Skip to Main content Skip to Navigation

Modélisation multi-facteurs pour l’aide à la décision dans le traitement par chimiothérapie des tumeurs cérébrales de type gliome diffus de bas grade

Abstract : Diffuse Low-Grade Glioma (DLGG) is defined by the WHO as a primary tumour of the central nervous system and represents 15% of all glial tumours combined. A DLGG grows slowly, and inevitably evolve into a much more aggressive (grade III) glioma, which eventually leads to the death of the patient. Three types of treatment are available: surgery, chemotherapy and radiotherapy. Today, the median survival rates reported in studies varies from 10 to 15 years. Unfortunately, the prognosis for DLGG is highly variable, with a high standard deviation of total survival, and some patients are surviving only a few years. Within the framework of DLGG management at Nancy University Hospital, chemotherapy is one of the most widely used treatments and there are very variable responses in terms of intensity, duration and response profiles. The thesis work is located in this context. It concerns the study of the response to chemotherapy and consists in developing decision-making tools for the neuro-oncologist in the follow-up of patients. The first part of this thesis work therefore focuses on the choice of the volumetric method. The volume response curve can then be characterised in terms of response intensity. The second part of this work concerns response modelling using statistical learning techniques. Many explanatory variables (epidemiological, genetic) are under study. A new variable called ESVR, which is an original measure allowing us to quantify the infiltrating DLGG phenotype, will also be used. The factorial analysis and machine learning methods initially make possible to define the variables that provide the most information. Exploratory analyses of the data reveal a redundancy of information among certain genetic and epidemiological factors. The models show a greater influence of quantitative variables on the response to chemotherapy compared to qualitative variables. A discussion is finally produced on the importance of the variables used in the prediction of the response to chemotherapy. The aim of this thesis is to produce a set of rules which will enable clinicians to anticipate, before administering the treatment, its effect on tumour volume, which will allow a more advised choice of therapeutic strategy than possible nowadays.
Complete list of metadata
Contributor : Thèses UL Connect in order to contact the contributor
Submitted on : Wednesday, June 30, 2021 - 2:52:43 PM
Last modification on : Saturday, October 16, 2021 - 11:14:17 AM
Long-term archiving on: : Friday, October 1, 2021 - 6:37:04 PM


Files produced by the author(s)


  • HAL Id : tel-03274856, version 1


Cyril Brzenczek. Modélisation multi-facteurs pour l’aide à la décision dans le traitement par chimiothérapie des tumeurs cérébrales de type gliome diffus de bas grade. Cancer. Université de Lorraine, 2021. Français. ⟨NNT : 2021LORR0095⟩. ⟨tel-03274856⟩



Record views


Files downloads