Management of high-flow arteriovenous access

Article information

Kidney Res Clin Pract. 2025;44(1):58-68

Publication date (electronic) : 2025 January 31

doi : https://doi.org/10.23876/j.krcp.23.196

Hoon Suk Park¹, Seok Joon Shin^,2

¹Division of Nephrology, Department of Internal Medicine, The Catholic University of Korea, Seoul St. Mary’s Hospital, Seoul, Republic of Korea

²Division of Nephrology, Department of Internal Medicine, The Catholic University of Korea, Incheon St. Mary’s Hospital, Incheon, Republic of Korea

Correspondence: Seok Joon Shin Division of Nephrology, Department of Internal Medicine, The Catholic University of Korea, Incheon St. Mary’s Hospital, 56 Dongsu-ro, Bupyeong-gu, Incheon 21431, Republic of Korea. E-mail: imkidney@catholic.ac.kr

Received 2023 July 27; Revised 2024 March 3; Accepted 2024 March 12.

Abstract

Introduction

For hemodialysis (HD), HD vascular access is absolutely necessary. It is well-known that HD vascular access includes a catheter, an artificial arteriovenous graft (AVG), and an autogenous arteriovenous fistula (AVF). In the case of a catheter, there is a high risk of infection when it is used for a long time. Thus, the use of an arteriovenous access is recommended as a permanent vascular access. Problems related to blood flow in arteriovenous access that might occur during its use can be classified into two categories: cannulation-related and non-cannulation-related problems. Cannulation-related problems arise from low blood flow within the arteriovenous access. They are commonly encountered in clinical practice. Examples of cannulation-related problems include arteriovenous access stenosis or obstruction. Non-cannulation-related problems include steal syndrome and high-output cardiac failure resulting from the use of arteriovenous access. Both steal syndrome and high-output cardiac failure are frequently associated with increased blood flow through the arteriovenous access. However, it is common for diabetic and elderly patients to have compromised systemic vascular status or impaired heart functions [1–3]. In such cases, even if the blood flow through the arteriovenous access is not substantial, steal syndrome or high-output heart failure can still be problematic. In recent years, diabetes has become the leading cause of end-stage kidney disease (ESKD), accounting for more than 50% of cases [4]. Additionally, HD treatment among individuals over the age of 65 years constitutes more than half of total cases [5]. Given the rising prevalence of diabetes and the increased occurrence of HD in the elderly population, it is crucial to adopt a more meticulous approach to managing blood flow in HD vascular access [6].

Arteriovenous access versus venovenous access

Under normal circumstances, without an arteriovenous access connecting arteries and veins, arterial blood delivered to tissues will return to the heart through veins, completing the regular tissue circulation. However, in the presence of arteriovenous access for HD, a different pathway is followed. Arterial blood is directed into the HD machine, where waste products and excess fluids are removed, resulting in purified blood. This purified blood is then directed back to the heart through a vein, bypassing the normal tissue circulation. Upon reaching the heart, it continues its journey through pulmonary circulation where oxygenation takes place, similar to regular circulation (Fig. 1). Therefore, cardiopulmonary recirculation (CPR) in patients with ESKD who have arteriovenous access for HD is defined as the degree of purified blood that bypasses tissue circulation and returns to the heart before it is passed through pulmonary circulation for oxygenation [7,8].

Figure 1.

Flow diagram during hemodialysis using arteriovenous access.

Cardiac output is divided into systemic flow (Qsys) and access flow (Qac). Systemic flow passes the normal tissue circulation while access flow (Qac) passes the dialysis circuit. These two parallel flows combine and return to the heart. The concentration of any substance in the access inflow (Cac,in) is lower than that in systemic flow (Csys) because Csys gets mixed with the concentration in the access outflow (Cac,out).

Qb, dialyzer blood flow; Cd, out, dialyzer outflow concentration; Cd, in, dialyzer inflow concentration.

CPR=QacCO, where Qac means vascular access flow volume and CO means cardiac output.

