Preclinical Development of Non-Edited CD7 CAR-Modified Invariant NKT Cell Therapy for T-Cell Malignancies and Acute Myeloid Leukemia

Abstract

The major hurdle in developing chimeric antigen receptor T (CAR-T) cell therapy for T-cell malignancies is CAR-T cell fratricide. This refers to the self-killing of the CAR-T cells due to shared expression of targetable antigens (e.g., CD7) between CAR-T and malignant target T cells. This CAR-T fratricide creates difficulties in producing sufficient numbers of CAR-T cells for infusion and can limit survival of the cells in vivo. Nuclease (e.g., CRISPR)-mediated genomic disruption of CD7 or T cell receptor alpha chain expression provides an alternative approach to enabling expansion of CD7 CAR transduced T cells from allogeneic healthy donors. However, the safety and clinical implications of CRISPR-based genome editing remain unclear. Thus, other approaches to development of allogeneic, non-edited CD7 CAR-T cell therapy for relapsed and refractory T-cell malignancies is warranted. We unexpectedly found that human invariant natural killer T (iNKT) cells in normal blood donors displayed on average approximately 56% CD7 positivity (range 38 - 74%, n = 9 donors). Therefore, we hypothesize that CD7- subsets in iNKT cells expressing CD7 CAR can partially avoid fratricide without nuclease-mediated genome editing to delete CD7 and be expanded for adoptive cell therapy of CD7+ malignancies. By leveraging our proprietary 6000-10000-fold iNKT cell expansion platform, lentiviral transduced CD7 CAR iNKT (CD7 CAR-iNKT) cells can be expanded to clinical scale (n=2) (Fig. 1A). Based on CD19 CAR- and CD7 CAR-iNKT expansion rate (n=5), 100-500 doses (assuming 1-3 infusions per patient) of CAR-iNKT cells can be manufactured from a single donor-derived leukocyte concentrate. CD7 CAR-iNKT cells (n=5, 64-98% CAR+) specifically killed luciferase-expressing CD7+ T-cell acute lymphoblastic leukemia (T-ALL) (HSB2, Jurkat, MOLT13) and CD7+ AML (KG1) at E/T ratios ranging from 20:1 to 2.2:1 compared to mock iNKT cells and CD19 CAR-iNKT cells (p < 0.0001). Control CD19 CAR-iNKT cells (n=3) specifically killed CD19+ B-cell tumor cells but not CD19- T-ALL and AML cells (p < 0.0001). Mock iNKT cells showed minimal cytotoxicity to all target cells except for CD1d+ Jurkat. CD7 CAR-iNKT cells (n=5) also produced significant amounts of IFN-γ in response to CD7+ T-ALL (HSB2, Jurkat, MOLT13) and CD7+ AML (KG1, Kasumi-6) compared to mock (n=5, p < 0.001) and CD19 CAR-iNKT cells (p < 0.05 except for Jurkat, p < 0.01). CD19 CAR-iNKT cells (n=3) produced significant amounts of IFN-γ to CD19+ B-cell leukemia and lymphoma cell lines (Nalm6, Daudi, Raji) compared to mock and CD7 CAR-iNKT cells (p < 0.0001). As expected, CD7 CAR-iNKT cells significantly eliminated almost all CD7+ cells (mean ± SD 0.4% ± 0.1, n=4) compared to mock (61.6% ± 24.4, n=4) and CD19 or ROR1 CAR-iNKT cells (69.9% ± 24.4, n=3) (p < 0.0001). Furthermore, expanded CD7 CAR-iNKT cells were >95% CD3+TCRVα24-Jα18+ and specifically responded to a-galactosylceramide or 7DW8-5 lipid antigens presented by CD1d molecules. The anti-leukemic effect of CD7 CAR-modified iNKT cells was assessed in HSB2 T-ALL and KG1 AML xenograft models (n=5 per group). Three i.v. infusions of thawed CD7 CAR-iNKT cells (5x106 cells/mouse) at days 0, 3, and 6 into leukemia (5x105/mouse, i.v. at day -2)-bearing mice significantly suppressed T-ALL growth by bioluminescent imaging analysis at days 3 (p < 0.0001), 6 (p=0.0001-0.0005), 10 (p=0.0012-0.0024) and 17 (p < 0.0001) compared to PBS and mock controls (Fig. 1B). A significant survival advantage was observed after CD7 CAR-iNKT cell therapy compared to PBS and mock controls (p = 0.0009, Fig. 1C). Two i.v. infusions of thawed CD7 CAR-iNKT cells into pre-established AML mice significantly suppressed AML growth at days 3 (p = 0.0096), 6 (p < 0.0001), 10 (p = 0.0026), 17 (p < 0.0001), 24 (p < 0.0001), 31 (p < 0.0001), 38 (p < 0.0001) and 45 (p = 0.0003) compared to mock control (Fig. 1D). CD7 CAR-iNKT cell therapy also significantly prolonged the animal survival (p = 0.0069, Fig. 1E). In conclusion, these results demonstrate that CD7 CAR-iNKT cells without genome editing can be manufactured to therapeutic doses and mount anti-CD7+ tumor responses in vitro and in vivo. Our CD7 CAR-iNKT cells can be used as allogeneic, off-the-shelf products to treat all subtypes of T-ALL and 30% of CD7+ AML patients. In addition, our CD7 CAR-iNKT cells are likely more effective in 70% CD7+CD1d+ T-ALL patients as they provide dual target specificity and reduce the likelihood of relapse. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal