In order to design a feasible somatic cell gene delivery sysgogues. While glucose responsiveness commenced at a tem for the treatment of type I diabetes, a suitable cell type lower concentration than normal islets, a secretion curve needs to be determined. We have previously shown that approaching normal physiological conditions was generthe stable transfection of the full-length insulin cDNA into ated. Immunoelectron microscopy revealed the presence the human liver cell line, (HEP G2ins) resulted in synthesis, of insulin-containing granules, similar in size and appearstorage and acute regulated release of insulin to analogues ance to those of the normal beta cell. These results demof cAMP, but not to the physiological stimulus glucose. In onstrate that while it is most likely that the HEP G2ins/g attempting to explain the lack of glucose responsiveness cell line predominantly secretes insulin via the constitutive of the HEP G2ins cells we have stably transfected these pathway, significant acute regulated release was seen in cells with the human islet glucose transporter GLUT 2 response to glucose, and thus represents significant pro-(HEP G2ins/g cells). The HEP G2ins/g cell clones exhibit gress in the creation of a genetically engineered 'artificial glucose-stimulated insulin secretion and glucose potentibeta cell' from a human hepatocyte cell line. ation of the secretory response to nonglucose secreta-

No abstract

Up to 60% of patients (pts) undergoing LDLT develop biliary complications, most within the fi rst 3 months postoperatively (PO). We describe a change in anastomotic technique that signifi cantly decreased the incidence of biliary complications at our center. This retrospective is a longitudinal cohort analysis of 77 LDLT completed at our institution since 2001. For the fi rst 59 LDLT, the biliary anastomosis, either duct-to-duct (49) or Roux-en-Y (10) reconstruction, were completed with a running absorbable suture (Group I). Since 2006 all our LDLT to date (16) have been completed with a running posterior line and an interrupted anterior row of stitches also using absorbable sutures (Group II). Since then, the rate of biliary complications (leak and stricture), has signifi cantly decreased. In Group I bile leaks occurred in 44% of pts and strictures in 63%, many of the leaks occurring after endoscopic balloon dilatation of strictures. Pts in Group II had a 6% incidence of both biliary leaks and strictures (p=0.0068 and p=0.0001 respectively) ( Table 1). The mean follow-up was 47 months for Group

No abstract