空间解析基因组学揭示胶质母细胞瘤的演化模式及影像学相关性

The spatial evolutionary dynamics of glioblastoma remain poorly understood. Here, we performed spatially resolved molecular profiling of 78 multi-regional tumor specimens from 24 GBM patients by integrating neuro-navigation-guided intraoperative sampling with deep whole-exome sequencing. This dataset represents one of the most spatially detailed genomic resources for GBM to date. By coupling spatial coordinates with genomic complexity, we delineated two distinct evolutionary trajectories: an “Expansive” model, in which three-dimensional growth and molecular diversification proceed in parallel, and a “Stochastic” model, where genomic diversification occurs independently of spatial expansion. These models were supported by phylogenetic reconstruction and radiogenomic analyses, revealing how spatial architecture constrains clonal dynamics. Quantitative integration of molecular and physical distances uncovered that tumors with greater spatially correlated genomic diversity exhibited worse clinical outcomes. Patients harboring “Stochastic” tumors demonstrated inferior survival probabilities compared to those with “Expansive” tumors, suggesting that spatially derived molecular metrics may serve as prognostic indicators of tumor aggressiveness. Furthermore, MRI-derived radiomic features, particularly texture-based metrics from T1-contrast enhanced images, mirrored underlying genomic complexity and aligned with evolutionary modes, establishing a link between intratumoral heterogeneity and noninvasive imaging phenotypes. While prior genomic studies have characterized GBM evolution at the molecular level, most lacked spatial resolution and failed to capture the three-dimensional architecture of tumor growth within the human brain. Our study overcomes this limitation through an international collaboration between the Medical University of Innsbruck and Korea University College of Medicine. Together, these results define anatomically distinct evolutionary trajectories of GBM and underscore how spatial context shapes molecular diversity, clinical behavior, and imaging manifestations. This spatially integrated framework provides a foundation for precision oncology approaches that incorporate spatial evolutionary constraints into therapeutic stratification.