Rational development of multi-targeted CAR-T cell constructs in pediatric acute myeloid leukemia
Terry Fry
MDUniversity of Colorado-Denver
Project Term: July 1, 2019 - June 30, 2022
We propose to rationally select targets for and design multi-antigen CAR constructs for pediatric acute myeloid leukemias with the intent of generating durable responses and reducing off-tumor toxicity. This will be achieved using two complimentary approaches 1) the identification of markers expressed on pediatric AML stem cells (LSC); and, 2) phenotypic interrogation of AML evolution under single antigen CAR targeting in both preclinical models and in patients treated with a CD33 CAR trial.
Acute myeloid leukemia (AML) is a type of leukemia that affects both children and adults of all ages. It is also a particularly difficulty type of leukemia to treat, as the disease usually comes back (relapses) after initial therapy and becomes more resistant to additional therapy at that stage. One form of immunotherapy for acute lymphoid leukemia (ALL), known as chimeric antigen receptor T cells (CART), has had impressive success in treating this disease despite patients receiving multiple prior lines of therapy without benefit and is likely curative for some patients. CART are T-cells that are engineered to express a receptor that both identifies a surface protein (antigen) on the leukemia cells and generates a “kill” signal for the T-cells to attack the target cells. While promising, CART therapy to date has not been a magic bullet, since the leukemia can remove its surface proteins to escape the T cells over time in half of paitients who enter remission with CART. A possible means to overcome these limitations of CART is to design receptors that target two different proteins at once – so-called dual antigen targeting CART. Leukemia cells being attacked at two different points are less likely to be able to develop resistance to these cells. This approach is being tested in patients with ALL. Because of differences in the surface proteins on AML in the same patient and between patients as well as the ability fo AML cells to change expression of these proteins during therapy, it is necessary to explore the dual antigen targeting when designing CART for this disease in order to provide the best long-term outcomes possible. We intend to take a rational approach to designing dual antigen targeting CART for AML in this project. We will study a panel of pediatric patient samples to identify the surface proteins that are most highly expressed and which can serve as reliable markers of leukemia stem cells (LSC), which lead to chemotherapy resistance and relapse in adults with AML. AML LSCs are know to exist in pediatric AML but have not been well studied. Established differences in the genetics of AML in children compared to adults indicate that LSCs will also be different. We will also study samples in the lab and from clinical trial patients treated with single antigen CART, to understand how relapse happens and what surface proteins change as leukemia develops resistance. Finally, we will use all of this information to design dual antigen targeting CART that (1) target LSC and (2) anticipate the ways in which leukemia develops resistance and prevent it. We will then test these CART to see how effective they are at killing leukemia cells in mice. With this information we will gain knowledge that can be used to develop future clinical trials for children with AML using these optimized CART constructs with the goal of inducing durable remissions.