If an arteriovenous access is created, a CPR value of ≤20% is usually not clinically significant. When the CPR is between 20% and 30%, there is a risk of high-output cardiac failure due to increased CO. Thus, caution is required. If the CPR exceeds 30%, there is a high possibility of high-output cardiac failure or ischemic steal syndrome below the arteriovenous access formation as blood flow through the arteriovenous access increases [9–11]. At this point, a thorough evaluation is necessary. This thorough evaluation for cardiovascular performance includes measurements of CO, ejection fraction (EF), systemic vascular resistance (SVR), and left ventricular (LV) end-diastolic and systolic dimensions. The increased hemodynamic load from the creation of arteriovenous access creation is initially well compensated through cardiac chamber dilatation and a corresponding decrease in SVR. During this compensatory phase, patients typically remain asymptomatic, and EF is generally preserved. However, under certain stresses such as myocardial ischemia, the inability to maintain this increased CO can lead to decompensated high-output cardiac failure. In this decompensated state, tachycardia may develop as the heart rate increases to maintain CO due to lower-than-expected systolic volume. EF may remain within normal limits or decrease. Additionally, hypertension may occur as a consequence of increased SVR during this period.

Pulmonary hypertension (PH) may manifest when the pulmonary circulation is unable to accommodate the increased CO, particularly when nitric oxide (NO, vasodilator) levels decrease and endothelin-1 (ET-1, vasoconstrictor) levels increase. Symptoms may include exertional dyspnea and chest discomfort. Confirmation of PH requires right heart catheterization, which typically reveals elevated pulmonary artery pressure and pulmonary vascular resistance with normal pressures on the left side. Doppler echocardiography, with or without exercise, is useful for screening subclinical PH.

For patients with high-flow access, it is recommended to perform an echocardiogram biannually or at least annually. Monitoring should include LV end-diastolic and systolic dimensions, LV mass index, CO, and indices of contractility (e.g., EF). Patients exhibiting worsening parameters should consider undergoing active flow reduction.

In some cases, closure of the arteriovenous access might also need to be strongly considered. On the other hand, in the case of a catheter, it is positioned within the central vein or right atrium, enabling the acquisition of a significant blood flow volume suitable for HD. This catheter functions by introducing blood from the vein and returning it to the same vein without being connected in parallel with the entire circulatory system like arteriovenous access. Consequently, the use of a catheter does not impact CO (Fig. 2).

Figure 2.

Flow diagram during hemodialysis using arteriovenous access (A) and catheter (B).

In terms of clearance, the arteriovenous access and systemic tissue compartment (V) are connected in parallel, while the catheter and systemic compartment (V) are connected in series. In the series connection, there is no occurrence of cardiopulmonary recirculation.

Cac, out, access outflow concentration; Cd, out, dialyzer outflow concentration; Cd, in, dialyzer inflow concentration; Cven, venous concentration; Cart, arterial concentration.

Cardiac output/cardiac index

CO refers to the volume of blood pumped by the heart per minute. It is typically measured using echocardiography or the saline dilution method (normal range, 4–8 L/min). However, the significance of the same CO value might vary depending on a person’s physique. To address this, CO is adjusted according to body surface area, resulting in a cardiac index (CI) (normal range, 2.5–4.2 L/min/m²). Both CO and CI can show differences before and after HD due to the impact of ultrafiltration amount [12]. In the late 1990s and early 2000s, several studies investigated the relationship between CI and sudden cardiac death in HD patients. Sudden cardiac death in HD patients often occurs right before or after dialysis [13]. Researchers have explored the association between sudden cardiac death and CI in HD patients, seeking possible solutions. Various clinical scenarios related to high or low CI are presented in Table 1. Based on these scenarios, diagnostic workup recommendations for preventing sudden cardiac death were made in advance [14].

Table 1.

Abnormal cardiac index and its possible clinical condition

Definition of a high-flow access

Currently, there is no universally accepted definition for high-flow access. Nonetheless, the consensus from the studies so far suggests that Qac greater than 2 L/min or CPR greater than 20% can be used as criteria for high-flow access. Therefore, if Qac exceeds 2 L/min, a work-up for high-output cardiac failure should be initiated promptly.

