GVHD in HLA-A2 Mismatched Transplantation Caused by a Combined CD8 Response Directed Against HLA-A2 and a CD4 Response Recognizing an HLA-A2 Derived Peptide in HLA-DR1.

HLA mismatched Stem Cell Transplantation (SCT) can be performed in patients with leukemia if no HLA identical donor can be found. Although HLA class I is expressed by almost all recipient cells and the frequency of allo-HLA reactive T-cells is high in all normal donors, the risk of GVHD after single locus mismatched SCT is comparable to that after HLA identical SCT. Thus, the occurrence of GVHD may not simply be explained by recognition of the mismatched HLA by allo-HLA reactive T-cells. Therefore, we characterized in detail the nature of the allo-immune response in an HLA mismatched setting. A patient with acute myeloid leukemia was treated with T-cell depleted SCT from a sibling donor who was HLA identical except for an HLA-A2 crossover. 6 Months after SCT, Donor Lymphocyte Infusion (DLI) of 2.5*10e6 CD3+ T-cells/kg was given for mixed chimerism caused by persistence of patient T-cells. No clinical response and no GVHD developed. 12 months after SCT the leukemia relapsed with 9% blasts in bone marrow, and a second DLI of 7.5*10e6 CD3+ T-cells/kg was given. The patient died of grade IV GVHD 5 weeks after DLI. durign the GVHD flow cytometry of PBMC’s showed conversion of patient to donor type T-cells. 80% Of the CD8 and 40% of the CD4 T-cells were activated, as determined by co-expression of HLA-DR. These activated T-cells were single cell sorted, non-specifically expanded, and tested for alloreactivity using cytotoxicity and cytokine production assays. 46 Out of 56 isolated CD8 clones and 7 out of 88 CD4 clones recognized patient but not donor target cells, indicating that at the time of the GVHD almost 70% of circulating T-cells were alloreactive. The response was highly polyclonal as shown by usage of at least 13 different TCR Vβs by the CD8 clones, and 6 by the CD4 clones. HLA restriction of the clones was tested with HLA blocking antibodies, a panel of HLA-typed target cells and donor EBV-LCL transduced with HLA-A2. All alloreactive CD8 clones were HLA-A2 specific. To further characterize the specificity, CD8 clones were tested against T2 cells loaded with HPLC fractions of peptides eluted from HLA-A2. Some CD8 clones recognized HLA-A2 with all different HPLC fractions, indicating peptide-independent recognition. Other clones recognized one fraction indicating peptide specificity, or several fractions indicating “promiscuous” peptide recognition. The CD4 clones were HLA-DR1 restricted and recognized donor EBV-LCL transduced with HLA-A2, indicating that the peptide recognized in HLA-DR1 was derived from the mismatched HLA-A2 molecule. Therefore, CD4 clones were tested against different peptides covering the whole HLA-A2 sequence. All clones recognized epitope 101–122 derived from a hyper variable region of HLA-A2. These results indicate that the GVHD in this HLA-A2 mismatched transplantation was caused by a combined highly polyclonal CD8 response directed against the HLA-A2 molecule and a CD4 response recognizing an HLA-A2 derived peptide presented by HLA class II. We speculate that the absence of an immune response observed after the first DLI despite high frequency of allo-HLA reactive T-cells, indicates that CD8 anti-HLA-A2 T-cells are insufficient to cause severe GVHD. We hypothesize that the rise of HLA-DR expressing leukemic blasts presenting HLA-A2 derived peptide in HLA class II triggered the CD4 response which was necessary to initiate the CD8 alloresponse resulting in this clinical outcome.