Abstract

Immune checkpoint inhibitors (ICIs) have transformed cancer treatment by significantly improving clinical outcomes across a range of malignancies. However, systemic immune-related adverse events (irAEs) remain a major limitation of current ICI strategies. To address this challenge, we developed a bispecific antibody (bsAb) to selectively target tumor cells co-expressing hepatocellular carcinoma-associated antigen glypian-3 (GPC3) and programmed Cell Death Ligand 1 (PD-L1). The GPC3 × PD-L1 bsAb retained high affinity for GPC3 while exhibiting a ~ 300-fold reduced affinity for membrane PD-L1 compared to atezolizumab, enabling GPC3-dependent PD-1/PD-L1 blockade. The GPC3 × PD-L1 bsAb demonstrated potent and selective immune checkpoint inhibition in GPC3 high tumor cells in vitro and elicited robust antitumor responses in GPC3-positive tumor models. Importantly, the GPC3 × PD-L1 bsAb, unlike the anti-PD-L1 monoclonal antibody, exhibited selectivity for GPC3/PD-L1 double-positive tumors while sparing PD-L1-positive but non-targeting tissues. Biodistribution studies further confirmed preferential accumulation of the GPC3 × PD-L1 bsAb in GPC3-positive tumors. Collectively, these data provide proof-of-concept (POC) validation for the efficacy and safety of a tumor-targeted immune checkpoint inhibitor, highlighting its potential to enhance tumor specificity and reduce systemic toxicity for GPC3-expressing malignancies such as hepatocellular carcinoma.