Cancer therapy using Chimeric Antigen Receptor (CAR) T cells has yielded impressive clinical outcome in B cell malignancies, but this success has not yet been translated to other types of cancer. One of the main bottlenecks to generate efficient CAR-T cell therapies is the identification of targets exclusively and homogeneously present on cancer cells. NKG2D is an NK cell activating receptor binding to 8 different ligands induced by cellular stress, including malignant transformation, while being rarely detectable on the surface of healthy cells. Celyad's lead CAR-T product CYAD-01 is based on the full length human NKG2D fused to the intracellular domain of CD3ζ. The endogenous accessory protein DAP10, that associates with NKG2D and stabilizes the complex on the cell surface, provides co-stimulatory signaling upon binding of NKG2D with its ligands. Data from preclinical models and clinical trials have shown that CYAD-01 cells are effective against a variety of tumor types, including Acute Myeloid Leukemia (AML).
The aim of this study is to simulate the dosing and injection scheme of CYAD-01 cells in a mouse model of AML and to monitor CAR-T cell persistence and biodistribution in and without the presence of tumors.
In this study, we used xenografts of the human AML cell line THP-1-luc-GFP in NGS mice. This is an aggressive preclinical AML model, with mouse survival of barely over 2 weeks without treatment. The antitumor efficacy of single or multiple injections of distinct doses of CYAD-01 CAR T cells was evaluated by in vivo bioluminescence imaging. The biodistribution and persistence of CAR T cells in the peripheral blood, bone marrow and spleen were monitored by flow cytometry.
Single injection of CYAD-01 cells in the THP-1 mouse model resulted in transient tumor regression with 2-fold prolongation of animal survival. CYAD-01 cells exhibited limited persistence in this animal model, indicating that dose control can be achieved by multiple injections. To this end, 3 injections at weekly intervals enhanced the anti-tumor efficacy of CYAD-01 T cells, resulting in 4-fold increase in animal survival compared to control, untreated animals. To evaluate the minimal dose of CYAD-01 cells that can be used to detect anti-tumor efficacy against the THP-1 AML animal model, four different doses of CYAD-01 T cells were compared using the multiple injection scheme (0.3, 1, 3 and 10 million CAR-T cells per injection). While all doses had an effect in reducing tumor load, tumor control was evident in the 2 higher doses. Importantly, biodistribution of CYAD-01 CAR-T cells differed between control mice and animals bearing AML tumor xenografts. When CYAD-01 cells were injected in tumor-bearing mice, CAR-T cells quickly disappeared from the peripheral blood, indicating homing to the tissues where the cancer cells are located, such as the bone marrow. These experiments are ongoing and the results will be discussed in the conference.
Our study provides proof of principle that multiple injections of relatively low doses of CYAD-01 CAR-T cells exhibit effective anti-tumor activity in an aggressive animal model of AML. This has been verified by the promising preliminary results from the Phase I THINK trial (NCT03018405). Importantly, the persistence of CAR-T cells in the peripheral blood differs between healthy mice and AML models which requires much further investigation.