Particularly, while high-output cardiac failure secondary to an AVF has been well-documented in dialysis patients since the 1970s, the exact prevalence of this condition remains elusive, as many cases are either underrecognized or unreported. Unfortunately, even the exact prevalence of high-flow access (Qac greater than 2 L/min) has seldom been reported thus far. For this review, we investigated the prevalence of high-flow access in our affiliated hospitals. According to our unpublished data, the prevalence of high-flow access is 14% in the entire population, with 12% in forearm access and 16% in upper arm access.

Flow reduction methods

If high-flow arteriovenous access is suspected to be causing high-output cardiac failure, an intervention or surgery is performed to reduce its blood flow. If the blood flow through the arteriovenous access exceeds 2 L/min, potential cardiotoxicity associated with that arteriovenous access should be considered, even if the patient may not exhibit symptoms. When symptoms are present, the patient understands the necessity for blood flow reduction intervention or surgery and agrees to undergo subsequent treatment. This is because the occurrence of symptoms of high-output cardiac failure is influenced by the amount of blood flow through the arteriovenous access, as well as the patient’s cardiac function, age, and diabetes status (Fig. 3).

Figure 3.

Work-up diagram for high-flow hemodialysis vascular access.

HD, hemodialysis; FU, follow-up; VA, vascular access; PH, pulmonary hypertension; Qac, vascular access flow volume; Tx, treatment.

For reducing blood flow, several techniques (Fig. 4) are used: 1) surgical banding, which involves using a dilator or graft; 2) MILLER (Minimally Invasive Limited Ligation Endoluminal-Assisted Revision) banding, which utilizes a balloon catheter; 3) plication with or without anastomotic revision; 4) graft interposition method, which closes the existing anastomosis site and inserts a small artificial blood vessel to create a new anastomosis for controlling inflow blood flow; and 5) graft inclusion technique, which involves opening the anastomosis site and inserting a small artificial blood vessel inside the preexisting vein to control the blood flow from the preexisting artery.

Figure 4.

The following are various endovascular and surgical techniques used for flow reduction.

MILLER, Minimally Invasive Limited Ligation Endoluminal-Assisted Revision.

Among these options, banding is a simple procedure. However, in many cases, blood flow returns to its original state in medium to long term, resulting in a low success rate. Plication is performed when there are severe aneurysmal changes in blood vessels. It involves reducing the size of the anastomotic site to effectively reduce blood flow. However, caution is required when dealing with large aneurysms as the surgical field can be extensive. Graft interposition is a surgery that involves closing the existing anastomotic area and inserting a small-sized graft to control blood flow, it is the most preferred method among surgeons. However, because a graft is introduced, there is an increased risk of infection. On the other hand, a graft inclusion technique is known to reduce the risk of infection by inserting an artificial blood vessel into the autologous vein. It has recently gained attention [15]. In addition, there is a T-banding technique that can reduce the inflow of blood by wrapping and tightening the artificial blood vessel in a T-shape from the outside of the entire artery and vein. This method can effectively reduce blood flow immediately after surgery because the entire anastomosis area is wrapped and tightened with artificial blood vessels (Fig. 5) [16]. However, over time, veins enclosed within the artificial blood vessel can undergo remodeling, leading to widening of the lumen and recovery of blood flow. Therefore, it is important to consider the possibility of recurrence even after T-banding (Fig. 6).

Figure 5.

T-banding technique for flow reduction, which involves wrapping both the anastomosis and the venous portion.

Figure 6.

Outcomes of T-banding in venous access remodeling.

(A) The venous portion wrapped by T-banding is shown. (B) The remodeled venous portion, which leads to lumen widening and a return to high-flow access, is shown.

Drawback of flow reduction methods

Compared to the surgical methods mentioned above for reducing blood flow, the banding procedure, which uses a balloon catheter or dilator to accurately reduce blood flow to a desired size, has the advantage of being a simple procedure. However, in cases where the aneurysm has already advanced significantly, it is not uncommon for reduced blood flow to recover within several months. Previous case reports have demonstrated that after the banding procedure, blood flow is restored, and surgical reduction is performed as planned, the ligated banding material is found to be embedded within blood vessels in the surgical field (Fig. 7) [17,18]. Therefore, the cause of blood flow recovery after the banding procedure might be attributed to this mechanism.

Figure 7.

Complications of balloon catheter banding in high-flow venous access.

