Skip to Main content Skip to Navigation

Mécano-biologie avec approche multiphasique de la croissance tumorale : application clinique au glioblastome IDH positif.

Abstract : The basis of this work is the reactive multiphase poromechanical model developed since 2013 by Sciumè et al. A porous system is modelled as a porous solid saturated by two or three fluid phases, interstitial fluid, healthy cells and tumor cells respectively. These fluids have their own constitutive relationships, designed by the same authors. The porous system mimic a living tissue by the addition of a reactive species, denoted nutrient, but solely considered as oxygen. Indeed, the acute deprivation of oxygen - also known as hypoxia - due to the tumor cells proliferation will cause the necrosis of the tumor phase, changing its properties. This model were enhanced in 2014 by the same authors with a deformable porous solid, then allowing for modeling the interplay between extra-cellular matrix (ECM) and tumor cells. The model outputs were qualitatively compared to in vitro experiments of Chignola et al. and histological cuts of skin cancer of Chung et al., showing promising results. Additionally, by replacing the tumor phase by a non-proliferative cell phase under hypoxic condition, the model could qualitatively reproduced the mechanism of diabetic foot. Thus, this model proved its potential in cancer modeling and translation in other tissue pathologies.The motivation of this thesis was the translation of the model of Sciumè et al. into a clinical context.We begun with a wide state of the art of collaboration between mechanics and clinical oncology. The research were driven by two questions: which mechanical phenomena are of interest? How to translate mechanical-based modeling into clinical applications? Regarding the first question, we selected the mechanically-inhibited tumor growth and, as a consequence of stroma mechano-biology, the mechanically-induced phenotype switch. Regarding the second, we chose the image-informed modeling framework, firstly designed in 2002 by Swanson et al., enhanced by Yankeelov et al. since 2013 and thereafter intensively used. During the first year of the thesis, we quantitatively validated the model against the in vitro experiment of Alessandri et al. in, termed as cellular capsule technology. With these first results in our hands, we were honored to start a collaboration with the Toulouse Neuro-imaging center and M.D. Lubrano, neurosurgeon. The aimed clinical application would be the modeling of a non-operable glioblastoma common subtype, the isocitrate dehydrogenase wild-type. At this stage, the SRAS-Cov2 pandemic started to impact the progress of our work. Imaging staff were transferred to pandemic-related tasks, non-essential clinical collaboration were shut down. During this period, our first move were to build a model of healthy brain tissue, using literature and ex-vivo experiments to validate it. To build a code of image-informed modeling, we used public atlases. M.D. Lubrano and his collaborators managed to provide us a patient imaging dataset with two time points, the first at diagnosis, the second after 6 weeks of concomitant radio-chemotherapy. This allowed for testing our hypotheses, now adapted to brain tissue, on hypoxia and cell-ECM interaction, and proposed a calibration of our model. This work was only a first step of the inclusion of poromechanics in patient-specific brain cancer modeling. We hope this inspiring framework will lead to new understanding i n the physical description of cancer.
Document type :
Complete list of metadata
Contributor : ABES STAR :  Contact
Submitted on : Tuesday, May 10, 2022 - 2:34:24 PM
Last modification on : Friday, August 5, 2022 - 2:54:01 PM


Version validated by the jury (STAR)


  • HAL Id : tel-03663836, version 1


Stéphane Urcun. Mécano-biologie avec approche multiphasique de la croissance tumorale : application clinique au glioblastome IDH positif.. Human health and pathology. HESAM Université; Université du Luxembourg, 2022. English. ⟨NNT : 2022HESAE016⟩. ⟨tel-03663836⟩



Record views


Files downloads