PURPOSE: To study the hemodynamic effects of an arteriovenous fistula (AVF) used as an adjunct to venous reconstructions and to determine the optimal size for such a fistula. METHODS: A model of limb circulation with an AVF (in vitro system) was constructed with silicon elastic tubes and 40% glycerin solution as the fluid medium. Pulsatile arterial flow and venous return was maintained with a roller pump and a centrifugal assist device. Flows and pressures were measured for three different fistula diameters (3, 4, and 5 mm). A canine model of venous hypertension with outflow obstruction was constructed in 15 adult mongrel dogs. After 7 to 13 days an externally supported 8-mm expanded polytetrafluoroethylene femorocaval graft was implanted with a distal AVF (3 mm, n = 5; 4 mm, n = 5; 5 mm, n = 5). Arterial and venous flows and venous pressures were measured proximal and distal to the fistula before and after exercise. RESULTS: In the in vitro system, flows through the venous graft increased with increasing fistula size, but venous return decreased progressively, increasing the distal venous pressure. In the canine model, flow in the venous graft increased with each AVF (p < 0.01). Only the 3-mm AVF resulted in decreased distal femoral vein pressure (p < 0.01), orthograde flow, and improved venous return with exercise. CONCLUSION: AVFs increased flow through the femorocaval grafts, yet they impeded venous return. The ideal AVF-to-graft ratio used in our study was 0.375, because it increased graft flow, permitted forward flow in the femoral vein while reducing pressure, and improved venous return with exercise.
PURPOSE: To study the hemodynamic effects of an arteriovenous fistula (AVF) used as an adjunct to venous reconstructions and to determine the optimal size for such a fistula. METHODS: A model of limb circulation with an AVF (in vitro system) was constructed with silicon elastic tubes and 40% glycerin solution as the fluid medium. Pulsatile arterial flow and venous return was maintained with a roller pump and a centrifugal assist device. Flows and pressures were measured for three different fistula diameters (3, 4, and 5 mm). A canine model of venous hypertension with outflow obstruction was constructed in 15 adult mongrel dogs. After 7 to 13 days an externally supported 8-mm expanded polytetrafluoroethylene femorocaval graft was implanted with a distal AVF (3 mm, n = 5; 4 mm, n = 5; 5 mm, n = 5). Arterial and venous flows and venous pressures were measured proximal and distal to the fistula before and after exercise. RESULTS: In the in vitro system, flows through the venous graft increased with increasing fistula size, but venous return decreased progressively, increasing the distal venous pressure. In the canine model, flow in the venous graft increased with each AVF (p < 0.01). Only the 3-mm AVF resulted in decreased distal femoral vein pressure (p < 0.01), orthograde flow, and improved venous return with exercise. CONCLUSION: AVFs increased flow through the femorocaval grafts, yet they impeded venous return. The ideal AVF-to-graft ratio used in our study was 0.375, because it increased graft flow, permitted forward flow in the femoral vein while reducing pressure, and improved venous return with exercise.