Computational modeling of the cephalic arch predicts hemodynamic profiles in patients with brachiocephalic fistula access receiving hemodialysis

Abstract

Background The most common configuration for arteriovenous fistula is brachiocephalic which often develop cephalic arch stenosis leading to the need for numerous procedures to maintain access patency. The hemodynamics that contributes to the development of cephalic arch stenosis is incompletely understood given the inability to accurately determine shear stress in the cephalic arch. In the current investigation our aim was to determine pressure, velocity and wall shear stress profiles in the cephalic arch in 3D using computational modeling as tools to understand stenosis. Methods Five subjects with brachiocephalic fistula access had protocol labs, Doppler, venogram and intravascular ultrasound imaging performed at 3 and 12 months. 3D reconstructions of the cephalic arch were generated by combining intravascular ultrasounds and venograms. Standard finite element analysis software was used to simulate time dependent blood flow in the cephalic arch with velocity, pressure and wall shear stress profiles generated. Results Our models generated from imaging and flow measurements at 3 and 12 months offer snapshots of the patient’s cephalic arch at a precise time point, although the remodeling of the vessel downstream of an arteriovenous fistula in patients undergoing regular dialysis is a dynamic process that persists over long periods of time (~ 5 years). The velocity and pressure increase at the cephalic bend cause abnormal hemodynamics most prominent along the inner wall of the terminal cephalic arch. The topology of the cephalic arch is highly variable between subjects and predictive of pathologic stenosis at later time points. Conclusions Low flow velocity and wall pressure along the inner wall of the bend may provide possible nidus of endothelial activation that leads to stenosis and thrombosis. In addition, 3D modelling of the arch can indicate areas of stenosis that may be missed by venograms alone. Computational modeling reconstructed from 3D radiologic imaging and Doppler flow provides important insights into the hemodynamics of blood flow in arteriovenous fistula. This technique could be used in future studies to determine optimal flow to prevent endothelial damage for patients with arteriovenous fistula access.