Optimisation of in vitro assays for accurate risk assessment of T-cell responses to biologics with potential immune liabilities.

Biologics are increasingly used to treat disease, but their clinical efficacy is often hindered by anti-drug antibodies and immunological adverse reactions. Accurate prediction of immunogenicity to biologics at a preclinical development stage would improve safety and prevent financial loss to pharmaceutical companies. A stepwise approach to explore immunogenicity to biologics was developed by; (1) assessing the sensitivity of existing PBMC stimulation assays, (2) optimisation of a DC: T-cell co-culture assay, (3) definition of immune epitopes using rituximab MHC class II-binding peptides. Lymphocyte transformation tests and IFN-γ ELISpot assays were performed using healthy donor PBMC to assess the T-cell stimulatory capacity of a panel of biologics. DC: T-cell co-culture methods were optimised for timepoints, cell numbers and immune dysregulation and assessed using the panel of biologics. Sixteen 15-mer peptides derived from the variable region of rituximab were used to generate and characterise T-cell clones. Initial PBMC experiments showed low sensitivity, presenting frequent false negatives. DC: T-cell assays showed increased sensitivity to biologics, however, high background responses were observed in vehicle control cultures. Optimisation of assay conditions reduced the background counts and increased the dynamic range and sensitivity of the assay. CD4 + T-cell clones were responsive to two 15-mer rituximab peptides in proliferation and cytokine release assays. These data indicate that DC: T-cell co-cultures and cloning methods should be used alongside epitope prediction tools to assess T-cell responses to biologics with potential immune liabilities. Implementing the stepwise approach outlined could identify high risk molecules early in development and highlight "hotspot" regions of molecules for risk mitigation.