An in situ evaluation of distal splenic arteriovenous fistula on pancreas function in an isolated pancreas segment

P. J. Garvin, M. A. Castaneda, M. L. Niehoff, K. A. Mauller and J. J. Brems

To determine the effects of a distal splenic arteriovenous fistula on endocrine function and pancreatic blood flow, 25 dogs underwent proximal pancreatectomy with the pancreatic tail left in situ and free intraperitoneal drainage of the pancreatic duct. Group A served as controls. In groups B through E, ligation of all nonpancreatic splenic vessels was accomplished. In group B, no further manipulations were performed. In group C, an arteriovenous fistula was created. Groups D and E were identical to groups B and C, respectively, except for the induction of bile pancreatitis. During intravenous glucose tolerance testing, the mean (+/- SEM) basal-to-peak insulin difference was 10.1 +/- 3.5 microU/mL in group A, 16.3 +/- 3.6 microU/mL in group B, 14.8 +/- 5.1 microU/mL in group C, 16.4 +/- 3.1 microU/mL in group D, and 13.0 +/- 4.4 microU/mL in group E. Corresponding mean (+/- SEM) glucose clearance values were as follows: -0.907% +/- 0.24%/min, -0.867% +/- 0.14%/min, -1.056% +/- 0.21%/min, -1.365% +/- 0.26%/min, and -0.887% +/- 0.20%/min. These values were not significantly different. Ligation of all splenic arterial and venous branches resulted in a 64.8% to 78.3% reduction in splenic artery blood flow that was restored to 60.9% to 84.9% of basal flow by an arteriovenous fistula (groups C and E). In conclusion, the creation of a splenic arteriovenous fistula was not beneficial in this model and other factors (rejection or technical) should be considered in vascular thrombosis following segmental pancreatic transplantation.