Computational approaches to modelling and optimizing cancer treatment

Computational models can be applied to optimize treatment schedules and model treatment responses in cancer therapy. In this Review, we provide an overview of such computational approaches, including deterministic models, such as those based on ordinary and partial differential equations, stochastic models, spatially explicit agent-based approaches as well as control theory and machine learning methods. We discuss their advantages and current limitations in different scenarios. We outline how therapeutic decision-making can be aided by mathematical and computational approaches and how patient-specific responses can be assessed and incorporated into such methods. We also survey models that can incorporate adaptive changes throughout the course of treatment and discuss data and parameter estimation approaches. Finally, we highlight how such methods can lead to the identification of optimum treatment options for individual cancer and treatment types, and examine the challenges that remain to be addressed to enable the clinical translation of computational models in cancer therapy. Mathematical modelling provides a means of understanding cancer evolution and optimizing cancer treatment response. This Review outlines different computational methods of modelling cancer treatment response and identifying optimal treatment strategies.