DECIPHERING MECHANISMS OF RESISTANCE TO TARGETED THERAPY IN PAEDIATRIC GLIOMA

Objective

Paediatric gliomas are the most common malignant brain tumours in children, and despite advances in molecular classification and targeted therapies, resistance to treatment remains a major clinical challenge. In BRAFV600E-mutant gliomas, dual BRAF/MEK inhibition has improved outcomes, but more than half of patients relapse within two years. Increasing evidence suggests that resistance is often driven not by genetic mutations, but by reversible, non-genetic mechanisms such as epigenetic reprogramming. However, these processes remain poorly understood in paediatric glioma. This project addresses this gap by integrating single-cell transcriptomic and epigenomic data with proteomic data in a multimodal, time-resolved framework. Patient-derived explants and xenografts treated with targeted therapy are being profiled at multiple time points using cutting-edge assays that jointly capture gene expression and histone modifications, alongside direct measurements of kinase activity. We will develop and implement computational tools for multimodal data integration, including optimal transport models to track population dynamics over time and causal network inference to connect kinase signalling, transcription factor activity, and downstream gene regulation. This design will allow us to capture the mechanisms by which persister and resistant cells emerge and survive with unprecedented resolution. The project builds on an existing collaboration between groups at Institut Curie (Paris), the University Children’s Hospital Zurich, and the University of California, San Francisco. By integrating data across models, we aim to create the first single-cell multi-omic atlas of paediatric glioma under treatment. The expected outcome is the identification of actionable kinases and pathways driving relapse, informing preclinical testing and potentially guiding early-phase clinical trials. The approach can be extended to other BRAFV600E tumours or cancers with similar resistance mechanisms.