Combining a systems immunology approach with in vitro data to evaluate drug-induced changes in the immune response after Influenza A Virus infection – A blueprint for the preclinical assessment of infection risks for biologics

Biological modalities, such as antibody-based biologics, have revolutionized the treatment of many neoplastic and inflammatory diseases over the last years. Although being considered as less promiscuous to off-target interactions than synthetic immunosuppressive molecules, infections remain a key safety consideration. The immunosuppressive modulation of targeted immune system components allows infectious organisms a greater chance to spread or reactivate and cause disease in treated patients. Understanding how the modulated immune system combats infection is crucial to assess and predict potential complications of therapeutic biologics. The immune response to infections is a complex biological process, involving multiple cell types, pathways, and anatomical compartments, and is difficult to interpret solely at the level of in vitro assays. Mathematical models and computer simulations have been widely used to study immune responses to infections, e.g., in case of Influenza A Virus (IAV). We used a mathematical model [Lee et al. (2009), Journal of Virology] which represented published experimental findings on IAV infection to study the effect of a preclinical molecule on an IAV-specific immune response as compared to clinical or marketed competitor drugs with overlapping mechanism of action. The drugs are intended to inhibit autoreactive T cells in autoimmune diseases via inhibition of co-stimulatory T cell receptors leading to a decreased proliferation of effector CD4+ and CD8+ T cells. For each drug, in vitro data derived from mixed lymphocyte reaction assays were used to quantify the concentration-dependent inhibition of T cell proliferation and were integrated into the model. The drug-specific immune responses to IAV infection were simulated for two scenarios, i.e., primary IAV infection occurring during the treatment, and secondary IAV infection occurring during the treatment while IAV memory had already been established. Two in silico biomarkers of the altered immune responses were generated and compared, i.e., concentration-dependent changes in the time to viral clearance and in the antibody levels up to 30 days. As a result, the drug-dependent modulation of immune responses to IAV infection were consistent for primary and secondary infection. Furthermore, modulation of the immune response by the preclinical candidate could be ranked between those of two already marketed drugs. This allowed us to exclude the burden of an overt risk associated with the preclinical candidate as compared to the competitors which have an established clinical safety profile. This in silico study is generic since the systems immunology model is neither tailored to a specific modality, nor to the precise mode of action of the new biological or chemical entity. Instead, drug-specific actions on the immune system are represented by in vitro data integrated into the mechanistic description of model species. Furthermore, the study serves as a blueprint for the quantitative assessment of various common infections and their associated risks, including reactivation of Tuberculosis.