Functional diagnostic parameters for arteriovenous fistula.

Abstract

The inability to detect the arteriovenous fistula (AVF) dysfunction in a timely manner under the current surveillance programs, which are based on either diameter (d), flow rate (Q), or pressure (p) measurements, is one of the major challenges of dialysis treatment. Thus, our aim is to introduce new functional diagnostic parameters that can better predict AVF functionality status. Six AVFs were created between the femoral arteries and veins of three pigs, each pig having two AVFs on either limb. Flow fields and pressure drop (Δp) in AVFs were obtained via numerical analysis utilizing the CT scan and Doppler ultrasound data at 2D (D: days), 7D, and 28D postsurgery. The dataset included 16 (two pigs [four AVFs] for three time points, and one pig [two AVFs] for two time points) repeated measurements over time, and the statistical analysis was done using a mixed model. To evaluate the nature of pressure drop-flow relationships in AVFs, the Δp was correlated with the average velocity at proximal artery (v) and also the corresponding scaled velocity (v*) by the curvature ratio of anastomotic segment. Based on these relationships, two new functional diagnostic parameters, including the nonlinear pressure drop coefficient (Cp ; pressure drop divided by dynamic pressure at proximal artery) and the linear resistance index (R; pressure drop divided by velocity at proximal artery), were introduced. The diagnostic parameters that were calculated based on scaled velocity are represented as R* and Cp *. A marginal (P = 0.1) increase in d from 2D (5.4 ± 0.7 mm) to 7D (6.8 ± 0.7 mm), along with a significant increase in Q (2D: 967 ± 273 mL/min; 7D: 1943 ± 273 mL/min), was accompanied by an almost unchanged Δp over this time period (2D: 16.42 ± 4.6 mm Hg; 7D: 16.40 ± 4.6 mm Hg). However, the insignificant increase in d and Q from 7D to 28D (d = 7.8 ± 0.8 mm; Q = 2181 ± 378 mL/min) was accompanied by the elevation in Δp (24.6 ± 6.5 mm Hg). The functional diagnostic parameters, R and Cp , decreased from 2D (R = 22.4 ± 2.8 mm Hg/m/s; Cp = 12.0 ± 2.6) to 7D (R = 20.8 ± 2.8 mm Hg/m/s; Cp = 8.1 ± 2.6), and then increased from 7D to 28D (R = 35.5 ± 5.7 mm Hg/m/s; Cp = 17.5 ± 3.6) with a marginal significance. However, when the scaled velocity was used to calculate R* and Cp *, the increase in diagnostic parameters from 7D to 28D achieved statistical significance (P < 0.05). In summary, although the differences in the hemodynamic parameters (d, Q, and Δp) from 7D to 28D were insignificant, changes in their combined effects in the form of diagnostic parameters were significant. Therefore, the functional diagnostic parameters are capable of better distinguishing changes in the hemodynamic variations, and thus, could be promising endpoints to diagnose the functionality of AVFs over time.