« Evaluation non-invasive des Gliomes par Imagerie Résonance Magnétique :
Effets des traitements anti-angiogéniques (Avastin) sur la microvascularisation
et la microarchitecture tumorale et péritumorale »
Place : Mazaré room, Boucherle building, Faculté de Médecine et Pharmacie, 38700 La Tronche
- Boudewijn VAN DER SANDEN, Chargé de Recherche, INSERM, laboratoire TIMC - UMR 5525, Grenoble (Director)
- Teodora-Adriana PERLES-BARBACARU, Maître de conférence, CRMBM - UMR 7339, Marseille (Co-director)
- Dr. Sylvain Miraux, Directeur de recherche, CNRS, CRMSB, Bordeaux (Reporter)
- Dr. Samuel Valable, Directeur de recherche, CNRS, ISTCT, Caen (Reporter)
- Pr. Isabelle Berry, Professeur des Universités - Praticien Hospitalier, CHU de Toulouse (Examiner)
- Pr. Antoine Delon, Professeur des Universités, Université Grenoble Alpes, CNRS, Liphy, Grrenoble (Examiner)
In neuro-oncology, the effects of anti-angiogenic treatments are not predictable, as observed in glioblastomas treated with Avastin. Only a minority of patients show a significant response to treatment. Conventional imaging modalities are not able to evaluate the efficiency in the early phase of the treatment. Thus, the challenge remains to find and to validate new biomarkers that are able to predict the early response to such therapies.
The aim of this work is to develop and implement a preclinical multiparametric magnetic resonance imaging (MRI) protocol for the characterization and follow up of early microvascular and microstructural changes in the tumor and its peritumoral regions after treatment with Avastin. For this purpose, the quantification of the blood volume and Kmodel (an apparent coefficient that is related to the contrast agent (CA) uptake rate), and evaluation of brain microarchitecture by diffusion tensor imaging were developed and evaluated as biomarkers.
The Rapid Steady State T1 (RSST1) method was initially developed for blood volume quantification in the absence of CA extravasation. In the first part of this thesis, we have implemented and adapted this MRI technique for the quantification of both blood volume and Kmodel in tumors where the CA extravasates. We developed a mathematical model for the RSST1 signals that accounts for the unidirectional bi-compartmental exchange of CA from the vascular towards the extravascular compartment. This development allows to the quantification of vascular parameters in a rat glioma model (C6). The results were confirmed using another MRI modality, the steady state magnetic susceptibility method, and quantitative histology.
In the second part, we studied the sensitivity of the RSST1 method for the follow up of the glioma response to anti-angiogenic treatment under clinical conditions. In this study, the effect of Avastin treatment in a murine orthotopic U87 MG glioma model was analyzed. The RSST1 method demonstrated a high reproducibility in the blood volume quantification and a superior sensitivity in comparison to CA enhanced T1-weighted imaging (T1W-Gd-DOTA) for the detection and follow-up of the tumor response to Avastin, especially in early stages of tumor progression. Blood volume quantification by MRI was correlated to measures obtained by two-photon microscopy.
In the last part of this thesis, we have studied the capacity of diffusion tensor imaging (DTI) coupled with FLAIR (fluid-attenuated inversion recovery) MRI and T1w-Gd-DOTA, to characterize tumor, peritumoral, and contralateral regions of the U87MG glioma model. We quantified DTI parameters before and during the invasion of tumor cells induced by Avastin in the peritumoral zone for different administration modes: intravenous and intratumoral via Convention-Enhanced Delivery. Therefore, the delineation of peritumoral regions for each tumor in an early stage was based on anatomical images, that took into account the individual tumor progression at later stages. Significant differences were detected for DTI parameters between the tumor, peritumoral, and contralateral regions and a different evolution of these parameters was noticed according to the Avastin injection mode.