Incidence of pancreas graft thrombosis in portoiliac and portocaval venous anastomosis

Background. Pancreas graft thromboses represent more than 70% of all technical failures; multiple risk factors have been implicated. We analyzed the thrombosis rates using portoiliac versus portocaval vein anastomoses. Patients and methods. The series includes 53 patients who underwent pancreas transplantation: 49 simultaneous pancreas-kidney and 4 pancreas after kidney. There were 27 men and 26 women, of mean age of 37.2 +/- 7.0 years. We compared two groups of recipients that were classified according to venous anastomosis: (A) portoiliac (n = 30), and (B) portocaval (n = 23). Results. The recipients did not show significant differences in age, gender, or duration of diabetes mellitus, but body mass index was significantly higher among the portocaval group. A bladder-drained pancreas technique was more frequently performed in the portoiliac group (93% of patients) versus an enteric-drained pancreas in the portocaval group (81%; P < .001). Heparinization was performed in 12 recipients: 11 (36.6%) in the portoiliac group and 1 (4.3%) in the portocaval group (P < .01). Vascular graft thrombosis (venous in six and arterial in one) developed in seven patients (13.2%) all in the portoiliac group (23%) (P < .02). Two-year patient survival was 93% in the portoiliac group and 94% in portocaval group (P = NS). Two-year graft survival was 66.6% in the portoiliac group and 85.9% in portocaval group (P = .07). Conclusion. There was no graft thrombosis among patients with a portocaval vein anastomosis.