying genes.⁷⁷ It is likely that chronic kidney disease (CKD) and the comorbidities seen in ESRD patients also result in epigenetic changes that affect fistula functionality.⁷⁸ In addition to these changes, several polymorphisms have been linked to AVF patency rates, specifically VEGF-A, HO-1, and TGF-β.^{12 79 80 81}

There have been genetic studies focused on thrombosis. For example, polymorphisms in the TNF-α promoter region have been linked to AVF thrombosis.⁸² There have also been studies assessing platelet function. Platelet antigens such as human platelet antigen-3aa (HPA-3aa) have been found to be linked to AVF failure. This gene codes for a part of glycoprotein IIb (GbIIb). Certain alleles of these genes have been correlated with higher rates of access thrombosis.⁸³ Studies examining genetic predispositions to AVF failure lend credence and support to the current understanding of the molecular changes leading to AVF failure. Future work aiming at gene-specific therapy may help alleviate IH.

Conclusion

Hypoxia, uremia, shear stress, and thrombosis are the major factors implicated in AVF failure. These forces act on fibroblasts, macrophages, ECs, and platelets. They induce many interrelated genes implicated in inflammation and proliferation. These cytokines, along with the stressors of surgical anastomosis, repeat needle stick injury, and dialysis itself leads to IH and thrombosis.