Researchers at Uppsala University developed a new method combining artificial intelligence and live brain tissue imaging to track how individual glioblastoma cells spread.
Glioblastoma stands out as one of the most aggressive brain cancers. Tumor cells routinely migrate past the surgical margin into healthy tissue, triggering recurrence after treatment.
To capture this movement, scientists published a study in Nature Communications detailing a tool called GlioTrace. The system monitors thousands of individual cancer cells in real time.
“Until now, we have mostly seen snapshots of how glioblastoma spreads,” said Hitesh Mangukiya, a researcher at the Department of Immunology, Genetics and Pathology at Uppsala University and a lead author, News-Medical reported.
The research team cultured glioblastoma cells from six patients inside mouse brains. They sliced the brain tissue thin and kept the samples alive in a laboratory setting.
Automated microscopes photographed the slices every two hours for up to five days. A machine-learning system tracked individual tumor cells across the entire brain slice.
The analysis revealed distinct movement patterns between tumors. Some cells migrated primarily along blood vessels, while others dispersed diffusely through the brain tissue.
Individual cells also altered their physical shape and speed over time. They switched dynamically between vessel-bound travel and independent movement.
“What surprised us most was how flexible the tumor cells are,” said Madeleine Skeppås, a PhD student and co-lead author at Uppsala University. “The same cell can change how it moves over just a few days.”
GlioTrace also tested how specific pharmaceutical substances altered tumor cell movement. Two tested drug substances successfully slowed the migration of cancer cells within the brain slices.
However, treatment responses varied widely between patient-derived tumors and different migrating cell types. Radiation and chemotherapy combined to increase the proportion of a specific, fast-moving cell shape in certain tumor models.
These fast cells occupied central positions in the transition network between cell states. Further research must determine whether these populations survive treatment to drive tumor recurrence.
“Using this method allows us to monitor, for the first time, how thousands of individual tumor cells move through living brain tissue and to see how they adapt their behaviour in real time,” said Sven Nelander, Professor of Integrative Cancer Research at Uppsala University.
Skeppås developed the computational system alongside colleagues at Chalmers University of Technology and the University of Gothenburg. The software processes microscopy images, tracks cell morphologies, and calculates transition timelines.
The analysis software is available freely to other researchers. Researchers have similarly mapped glioblastoma using the GBM-Space atlas.
“We hope that the method will help us understand how tumor cells interact with their surroundings and identify treatments that can prevent cancer from spreading through the brain,” Mangukiya said.
Tumor cells invade surrounding healthy brain tissue far beyond the visible tumor boundary. Surgeons cannot remove this extended region without damaging vital brain functions.
Researchers slice mouse brains containing patient-derived tumors and keep the tissue alive. Automated microscopes capture images every two hours while artificial intelligence tracks individual cell trajectories.
No, Mangukiya said that GlioTrace remains strictly a research tool. Further studies are required before the software can guide direct patient treatment decisions.
“For now, GlioTrace is a research method, and further studies are needed before it could be used to guide treatment decisions in patients,” Mangukiya said.