Hybrid arteriovenous graft for hemodialysis vascular access in a multicenter registry.

ABSTRACT

OBJECTIVE: The aim of our study was to identify patients' characteristics that predicted a higher chance of arteriovenous graft patency in patients undergoing Gore Hybrid Vascular Graft (GHVG; W. L. Gore & Associates, Flagstaff, Ariz) implantation for hemodialysis access. The GHVG is a polytetrafluroethylene (PTFE) prosthesis with a nitinol-reinforced section (NRS) at the venous end. METHODS: All consecutive patients undergoing GHVG implantation for hemodialysis access at 10 tertiary referral centers between December 2013 and January 2018 were included in the study and compared with a control group of patients undergoing standard PTFE graft implantation. Selection of patients for hybrid graft implantation was based on the impossibility of autogenous vascular access creation. RESULTS: There were 145 patients included in the GHVG group and 218 in the PTFE group. In the GHVG and the PTFE groups, the mean age was 67 ± 13 years and 65 ± 13 years, and male patients totaled 52% and 46%, respectively. The technical success was 99%. The mean duration of the intervention was 100 minutes (median, 95 minutes; interquartile range, 80-120 minutes). The brachial-axillary configuration was used in the majority of cases (n = 78 [54%]). The 5-cm NRS length was prevalent (n = 108 [75%]). The median NRS oversize was 14% (interquartile range, 0%-21%). Mean follow-up was 13 months (range, 0-55 months). Seventy-one patients (49%) underwent at least one reintervention. Primary, assisted primary, and secondary patency estimates at 12 months were 44% ± 5%, 47% ± 5%, and 65% ± 4% for the GHVG group and 41% ± 4%, 53% ± 4%, and 75% ± 3% for the control group, respectively (P = NS). One-year survival was 90% ± 3%. On multivariable Cox regression analysis, hypotension (P < .001; hazard ratio [HR], 5.8; confidence interval [CI], 2.6-13) and diabetes (P = .024; HR, 1.9; CI, 1.1-3.2) were significant predictors of GHVG loss. A larger graft size was protective against GHVG loss (P = .042; HR, 0.73; CI, 0.54-0.99). The 10-cm-long graft showed a tendency toward improved patency but did not reach statistical significance (P = .074; HR, 0.48; CI, 0.21-1.07). CONCLUSIONS: Diabetes and hypotension were predictors of loss of hybrid arteriovenous access. Smaller diameters of NRS were more prone to thrombosis, whereas the 10-cm length seemed to perform better than the 5-cm one.