(A, B) Banding of high-flow access using a balloon catheter. (C) Over time, due to strong, repetitive, pulsatile forces, the banding tie may erode the venous wall and subsequently become embedded. (D) Additionally, it could potentially be found hanging within the vein.

While there is no established target value for reducing Qac in high-flow access, the primary goals are to alleviate the patient’s symptoms, maintain efficient HD, and prevent thrombosis of the vascular access. The reduction of Qac can be as substantial as possible without increasing the risk of thrombosis. A Qac of 500 to 600 mL/min can be considered safe for AVF, while a flow of at least 1,000 mL/min is typically safe for AVG. In symptomatic patients with high-flow access, the primary aim is to alleviate symptoms through flow reduction. For asymptomatic patients, objective measures such as CO assessed by echocardiogram are important. These measures help to confirm the effectiveness of flow reduction and guide management to prevent potential cardiac complications associated with high-flow states.

Why now suddenly?

Autogenous AVF formation was first reported by Brescia, Cimmino, and Appel in the 1960s. Since then, autogenous AVF has become the preferred choice for HD vascular access due to its excellent long-term patency and minimal long-term complications. However, why has there been a sudden increase in interest in high-flow arteriovenous access and high-output cardiac failure in recent years? As mentioned earlier, unlike in the 1960s when HD was first started, one of the main reasons is the increasing age of patients undergoing HD for ESKD. Interestingly, the article by Brescia and Cimino also discussed the formation of autogenous AVF and its potential effect on CO and the risk of heart failure. They suggested that while CO might increase, it would not be clinically significant. They supported this claim by referring to cases with spontaneous AVF formed in the affected leg of polio patients without resulting in heart failure symptoms. Additionally, they mentioned orthopedic studies, demonstrating the effectiveness of increasing the length of a shorter leg by creating an AVF, did not report any cases of heart failure in patients who underwent the AVF surgery. However, upon closer examination of those orthopedic papers [19–21], it is interesting to note that the occurrence of ventricular hypertrophy on electrocardiogram and murmurs on auscultation were mentioned following the AVF formation despite the absence of echocardiographic results. Considering that patients with polio or those with different leg lengths were non-elderly in those studies, it is clear that AVF formation can increase CO. Nonetheless, it is evident that there has been a recent surge in interest regarding the cardiotoxicity of AVF, particularly in relation to creating AVF for HD in elderly patients [22].

Prevention

What can be done then to prevent high-flow arteriovenous access and high-output cardiac heart failure in HD patients? Treatments mentioned above are methods that can be performed when blood flow increases after the formation of arteriovenous access. For prevention, the expertise and technique of the surgeon who creates the arteriovenous access play a significant role [23]. For instance, actively considering the formation of a lower arm arteriovenous access can be a preventive strategy since the blood flow in the lower arm is generally lower compared to the upper arm [24]. However, it is important to note that arteries and veins in the lower arm are smaller in diameter and can be affected by vascular calcification. In such cases, using a microscope instead of a loupe for the formation of a lower arm arteriovenous access can be helpful. Microscopes are commonly used in various surgical fields such as plastic surgery, orthopedics, neurosurgery, and transplant surgery for precise anastomosis of small blood vessels. When it comes to forming arteriovenous access for HD, previous reports have highlighted the usefulness of microscopes, particularly in pediatric patients with ESKD, whose blood vessels are relatively smaller than those of adults. What is interesting is that in the case of arteriovenous access formed in the forearm of elderly patients, inflow artery calcification is often detected on ultrasound or general X-ray. It is observed that the increase in the size of such an inflow artery over time is small or insignificant. Consequently, when the increase in the size of the inflowing artery remains limited due to calcification, the corresponding arteriovenous access blood flow does not significantly increase. As a result, the long-term risk of high-output cardiac failure is low. With a limited increase in artery size, the arteriovenous access maintains a stable blood flow, which helps mitigate the risk of developing high-output cardiac failure over time [25].

