[en] We propose the modeling of glioblastoma isocitrate dehydrogenase wild-type (GBMwt) build on
the following hypotheses: the brain tissue is a porous medium, the coupling of hypoxia
consequences and mechanical interplay between extra-cellular matrix and tumor cells is the driver
of the malignant evolution of the disease. In this thesis, a poromechanical model is developed with
the aim of a clinical application in oncology. A review, with a large scope, is done on mechanical
applications in clinical management of cancer. The model is first validated on in vitro experimental
data of encapsulated multi-cellular spheroids. Then, a clinical collaboration is initiated with the
Neuro-imaging center of Toulouse, and the targeted clinical application is the modeling of non-
operable GBMwt. To this end, the model is first adapted to the specificity of brain tissue
mechanics. Characteristic features of the disease are modeled: necrotic core, modified extra-cellular
matrix production, emerging malignant phenotype and invasion. Clinical imaging data are pre-
treated to inform the model in a patient specific basis. A proposition of modeling is provided with
an evaluation against clinical data.
