Discovery and Immunotherapeutic Targeting of the Tumor-Specific Target LAMP5 in Acute Myeloid Leukemia

Acute myeloid leukemia (AML) is a heterogeneous leukemia with dismal outcomes. Use of immunotherapy in AML remains difficult in large part due to overlap of putative targets on AML blasts and normal hematopoietic cells. Through transcriptomic analysis of over 2000 AML samples, LAMP5 (Lysosomal Associated Membrane Protein Family Member 5) was discovered as a potential target in AML. LAMP5 is silent in normal hematopoiesis but highly expressed in subsets of AML, making it an ideal therapeutic target. More importantly, LAMP5 is highly expressed in KMT2A-rearranged (KMT2A-r) leukemias, KAT6A-CREBBP and NUP98 fusions which are associated with poor outcomes (Panahi, Blood. 2023). Further, in those with normal karyotype, LAMP5 expression is highly associated with poor outcome. Cumulatively, combination of expression in high-risk AML and lack of expression in normal hematopoiesis makes LAMP5 an ideal therapeutic target in AML.

To evaluate the preclinical efficacy of LAMP5 as an immunotherapeutic target in AML, we developed a novel LAMP5 CAR T cell by cloning a LAMP5 targeting scFv into our CAR T backbone which includes an IgG4 hinge with a 41BB/CD3z costimulatory domain. To assess in vitro efficacy and specificity, we co-incubated LAMP5-positive (OCI-AML3) and negative (Kasumi-1) cell lines with our LAMP5 CAR T or unmodified T cells for 24 hours then assessed target specific lysis and found our LAMP5 CAR T cells resulted in over 80% cell killing in the LAMP5-positive cell line with no cytotoxicity seen in the LAMP5-negative cell line highlighting both the robust efficacy and target-specificity of our novel LAMP5 CAR T cells. To assess reactivity of our LAMP5 CAR T we assayed pro-inflammatory cytokine production including IFN gamma, IL2, and TNF alpha following 24 hours of co-incubation of CD8 or CD4 LAMP5 CAR T cells with OCI-AML3 cells. This led to over 300-fold increase in cytokine production in the CD8 CAR T and over 1000-fold increase with the CD4 CAR T compared to negligible cytokine production when co-incubated with unmodified T cells.

To assess in vivo efficacy of our LAMP5 CAR T cells we developed an aggressive in vivo cell-derived xenograft (CDX) model using OCI-AML3, transplanted into NSG mice. After leukemic engraftment, the leukemia-bearing mice were treated with either unmodified T cells or LAMP5 CAR T cells at 5x10⁶ cells (1:1 CD4:CD8) per mouse. Leukemia burden was assessed by bioluminescent imaging and peripheral blood analysis. All control mice who received unmodified T cells rapidly developed progressive leukemia requiring euthanasia by Day 20 post-T cell infusion. Disease burden in the peripheral blood showed an average of >20% leukemia in all unmodified T cell treated mice whereas no detectable leukemia was present in any LAMP5 CAR T treated mice with all LAMP5 CAR T cell mice showing clearance of disease by bioluminescent imaging. Treatment with LAMP5 CAR T cells significantly prolonged survival with average survival of 96 days post-CAR T cell infusion with one mouse living to >160 days post-CAR T infusion (p=0.0015).

Here we demonstrate the robust and target-specific cell killing of our novel LAMP5 CAR T cells against LAMP5 positive AML cell lines in vitro and a significant reduction in leukemic burden and prolonged survival in an aggressive in vivo AML CDX model. Most notably LAMP5 is silent in normal hematopoiesis, reducing the potential for on-target, off-tumor toxicity such as myeloablation. LAMP5 serves as a highly attractive immunotherapeutic target for multiple high-risk fusion subtypes of AML, notably KMT2A-r and NUP98 fusion AML which are associated with dismal outcomes and only limited therapeutic options available to date.