GBM is the most frequent and malignant subtype of glioma and is associated with dismal prognosis. It exhibits great heterogeneity in terms of biological characteristics and macroscopic growth phenotypes: from nodular tumours with well-formed boundaries, to tumours with varying degrees of healthy-tissue infiltration. GBM growth often causes compression to the surrounding tissue and induces mechanical stress, increasing the intra cranial pressure and impairing neurological function. At the same time, this triggers changes in the micro-environment of the tumour that enhance GBM aggressiveness.
Modelling GBM growth
The GlimS project aimed to study the role of the biomechanical forces generated by GBM growth. For this purpose, scientists developed a framework for estimating tumour characteristics from a single pre-operative MRI data set. The research was undertaken with the support of the Marie Skłodowska-Curie (MSC) programme at the University of Bern in Switzerland in collaboration with the Beckman Research Institute in the United States. “I employed inverse modelling to estimate macroscopic growth characteristics of GBM from routine clinical MR-imaging,” explains Daniel Abler, MSC research fellow. Abler worked under the assumptions that the infiltrative growth of GBM can be described mathematically as a reaction-diffusion process and that an increase in local tumour cell concentration results in volumetric growth which in turn induces mechanical stresses in the tissue. Measuring GBM growth characteristics proved challenging as the patient’s healthy brain anatomy – the situation prior to the tumour invasion – is typically unknown. By the time of diagnosis, the mechanical impact of the tumour known as medicine (mass-effect) may have already distorted brain structures. For this purpose, Abler reconstructed the past evolution of the tumour in the patient’s brain from the images acquired at the time of diagnosis.
Putting the GlimS modelling framework to the test
The scientific team applied the developed inverse modelling approach to pre-operative images of patients diagnosed with GBM to obtain patient-specific estimates of tumour growth parameters. Next, they compared the simulated tumour shape, tumour cell distribution and simulated tissue deformation with pre-operative MRI assessment. According to Abler, “our inverse modelling approach allowed us to discriminate different tumour regions, approximate tumour borders, and qualitatively reproduce tumour-induced healthy-tissue deformation.” Importantly, the GlimS approach offers additional information not available from MRI: By simulating GBM growth, it provides spatially resolved estimates of tumour cell density and tumour-induced mechanical stresses. The GlimS tool enables detailed non-invasive quantification of the macroscopic biomechanical forces acting during tumour growth. From a fundamental perspective, this will advance understanding the clinical consequences posed by these tumour induced forces, and enable further investigation into the role of distinct GBM growth phenotypes for clinical outcome. Future activities of the GlimS team include the correlation of model predicted growth characteristics with clinical data, paving the way towards the identification of mechanically-informed GBM biomarkers. Given the extremely poor prognosis of GBM, these biomarkers may help further stratify patients and make more informed treatment decisions.
GlimS, GBM, MRI, inverse modelling, glioblastoma multiforme, tumour growth, brain cancer, biomechanical forces, mechanical stress, prognostic biomarkers