Vascular access for dialysis patients

An irreversible loss in kidney function results in accumulation of fluid and metabolic waste products in the human body that finally leads to death. In Europe, more than half a million patients suffering from end-stage renal disease (ESRD) are treated by hemodialysis. Hemodialysis patients are 3-4 times a week for 3-4 hours connected to an artificial kidney to replace the lost kidney function. For proper hemodialysis, a well-functioning vascular access (VA) is of paramount importance. Such a VA should be easily accessible for repeated cannulation over time, should have sufficient vessel caliber and should have a blood flow rate of at least 600 ml/min. Unfortunately, the human body has no such vessel and therefore a permanent VA needs to be created surgically. An arteriovenous fistula (AVF), created by connecting an artery and vein in the arm, is the preferred VA. Directly connecting the artery and vein results in a low resistance conduit and consequently to a five- to thirtyfold increase in local blood flow. The proximal vein normally adapts its caliber to this increased flow and becomes a suitable VA after six weeks (i.e. maturation). Forearm AVFs have the most ideal characteristics, but have high requirements on the vein and artery that cannot always be met. As a result this type of AVF has also a high risk of nonmaturation and early thrombosis. Upper arm AVFs are less hampered by nonmaturation, but have a higher risk on high flow and corresponding demand on the heart. The choice between these two configurations is especially problematic in the increasingly older and comorbid patient entering hemodialysis treatment.

Another option for VA creation is the arteriovenous graft (AVG), for which the artery and vein are connected using a synthetic graft. In contrast to AVFs, the AVG is earlier suitable for HD as it does not require the aforementioned vascular adaptation. However, thrombotic occlusions and higher infection rates hamper their long-term outcomes in comparison to AVFs. Furthermore, AVGs have comparable cardiac demands as AVFs. A central venous catheter (CVC) is a third option for VA creation. A catheter enables the immediate start of hemodialysis and is therefore mostly used in case of acute renal failure. However, catheters suffer to a large extent from infections (bacteremia and septicemia) and thrombotic occlusions. Moreover, CVCs have an increased risk on recirculation, compromising dialysis efficiency. As a result of these drawbacks, permanent use of CVCs is only opted when all other options are exhausted, albeit those CVCs have the lowest cardiac demands of the three types of VA.

Research line

 In this research line, we develop mathematical models to support patient-specific AVF surgery planning (PhD Theses: Maarten Merkx and Wouter Huberts). In addition, we develop both fluid dynamic (CFD) and fluid-structure interaction (FSI) models to design hemodynamically optimized AVGs (PhD-project: Sjeng Quicken). All these projects are in close collaboration with the Department of Biomedical Engineering (chaired by Tammo Delhaas) and the Department of Vascular Surgery (vascular surgeons Jan Tordoir and Barend Mees) of the Maastricht University Medical Center.

Supporting AVF surgery planning using a patient-specific computer simulation model

Selecting the optimal location for AVF creation (upper- or lower arm), tailored to the individual patient, is challenging. Prior to surgery the selection of the optimal location for vascular access creation in hemodialysis patients is usually based on physical examination, the patient’s medical history, preoperative vessel assessment by duplex scanning and on the experience of the vascular surgeon. Despite this extensive preoperative workup access-related complications (nonmaturation and early thrombosis in 30% of patients, stenosis development, high-flow complications like cardiac failure and hand ischemia in 5% of patients) remain predominant and are caused by the induced changes in hemodynamics due to the connection of a vein to an artery (arteriovenous anastomosis).

In the European ARCH-project (FP7-ICT-224390) we have developed and validated, in close collaboration with amongst others the Maastricht University Medical Center (MUMC+), a patient-specific computational model that was able to predict postoperative AVF flow for different AVF configurations. 

Recently, the MUMC+ initiated a multicenter randomized-controlled clinical trial to examine the efficacy of the model to support VA surgery planning. The results of the RCT should answer the question whether the additional information provided by the model (i.e. a preoperative estimation of the postoperative flow) will also improve clinical outcomes such as reduced nonmaturation rates, high flow and hemodialysis access-induced distal ischemia (HAIDI). The RCT is funded by the Dutch Kidney Foundation (DKF NT12.01) and will include approximately 350 patients in 9 Dutch dialysis facilities. The results of this RCT are expected to be published in 2018.

Hemodynamically optimized AV-graft for hemodialysis

A large portion of patients with kidney failure are dependent on an arteriovenous graft to facilitate hemodialysis. These grafts typically function for only two years, after which a new vascular access has to be created surgically. During this period several reinterventions are typically required to ensure graft function. The main reason for AVG dysfunction is neointimal hyperplasia, which is caused by a post-surgical pathological inward remodeling process and results in stenosis and finally AVG occlusion.

It is established that the development of NIH is primarily caused by hemodynamics, especially by changes in wall shear stress (WSS) measures and flow disturbances, which are induced by AVG creation. Mechanistic understanding of the process of NIH development is however lacking. The aims of this project are to:

  1. Provide better understanding of the process of AVG patency loss;
  2. Propose a hemodynamically optimized graft geometry that facilitates favorable flow patterns, which might support in the development of a new generation of AVGs that exhibit longer patencies.

 In this project we aim to study graft hemodynamics with the ultimate goal of proposing a hemodynamically optimized graft. This project is performed in collaboration with Chemelot InSciTe via the XS-graft project.


Projects for bachelor-end projects, internships and MSc projects are available.

  • 3-D modeling for optimizing the anastomosis model of the PWPM
  • Modeling patient-specific vascular remodeling
  • Modeling structural and functional cardiac changes after AVF creation in individual patients
  • Using inverse modeling approach to assess patient-specific model parameters and their uncertainty
  • 3-D fluid-structure interaction modeling of graft-vein interaction
  • 3-D modeling of AV-graft hemodynamics
  • Modeling of needle hemodynamics
  • Ultrasound based 3D reconstruction of the vascular system

Other projects can be designed in consultation with the supervisors (Wouter Huberts, Richard Lopata, Frans van de Vosse and Barend Mees)

Students working on this project:

Sjeng Quicken (PhD project), Niek Zonnebeld (PhD project), Pamir Sawo (PhD project), Judith Muijsers (MSc project), Roy van Hees (MSc)

Completed PhD projects: 

Aron Bode, Maarten Merckx, Wouter Huberts, Bogdan Ene-Iordache

Completed QME projects:   

Wouter Huberts, Tommy Maas

Completed MSc projects:

Lars Pustjens, Siske Bos, Kujtim Gashi, Sjeng Quicken, Wouter Huberts