INTRODUCTION: Acute myeloid leukemia (AML) is the most common form of adult acute leukemia. Currently, allogeneic hematopoietic stem cell transplant (HCT) is the standard of care for high-risk patients yet ~40% patients will relapse, emphasizing the need for new therapeutic approaches. Effective targeted therapy in AML has been hampered by a lack of cancer specific targets. One described AML target is CD33, a glycoprotein expressed predominantly on normal myeloid cells and the majority of AML cells. However, CD33-directed therapies have been challenged by cytopenias due to on-target, off-tumor toxicity and, in the case of bispecific antibodies, cytokine release syndrome (CRS). Tremtelectogene empogeditemcel (trem-cel) is a gene-edited hematopoietic stem and progenitor cell (HSPC) transplant product currently in clinical trial (NCT04849910) whereby CD33, shown to be dispensable for normal hematopoiesis and function, is deleted to allow for post-HCT treatment with CD33-directed immunotherapies with reduced myelotoxicity as the reconstituted hematopoietic compartment lacks CD33. JNJ-67571244, a bispecific antibody that binds to both CD3 and CD33, is capable of inducing T cell recruitment for CD33-directedtumor cytotoxicity. This study examines whether CD33 neg hematopoietic cells are protected from JNJ-67571244 and if CD33 neg human HSPC (hHSPC) xenotransplanted mice have reduced levels of inflammatory cytokines associated with CRS when treated with JNJ-67571244. METHODS: In vitro proof-of-concept (POC) studies were carried out using CD3 + T cells co-cultured with JNJ-67571244 and AML target cell lines. Further studies were conducted with CD34 + hHSPCs that were either CRISPR/Cas9 gene edited to delete CD33 or treated with a non-targeting gRNA control (gCtrl) followed by differentiation towards the monocytic lineage. These in vitro differentiated monocytes (MIVD) were assayed for editing frequency, CD33protein expression and myeloid phenotype by flow cytometry prior to co-culture with donor matched T cells and JNJ-67571244 ranging from 0.5 pM to 500 nM. Cytotoxicity was analyzed using flow cytometry at 48 and 72 hr and cytokine concentrations in the supernatants were measured using Luminex. For in vivo POC studies, CD33 or gCtrl-edited CD34 + hHSPCs were xenotransplanted into NSG-SGM3 mice followed by treatment with either control bispecific (CD3xNull) or JNJ-67571244 at 0.5 mg/kg. Plasma cytokine concentration were measured via Luminex and hematopoietic tissues were analyzed by flow cytometry to measure protection from JNJ-67571244 cytotoxic killing. RESULTS: Annexin V and Live/Dead staining showed that JNJ-67571244 was able to induce AML target cell killing by human T cells. Importantly CD33 neg MIVD cells co-cultured with T cells and JNJ-67571244 were significantly more viable at all doses compared to gCtrl cells. The EC 50 of gCtrl MIVD was 0.0208 nM whereas the CD33 neg MIVD EC 50 was at least 20,000-fold higher (>430.8 nM). T cell activation, measured by CD25 and CD69 positivity, and release of cytokines that are associated with CRS such as IFN-γ, TNF-α, IL-2, IL-6 and IL-10, were all significantly decreased in the CD33 neg MIVD cells treated with JNJ-67571244 compared to the gCtrl group at all doses. Analysis of blood and bone marrow from xenotransplant studies revealed that JNJ-67571244 was effective at killing CD33 + cells as well as CD14 + monocytes in the gCtrl group treated with JNJ-67571244. Conversely, CD14 + monocytes and CD15 + neutrophils derived from CD33-edited hHSPCs, were not significantly reduced in cell number when treated with JNJ-67571244 suggesting that they were protected from JNJ-67571244 specific killing. IFN-γ, TNF-α, IL-2, IL-6 and IL-10, were all significantly decreased in mice transplanted with CD33-edited hHSPCs compared to gCtrl group in response to JNJ-67571244. CONCLUSIONS: Taken together, these studies demonstrate that CD33 deleted hematopoietic compartment is protected from the CD33 directed immuno-therapy JNJ-67571244 both in in vitro cytotoxicity assays and preclinical xenotransplantation studies, with decreased concentrations of inflammatory cytokines associated with CRS. These findings enable the development of a next-generation AML treatment strategy by pairing trem-cel transplant with a subsequent CD33-directed bispecific compound to potentially improve safety and efficacy while minimizing myelotoxicity.