Predictive parameters for in vivo alloreactivity

Despite optimal HLA matching and a negative serological crossmatch, confrontation with allogeneic cells by organ- or stem-cell transplantation or platelet transfusion, can lead to an alloimmune response resulting in graft rejection, graft vs. host disease or platelet refractoriness. It would be attractive to be able to select beforehand those donor/recipient combinations, that do not lead to a destructive alloimmune response and exclude transplantation or transfusion with donors, that induce a strong alloimmune reaction. Many predictive parameters have been identified on the basis of retrospective analysis of graft survival data, and the results of in vitro tests to measure T and B cell alloreactivity. However, most of these parameters have shown to be relevant when a statistical analysis is performed on the population level but do not have a direct impact for the individual patient. An exception is a molecularly based algorithm, called HLA matchmaker, which seems to predict which HLA mismatches do not lead to alloantibody formation in a particular individual. Prediction of T cell alloreactivity is more difficult and will need the development of additional in vitro tools or adaptation of the HLA matchmaker program. Although the direct clinical implication of NK cell mediated allorecognition is not clear yet, this may be a complicating factor when establishing solid parameters for the prediction of an alloimmune reaction in vivo. (C) 2002 Elsevier Science B.V. All rights reserved.