Pre-Clinical Activity of Allogeneic Anti-CD22 CAR-T Cells for the Treatment of B-Cell Acute Lymphoblastic Leukemia

Autologous T-cells engineered with chimeric antigen receptors (CARs) against CD19 are proving to be an efficacious immunotherapy for patients with relapsed or refractory B-cell acute lymphoblastic leukemia (B-ALL). At present, CAR technology is administered through the custom-made manufacturing of therapeutic products from each patient's own T-cells. However, this patient-specific autologous paradigm is a significant limiting factor in the large-scale deployment of CAR technology. In this study, we utilized allogeneic "off-the-shelf" engineered CAR T-cells from third-party healthy donors. The CD22 surface antigen is commonly expressed in B-ALL patients as well as in healthy B-cells. Here, its potential as a CAR target was investigated using allogeneic off-the shelf engineered CAR T-cells against human CD22 (UCART22).

UCART22 cells harbor surface expression of an anti-CD22 CAR (CD22 scFv-41BB-CD3z) and the RQR8 ligand, a safety feature rendering the T-cells sensitive to the monoclonal antibody rituximab. To reduce the potential for alloreactivity, the cell surface expression of the T-cell receptor (TCR) is abrogated through the inactivation of the TCRα constant (TRAC) gene using Cellectis' TALEN^® gene-editing technology.

The level of CD22 cell surface molecules was measured using BD Quantbrite beads for both patient peripheral blood samples and B-ALL cell lines. B-ALL cell lines (n=8) expressed a greater amount of CD22 molecules per cell than patient samples (n=14) (5,028 +/- 1,342 compared to 951 +/-160 molecules/cell, p=0.044), with highest expression of CD22 in two Ph-like B-ALL cell lines (MUTZ5, shown in Figure1A and MHH-CALL4).

The in vitro cytotoxic activity of UCART22 cells was evaluated by co-culturing UCART22 or non-transduced CAR(-) TCRαβ(-) control T-cells (NTD) with B-ALL cell lines and primary human samples, at a maximum 10:1 effector to target ratio (represented in Figure1B). Using flow cytometry, significant antigen-specific cytotoxic activity of UCART22 cells was found compared to NTD controls and correlated with CD22 expression factored by the %kolmogorov-smirnov max difference in CD22-PE fluorescence compared to unstained controls (Pearson correlation r-squared for cell lines= 0.6850, p=0.0001 and r-squared for patient samples=0.6204, p=0.0008).

Secretion of 13 cytokines was measured after 1:1 co-incubation of effector and target cells. UCART22 cells stimulated by CD22(+) B-ALL, but not NTD cells, secreted high levels of IFNγ, TNFα, IL-5, IL-17A and IL-17F in the culture supernatants, with cytokine levels being proportionate to CD22 abundance (represented in Figure1C).

In addition, immune compromised mice engrafted with Daudi cells, a CD22(+) expressing Burkitt's lymphoma cell line, were treated with UCART22 cells. Treatment doses of 1-10x10^6 cells per mouse reduced disease burden (Figure 1D), measured by bioluminescence imaging, and extended survival in a dose-dependent fashion compared to saline or NTD treated controls. Additional PDX studies using B-ALL patient derived xenografts are ongoing and will be presented. Altogether, these results show supporting evidence for the future use of allogenic UCART22 in B-ALL immunotherapy.

View largeDownload slide

View largeDownload slide

Close modal