Abstract
BACKGROUND: The cost of islet procurement from donor pigs is increased by the use of organs that produce low yields. We developed an assessment system using dithizone-stained pig pancreas biopsies to enable the preselection of donor organs. METHODS: Pig pancreas biopsy slices were soaked in dithizone solution. The islets were evaluated before islet isolation by converting the islet counts (IC) to islet equivalents (IE), and then determining the IE/cm(2), IE/IC, % islets >150 microm, and % islets >200 microm. These parameters were evaluated in 3 different areas of the pancreas (duodenal, splenic, and connecting lobe; n = 42 each). Stepwise multivariate linear regression analysis was performed to assess for correlations with islet yield and decide which area of the pancreas had the most predictive value. To identify other predictors, including donor and islet isolation variables, we performed binary logistic regression analysis with significant variables from the univariate analysis (n = 67). For this analysis, the pigs were categorized into high (n = 23) and low (n = 44) yield groups. RESULTS: Stepwise multivariate linear regression analysis revealed that IE/cm(2) of the splenic lobe significantly predicted islet yield. Binary logistic regression analysis indicated that the IE/mm(2) of the splenic lobe was the only parameter that significantly correlated with successful pig islet isolations (P = .01; odds ratio 3.605). Variables associated with donor and islet isolation, such as age, gender, ischemic time, or enzyme lot, were not significantly correlated with islet yield. CONCLUSION: Our study suggests that the islet distribution of splenic lobe biopsies can be a reliable predictor of islet yield from pig pancreata.