In addition, preferring a proximal radio-upper arm cephalic AVF with the proximal radial artery as the inflow artery rather than a brachiocephalic fistula is another preventive method for high-flow access and subsequent high-output cardiac failure. This involves creating an anastomosis between the vein and the proximal radial artery at the elbow, below the brachial artery, after branching of the radial artery and the ulnar artery [26]. Another similar AVF is the Gracz one, which involves connecting the antecubital vein to the proximal radial artery. This type of AVF utilizes a nearly identical anastomosis pattern as endovascular AVF, which has recently been introduced with its low surgical trauma. In the case of endovascular AVF, the proximal radial or ulnar artery is connected to the adjacent antecubital vein using thermal ablation, a nonsurgical technique. It is particularly noteworthy that the occurrence of cephalic arch stenosis, known as the Achilles tendon of brachiocephalic AVF, is relatively low in endovascular AVF. This can be seen as a natural outcome considering the concept of inflow-outflow mismatch. Surgically created proximal radio-upper arm cephalic AVF, not endovascular AVF, was created as a result of preference for patients with a high possibility of steal syndrome in the past. When the proximal radial or ulnar artery is used as the inflow artery, the blood flow is smaller compared to that when the brachial artery is used as the inflow artery. Therefore, the possibility of high-output cardiac failure is expected to be low.

Recently, with the increasing number of elderly patients undergoing HD, the choice of vascular access for HD depends on the severity of heart failure (Table 2). In such cases, it is recommended to use only AVF in the lower arm as the preferred vascular access for HD. Alternatively, a catheter may be recommended as a permanent long-term vascular access for HD. According to the recommendation, if severe heart failure is observed on echocardiography when initiating HD, delaying the creation of permanent HD vascular access is suggested. Instead, the focus should be on improving cardiac function through volume control by HD. Afterward, another subsequent echocardiographic evaluation should be performed to reassess cardiac function after volume control. If there is an improvement, it is recommended to create an arteriovenous access for permanent vascular access for HD. Furthermore, it is advised to periodically evaluate cardiac function using echocardiography even after the creation of arteriovenous access (Table 3) [27,28].

Table 2.

Considered vascular access when beginning HD in ESKD patients with heart failure

Table 3.

Recommended follow-up schedule in ESKD patients with heart failure

Other things nephrologist should keep in mind as for Qac

A previous study has found that the formation of an AVF before HD can improve the glomerular filtration rate in patients with advanced chronic kidney disease. This improvement potentially leads to a delay in the initiation of dialysis [29]. The authors proposed two possible explanations for this phenomenon. Firstly, they suggest that antioxidants released from the ischemic area, which occurs below the AVF after its formation, could provide protective effects on kidneys. This concept is well-known as remote ischemic preconditioning. Secondly, the authors mentioned that the increase of CO and subsequent rise in renal blood flow resulting from the formation of the AVF might have contributed to renoprotective effects.

In the field of kidney transplantation, research has been conducted on the occlusion of AVF after transplantation. Effects of AVF in the context of kidney transplantation can be divided into their impact on the heart and potential effects on the transplanted kidney. Studies investigating the effects of AVF after transplantation on the heart support the hypothesis that AVF can potentially lead to heart failure. Those studies have suggested that closure of AVF might be considered as a means to reduce the occurrence of cardiovascular complications following transplantation. In contrast, one study has reported an intriguing result regarding the effect of an AVF on the transplanted kidney. According to that study, when the AVF was closed, the glomerular filtration rate of the transplanted kidney was observed to decrease more rapidly [30]. This suggests that caution is needed when considering AVF occlusion in kidney transplant recipients.

Therefore, apart from serving as vascular access for HD, nephrologists should be aware that arteriovenous access might have additional benefits or potential harm for patients with ESKD [31].

Conclusions

An arteriovenous access is a crucial component for vascular access in HD, which is necessary for maintaining lifelong HD treatment. However, it is important to be aware that arteriovenous access can theoretically lead to cardiotoxicity in patients with ESKD, potentially reducing the survival rate of these HD patients, depending on the blood flow volume.

Close monitoring of blood flow volume of arteriovenous access and cardiac function with appropriate interventions, when necessary, can help prevent high-flow arteriovenous access and high-output cardiac failure in HD patients. Nephrologists, serving as primary care providers for HD patients, should recognize the importance of managing high-flow HD vascular access.

Article information Continued