Abstract 1555: Enhanced HER2-dependent immune activation by margetuximab, an investigational Fc-engineered anti-HER2 mAb, supports combination with checkpoint blockade

Abstract

Abstract Background: Margetuximab (M) has demonstrated anti-tumor activity in patients with advanced HER2+ gastric cancer and PFS superiority to trastuzumab (T) in pre-treated metastatic HER2+ breast cancer patients (1). Similar to T, M inhibits HER2 signaling; additionally, M enhances engagement with the activating Fcγ receptor (FcR)IIIA, while diminishing interaction with the inhibitory FcRIIB. Given the role of FcRs in coupling innate and adaptive immune responses, clinical studies of M in combination with PD-1 or PD-1 plus LAG-3 blockade have been initiated (2,3). The combination of M with tebotelimab, an investigational bispecific DART® molecule blocking PD-1 and LAG-3, has demonstrated encouraging early evidence of anti-tumor activity in patients with advanced HER2+ malignancies, including those with PD-L1-negative tumors (3). Here we present data supporting the potential mechanisms underlying the anti-tumor activity elicited by M and tebotelimab combination therapy. Methods: PBMCs were challenged with HER2-expressing tumor cell lines in the presence of M or T. Immune modulation was assessed by analysis of immune cell phenotype, gene expression profiles, and cytokine secretome. Functional activity of immune cells following exposure to the combination of M with tebotelimab was evaluated. Results: M induced greater upregulation relative to T of both co-stimulatory (eg 4-1BB) and/or checkpoint molecules (eg PD-L1 and LAG-3) on NK cells and/or monocytes in PBMCs cocultured with HER2-expressing tumor cells. Compared with T, M also induced greater proinflammatory cytokine release, including IFNγ that in turn upregulates PD-L1 and MHC class I expression on co-cultured tumor cells. These effects were not observed with an Fc-null version of M or were inhibited by the addition of FcR-blocking Abs, demonstrating Fc:FcR dependency of M-induced immune activation. Interestingly, CD8 T cells, which lack FcR, also exhibited LAG-3 upregulation in the above cultures, suggesting the effects of M can also be mediated via a cascade mechanism. An IFNγ neutralizing mAb reduced tumor cell-associated PD-L1 and MHC class I upregulation, but not LAG-3 induction on T-cells. PBMCs, pre-treated with a combination of M and tebotelimab in the presence of HER2-expressing tumor cells, demonstrated enhanced cytolytic activity against both the canonical NK cell target, K562, or M-opsonized HER2-positive tumor cells. Conclusions: M, through an Fc-dependent mechanism, can upregulate checkpoint molecules on NK-cells, CD8 T cells and tumor cells, thereby sensitizing them to immune checkpoint blockade. In turn, PD-1 and LAG-3 blockade by tebotelimab enhances M-mediated NK cell cytolytic activity in vitro.