In vitro hemodynamic model of the arm arteriovenous circulation to study hemodynamics of native arteriovenous fistula and the distal revascularization and interval ligation procedure.

Background: Experimental modeling of arteriovenous hemodialysis fistula (AVF) hemodynamics is challenging. Mathematical modeling struggles to accurately represent the capillary bed and venous circulation. In vivo animal models are expensive and labor intensive. We hypothesized that an in vitro, physiologic model of the extremity arteriovenous circulation with provisions for AVF and distal revascularization and interval ligation (DRIL) configurations could be created as a platform for hemodynamic modeling and testing.

Methods: An anatomic, upper extremity arteriovenous model was constructed of tubing focusing on the circulation from the subclavian artery to subclavian vein. Tubing material, length, diameter, and wall thickness were selected to match vessel compliance and morphology. All branch points were constructed at physiologic angles. The venous system and capillary bed were modeled using tubing and one-way valves and compliance chambers. A glycerin/water solution was created to match blood viscosity. The system was connected to a heart simulator. Pressure waveforms and flows were recorded at multiple sites along the model for the native circulation, brachiocephalic AVF configuration, and the AVF with DR without and with IL (DR no IL and DRIL).

Results: A preset mean cardiac output of 4.2 L/min from the heart simulator yielded a subclavian artery pressure of 125/55 mm Hg and a brachial artery pressure of 121/54 mm Hg with physiologic arterial waveforms. Mean capillary bed perfusion pressure was 41 mm Hg, and mean venous pressure in the distal brachial vein was 17 mm Hg with physiologic waveforms. AVF configuration resulted in a 15% decrease in distal pressure and a 65% decrease in distal flow to the hand. DR no IL had no change in distal pressure with a 27% increase in distal flow. DRIL resulted in a 3% increase in distal pressure and a 15% increase in distal flow to the hand above that of DR no IL. Flow through the DR bypass decreased from 329 mL/min to 55 mL/min with the addition of IL. Flow through the AVF for both DR no IL and DRIL was preserved.

Conclusions: Through the construction and validation of an in vitro, pulsatile arteriovenous model, the intricate hemodynamics of AVF and treatments for ischemic steal can be studied. DR with or without IL improved distal blood flow in addition to preserving AVF flow. IL decreased the blood flow through the DR bypass itself. The findings of the AVF as a pressure sink and the relative role of IL with DR bypass has allowed this model to provide hemodynamic insight difficult or impossible to obtain in animal or human models. Further study of these phenomena with this model should allow for more effective AVF placement and maturation while personalizing treatment for associated ischemic steal.

Clinical Relevance: The complications of arteriovenous fistula (AVF)-associated steal with its concurrent surgical treatments have been clinically described but have relatively little published, concrete hemodynamic data. A further understanding of the underlying hemodynamics is necessary to prevent the occurrence of steal and improve treatment when it occurs. Specific objectives are to study the blood flow through an AVF with varying anatomic and physiologic parameters, determine what factors contribute to the development of arterial steal distal to an AVF, and create optimal interventions to treat arterial steal from an AVF when it occurs. The long-term goal is creation of AVF tailored to patient-specific parameters, resulting in higher rates of functional fistulas with decreases in fistula-related complications. The ability to study fluid dynamics using a unique, in vitro, upper extremity pulsatile arteriovenous circulation simulator creates the ideal platform for this work.