Back

Cerebral vasculature shapes the spatial patterning of brain metastases

Farahani, A.; Liu, Z.-Q.; Bazinet, V.; Couturier, C. P.; Porter, E.; Capaldi, D. P. I.; Morin, O.; Dagher, A.; Misic, B.

2026-07-22 neuroscience
10.64898/2026.07.22.739598 bioRxiv
Show abstract

Primary cancer cells that originate in diverse tissues in the body can spread to the brain through various physiological pathways and form metastatic tumors. The spatial patterning of brain metastases is highly stereotyped across individuals, but the physiological factors that confer regional vulnerability to tumor colonization are poorly understood. Here we map the brain metastases associated with primary breast cancer, lung cancer, and malignant melanoma to the multiscale organization of the brain. We analyze structural magnetic resonance imaging (MRI) data from more than 2, 300 cancer patients with approximately 10, 000 brain metastases to estimate metastatic frequency maps. We then examine whether transcriptional (microarray profiling), functional (functional MRI), and vascular (arterial spin labeling) features can predict the spatial pattern of each brain metastasis type. Our analyses highlight vascular anatomy as an integral determinant of metastatic patterning, particularly in breast and lung cancers. We find arterial border-zones as sites of elevated vulnerability to metastatic invasion. These areas, characterized by slow flow and small-caliber vessels, create a hemodynamic environment that favors the arrest and extravasation of circulating tumor cells. Together, these results provide quantitative support for the long-standing hypothesis that vascular architecture and its biomechanical properties constrain metastatic seeding in the brain. Identifying vascular topology as a key determinant of metastatic patterning suggests that systemic vascular and metabolic factors may contribute to metastatic risk.

Matching journals

The top 3 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.