(T1130-02-11) Rapid and Predictive Immunogenicity Risk Assessments Using In Silico Methods

Purpose: Immunogenicity risk assessment is an essential component of appraising the development potential of biologic drug candidates. The incidence of anti-drug antibody (ADA) formation is correlated with CD4+ T cell epitope content. Our group has developed an in silico immunogenicity risk assessment platform (ISPRI) and comprehensive in silico methods for identifying (CD4+) T effector and regulatory T cell epitopes (Tregitopes) in monoclonal antibodies, enabling a rapid and accurate forecast of immunogenic potential. Recently, drug developers have emphasized in vitro methods, such as MAPPs and DC:CD4+ T cell restimulation assays to identify potentially immunogenic peptides processed and presented by human antigen-presenting cells (APCs). In this study, our immunoinformatic tools were retroactively applied to a set of 24 biotherapeutics originally presented by Siegel M, et al. We compared reported rates of ADA formation to rates predicted by our in-silico algorithms and to rates observed in DC:CD4+ T cell restimulation assays and also compared the estimated cost and speed of the two types of assays.

Methods: We have previously evaluated the amino acid sequences of a panel of biotherapeutics with varying clinical ADA response rates. Using models for globally representative HLA-DR supertype alleles covering >95% of the human population, ISPRI generated Tregitope-adjusted Immunogenicity Score, which were highly correlated with clinical immunogenicity; an “ADA prediction” tool using this model is now available in the ISPRI toolkit. For this case study, we conducted a retrospective comparison of ADA response rates for 24 biologics using the ISPRI ADA prediction tool, and ADA response rates obtained through the DC:CD4+ T cell restimulation assay published in Siegel at al. (2022). We performed a correlation analysis between predicted ADA response rates and observed ADA for 24 biotherapeutics, encompassing both ISPRI-predicted ADA response rates and in vitro immunogenicity rates that were assessed using the DC:CD4+ T cell restimulation assay. Furthermore, we assessed the predictive efficiency of ISPRI-predicted ADA rates and ADA rates obtained in the DC:CD4+ T cell restimulation assay.

Results: The ADA response rates generated by ISPRI for a diverse panel of 24 biotherapeutics revealed noteworthy correlations with observed clinical ADA rates. 19 out of 24 biotherapeutics were correctly categorized as either high or low, resulting in a remarkable 79% alignment. The DC:CD4+ T cell restimulation assay correctly categorized 16 biotherapeutics out of the 24, achieving a 67% alignment rate. Both methods correctly categorized the majority of cases. However, the in-silico process was completed in less than 24 hours using ISPRI, whereas the in-vitro approach would likely require many weeks, if not months, to complete.

Conclusion: As this case study illustrates, the ISPRI approach provides a rapid and lower cost method of assessing immunogenicity risk potential and is closely correlated with clinical observations. In this study, ISPRI-predicted ADA rates exhibited a robust alignment of approximately 70% with clinical rates, further aligning with ADA rates derived from in-vitro assays. Despite the significant correlation between the DC:CD4+ T cell restimulation assay and ISPRI, certain differences persisted. These variations could arise from potential limitations in the in-silico method or, alternatively, from information gaps in the in-vitro assays (29% of the variance between the invitro results and the clinical results was ‘unexplained’). The ISPRI toolkit provides an important ‘first look’ at risk that can be obtained more rapidly and at lower cost than in vitro assays.

References: Siegel M, et al. Validation of a Dendritic Cell and CD4+ T Cell Restimulation Assay Contributing to the Immunogenicity Risk Evaluation of Biotherapeutics. Pharmaceutics. 2022 Dec 1;14(12):2672. doi: 10.3390/pharmaceutics14122672. PMID: 36559166; PMCID: PMC9781343.