First step towards predicting clinical immunogenicity of biologics using in vitro based readouts as animal trial alternatives

Immunogenicity risk assessment is a critical step in the therapeutic development process of biologics. However, anti-drug antibodies continue to occur in the clinic, compromising safety and efficacy and contributing to discontinuation of otherwise effective drugs. In order to evaluate the predictive power of current preclinical tools of clinical immunogenicity, we collated, normalized, and analyzed published in vitro immunogenicity data across therapeutic proteins and monoclonal antibodies and correlated with clinical ADA frequency. Overall, combined in vitro readouts showed weak correlation with clinical immunogenicity, consistent with the limited translational performance of conventional assays. Bridging the mechanistic gaps in presentation and processing of biologics—for example, through assays that capture dendritic cell migration—substantially improved predictive power. Accordingly, New Approach Methodologies (NAMs) should integrate biologically relevant mechanisms, micro physiological systems (MPS), and advanced Artificial Intelligence/Machine Learning tools combined with data science to enhance the accuracy and translational relevance of immunogenicity prediction.