Integration of immunogenicity risk assessment to facilitate a Quality by Design (QbD) approach to biologic discovery and development phases of therapeutic proteins enables mitigation of the clinical impact of immunogenicity. Systematic evaluation of potential immunogenicity risks enables optimal molecule design, selection of candidates, and adaptive CMC strategies, whilst maintaining developability and potency. A QbD approach accelerates development by avoiding cumbersome deimmunization procedures, and enables smooth regulatory interactions through the product lifecycle.

The Immunogenicity Database Collaborative (IDC) was established to address fragmented immunogenicity data and provide open-source resources for the industry. Following the release of IDC DB V1, a curated clinical dataset integrating therapeutic, sequence, and trial-level features, the initiative is expanding to V2 through coordinated working groups. This presentation will highlight V1 insights and progress toward V2, including the integration of nonclinical data, automated clinical data capture, and expanded annotation of clinical ADA impacts.

Antigen-presenting cell (APC) internalization is a critical early event in eliciting anti-drug antibody (ADA) responses, making it a key parameter in immunogenicity risk assessment. We have developed a novel primary APC internalization and activation platform to predict and assess immunogenicity risk for monoclonal antibody (mAb)-based therapeutics. Internalization and activation (Signal 2) metrics demonstrate significant correlation across a panel of clinical mAbs, enabling mechanistic risk stratification. Complementing this, ongoing in silico predictive modeling efforts aim to predict APC internalization at the discovery stage, enabling high-throughput triaging and supporting data-driven immunogenicity decision-making early in biologics development.

This talk reviews the evolving machine learning models for the prediction of TCR-pMHC interaction, assessing state-of-the-art methods and their performance on out-of-domain distributions. Key focus areas include the impact of feature representation, the size and diversity of available datasets, and the necessity of rigorous and independent benchmarking. We will discuss in detail the generalization of existing models to unseen epitopes and the selection of unbiased negative data as major challenges in the field.

The B Cell Immunogenicity Risk Assessment (BIRA) working group is an open forum advancing risk assessment of the B cell component of unwanted immune responses to biologics. Associated with the European Immunogenicity Platform (EIP) and the American Association of Pharmaceutical Scientists (AAPS), BIRA has established a collaborative framework to assess current knowledge, identify gaps, and harmonize concepts and terminology of B cell immunogenicity risk assessment. This presentation will provide an overview of BIRA’s objectives and progress to date.

Building on our human B cell assay using healthy donor blood, we are developing an ex vivo splenocyte culture (immune organoid) system. This exploratory platform aims to provide deep mechanistic insights into the immunogenicity of complex therapeutics, such as immunocytokines and multispecific antibodies. By capturing drug-specific plasmablast expansion and cellular crosstalk, these organoids facilitate the early identification of biological drivers behind immune responses in next-generation biologics.

New approach methodologies (NAMs) are increasingly used to support regulatory decision-making by generating fit-for-purpose, human-relevant evidence to complement—and, where justified, reduce—animal studies. We will review current NAM approaches for preclinical immunogenicity, immunosafety, and DMPK, including in silico prediction and established in vitro microphysiological systems (MPS) such as organoids and organ-on-chip, alongside targeted assays. We will discuss context of use, validation/qualification considerations, and fit-for-purpose reporting elements to support IND-enabling, decision-quality packages.

Immunogenicity can compromise the safety and/or efficacy of therapeutic protein products, and is a priority issue for regulatory agencies. In addition to poor patient outcomes, the social and economic costs associated with neutralizing antibodies are considerable. In this presentation, I survey the immunogenicity of approved therapeutic proteins, discuss strategies for clinical management of immunogenicity, and identify challenges associated with circumventing the immune responses to approved protein therapeutics.

Oral cyclic peptides with noncanonical amino acids (NCAAs) are an emerging modality combining protease resistance and permeability with new immunogenicity questions. This talk presents an end-to-end immunogenicity risk assessment framework, highlighting key drivers of ADA risk and clinical relevance. Limitations of protein-centric in silico MHC II tools for NCAA macrocyclic peptides and the role of in vitro assays are discussed, alongside a tiered risk-based ADA strategy emphasizing clinically meaningful effects.

The therapeutic peptide field is expanding rapidly, driven by high target specificity, selectivity, favorable safety profile, and market demand. To date, generic complex peptides have largely been limited to synthetic manufacturing, regardless of the reference product’s origin. As FDA considers broadening the 505(j) pathway, this presentation explores key risk differences between synthetic and recombinant peptides and reviews available nonclinical strategies, highlighting strengths and gaps in immunogenicity risk assessment for recombinant peptides in the generic space.

Positive control selection is critical to ensuring neutralizing antibody assays generate robust, interpretable data suitable for regulatory decision-making. Poorly chosen controls can mask assay limitations, obscure clinically relevant neutralization, and introduce risk as programs advance. This presentation highlights practical strategies for selecting and qualifying biologically relevant positive controls that reflect biological relevance, improve assay performance, and reduce development risk, ultimately strengthening confidence in immunogenicity assessments across biologics.

Preclinical vaccine development typically relies heavily on animal models, which are poorly predictive of human responses and do not incorporate age, a key determinant of immunity. To address this unmet need, we have developed a suite of complementary human in vitro modules to profile age-specific innate and adaptive immune activation potential of candidate vaccine formulations. This suite of platforms is named MEMPHIS (Modular Evaluation of iMmunogenicity using multi-Platform Human In vitro). Our published and unpublished observations confirm that MEMPHIS can rapidly and accurately down-select candidate vaccine formulations in vulnerable populations where participants have suboptimal immunity to immunization.