CLEC2A, a validated cancer-specific antigen, is an ideal candidate for immunotherapeutic targeting in KMT2A-rearranged AML

Abstract

Abstract Through an expansive proteo-genomic screening of over 3000 AML samples and over 200 normal hematopoietic counterparts, we discovered CLEC2A transcript to be expressed in AML, but entirely silent in all normal hematopoietic cells including cord blood stem cells (CBSC), CD34+ peripheral blood stem cells (PBSC), bulk normal bone marrow (NBM) as well as myeloid and lymphoid progenitors. Uniform and homogeneous cell surface expression of CLEC2A antigen was validated in AML. To further validate CLEC2A as a novel cancer-specific antigen, CLEC2A expression was characterized in AML and normal tissues. CLEC2A-targeting CAR T cells were developed and underwent extensive preclinical in vitro and in vivo evaluations. CLEC2A is a Cancer-Specific Antigen enriched in KMT2A-r AML: Interrogation of CLEC2A expression in AML demonstrated that CLEC2A is highly enriched in KMT2A-rearranged (KMT2A-r) AML with uniform expression in the highest risk fusion partners including MLLT4 and MLLT10. A causal link between KMT2A fusions and CLEC2A expression was established through demonstration of physical interaction of the KMT2A fusion protein with the CLEC2A promoter. As KMT2A fusions are considered to be leukemia initiating events, the causal link provides a critical rationale for targeting CLEC2A. Interrogation of the GTEX database demonstrated lack of transcript expression in all tissues except for low expression in the skin. Good laboratory practice-compliant (GLP) tissue microarray (TMA) cross-reactivity studies (Charles River Laboratories) demonstrated that cell surface CLEC2A protein is not detected in any healthy tissue (including skin), validating this target as a true cancer-specific antigen. Development of CLEC2A CART: A library of fully human CLEC2A binders were screened for binding against target positive and negative cell lines. The top 2 binders were reformatted and cloned into our second-generation CART construct backbone (IgG4 hinge with a 41BB costimulatory domain and CD3z signaling domain) for preclinical evaluations to identify the binder of choice for advancing into in vivo models. In vitro cytotoxicity and cytokine production of CLEC2A CART were assessed with each binder in CLEC2A-positive and negative cell lines compared to unmodified control T cells. The CLEC2A-3H10 binder emerged as the top candidate and was advanced for in vivo preclinical evaluations. Evaluation of CLEC2A CART Efficacy: CLEC2A-3H10 CART was evaluated in target positive and negative cell line derived xenograft (CDX) models, as well as in a highly refractory KMT2A-MLLT4 patient derived xenograft (PDX) model. NSG mice were engrafted with target positive and negative AML cells then treated with 5e6 CLEC2A CART or unmodified T cells. CLEC2A CART treatment significantly improved disease-free survival (median 71 days vs. 31 days, p=0.0012). KMT2A-MLLT4 PDX cells were engrafted in NSG-SGM3 mice and treated with an optimized cell dose of 10e6 CLEC2A-3H10 CART or unmodified T cells. CLEC2A CART treatment eradicated leukemia in PDX leukemia-bearing mice and significantly improved survival (median 141 days vs. 36 days, p=0.0002). Advancing to Clinical Trial: CLEC2A meets all criteria as a promising candidate for therapeutic targeting in KMT2A-r AML. Uniform expression in very high-risk AML and a causal link between the fusion and induction of target protein provides a rationale that effective targeting of the CLEC2A antigen may eliminate fusion positive leukemia-initiating cells. Further, lack of expression in normal tissue reduces the potential risk of on-target, off-tumor toxicity and decreases likelihood of T-cell exhaustion due to high antigen exposure. CLEC2A-3H10 CART has advanced into clinical development at Seattle Children's Hospital/Fred Hutch Cancer Center (Seattle, Washington) with lentiviral vector and at Bambino Gesù Children's Hospital (Rome, Italy) with clinical grade retroviral vector with parallel trial activation anticipated in early 2026.