EARLY POSTOPERATIVE FLEXION OF LOOP-CONFIGURED ARTERIOVENOUS GRAFTS: FOUR CASE ANALYSIS

Arteriovenous grafts (AVGs) for hemodialysis in a loop configuration offer the advantage of providing a wide puncture area. The elasticity of an AVG can differ depending on the implantation technique, even when the same material is used. Kinking or stenosis at the loop apex may occur intraoperatively or within two days postoperatively, typically in almost two patterns. One is caused by excessive curvature at the loop apex, and the other results from excessive redundancy (approximately 1–2 cm) at the artery and AVG anastomosis that secondarily affects the apex, leading to kinking. It is generally considered that kinking rarely occurs at the loop apex with expanded polytetrafluoroethylene (e-PTFE) grafts. However, in our institution, the incidence of kinking-related stenosis was seven times higher with e-PTFE grafts than with polyurethane (PU) grafts. These findings underscore the importance for surgeons to use vascular grafts whose handling characteristics they are familiar with and to match the graft properties appropriately to the surgical technique. (Fig.1)

We defined a “new AVG implantation” as the creation of an artificial graft anastomosed to an autologous artery on one side and an autologous vein on the contralateral side. Between September 2020 and September 2025, a total of 378 graft-related procedures were performed at our institution. Among these, 174 cases involved new loop-type AVG implantations: 167 cases (96.0%) used polyurethane (PU) grafts, and 7 cases (4.0%) used expanded polytetrafluoroethylene (e-PTFE) grafts. Tunneling was performed through a single relay point. Kinking at the loop apex occurred in 4 cases, for which the causes and clinical courses were analyzed.

Among the 4 cases of kinking, 3 cases (1.8%) involved PU grafts and 1 case(14.3%) involved an e-PTFE graft. All kinks occurred at the loop apex. Only 1 case of PUs was successfully corrected by balloon angioplasty, while the remaining 3 cases required surgical revision. During reconstructive surgery, approximately 1 cm of AVG was excised, focusing on the flexion site, followed by end-to-end anastomosis. The operative times were 18 and 32 minutes, and 41 minutes in one case combined with thrombectomy. When more than 1 cm of redundant graft required correction, maintaining patency by balloon angioplasty alone proved difficult. (Fig.2)

Implantation with two-relay-point-operation may reduce the risk of AVG kinking; however, it has the drawback of overlapping two incisions with the AVG, which makes puncture near the scar areas more difficult. In the 3 cases that required reconstruction, the minor axis of the loop was narrower than average arm width cases. Even in patients with thin arms, a loop shape with a broader minor axis should be selected. Forearm loop AVGs operated between the elbow and forearm rarely shorten postoperatively. Whereas brachio-forearm loop AVGs, which form their apex at the middle forearm, often show postoperative shortening. Therefore, a slightly shorter AVG length is preferable in such cases. To prevent kinking or stenosis in loop-type AVG surgeries, it is crucial to adjust the AVG length carefully according to both the loop placement and the anatomical conditions, even when using AVGs made of the same material.