Multispecific antibodies, especially TCEs, promise powerful tumor cell killing ability but face practical hurdles: high-throughput assembly for testing, tunable CD3 potency, costimulatory control, and precise heavy/light chain pairing for manufacturability. This poster showcases an integrated toolkit that addresses those bottlenecks with a chain exchange–based screening process, functionally profiled anti-CD3 and anti-CD28 panels, high-fidelity pairing mutations, and low polyreactivity to ensure developability. 

Approach and outcomes

• Proprietary HC-LC and HC-HC pairing mutations (in CH1/Cκ and CH3) direct correct assembly of four-component IgG-like bispecifics; a therapeutic using these mutations (HS-20117) is in clinical trials.
• Adimab’s chain-exchange (ChEx) workflow can convert 24 monospecific inputs into as many as 144 multispecifics, and the HIC retention time of outputs has been shown to match the mean of their inputs, supporting predictable biophysical behavior.
• In vivo discovery, humanization, and yeast-based optimization produced a broad-affinity anti-CD3 panel with human-cynomolgus cross-reactivity and low polyreactivity relative to clinical comparators.
• When formatted as CD3 × HER2 bispecific IgGs, anti-CD3 variants drove potent redirected T cell cytotoxicity (RTCC) on HER2-high (SKOV3) and HER2-low (OVCAR3) cells, with efficacy tracking with the CD3 monovalent affinity.
• A format-diverse anti-CD28 panel was affinity-optimized to provide tunable costimulation; coadministration enhanced cytotoxicity and shifted IL-2 EC50s in line with increased activity.

These results directly address the assembly, throughput, and functionality-tuning challenges that slow TCE development. The data support a practical route to rapidly build large, developable multispecific panels while maintaining predictable biophysical properties and controllable T cell activation.

View poster