RESUMEN
The purpose of this retrospective study was to analyze 2 sizes of expanded polytetrafluoroethylene (PTFE) upper arm grafts for dialysis: 8 millimeters, tapered to 6 mm at the arterial side, and 6 millimeters. All upper arm PTFE grafts (Gore-Tex(R)) were performed between January 1981 and April 1997. Patient characteristics, complication rate, and patency rates were compared for both kind of grafts. Five hundred and seven PTFE grafts were analyzed (183 6-mm grafts and 324 6- to 8-mm grafts). Early failure was found in 5 grafts (2.7%) in 6-mm grafts, and in 5 grafts (1.5%) in 8-mm grafts (not significant). Steal syndrome was found in 1 patient (0.5%) of the 6-mm group, and in 11 (3.4%) of the 8-mm grafts (p=0.085). The rate of late complications requiring surgical repair was 0.56 episode per graft-year in the 6-mm grafts group, and 0.33 in the 8-mm grafts (p<0.001). Primary patency rates of 6-mm grafts were 72%, 33%, and 19% at 1, 3, and 5 years; and secondary patency rates were 86%, 68%, 56%, and 44% at 1, 3, 5, and 6 years, respectively. In the 8-mm grafts group, primary patency rates were 77%, 52%, and 39% at 1, 3, and 5 years; and secondary patency rates were 92%, 84%, 73%, and 66% at 1, 3, 5, and 6 years, respectively. Comparison of patency rates of 6-mm and 8-mm grafts were statistically significant (p<0.001) for both primary and secondary curves. However, secondary survival curves were similar for both kind of grafts in a subpopulation of diabetic patients. The authors conclude that the 8-mm graft, tapered to 6 mm at the arterial side, is a dialysis graft with fewer complications and a better patency rate than grafts of 6 mm placed in the same anatomical position, at least in a population of nondiabetic patients. Steal syndrome was observed in some cases of diabetic and older patients with a large-bore graft. Thus, this kind of prosthesis should be avoided in this population. On the other hand, this is not a prospective, randomized study made with any intention for comparison. Therefore, the aforementioned conclusions must be cautiously considered.