Prediction of clinical outcome in islet allotransplantation.

R ecent progresses in the pancreas’ enzymatic digestion process along with novel immunosuppression strategies have led to successful clinical trials of islet transplantation in humans. On the other hand, clinical outcome remains variable and unpredictable in centers with limited experience. The possibility of predicting in vivo islet graft function should allow the selection of preparations on the basis of their potential success, thus improving the overall results and making the processes more consistent and reproducible. Graft function prediction is a work in progress. Initially, the best parameters representative of engrafted islet mass in recipients should be defined. Fasting Cpeptide and exogenous insulin requirements are commonly used, although other methods for a more complete characterization of graft function (i.e., -score) have recently been described but not validated in a large number of patients. In addition, some of these data were shown to predict long-term graft function and might be used to establish whether recipients require further islet infusions. Many pretransplant parameters representative of islet preparations and predictive for in vivo function have been proposed. C-peptide values as well as the exogenous insulin requirements of recipients were shown to be directly correlated with the number of transplanted islets, but there are many exceptions to this association. Other methods to define the quality of an islet preparation include analyses of islet morphology, cell composition, response to glucose, and viability and production of proinflammatory molecules. The most promising appear to be those that simultaneously analyze more than one aspect of islet physiology. Islet transplantation has great potential for the normalization of the main parameters of glucose metabolism in diabetic patients. A sufficient number of islets can now be obtained from good quality pancreata more frequently than in the past, and diabetes can generally be reversed, normalizing A1C levels and eliminating severe hypoglycemic episodes (1). Clinical data produced by some centers on patients 1 year after the transplant report that the percentage of normal C-peptide secretion is 100% and that of insulin independence 80%, with an improvement of glycemic compensation (1,2). In fact, data from the Islet Transplant Registry (Giessen, Germany) and the Collaborative Islet Transplant Center and the conclusions of a multicenter trial sponsored by the National Institutes of Health, produced to verify the reproducibility of the Edmonton Protocol, report that the comprehensive percentage of success (in terms of insulin independence) is only 50% (3,4). This means that results are not reproducible to the same extent among different centers (5). Even in the centers with the highest percentage of transplant success, insulin independence is reached in only a few cases by a single transplant; in most cases it is necessary to repeat two or even three infusions (6,7). Therefore, it is clear that the efficacy of a transplant preparation is variable and not always predictable. The possibility of predicting whether a preparation can work in vivo represents a difficult goal. There are many factors that interfere with islet function in vivo, including quality and number of transplanted islets and their engraftment, preand posttransplant immunological conditions (both in terms of autoimmunity and alloimmmunity), recipient immunological condition, and toxicity of administered drugs (1,8–10). The prediction of transplant success, however, represents an important objective to be reached. The possibility of predicting in vivo islet graft function should allow the selection of preparations on the basis of their potential success, thus improving overall results. Furthermore, it should result in more consistent, reproducible procedures and permit a proper evaluation of costs and benefits. Altogether, these represent prerequisites for the evolution of islet transplantation from a research procedure to a therapeutic option available to diabetologists. To define the relation between in vivo parameters and transplant function, it is necessary to define evaluation criteria for in vivo functionality.