Identification and generation of tumor antigen-specific T cells directed against AML

Abstract

Background & Aim Acute myeloid leukemia (AML) is an aggressive hematological malignancy characterized by aberrant expression of leukemia associated antigens (LAA) that can be presented in context of HLA molecules, to elicit anti-tumor T cell response and improve clinical outcomes. This study aimed to 1) identify targets of endogenous LAA-specific T cells in AML patients post allogeneic hematopoietic stem cell transplant (allo-HSCT) and whose detection was associated with favorable clinical outcomes and 2) assess the feasibility of generating LAA-specific T cells targeting these antigens for adoptive immunotherapy. Methods, Results & Conclusion We obtained blood from patients (n=8) with relapsed and/or refractory AML who were in remission post allo-HCST, and measured the frequency of T cells reactive against known LAAs (e.g. WT1, CCNA1, survivin, PRAME). These patients were followed for 1 year post-sampling. Of the tested antigens, we identified that the best predictor of those who would eventually relapse was the magnitude CCNA1-directed T cell activity (clinical remission vs relapse: 171 vs 7.13 mean spot forming colonies/1 × 10e5; p=0.03). We next hypothesized that outcomes for AML could be improved by the ex vivo preparation and adoptive transfer of stem cell donor-derived CCNA1-specific T cells to patients at risk of relapse post allo-HSCT. Thus, we exposed autologous peripheral blood mononuclear cells from healthy donors (n=14) with CCNA1 pepmix and were able to induce specific responses in 13 of 14 donors tested. The cultures expressed both central (30±3%) and memory markers (30±6%) and were polyclonal though the dominant CCNA1-directed response was CD8-mediated. Reactive cells were polyfunctional (dual IFNg+/TNFa+ secretion), and capable of lysing HLA-matched CCNA1+ myeloid leukemia cell lines and primary AML blasts, while CCNA1 negative and/or HLA mismatched targets were spared (mean±SEM %lysis =42.6±10 vs 4.3±1.7; p=0.03) Thus, by measuring endogenous immune activity and correlating this data with clinical outcomes we identify clinically relevant target antigens that may be suitable for future T cell therapy studies in AML. As a proof of principle, this study demonstrated the importance of CCNA1-targeted T cells and established the preclinical feasibility of expanding Th1-polarized reactive cells from healthy donors that are suitable for adoptive transfer. Our future studies involve identification of other novel targets in AML including neoantigens and clinical translation of these findings.