Structural connectome quantifies tumor invasion and predicts survival in glioblastoma patients

Yiran Wei, Chao Li (Lead / Corresponding author), Zaixu Cui, Roxanne C. Mayrand, Jingjing Zou, Adrianna L.K.C. Wong, Rohitashwa Sinha, Tomasz Matys, Carola-Bibiane Schönlieb, Stephen John Price

Research output: Working paper/PreprintPreprint

2 Downloads (Pure)


Glioblastoma widely affects brain structure and function, and remodels neural connectivity. Characterizing the neural connectivity in glioblastoma may provide a tool to understand tumor invasion. Here, using a structural connectome approach based on diffusion MRI, we quantify the global and regional connectome disruptions in individual glioblastoma patients and investigate the prognostic value of connectome disruptions and topological properties. We show that the disruptions in the normal-appearing brain beyond the lesion could mediate the topological alteration of the connectome (P <0.001), associated with worse patient performance (P <0.001), cognitive function (P <0.001), and survival (overall survival: HR: 1.46, P = 0.049; progression-free survival: HR: 1.49, P = 0.019). Further, the preserved connectome in the normal-appearing brain demonstrates evidence of remodeling, where increased connectivity is associated with better overall survival (log-rank P = 0.005). Our approach reveals the glioblastoma invasion invisible on conventional MRI, promising to benefit patient stratification and precise treatment.
Original languageEnglish
Number of pages32
Publication statusPublished - 3 Jul 2021


Dive into the research topics of 'Structural connectome quantifies tumor invasion and predicts survival in glioblastoma patients'. Together they form a unique fingerprint.

Cite this