Quantitative evaluation of retrograde arterial flow in chronic experimental arteriovenous fistulas.

A 2 cm long iliac av. fistula is established in 13 dogs. Flow measurements are performed in the fistula limbs and the adjacent arteries immediately (stage I), as well as 3 months (stage II) and 6 to 12 months (stage III) post-shunt. The following results are obtained: 1) In the acute stage, average flow in the distal fistula artery is very small but directed towards the periphery. In stages II and III a progressively increasing retrograde flow directed towards the fistula is regularly measured in the distal fistula artery. 2) Even in the acute stage a considerable retrograde flow is measured in the distal fistula artery when the proximal artery is clamped. Likewise, in stages II and III, this "free" retrograde flow (after occlusion of proximal fistula artery) is 40 to 60% higher than the spontaneous retrograde flow. 3) In chronic av. fistulas, the "free" retrograde flow in the distal fistula artery decreases significantly after occlusion of the adjacent tail artery and the contralateral iliac artery. It can thus be calculated that in chronic iliac av. fistulas the tail artery contributes approximately 44% and the contralateral iliac artery 22% to retrograde arterial fistula flow. The remaining 34% of retrograde flow are mainly derived from collaterals connecting side-branches of the abdominal aorta with the distal arterial limb. 4) These extensive collaterals adjacent to a chronic iliac av. fistula are demonstrated angiographically and by post-mortem vascular casts. 5) In all stages, occlusion of the proximal fistula artery simultaneously increases flow in the tail artery and contralateral iliac artery by more than 50%. This stresses the major contribution of these arteries to retrograde arterial fistula flow. 6) Successive occlusion of the tail artery and the contralateral iliac artery in stages II and III results in a nearly identical per cent decrease in "free" retrograde flow and in pressure (measured in the distal fistula artery). Even in the pre-fistula control stage, this occlusion experiment induces the same relative pressure-drop in the distal iliac artery. The proportional contribution of the collateral pathways to retrograde arterial fistula-flow and -pressure remains unchanged in the chronic stages post-fistula in spite of the tremendous expansion of collateral vessels within one year. These findings indicate that the three main collateral networks adjacent to the iliac artery all expand proportionally to their pre-fistula status in the months following construction of a large iliac av. fistula. This strongly suggests that the extensive interarterial anastomoses in a chronic av. fistula are all preformed collaterals which have gradually expanded.