In Vitro Generation of Multi-Epitope Specific CD4+ T Helper Cells for Adoptive Immunotherapy of Leukemia

Introduction: Acute myeloid leukemia (AML) and other hematological malignancies that constitutively express MHC class II molecules represent the ideal target for leukemia-specific CD4⁺ T helper (Th) cells. CD4⁺ Th cells are central to the functioning of the immune system. They regulate adaptive immunity against infections and drive pathogenic responses in many autoimmune diseases. Our preclinical studies indicate that adoptively transferred antigen specific Th cells efficiently eradicate even advanced tumors in mice, showing functional advantage over their better characterized classical CD8⁺cytotoxic counterparts. Unfortunately, the production of human class II-restricted Th cells is complex and challenging. Consequently, the clinical activity of tumor-specific Th cells has not been systematically evaluated and very little is known about their therapeutic potential in humans. Striking evidence of their enormous power comes from some anecdotal clinical reports of complete regression of metastatic cancers upon transfer of antigen-specific Th cells.

Methods and Results: We investigated a reliable GMP-compatible method for in vitro expansion of antigen-specific CD4⁺ T cells targeting common leukemia associated antigens (LAAs) Willm’s Tumor antigen 1 (WT1) and Preferentially Expressed Antigen in Melanoma (PRAME) for future adoptive immunotherapy in the setting of allogeneic stem cell transplantation (SCT). We hypothesized that the naïve CD4⁺ T cell compartment, rather than the bulk Th cell population could be a superior source for generating a tumor antigen specific cell product. To test this hypothesis bulk and naïve CD4⁺ T cells were isolated from the peripheral blood of normal donors by magnetic bead separation. Purified T cells were stimulated in vitro with autologous dendritic cells (DCs) pulsed with overlapping 15 amino-acid long peptide (pepmixes) spanning the length of both proteins. Two rounds of stimulation were performed in presence of IL-7, IL-15 and later +/- addition of IL-1, IL-6 and IL-23. IL-2 in low concentration was supplemented during the second round of stimulation.  At the end of the second expansion the cells were tested for reactivity by intracellular cytokine production using flow cytometry (FACS) upon stimulation with cognate LAAs or irrelevant control pepmixes. In T cell cultures derived from naïve CD4⁺ T cells we observed robust induction of PRAME reactivity from majority of tested normal donors, while reactivity against WT1 was donor dependent. The frequency of LAAs in bulk CD4⁺ T cells was significantly lower in all cases, supporting the notion that pre-exiting memory Th cells have a competitive advantage over the LAA-specific precursor. T cell cultures supplemented with inflammatory cytokines demonstrated further enhancement of antigenic reactivity. Next we tested if LAA pepmix –stimulated T cells can recognize tumor targets. Naïve-derived PRAME and WT1 Th cells generated from normal SCT sibling donors produced IFN-γ, IL-2 and TNF-α upon exposure to fully HLA-matched AML blasts while no reactivity was seen in control CMV pp65-specific Th cells from the same donors. This observation suggests that LAA-specific CD4⁺T cells induced with pepmixes have the ability to recognize physiologically-relevant tumor antigens.

Conclusions:  Here we report the feasibility of generating naïve-derived anti-leukemia CD4⁺ T cells from majority of normal donors.  Removal of competing memory Th cells unmasks the LAA-specific reactivity, thus improving the reliability of the process. Importantly, these Th cells demonstrate highly-specific recognition of the tumor epitopes naturally processed by HLA-matched leukemic blasts, establishing the foundation for a future adoptive immunotherapy clinical trial in patients with hematological malignancy.