Effect of anastomosis angle on the localization of disturbed flow in 'side-to-end' fistulae for haemodialysis access.

BACKGROUND: Early failure of the vascular access for haemodialysis (HD) after the surgical creation of a radial-cephalic arteriovenous fistula (AVF) occurs mainly due to a juxta-anastomotic stenosis. Even if elevated blood flow induces high wall shear stress, we have recently shown that disturbed flow, characterized by low and reciprocating flow, may develop in zones of the AVF where it can provide a good indication of the sites of future stenoses. The present study was aimed at investigating whether the anastomosis angle influences disturbed flow in radial-cephalic 'side-to-end' AVF. METHODS: By means of a parametric AVF model we created four equivalent meshes with anastomosis angles of 30°, 45°, 60° and 90°, respectively. We then performed transient, non-Newtonian computational fluid dynamics simulations using, as boundary conditions, previously measured blood volume flow and division ratio in subjects requiring primary access. The relative residence time (RRT), a robust indicator of disturbed flow, was calculated for the overall wall surface and disturbed flow was localized as areas having RRT > 1. Quantitative characterization and statistical tests were employed to assess the difference in RRT medians between the four anastomosis angle cases. RESULTS: Disturbed flow was located in all AVF models in the same areas where flow recirculation and stagnation occurred, on the inner wall of the swing segment (SS) and on the arterial wall at the anastomosis floor (AF). A smaller angle AVF had smaller disturbed flow areas with lower RRT peak values, either on the venous or the arterial limb. There were significant differences in the RRT medians on the SS and on the AF between sharper (30° and 45°) and wider (60° or 90°) angles. CONCLUSIONS: We have found that in 'side-to-end' radial-cephalic AVFs for HD, the anastomosis angle does impact on the local disturbed flow patterns. Among the four geometries we considered in this study, the smaller angle (30°) would be the preferred choice that minimizes the development of neointima. Clinicians should consider this at the time of AVF creation because the anastomosis angle is in part amenable to surgical manipulation.