Abstract:Glioblastoma is a high-grade brain cancer that is marked by considerable tumor heterogeneity and poor outcomes. Molecular profiling has progressed from analyzing bulk samples to identify transcriptomic and DNA methylation subtypes enriched for characteristic genetic lesions, to single-cell studies confirming that these bulk-derived classifiers are an average of intratumoral cellular heterogeneity. It has long been posited that a complete understanding of the mechanisms underlying this heterogeneity will offer new therapeutic opportunities. Madhur et al. used 3D surgical neuronavigation to collect 103 spatially mapped samples from 10 patients that were then profiled for exomes, RNA, 3D genome (Hi-C) and chromatin accessibility (single-nucleus or bulk ATAC-seq). Applying a suite of analytical methods, the authors describe substantial heterogeneity at every molecular level, across patients and spatial locations within each tumor, reinforcing past conclusions. Although there are some interesting, specific findings — such as the association of AP-1 activity with a 'mesenchymal' signature across both normal and tumor cell types — the most notable takeaway is that this kind of exhaustive, multi-modality molecular profiling seems primarily a starting point in the long journey from understanding the biology of cancer to the improvement of patient outcomes. Original reference: Cell 187 , 446–463.e16 (2024)

Spatial and molecular exploration of glioblastoma heterogeneity