“Prognostic value of 18F-FDG PET (CT and/or MRI) radiomic features combined with omics data in early-stage poor-prognosis breast cancer treated with neoadjuvant chemotherapy: A deep learning analysis.”
Objectives:
The project aims to identify early predictors of pCR in triple-negative (TN) and HER2+ breast cancer to establish a foundation for personalized therapy before evaluating response quality. Large and heterogeneous data, i.e., a multi-omic approach, will be extracted from tumor biopsies before any treatment. The utility of multi-omics, including radiomics, in precision oncology has been demonstrated in metastatic breast cancer. We will extend this approach to early-stage breast cancer in our NEO-EPICURE study. We will refine patient classification through transcriptional characterization of critical sub-populations for breast cancer subtypes and infer their presence through deconvolution approaches. Additionally, the acquisition of PET/MRI equipment by our consortium provides an opportunity to evaluate this new tool within multi-omic approaches. PET/MRI may enhance tumor characterization, especially with new MRI sequences. Imaging, which allows non-invasive, reproducible, and whole-body evaluation, has the potential to provide promising image-derived biomarkers in breast cancer. Soluble markers also help predict risk and can provide fine longitudinal biological information documenting patient response to treatment non-invasively.
You can find all the information in the presentation document.
Contacts:
PhD supervisors: Nicolas Normand, nicolas.normand@polytech.univ-nantes.fr; Caroline Rousseau, caroline.rousseau@ico.unicancer.fr
Co-supervisor: Ludovic Ferrer, ludovic.ferrer@ico.unicancer.fr
