Crossover Study on Effects of Duct Obliteration, Celiac Denervation, and Autotransplantation on Glucose- and Meal-Stimulated Insulin, Glucagon, and Pancreatic Polypeptide Levels

In segmental-pancreas transplantation the body and tail of the pancreas are used. In an experimental study in dogs, the effects of sequentially conducted removal of the right pancreatic lobe (pancreatic head), duct obliteration, celiac denervation, and autotransplantation were studied according to a crossover design. Two groups of dogs were studied. In both groups the right lobe of the pancreas was removed at primary operation, and the duct of the transected left lobe (body and tail) was injected with fibrin sealant. The left lobe was completely freed from surrounding tissue (celiac denervation) in group 1 (n = 9), and the innervation of the left lobe was left intact in group 2 (n = 8). At 12 wk, two dogs in group 1 and four dogs in group 2 underwent successful autotransplantation of the left lobe. Pancreatic hormone secretion was stimulated by intravenous glucose injection and test-meal administration before primary operation and at 11 and 18 wk thereafter. The combination of removal of the right lobe and duct obliteration led to a decrease in glucose tolerance at both stimulation tests and a decrease in peripheral insulin release after intravenous glucose injection. At test-meal administration, no change in insulin and glucagon levels was demonstrated. If ceiiac denervation was added, similar results were obtained based on the understanding that the peripheral insulin release after the test meal was significantly elevated. Meal-stimulated pancreatic polypeptide response was abolished in both groups. Removal of the right lobe leads to parasympathic denervation of the left lobe, and celiac denervation mainly interferes with α-adrenergic innervation. Cholecystokinin was probably responsible for the absence of a decrease in insulin secretion at test-meal administration after primary operation. Secondary autotransplantation led to elevation of peripheral insulin release at both tests but had no demonstrable influence on glucose tolerance expressed in K values.