Resisting cell death is one of the hallmarks of cancer and a major challenge in treating glioblastoma (GBM), one of the most aggressive and treatment-resistant cancers, with a median survival of only 15 months. Current treatment protocols, including surgery, radiotherapy, and temozolomide (TMZ) chemotherapy, often fail due to drug resistance, primarily driven by O-6-methylguanine-DNA methyltransferase (MGMT).

In this talk, we will explore mathematical models elucidating MGMT's role in TMZ resistance in GBM. We will initially show that stochastic gene expression models can explain TMZ-mediated resistance in GBM cells. Specifically, through the incorporation of stochastic elements of cell growth, division, and death, we will demonstrate that even minimal models produce phenotypic selection of resistant cells.

Further expanding on this, we will include the methylation status of the MGMT promoter, an epigenetic marker linked to patient outcomes. Using Approximate Bayesian Computation and relevant biological data, we will reveal different resistance modes: phenotypic selection of MGMT, shifts in promoter methylation status, or both. Our analysis identifies parameter regimes correlating with these modes, providing insights into their unique and shared requirements.

By targeting MGMT expression and regulatory mechanisms, we propose novel strategies to reduce viable GBM cells and improve outcomes. These findings highlight the importance of integrating mathematical modelling with biological data to address cancer resistance and develop effective interventions.