In silico clinical trials of BiTE expression by oncolytic viruses reveal the impact of patient heterogeneity on dosage protocol.

Immunotherapies have become a transformative therapeutic strategy for many cancer types in recent years. Bispecific T-cell engagers (BiTEs) are one promising immunotherapy that enhances cellular antitumour immunity by redirecting T cells towards cancer cells. Recent evidence suggests that BiTE efficacy can be augmented by encoding BiTEs in oncolytic measles virus vectors (MV-BiTE). Infection of cancer cells with MV-BiTE causes the local production of BiTEs and has shown safety and efficacy in murine tumour models. However, whether the observed efficacy of this treatment will translate to a heterogeneous human population is unknown. In this work, we generate an in silico clinical trial of MV-BiTE therapy using a system of ordinary differential equations. We capture potential heterogeneity of individual patients using variability in in vivo tumour volume and change in baseline (%) measurements from a Phase II clinical trial. In lieu of human MV-BiTE data, we use the oncolytic virus talimogene laherparepvec (T-VEC) as a surrogate oncolytic virus carrying an immunostimulatory payload. Our predictions imply that the main drivers of heterogeneity are the underlying effector T cell killing rate and BiTE pharmacokinetics. Furthermore, we find that if individuals are classified as non-responders to the Phase II T-VEC clinical protocol, they may respond to more frequent administration of lower dosages. This work highlights how in silico clinical trials can provide predictions for novel therapeutics to generate hypotheses and guide the design of treatment schedules for clinical translation.