Modelling the Heterogeneity of Glioblastoma

Glioblastoma multiforme (GBM) is a cancer of the brain (specifically of the glial cells), known for its potential for aggressive growth and diffuse invasion of the surrounding normal-appearing tissue in the brain. Usually, a patient with GBM will undergo surgical resection (surgery to remove as much of the tumour as possible), with complementary radiotherapy and/or chemotherapy, but this rarely acts as a cure. This is partly due to the extent to which the cancer is able to grow and invade the brain before a patient notes any symptoms. Combined with the diffuse nature of the tumour making treatment difficult, prognosis for GBM patients is poor, with an average survival time of 12-18 months and only 3% of patients living beyond 2 years after diagnosis.

After resection of GBM, recurrence of the tumour is commonly observed, typically around six months later. The ability to identify areas most at risk of this recurrence could be crucial in improving the treatment of GBMs. Different regions of a GBM tumour may be more aggressive, invasive or respond less well to radiotherapy or chemotherapy, and doctors are currently building datasets that would allow this brain cancer heterogeneity to be predicted from patient MRI scans. The key objective of this project is to develop novel mathematical models and simulation tools to forecast GBM growth based upon patient-derived data on brain cancer heterogeneity, which we hope could be used to predict which are the more "dangerous" regions of a patient's tumour, and to improve patient outcomes. These new models would use partial differential equations to describe the rate of growth and spread of cancer subpopulations, with the model parameters estimated from patients' MRI scans.

Another important phenomenon in brain cancer is the progression of lower-grade gliomas, which can maintain a relatively benign status for many years, into the higher-grade and more aggressive GBM. The ability to understand and predict this progression could be crucial. Using the datasets that characterise heterogeneity in low-grade glioma, and comparing with GBM data, this project hopes to use mathematical models to identify low-grade glioma features that may predict progression to GBM.