Since 1981 femoral arteryâveinring vascular access problems. ... All patients in whom a femoral artery-vein loop graft was ..... femoral artery and the draining femoral vein, as many central veins in the chest and neck free for insertion.
Nephrol Dial Transplant (1998) 13: 1215–1220
Nephrology Dialysis Transplantation
Original Article
The femoral artery–femoral vein polytetrafluoroethylene graft: a 14-year retrospective study Asher Korzets1, Yaacov Ori1, Shlomo Baytner2, Dina Zevin1, Avry Chagnac1, Talia Weinstein1, Michal Herman1, Morris Agmon2 and Uzi Gafter1 1Department of Nephrology and 2Vascular Surgery Unit, Rabin Medical Center, Golda (Hasharon) Campus, Petah Tikva, The Sackler School of Medicine, Tel Aviv University, Tel Aviv, Israel
Abstract Background. The use of the femoral vessels for permanent haemodialysis access has been neglected during the last two decades. Since 1981 femoral artery–vein loop polytetrafluoroethylene grafts have been constructed in our chronic haemodialysis patients. This study examines results obtained in patients with this particular graft over the last 14 years. Methods. This clinical study is retrospective in nature. Overall 35 patients, with 37 femoral grafts, are included. Inclusion and exclusion criteria for this type of graft are given and the surgical procedure detailed. Results. Seven patients had femoral grafts used as primary dialysis access. Twenty-eight patients had femoral grafts used after multiple access failures. There was no perioperative mortality. Immediate thrombotic non-function of the graft occurred in three patients. In the long term no patient death was related to the femoral grafts. Twenty-seven (73%) grafts had no longterm complications. The leading cause for graft ‘loss’ was patient death; in the first year 10 grafts were lost, eight because of patient death. All eight patients died with functioning grafts. Median graft survival was 21 months in all patients and 28 months in non-diabetic patients. Twenty-seven (73%) grafts were patent at the end of the first year, 33% of grafts were still patent after 5 years. Worsening claudication occurred in four patients; one diabetic required foot amputation. Four patients had late graft thrombosis; only two patients had bacteraemia originating from the femoral graft. Urea reduction ratio greater than 60% was measured in 87.5% of patients. Conclusion. The femoral artery–vein graft is a good primary and secondary haemodialysis access. Both infection and thrombosis rates are low and graft survival is comparable, if not superior to, that of upperlimb grafts. The graft is easy to cannulate, can be used early, is easily protected, and is cosmetically acceptable.
Correspondence and offprint requests to: U. Gafter MD PhD, Department of Nephrology, Rabin Medical Center—Golda Campus, 7 Keren Kayemet St, Petah Tikva 49372, Israel.
Introduction Many haemodialysis patients are plagued with recurring vascular access problems. An arteriovenous fistula remains the preferred choice for initial access, but expanded polytetrafluoroethylene (PTFE) grafts have become an accepted alternative [1]. PTFE grafts are easy to use, easy to revise, and withstand repeated cannulations for many years [2]. However, the literature concerning such grafts in the thigh is scarce. The use of the Thomas shunt in the femoral vessels has been completely abandoned because of a high rate of infection and the danger of secondary haemorrhage from either the arterial or the venous side of the shunt [3]. In 1980 Morgan et al. described their experience with 14 PTFE femoral triangle grafts, and they concluded that infection with these grafts was a major problem [4]. Since then only one further report in the English literature has described clinical experience with PTFE thigh grafts [5]. In 1981 a femoral artery–femoral vein loop PTFE graft was first used in our unit as a haemodialysis access. This retrospective study details the clinical results obtained in 37 thigh grafts performed until June 1995.
Subjects and methods All patients in whom a femoral artery-vein loop graft was constructed, between January 1982 to June 1995, entered this retrospective study. Inpatient and dialysis medical charts were analysed for the following parameters: patient characteristics, underlying renal disease, previous vascular accesses, reason for using the femoral artery–vein graft in each patient, immediate perioperative complications and their treatment, graft survival after ‘creation’ of the graft, late complications associated with it, adequacy of dialysis while using this graft, and cause of graft failure. All patients with overt peripheral vascular disease, or an echo-Doppler examination of the lower-limb arteries demonstrating proximal arterial occlusions, were excluded as possible candidates for a femoral vessel graft. The study end-point was arbitrarily established at June 1995. This enabled the study to have at least a
© 1998 European Renal Association–European Dialysis and Transplant Association
1216
12-month follow-up period until June 1996. Graft survival was defined as that period of time until the graft failed for whatever reason, or until the patient’s death. In this study no differentiation was made between primary and secondary patency. Surgical thrombectomy was performed whenever required. The Rabin Medical Center—Golda (Hasharon) Campus serves as a tertiary referral unit for haemodialysis access problems and femoral grafts were used in both patients subsequently dialysed by our unit and by other hospitals. Only patients dialysed in the Nephrology Department of the Rabin Medical Center—Golda (Hasharon) Campus were included in this study.
Surgical procedure Intravenous cefazolin 1.0 g is given 1 h preoperatively. The lower abdomen and the thigh down to the ipsilateral knee are prepared. A longitudinal incision of the skin (approximately 6 cm) is then made below the inguinal ligament, over the anteromedial aspect of the thigh. The superficial femoral artery, below its exit from the common femoral artery, and the femoral vein and its branches are exposed. A lateral longitudinal arteriotomy (1–1.5 cm long) is made into the superficial femoral artery. A PTFE graft (internal diameter 6 mm) is cut obliquely at 45°, and its end positioned to the arteriotomy opening. The graft is then tunnelled inferiorly in a subcutaneous plane. The distal end of the loop lies approximately 8–10 cm superiorly to the knee. At this area a further skin incision is made to ensure that no kinks are present in the graft. The graft is then turned superiorly in the subcutaneous plane until it reaches the exposed femoral vein. Average length of the loop is approximately 25–30 cm. A medial longitudinal venotomy (2 cm) is made, the graft cut obliquely, and the veno-graft anastomosis made. After both venous and arterial clamps are removed, an immediate thrill should be heard over the entire graft. Sutures of the superior longitudinal incision are removed after 14 days. Sutures of the small inferior incision are removed after 21 days. Cannulation of the graft is recommended after 10–14 days, but successful early cannulation, even after 48 h, can be performed. In this later situation haemodialysis without anticoagulation or with low molecular weight heparin is recommended. Postoperatively, dipyridamole, 75 mg thrice daily, was given indefinitely to all patients. Low-dose aspirin (100 mg) was also given indefinitely to all patients without a contraindication to this drug. A failing graft was suspected when venous pressures rose inexplicably, or when prolonged bleeding after decannulation was noted. Angiography was then performed. Routine screening angiography or sonographic examination of the graft were not performed if the graft was functioning well.
Statistical analysis Comparisons of non-diabetic vs diabetic median graft survival was made using the Mann–Whitney U test. Two-tailed P60 in 35 (87.5%) measurements. URR 60 was obtained. In the other two patients the constantly low URR (
