Background: Recombinant human (rh) IL-15, the homeostatic factor for natural killer (NK) cells, is being clinically developed, but it has little antitumor activity alone. B7H3 (CD276) is an immune checkpoint inhibitor that is associated with poorer prognosis and is highly-expressed on prostate cancer. NK cells can be given as allogeneic products and, unlike T cells, do not induce cytokine-release syndrome or neurotoxicity. Here we developed a B7H3 targeting Tri-Specific Killer Engager (TriKE) as a novel dual camelid (cam) TriKE containing WT IL-15 and two cam engagers targeting CD16 on NK cells and B7H3 on tumor targets, making NK cells antigen specific. We have previously demonstrated that NK cells infiltrate prostate cancer tumors. As proof of concept, a clinical trial of a CD33-targeted TriKE for AML (NCT03214666) induced endogenous NK cell expansion and activation in refractory AML patients. Methods: Prostate cancer cell lines or patient-derived xenografts (PDX) were incubated with healthy donor or prostate cancer patient NK cells with or without B7H3 TriKE. PDX were propagated in NSG mice and then homogenized for in vitro assays. NK cell function was measured by flow cytometry and IncuCyte live tumor imaging assays. Castration-sensitive and -resistant prostate cancer (CSPC and CRPC, respectively) patient or normal donor peripheral blood mononuclear cells (PBMC) were immunophenotyped using 42-marker NK specific or broad immune cytometry time-of-flight (CyToF) panels. Results: B7H3 TriKE resulted in a dose-dependent proliferation of NK cells, but not T cells. This was in marked contrast to rhIL-15, which stimulated both cell types. camB7H3 was broadly expressed on prostate, head and neck, ovarian and glioblastoma cancers as well as multiple myeloma. We observed a B7H3 TriKE dose-dependent increase in CD107a degranulation and inflammatory cytokines to all B7H3 positive targets that was highly specific, with no response seen with B7H3 negative hematologic targets and CRISPR KO controls. Compared to rhIL-15, B7H3 TriKE given at molar equivalent dosing induced B7H3+ target killing in in a dose-dependent manner above that seen with rhIL-15 induced natural cytotoxicity. Using CSPC and CRPC patient PBMC (n=11-15), we demonstrated that there is no significant loss in NK cell degranulation/interferon gamma production or target cytotoxicity compared to healthy age- and sex-matched donors when treated with B7H3 TriKE. CyToF analysis of CSPC and CRPC patient PBMC is ongoing. In vivo activity in xenogeneic models of human tumor is underway. Conclusions: B7H3 TriKE delivers an NK cell specific IL-15 signal to expand NK cells and is highly specific against B7H3+ prostate cancer cell lines and PDX. Clinical-grade B7H3 TriKE is undergoing validation and a Phase 1/2 clinical trial is planned to open in the 3rd Quarter of 2024 for CRPC patients progressing on one or more therapies the CRPC setting.