Background and Purpose

‍

Unanticipated central nervous system (CNS) neurotoxicity remains a major driver of late-stage clinical attrition, particularly for drugs that progress through traditional preclinical testing without revealing human-relevant liabilities. This challenge is further amplified for emerging modalities such as antisense oligonucleotides (ASOs), where CNS adverse events are difficult to predict based on sequence or target biology alone.

Human induced pluripotent stem cell (iPSC)-derived cortical organoids provide a physiologically-relevant system to model neuronal network activity and functional perturbations in vitro. In this study, we leverage a combination of 28bio CNS-3D Brain Organoids and AI-enabled predictive modeling to identify clinically observed seizure liability across both small molecules and ASOs.