| Abstract: |
| We present a stochastic agent-based model for the spatial dynamics of infection by oncolytic viruses in solid tumours. The model describes the interactions between uninfected and infected tumour cells and considers two alternative movement mechanisms, namely undirected random motion and pressure-driven displacement. For both cases, we obtain the corresponding continuum equations and carry out a systematic comparison between the individual-based and PDE descriptions in one and two space dimensions. We also study the associated one-dimensional travelling waves.
In the case of undirected motion, the agreement between agent-based simulations and the numerical and analytical results for the continuum model is good. For pressure-driven motion, instead, marked discrepancies emerge: over a wide parameter range, infection remains confined to the tumour core, even though the continuum model predicts travelling infection waves. These results highlight the significant impact of spatial constraints in the tumour microenvironment on virotherapy, as well as the central role of stochastic effects.
The same modelling framework also provides a natural basis for extensions including explicit viral particles, making it possible to assess when reduced quasi-steady descriptions are adequate and when viral kinetics significantly affect front propagation and transient behaviour.
This talk is based on results published in Bulletin of Mathematical Biology 85:92 (2023) and 88:66 (2026). |
|