Abstract
Acute myeloid leukemia (AML) originates from the hematopoietic stem and progenitor cell compartment and is associated with an overall poor clinical outcome. T cell-engaging bispecific antibodies (TCE), binding to a tumor-associated antigen and to CD3, redirect T cells to target antigen-expressing cells in an MHC/TCR-independent fashion. The development of TCEs for clinical application is facing several challenges. First, it is difficult to identify a tumor-selective, non-MHC-presented tumor target, not expressed on the tissue of tumor origin, thus limiting specificity. Second, CD3-directed TCEs do not provide a second, T cell-activating signal, such as the stimulation of CD28 or 41BB, thus possibly limiting T-cell efficacy. To address both aspects, we generated a CD33xCD28 IgG4-scFv2 with CD28 agonistic activity and tested it in combination with a previously by us reported CD117xCD3 TCE in AML. Combining these two bispecific antibodies significantly improved T cell-mediated lysis of AML cell lines and primary AML samples by enhancing T-cell activation, proliferation, and cytokine release in vitro. Furthermore, the addition of CD33xCD28 IgG4-scFv2 showed faster time to cell attachment, increased lytic events, and improved specificity toward double-target-expressing cells. In summary, the data indicate that combining costimulation via a second tumor-associated antigen to CD3-TCEs enhances T-cell lytic activity and simultaneously increases specificity against double-target-expressing AML cells. SIGNIFICANCE: Current TCEs lack costimulatory signaling. The here-described CD33xCD28 IgG4-scFv2 delivers target-selective costimulation and enhances activation, proliferation, cytokine release, and cytotoxicity when used in combination with CD117xCD3 TCE in AML in vitro. Furthermore, the dual tumor-associated antigen targeting improves therapeutic specificity by preferentially eliminating CD117+CD33+ cells, possibly allowing improved bispecific antibody-mediated AML therapy.