AI-powered imaging tracks thousands of brain cancer cells moving through living tissue

Researchers at Uppsala University have created a novel technique that employs live brain tissue imaging in conjunction with artificial intelligence to observe individual glioblastoma cells as they propagate through the brain. This groundbreaking method, known as GlioTrace, provides fresh insights into glioblastoma, an exceptionally aggressive type of brain cancer.

By tracking the movements of thousands of tumor cells, GlioTrace can reveal how these cells alter their behavior, transition between various modes of movement, and respond to cancer therapies. The study, published in Nature Communications, utilized GlioTrace to analyze glioblastoma cells sourced from six patients. The researchers cultivated these cells as tumors in mouse brains, which were subsequently sliced into thin sections and maintained alive in a laboratory setting.

Through automated microscopy, the researchers captured images of the slices at two-hour intervals for up to five days. A machine-learning algorithm identified and followed individual tumor cells, with statistical models employed to determine how cells shifted between different states. The findings uncovered significant disparities between tumors, with some cells predominantly migrating along blood vessels while others spread more diffusely through the brain tissue.

Furthermore, individual tumor cells exhibited the ability to change their shape and movement over time, even switching between migrating along blood vessels and independently traversing the tissue. Perhaps most notably, the study highlighted the remarkable adaptability of tumor cells, with a single cell capable of altering its movement patterns within just a few days.

To better comprehend and potentially thwart the spread of tumor cells, the researchers sought to understand this dynamic ability. They also investigated the impact of various drugs on tumor cell movement. Two drug substances were found to diminish both the movement and spread of tumor cells in the brain slices; however, the efficacy of these drugs varied between patient-derived tumors and among different types of migrating cells.

Additionally, the team discovered that a combination of radiation and chemotherapy in certain tumor models resulted in an increased proportion of cells with a specific shape, which were among the fastest-moving cells and played a central role in the network of cell state transitions. These findings suggest that such invasive cells may persist despite treatment, although further research is necessary to ascertain whether these cells are more resistant to treatment or contribute to tumor recurrence.

By enabling the real-time monitoring of thousands of individual tumor cells moving through living brain tissue, GlioTrace offers unprecedented insights into the invasive behavior of glioblastoma cells. The method also sheds light on how tumor cells interact with their surroundings and could potentially pave the way for identifying treatments that prevent cancer from spreading through the brain.

However, as GlioTrace is currently a research method, additional studies are required before it can be utilized to inform patient treatment decisions.