Successful Generation of Mature Professional Acute Myeloid Leukemia (AML) – Antigen Presenting Cells (APC) Does Not Predict Successful Induction of AML Reactive T Cells for Adoptive Immunotherapy.

Treatment of relapsed acute myeloid leukemia (AML) after allogeneic stem cell transplantation (alloSCT) with donor lymphocyte infusion (DLI) can induce a graft-versus-leukemia (GVL) reaction but also graft-versus-host disease (GVHD). GVHD may be avoided and the likelihood of a GVL response increased by infusion of in-vitro generated donor-derived cytotoxic T cell lines reactive with the AML cells. Treatment of patients with relapsed AML after alloSCT with DLI often fails, possibly due to a limited antigen presenting function of the primary AML cells. Therefore, to improve this antigen presenting function, we studied the ability to modify AML cells into maturated professional APC and the consequences for the subsequent generation and isolation of AML-reactive T cells using these AML-APC as stimulator cells. AML-APC were generated as follows:

1. primary AML cells were cultured in GM-CSF, SCF, TNFα and IL-4, and after 3–6 days maturation was induced by DKTP (diphtheria, whooping cough, tetanus and polio) vaccine, by addition of monocyte conditioned medium (MCM) mimic (containing IL-1β, IL-6 and TNFα) with PGE2, or by CD40-cross-linking, or

2. AML cells were cultured for 1–4 days with IL-4, GM-CSF and calciumionophore (CI).

The highest numbers of mature AML-APC could be generated after maturation by CD40 cross-linking or MCM mimic/PGE2. Induction or upregulation of CD80, CD86 and CD83 expression was determined in 5 -100% of the cells, IL-12 production was induced after CD40 cross-linking and the AML-APC did not produce IL-10. Cell-populations of leukemic cells containing more than 25% fully maturated AML-APC could be obtained in 8 out of 15 cases. AML-reactive T cell lines were generated in 9 HLA-identical donor-patient combinations. In 4 of these 9 combinations cell suspensions containing > 25% fully maturated professional AML-APC could be generated. Donor T cells were stimulated with AML-APC in a ratio of 10:1, and at day 7 and 11 of the culture, 10 IU/ml IL-2 was added. At day 14 responder cells were restimulated with AML-APC, and 16 hours after restimulation IFNγ producing AML-reactive T cells were stained using the IFNγ secretion assay (Miltenyi Biotech) and isolated by FACS sorting. Using a flowcytometry-based cytotoxicity assay in which the AML cells were labeled with CFSE we analyzed the isolated IFNγ positive T cells for their ability to kill the leukemic cells. In 2 of the 4 combinations in which optimal AML-APC could be induced, IFNγ producing T cells were capable of killing the leukemia. However, when we compared the capability of unmodified primary AML cells and professional AML-APC to induce AML-reactive T cells in these 2 responses, we found in both cases that the unmodified AML cells and AML-APC were equally capable to generate AML-reactive T cells, suggesting that the generation of professional AML-APC is not crucial for the induction of AML-reactive T cells. This is further supported by the fact that also in 2 of the 5 donor-patient combinations where no fully maturated professional AML-APC could be generated AML-reactive T cells could be induced. The expansion rate of responding T cells was higher after stimulation with fully maturated AML-APC. In conclusion, mature professional AML-APC could be generated in the majority of cases, but this modification of the malignant cells did not correlate with their ability to successfully induce AML-reactive T cell lines for adoptive immunotherapy.