Introduction: Enteric drainage is the most common method of duct drainage in pancreas transplantation. Enteric leak may develop in up to 5-8% of patients, frequently necessitating immediate pancreatectomy of a functioning pancreas for source control. Pancreatic allograft salvage has been described but little is detailed on outcomes after attempted salvage. We examined risk factors for enteric leak and outcomes after attempted graft salvage.

Methods: We performed a single center retrospective review of pancreas transplants performed between 1995-2016. Demographics and transplant characteristics were compared between those who developed an enteric leak and those who did not using t-tests, Chi-Squared tests, or Fisher’s exacts tests. Graft and patient outcomes were examined using Cox proportional hazards models with a time-dependent covariate for enteric leak status and Kaplan-Meier curves.

Results: A total of 1,153 patients were assessed with an enteric leak rate of 2.9%. Donors of allografts that developed leak were older (37.9 vs. 29.0, p=0.001), had a higher KDPI (37% vs. 24%, p<0.001), a higher pancreas donor risk index (1.83 vs. 1.32, p<0.001), and a longer cold ischemic time (16.5 vs. 14.8 hours, p=0.03). On multivariate analysis, the hazard of enteric leak was seen to increase 5% with every year of donor age (p<0.001) and 10% with every hour of pancreas cold ischemic time (p=0.02). Enteric leak increased the chance of pancreas graft loss within the first 6 months of transplant (HR 13.9 [CI 8.5-22.9], p<0.001, Fig. 1(a)). However, 50% of allografts undergoing attempted salvage survived and maintained long term function, with similar long-term pancreatic graft survival between patients who achieved successful salvage and patients who did not experience enteric leak (82.5% salvage vs. 81.5% control, p=0.76, Fig. 1(b)). Kidney graft survival was significantly better in patients who underwent successful salvage than in patients who lost their pancreas to immediate pancreatectomy or failed salvage (90.0% successful salvage vs. 38.5% pancreatectomy, p<0.001, Fig. 2). Among those with at least 6 months of patient and graft survival, five-year patient survival was similar between salvage and non-leak groups (90.9% leak vs. 93.5%, p=0.76). 

Conclusions: Pancreatic allograft salvage can be safely performed with a 50% rate of successful salvage and without increased mortality.