AGB101, a first-in-class Vbeta17 x DLL3 bispecific antibody, selectively redirects Vbeta17 T memory cells to drive robust DLL3 dependent anti-tumor activity

CD3-based bispecific antibodies (bsAbs) that recruit pan-T cells to target tumor-associated antigens (TAAs) have shown clinical benefit but their use is limited by high rates of immune-related toxicities such as cytokine release syndrome (CRS) and immune effector cell associated neurotoxicity syndrome (ICANS), arising from broad T cell activation following CD3 engagement. In addition, CD3 bsAbs recruit antagonizing T cell subpopulations such as regulatory or exhausted T cells that can narrow the therapeutic index and affect durability of response. To overcome these limitations, we developed a novel bsAb platform (anti-TRBV19 x anti-TAA) that selectively activates the TRBV19 (Vbeta17) chain of the T cell receptor. The Vbeta17 T cell subset is a memory population that constitutes 5% of circulating T cells in healthy individuals and cancer patients. Vbeta17 T cells exhibit cytotoxic activity, enhanced persistence, and reduced susceptibility to exhaustion. Their selective recruitment enables more efficient anti-tumor immune responses, sparing activation of non-effector T cells. AGB101, is a highly optimized Vbeta17 bsAb targeting the solid tumor lineage antigen DLL3 that is overexpressed in small cell lung cancer (SCLC) and other neuroendocrine tumor types. AGB101 demonstrates DLL3-dependent activation and selective expansion of Vbeta17 T cells while the broader population of T cells, including naïve T cells, remain unaffected. In contrast, treatment with the CD3 x DLL3 bsAb, tarlatamab, leads to a rapid expansion of the entire T cell population including non-cytolytic subsets. In human PBMC/SHP-77 (DLL3+) co-culture assays, AGB101 showed equivalent cytotoxicity to tarlatamab, despite engaging 15-20-fold fewer immune effector cells, indicating that effector cell type rather than their prevalence mediates anti-tumor cytotoxic responses. AGB101 selectively and robustly expanded CD8+ Vbeta17 T cells with minimal CD4+ T cell expansion and induced stronger effector cell activation compared to CD3 engagement. In addition, AGB101 significantly increased the expansion of CCR7+/CD45RO+ memory T cells compared to tarlatamab treatment. Notably, AGB101 treatment resulted in much lower levels of cytokine release, particularly pro-inflammatory cytokines (e.g., TNF-alpha, IL-6, IL-1beta), compared to tarlatamab treatment. Lastly, AGB101 demonstrated robust anti-tumor activity in vivo in two DLL3 expressing SCLC tumor models. Collectively, our data demonstrate that TCR-directed targeting improves the quality, magnitude, and durability of anti-tumor T cell responses and these features of AGB101 mediated anti-tumor activity highlight selective Vbeta17 T cell redirection as a promising next-generation immunotherapeutic approach to enable more efficient immune system engagement, with the potential to provide improved clinical benefit to patients.