Immunotherapeutic Targeting of TSLPR with Chimeric Antigen Receptor T Cells in AML

Adoptive transfer of T cells engineered to express chimeric antigen receptors (CARs) has achieved impressive outcomes in the treatment of refractory/relapsed B-ALL, providing potentially curative treatment options for these patients. The use of CAR T in AML, however, is still in its infancy with limitations due to the innate heterogeneity associated with AML and the lack of AML-specific targets for therapeutic development. The CRLF2 gene encodes for thymic stromal lymphopoietin receptor (TSLPR) and has previously been shown to be highly upregulated in a subset of children and adults with B-ALL. Targeting TSLPR with CAR T cells demonstrates potent anti-leukemia activity against TSLPR-positive B-ALL (PMID 26041741). Through Target Pediatric AML (TpAML), we profiled the transcriptome of nearly 3000 children and young adults with AML and identified CRLF2 (TSLPR) to be highly expressed in a subset of AML, including the majority of AML harboring KM2TA (aka MLL) fusions. TSLPR cell surface expression was validated in primary patient samples using flow cytometry, which showed uniform expression of TSLPR on AML blasts. Given that TSLPR is expressed in AML with confirmed cell surface expression, we developed TSLPR-directed CAR T for preclinical evaluation in AML.

We generated a TSLPR-directed CAR using the single-chain variable fragment (scFv) derived from an anti-TSLPR binder (clone 3G1, MD Anderson), IgG4 spacer and 41-BB/CD3zeta signaling domains. The in vitro cytotoxicity of TSLPR CAR T cells was evaluated against the REH-1 cell line and primary AML specimens. TSLPR CAR T cells demonstrated anti-leukemia activity against REH-1 as well as against primary AML specimens. To evaluate the in vivo efficacy of TSLPR CAR T cells, we developed a patient-derived xenograft (PDX) model using bone marrow cells from a TSLPR-positive patient. These cells provided a robust model system to evaluate the in vivo activity of TSLPR CAR T cells, as they produced an aggressive leukemia in humanized NSG-SGM3 mice. The PDX generated from these cells died within 2 months of transplant with significant leukemia infiltration into the bone marrow, liver, and spleen. In the in vivo study, the leukemia burden was assessed by flow cytometric analysis of AML cells in the peripheral blood and bone marrow aspirates following treatment with unmodified control or TSLPR CAR T cells given at 10x10 ⁶ T cells per mouse. After CAR T treatment, we detected a significant decrease in leukemia infiltration into the peripheral blood and bone marrow in the CAR T-treated mice compared to mice that received unmodified T cells.

In this study, we report that similar to B-ALL, CRLF2 (TSLPR) is overexpressed in a subset of AML, providing a strategy to eliminate AML cells with CAR T cell therapy. We validated the cell surface expression of TSLPR and showed that the expression is uniform across AML specimens. We further demonstrate that CAR T cells targeting TSLPR were effective in eliminating AML cells in vitro and in vivo. Given that TSLPR is highly expressed in the KMT2A-rearranged AML, a subtype that is associated with poor outcomes, TSLPR-directed CAR T cells represent a promising immunotherapy for this high-risk AML subset.