OBJECTIVEThis study assesses β-cell replication in human donor organs and examines possible influences of the preterminal clinical conditions.RESEARCH DESIGN AND METHODSβ-Cell replication was quantified in a consecutive series of n = 363 human organ donors using double immunohistochemistry for Ki67 and insulin. Uni- and multivariate analysis was used to correlate replication levels to clinical donor characteristics and histopathologic findings.RESULTSβ-Cell replication was virtually absent in most donors, with ≤0.1% Ki67-positive β-cells in 72% of donors. A subpopulation of donors, however, showed markedly elevated levels of replication of up to 7.0% Ki67-positive β-cells. β-Cell replication was accompanied by the increased replication of glucagon-, somatostatin-, and CA19.9-positive cells. Prolonged life support, kidney dysfunction, relatively young donor age, inflammatory infiltration, and prolonged brain death before organ retrieval were all found to be significantly associated with an increased level (≥90th percentile) of β-cell replication, with the first three risk factors being independent predictors. Increased β-cell replication was most often noted in relatively young donors (≤25 years) who received prolonged (≥3 days) life support (68%); in contrast, it was rare in donors with a short duration of life support regardless of age (1%). Prolonged life support was accompanied by increased levels of CD68+ and LCA/CD45+ infiltration in the pancreatic parenchyma.CONCLUSIONThese results indicate that preterminal clinical conditions in (young) organ donors can lead to increased inflammatory infiltration of the pancreas and to increased β-cell replication.

Aims/hypothesis. Pancreatic ducts are considered as potential sites for neogenesis of beta cells. In vitro studies have reported formation of islets from postnatal human and rodent duct tissue. We examined whether postnatal human duct-cell preparations can generate new beta cells after transplantation. Methods. Pancreatic duct cells were prepared from the non-endocrine fraction of human donor pancreases that were processed for islet-cell isolation. Grafts containing 0.5 million duct cells with 1% contaminating insulin-positive cells were implanted under the kidney capsule of normoglycaemic nude mice. At 0.5 and 10 weeks post-transplantation, implants were examined for their cellular composition and for the volumes of their composing cell populations, i.e. cytokeratin 19-positive duct cells, synaptophysin-, insulin-and glucagon-positive endocrine cells. Results. Between week 0.5 and 10, duct-cell volume decreased by at least 90% whereas the change in insulin-positive cell volume depended on donor age. Implants from donors over10 years had a threefold decrease in their insulin-positive cell volume, while those from donors under 10 years had a 2.5-fold increase. After 10 weeks, the implants from the younger donors consisted of 19% insulin-positive cells occurring as single units or small cell clusters. Three percent of these insulin-positive cells also expressed the ductal marker CK 19 and were consistently found in conjunction with ductal epithelia; up to 1% was positive for the proliferation marker BrdU and located in small endocrine cell clusters. Conclusions/interpretation. These data indicate that duct cell preparations from donors under 10 years can generate insulin-positive cells. This process might involve differentiation of CK 19-positive-insulin cells that are formed at the duct epithelia as well as proliferation of insulin-positive cells within endocrine cell aggregates. [Diabetologia (2003) 46:830-838]