Phase I clinical trial evaluating dual-targeting CAR-T cells (KD-496) against CLDN18.2 and NKG2DL in advanced gastrointestinal cancers and pancreatic cancers.

4224 Background: Chimeric antigen receptor T-cell (CAR-T) therapy has shown limited efficacy in gastrointestinal (GI) cancers. Previous studies indicate that bispecific CAR-T cells incorporating tandem single-chain variable fragments (scFvs) demonstrate enhanced therapeutic indices compared to monospecific constructs. We developed novel bispecific CAR-T cells (KD-496) targeting both NKG2D ligands and CLDN18.2, demonstrating promising preclinical antitumor activity and safety profiles. Methods: We conducted an open-label, single-arm, "3+3" dose-escalation study (NCT06134960) in patients with treatment-refractory advanced GI cancers. Following lymphodepletion with fludarabine, cyclophosphamide, and nab-paclitaxel, subjects received a single KD-496 infusion at three dose levels: 1×10⁸ (DL1), 3×10⁸ (DL2), or 5×10⁸ (DL3) CAR-T cells. Primary endpoints included safety and toxicity; secondary endpoints comprised efficacy, pharmacokinetics, and immunogenicity. Results: Between March 2024 and October 2025 , seven subjects (age range: 43-73 years) received KD-496 treatment (n=3 at DL1; n=3 at DL2; n=1 at DL3). Only one Grade IV possibly treatment-related adverse events (TRAEs) is Neutrophil count decreased (14%); Grade III possibly treatment-related adverse events (TRAEs) included Neutrophil count decreased (14%); Alanine aminotransferase increased (14%); Elevated aspartate aminotransferase (14%); anemia (29%); and White blood cell decreased (57%). Cytokine release syndrome (CRS) occurred in seven subjects, with one case (17%) reaching grade 3; no grade 4/5 CRS or neurotoxicity was observed. The dual-antigen requirement effectively prevented on-target off-tumor toxicity, with no patients experiencing severe gastrointestinal adverse events. No dose-limiting toxicities or serious adverse events were reported. Among evaluable subjects: Those with gastric cancer (n=4), achieved an objective response rate (ORR) of 75%; Among three subjects with pancreatic cancer, disease control rate (DCR) of 66.7% overall; Achieved an objective response rate (ORR) of 33% overall. Notably, the ORR reached 100% for both gastric cancer and pancreatic cancer advanced subjects in the medium-dose group. Conclusions: These initial data demonstrate a favorable safety profile for KD-496 CAR-T cells targeting NKG2DL/CLDN18.2. The preliminary efficacy signals warrant further investigation of KD-496, particularly in gastric cancer and pancreatic cancer. Clinical trial information: NCT06134960 .