Perkins, James D.

Biliary tract complications are a significant risk following orthotopic liver transplantation (OLT). We reviewed the biliary tract complications of the first 50 liver transplants in 44 recipients performed in our institution. The mean follow‐up was 8 months (range 1 to 16 months). Biliary tract complications that required surgical or instrumental intervention developed in 17 of the 50 (34%) transplants. No deaths were attributable to biliary tract complications. Hepatic artery thrombosis caused the biliary tract complication in 3 transplants (1 stenosis, 2 anastomotic leaks) and a cystocele of the cystic duct remnant caused 1 common bile duct obstruction. The other 13 complications were associated either with the type of anastomosis performed or type of stent used. An end‐to‐end choledochocholedochostomy was performed in 31 transplants and 2 developed biliary tract complications (6.5%). One developed a biliary leak and the other developed a stenosis at the anastomosis. Nineteen of these choledochocholedochostomies were stented with a straight tube and 6 developed stent‐related complications. Twelve were stented with a T‐tube and only 1 stent complication occurred. An end‐to‐side choledochojejunostomy with a Roux‐en‐Y loop was performed in 19 patients. Only 1 early biliary complication occurred (5%), which was an anastomotic leak. However, 2 developed a stenosis of the entire donor extra‐hepatic duct, 12 and 13 months following transplantation. All anastomoses were stented with a straight tube and only 1 leak developed at the entrance site of the stent into the Roux‐Y. We conclude that the choledochocholedochostomy and the choledochojejunostomy with a Roux‐en‐Y loop are safe methods of biliary reconstruction following OLT. We recommend that a T‐tube be used for stenting the choledochocholedochostomy, and a hepatic artery thrombosis must be excluded when biliary tract complications develop.

The importance of portal insulin delivery in the regulation of postprandial carbohydrate metabolism is uncertain. To address this question, three groups of dogs were studied: one group in which pancreatic venous drainage was transected and reanastomosed (portal insulin delivery), one in which the pancreatic drainage was transected and anastomosed to the inferior vena cava (peripheral insulin delivery), and one that received only a sham operation. Plasma insulin was greater (P < 0.05) during peripheral insulin delivery than in either the portal or sham groups, respectively, before and after meal ingestion. On the other hand, C-peptide concentrations did not differ between groups, resulting in a higher (P < 0.001) insulin to C-peptide ratio in the peripheral group. This indicated that the hyperinsulinemia in the peripheral group was due to decreased insulin clearance rather than increased insulin secretion. Isotopically determined splanchnic uptake of ingested glucose, postprandial suppression of hepatic glucose release, incorporation of CO2 into glucose (a qualitative measure of gluconeogenesis), and total-body glucose uptake were virtually identical in all groups. Similarly, plasma lipid, β-hydroxybutyrate, and lactate concentrations did not differ between groups. Our data indicate that, despite differences in systemic insulin concentration, portal and peripheral insulin delivery comparably regulate hepatic and extrahepatic carbohydrate metabolism after meal ingestion.