Abstract 4090Poster Board III-1025Allogeneic hematopoietic transplantation (allo-HCT) is the only curative option for patients affected by high-risk acute myeloid leukemia (AML). This is largely due to the ability of allogeneic immune system to eradicate leukemic stem cells (LSC). However, the fact that some patients still relapse after allo-HCT, suggests that strategies to increase LSC targeting by donor T cells are needed. For this purpose, we exploited the unique ability of myeloid blasts to differentiate into leukemic dendritic cells (LDC). We observed that a short (48h) exposure to calcium ionophore A23187 and IL-4 is able to induce LDC differentiation in 14/16 (86%) of AML that we studied, both de novo and secondary. Importantly, despite phenotypic and functional changes indicative of differentiation into DC-like cells, the process was accompanied by the maintenance of disease markers such as CD34 and CD117. Moreover, LDC maintained the expression of the oncogenic protein WT1, which is a putative LSC antigen. Thanks to these favourable characteristics, LDC proved to be superior to the original blasts in expanding leukemia-reactive T lymphocytes both in the autologous and allogeneic HCT setting (on average, 5-fold expansion of blasts-stimulated T cells vs 95-fold expansion of LDC-stimulated T cells, SEM=2,7 and 67,7 respectively, p=0,01). We observed that the level of T-cell expansion directly correlate with the percentage of LDC obtained upon treatment with A23187 and IL-4. Most importantly, LDC proved to be more potent than blasts in expanding central memory T lymphocytes (TCM), which are known to confer superior anti-tumor immunity (on average, 29% of TCM upon stimulation with blasts vs 53% TCM upon stimulation with LDC, SEM=7,2 and 5,7 respectively, p=0,01). LDC-expanded T lymphocytes were able to efficiently recognize and kill leukemic blasts in vitro (on average, 953 specific spots of IFN-g/50'000 effectors at E:T ratio of 10:1 -SEM=120- and 29% of specific killing at E:T ratio of 50:1 -SEM=7,4-). Importantly, analysis of different HLA-settings and different targets of patient origin, suggests that LDC can expand T lymphocytes with specificities against multiple antigens expressed by the original leukemia. In particular, we observed the expansion of WT-1 specific T cells upon LDC stimulation. Finally, when infused in NOD/Scid mice transplanted with the original leukaemia, LDC-stimulated T lymphocytes were able to induce long-term complete remissions (>16 weeks) in all mice analyzed, suggesting that this approach may be active against leukemic stem cells. These results show for the first time that LDC-stimulated human T cells could exert a strong GvL activity in vivo. Disclosures:Bordignon:Molmed Spa: Employment.