Preclinical study of armored claudin 18.2 nanobody CAR NK cells targeting gastric cancer

Introduction: Claudin 18.2 (CLDN18.2) is highly expressed in gastric cancer with minimal expression in normal tissues, making it an attractive target for engineered cell therapies. CAR NK cells offer a potentially lower risk of inciting inflammatory cytokines and may be easily sourced for allogeneic or “off-the-shelf” use. However, conventional CAR NK approaches often show limited persistence and reduced activity when engaging with solid tumors. To enhance antitumor activity and persistence, we developed CLDN18.2 nanobody-based CAR NK cells incorporating IL-15/IL15sushi armoring. This study evaluated the antitumor efficacy and persistence in vitro and in vivo of the armored CLDN18.2 CAR NK cell construct generated via a feeder-independent NK expansion platform. Methods: Alpacas were immunized with recombinant human CLDN18.2 extracellular domain protein. Peripheral blood mononuclear cell (PBMC) mRNA was extracted and converted into a VHH (nanobody) to generate a library. The library was panned against CLDN18.2 antigen to enrich for high-affinity binders, which were identified through ELISA screening and cellular binding analysis by flow cytometry. VHH clone sequencing were used to generate our cCAR.A CLDN18.2-specific nanobody was identified through screening and used to generate an armored CLDN18.2 CAR NK cell construct. NK cells were isolated from cryopreserved cord blood and expanded using a feeder-independent NK expansion platform. CAR expression was measured by flow cytometry. Cytotoxicity was assessed using SNU-601 and REH-C18.2xp target cells. NSG mice received CAR NK cells. Results: The armored CLDN18.2 CAR NK cells displayed strong CAR expression following NK cell transduction. CLDN18.2 CAR NK cells demonstrated remarkable in vitro cytotoxicity across multiple effector-to-target ratios, eliminating CLDN18.2-positive targets rapidly and completely. In vivo, the armored CLDN18.2 CAR NK cells were observed to impart strong anti-tumor activity resulting in regression in SNU-601 xenografts. In NSG mice, armored CLDN18.2 CAR NK cells resulted in rapid tumor reduction and measurable NK persistence in mouse peripheral blood. Conclusions: Armored CLDN18.2 nanobody CAR NK cells generated via a feeder-independent NK expansion platform showed strong antitumor activity and improved in-vivo performance in CLDN18.2-positive tumor models. CAR NK cells also demonstrated sustained persistence following infusion. Together, these findings support future clinical investigation of our platform in patients with CLDN18.2-positive gastric cancer.