Developing a hybrid agent-based mathematical model to study bladder infections

Abstract

Urinary Tract Infections (UTIs) are amongst the most common infections worldwide, affecting well over 150 million people each year. UTIs now account for 40% of all hospital-acquired infections and they are becoming harder to treat, with an estimated 1 in 3 being resistant to the most commonly used antibiotics. UTIs are also notorious for recurring. Using agent-based modelling techniques we have developed a model that describes the infection process in bladder infections. Gaining a better understanding of the pathophysiology can prove key to the development of new treatment strategies with greater success and less resistance rates. Data arising from animal and cell culture models are incredibly valuable but can be limited by ethical considerations and time constraints. Simulations from in silico models can be analysed and their results can complement clinical findings, adding to the understanding of how the infection behaves and how we might improve treatment. We simulate discrete agents in the system: E. Coli bacteria and the immune cell types reported to have critical roles in lower UTIs (LY6C macrophages, neutrophils and mast cells). The capacity of the bacteria to penetrate the bladder epithelial barrier and seek refuge in the bladder epithelial cells is a critical initiating step of infection and this process is simulated in our model. We also integrated in our model a treatment framework which we have used to investigate the effects of Trimethoprim, the most commonly used antibiotic for bladder infections, on the infection outcome.