Abstract 1728: Overcoming target heterogeneity in gastric cancer with an eKiH-engineered CLDN18.2×HER2 bispecific ADC

Abstract Background: Target heterogeneity, mosaic and cluster expression of CLDN18.2 and HER2 is common in gastric cancer (GC) and undermines the efficacy of single-antigen therapies. We previously developed DA3501 (AT-211), a CLDN18.2-targeting ADC currently advancing toward Phase 1 evaluation. The DA3501 antibody clone demonstrates enhanced affinity and internalization compared with zolbetuximab. To more comprehensively address heterogeneous tumors, we engineered a CLDN18.2×HER2 bispecific ADC built on Dong-A ST BsAb platform incorporating an enhanced knob-into-hole (eKiH) interface to improve heavy-chain pairing fidelity and bispecific assembly. Methods: Antibody discovery and optimization yielded lead clones for CLDN18.2 and HER2 with improved binding and internalization. Bispecifics were assembled using Dong-A ST BsAb platform. Conjugation to monomethyl auristatin E (MMAE), exatecan (a topoisomerase-I inhibitor), or dual-payload configurations was performed using site-selective chemistry. Target prevalence and co-expression were evaluated in human GC using single-cell RNA sequencing (scRNA-seq) of tumor specimens and multiplex immunohistochemistry (mIHC) on tissue microarrays (TMAs). Immune correlates were analyzed to assess combinatorial potential with immune checkpoint inhibitors (ICIs). Results: scRNA-seq revealed CLDN18.2 and HER2 expression in mutually exclusive as well as co-expressing tumor cell subsets, corroborated by mIHC showing intra-tumoral mosaic and clustered patterns. The DA3501 anti-CLDN18.2 clone exhibited superior binding kinetics and internalization versus zolbetuximab. ATS1002 retained high-affinity binding to both targets and demonstrated efficient internalization in dual-positive and single-positive (CLDN18.2 or HER2) models, supporting a “catch-all” approach for heterogeneous lesions. ADCs conjugated with MMAE, exatecan, or dual payloads showed potent in vitro cytotoxicity across CLDN18.2-high, HER2-high, and co-expressing GC lines. Immune-correlative analyses indicated potential synergy with ICIs in tumors exhibiting activated T-cell and myeloid signatures. Conclusions: This BsAb is a rationally engineered CLDN18.2×HER2 bispecific ADC designed to overcome target heterogeneity in GC. Together with the DA3501 program, it broadens therapeutic coverage of CLDN18.2/HER2-expressing tumors and combination-strategy studies, including ICI co-therapy. Citation Format: Kyoung-Ho Pyo, Seong-Hyun Park, Dongsop Lee, Haneol Kim, Younggyu Kong, Younggeun Lee, Hojin Yeom, Sowon Aum, Sun Hee Park, Huijo Oh, Cheyeon Kim,1 Hyeonseok Jin, Aera Lee, Hojeong Hong, Ju Hwan Kim, Hyungseok Choi, Mi-Kyung Kim, Taedong Han. Overcoming target heterogeneity in gastric cancer with an eKiH-engineered CLDN18.2×HER2 bispecific ADC [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 1728.