HLA and KIR gene polymorphism in hematopoietic stem cell transplantation

Sammanfattning: The major histocompatibility complex (MHC) in humans known as the Human Leukocyte Antigens (HLA) is localised to the short arm of chromosome six. The HLA class 1 antigens, HLA-A, -B and -C are highly polymorphic glycoproteins expressed on the cell surface of most nucleated cells in the body. The HLA class 11 antigens HLA-DR, -DQ and -DIP are expressed on the cells of the immune system, mainly B-cells, macrophages, activated T-cells and dendritic cells. Processed foreign antigens and self-antigens are presented to T-cells by the class 1 and class 11 molecules. The class 1 molecules interact with CD8 molecules on T-cells and the class 11 molecules with CD4. Matching for HLA class 1 and class 11 alleles is known to be important for the clinical outcome of hematopoietic stem cell transplantation (HSCT). However, the exact level of matching required to minimize the risk for immunological complications when using an unrelated donor is still not known. Natural killer (NK) cells interact with MHC class 1 molecules on target cells. In humans, NK cells are negatively regulated by killer cell immunoglobulin-like receptors (KIR) recognizing HLA class 1 antigens. Transplantation across HLA barriers may trigger donor cell alloreactivity, which can influence the result of the treatment. Until recently serological typing has been the primary technique used for HLA class 1 analysis. It has been assumed that HLA class 1 serological typing was more accurate than serological HLA-DR typing. However, several studies have shown that serological HLA-A, -B typing is poorer than expected. The first paper describes a systematic investigation of the accuracy of class 1 serological typing in all the groups of patients and healthy individuals that are routinely typed at our laboratory. Class 1 typing using PCR-SSP was more accurate and also gave a higher resolution, especially when typing patients with hematological disorders. We have retrospectively performed allele level typing for HLA class 1 and class 11 in unrelated donor/recipient pairs and correlated the degree of matching to the clinical outcome. We found that patients that had a donor with, at the time of transplantation, an unknown HLA-B allele level mismatch had a very poor transplant outcome with severe graft versus host disease (GVHD) and a high mortality rate. In addition we found that pairs mismatched for HLA-C but matched for the KIR ligand epitope had increased survival and disease free survival. The role of matching for HLA-DPB1 and -DPA1 is still unclear and debated. We found inferior survival and increased transplantationrelated mortality (TRM) in patients with an HLA-DPA1 incompatible donor. Extending the study, analysing a larger patient cohort, we found that the results are in concordance with our previous findings and that incompatibility for HLA-DPA1 is associated with inferior survival, increased infection related mortality and TRM. We also further investigated the role of KIR ligand incompatibility in unrelated HSCT. A total of 190 patients with hematological malignancies, transplanted with an unrelated donor were included in the study. We observed that KIR ligand mismatch was associated with increased TRM. The increased TRM was due to a higher rate of infections. The KIR genes located on human chromosome 19 and the HLA genes on chromosome 6 segregate independently. Thus HLA identical siblings are not KIR identical. Two recent studies have shown an impact of KIRHLA mismatch in HLA-identical HSCT, where no missing-self recognition can be expected. In paper V we present our results, of these effects in a single centre setting. A total of 80 patients with myeloid malignancies transplanted with an HLA-identical sibling donor were investigated. HLA and KIR genotypes were determined for the donor-recipient sibling pairs. We investigated if the combination of donor-inhibitory KIR and recipient HLA-C type could influence transplant outcome. Missing KIR-ligand was found for 51 of the 80 patients (63.7%). Lack of HLA ligand for donor inhibitory KIR had no effect on transplant outcome regarding disease free survival, overall survival, relapse or GVHD. In conclusion, the use of genomic typing techniques for HLA class 1 and 11 and KIR genes results in improved typing quality giving the opportunity to obtain a more precise matching in transplantation. Using these techniques, we here show a negative influence of HLA-B, HLA-DPA1 and KIR-ligand incompatibilities in allogeneic hematopoietic stem cell transplantation.