Immunogenicity Assessment of Biopharmaceuticals Using In Silico/In Vitro Combined Method

摘要

In recent years, the development of biological drugs has been made great strides, because of better specificity and less toxicity than chemical drugs. However, biological drugs can induce unwanted immune responses in patients leading to the development of anti-drug antibodies (ADAs). These ADAs can lead to reduction of drug efficacy through neutralization and rapid clearance of drugs, or toxicity due to cross-reaction with endogenous proteins. Therefore, eliminating the potential immunogenicity risk of biopharmaceuticals is essential for successful drug development. For the estimation of potential risk in clinical studies, in silico and in vitro PBMC-based analyses should be considered the first tools for the evaluation of immunogenicity, because of limitations in animal studies. Here we present an approach to improve the efficiency of both in silico prediction and in vitro assessment of immunogenicity. For the automated in silico prediction of MHC class I and II epitopes, we established an interface that visualizes the results from IEDB according to p values. The interface shows a potential average risk in three different colors depending on the p values (0.01, 0.05, and 0.1) of the region containing potentially immunogenic peptides in 27 HLA types. In silico prediction can be used to select HLA types with potential immunogenicity risk for in vitro evaluation of immunogenicity. We also found that the measurement of IFN-g levels in PBMC-based CD4 T cell assay systems using MSD is a cost-effective in vitro assay requiring less PBMCs at hi gher sensitivity than ELISA, CBA, or ELISpot assays. In summary, our study suggests that in silico/in vitro combined immunogenicity analysis is useful for assessing the potential immunogenicity risk of biopharmaceuticals in the early stages of drug development.