Charting the invasion routes of glioblastoma and their patient-specific vulnerabilities

The brain cancer glioblastoma (GBM) is characterized by rapid progression and a complete lack of effective therapeutic options. Unlike many fatal forms of cancer, glioblastoma causes death not by distant metastasis but by rapid local invasion.

This invasion, in turn, occurs via multiple routes, including infiltration of white matter tracts, and penetration of the perivascular or perineural spaces.

Here, we will test the hypothesis that the invasion routes of GBM depend on distinct cellular programs that can be systematically identified and effectively targeted.

To address this, we will use a newly established bank of 100 patient-derived xenograft (PDX) models, which can reliably model each invasive route.

First, we will characterize invasive populations by spatial transcriptomics of xenografts and evaluate new protein markers of invasion in Swedish and UK patients.

Second, we will identify key regulators of each invasive route, by a new strategy based on data mining, CRISPR-based gene ablation and single-cell sequencing in vivo.

Third, top candidates will be studied in mouse models, to test the hypothesis that specific modes of invasion can be directly targeted. Last, we will set up a computational resource for the assessment of invasion biomarkers and targeted therapies.

The project will deepen our understanding of the pathways that drive GBM invasion and their therapeutic vulnerabilities. Our innovative strategy may have further uses in aggressive forms of invasive cancer.