Direct Comparison of Anti-CD33 and Anti-CD123 Chimeric Antigen Receptor T Cells with CD28 or 41BB Co-Stimulatory Domains for Treatment of Acute Myeloid Leukemia

Introduction: Acute Myeloid Leukemia (AML) is an aggressive malignancy with very poor prognosis and limited treatment options. Chimeric antigen receptor expressing T (CART) cell therapy has been highly successful against refractory B cell acute lymphoblastic leukemia and lymphoma. Effective use of CART cells in AML has not yet been well established. Differences in efficacy and potency of CD28 or 41BB co-stimulatory domains of the CART cells have not been thoroughly studied particularly in the context of AML. Here, we systematically compare potency and efficacy of anti-CD123 and anti-CD33 CART cells with CD28 or 41BB co-stimulatory domains against AML both in vitro and in vivo .

Methods: anti-CD33 and anti-CD123 CART cells were developed using two different anti-CD33 and a single anti-CD123 single chain fragment variable sequences and cloning into a third-generation lentiviral plasmid containing a CD8 transmembrane domain, a co-stimulatory signaling domain (either CD28 or 41BB) and a CD3zeta domain. CART surface expression on T cell surface was measured by flow cytometry. Efficacy of CAR expressing T cells was tested against a panel of AML cell lines with relatively uniform expression of CD33 and variable expression of CD123 including MOLM14, MV4;11, THP1 and U937. Enzyme-linked immunosorbent assay kits for interleukin-2 (IL2) and interferon-gamma (IFN-gamma) were used to assess in vitro CART functionality, 10⁵ tumor cells were co-incubated with 10⁵ transduced CART cells or mock T cells for 16 hours in 96-well plate. In vitro killing of tumor by CART cells was evaluated by measuring green and red fluorescence by IncuCyte ZOOM, 10⁵ tumor cells that were co-incubated with 10⁵ transduced CART cells or mock T cells for 40 hours in 96-well plates. In vivo CART functionality was assessed by monitoring leukemia burden through bioluminescent imaging using mice engrafted with AML cell lines that received anti-CD33 and anti-CD123 CART cell therapy (tumor challenge of 1x10⁶ cells was given and CART cell therapy dose of 7x10⁶ transduced cells was given). Bone marrow and spleen of mice were assessed by flow cytometry for CART cell and leukemia presence.

Results: In vitro cytokine production of IL2 and IFN-gamma when anti-CD123 and anti-CD33 CART cells are co-incubated with various AML cell lines indicate greater production of cytokine in CART cells with a CD28 co-stimulatory domain. In vitro killing assay indicates no significant difference in killing capacity of anti-CD33 and anti-CD123 with CD28 or 41BB co-stimulatory domains, however there is some evidence that the CD28 co-stimulatory domain may yield more killing at earlier time points than 41BB co-stimulatory domain. In vivo activity of anti-CD33 and anti-CD123 CART cells against MOLM14, THP1 and U937 shows that clearance of AML occurs more potently and at earlier time points with the CD28 co-stimulatory domain when compared to the 41BB co-stimulatory domain. Bone marrow and spleen analyses show the lack of presence of leukemia but presence of CART cells with CD28 co-stimulatory domain.

Conclusions: CART cells targeting surface antigen CD33 and CD123 in AML demonstrate in vitro and in vivo activity against AML, showing clearance of leukemia and presence of CART cells. Additionally, in vitro and in vivo comparison of CD28 and 41BB co-stimulatory domains indicates that the CD28 may be more potent and efficacious in clearing AML. Further experiments are ongoing to more fully evaluate relative potency of CAR T cells incorporating different co-stimulatory domains.