Bold claim: Cancer’s first brain arrival isn’t a one-way trip to doom — it’s a tug-of-war where the brain’s microglia can either stop the invasion or, in some cases, let it slip past. And this is exactly what new real-time imaging reveals about how brain metastasis unfolds.
Metastasis happens when cancer cells break away from their original tumor, ride the bloodstream, and seed new tumors elsewhere. It’s the leading cause of cancer-related deaths, and brain metastases are especially serious, affecting roughly 10–30% of people with advanced lung, breast, or melanoma cancers. Treatments exist for established brain tumors, but preventing the initial seeding of cancer cells in the brain has proven far more challenging — until now.
Our brains host microglia, a frontline group of immune cells that can rapidly engulf and digest invading cells. Yet scientists had struggled to explain why microglia sometimes fail to destroy incoming cancer seed cells, mainly because researchers hadn’t been able to watch this crucial interaction in the living brain in real time.
Now, for the first time in this scenario, researchers captured the precise moment when microglia engulf seed cells attempting to establish a foothold in the brain. The study, published in Cancer Research, identifies two proteins that cancer cells use to dodge destruction by microglia during their initial brain arrival. When scientists removed these proteins, microglia became more effective at eliminating the invading cells, and brain tumor formation decreased. The human relevance of these findings is underscored by the discovery that CD24 and CD47 were present in about half of brain metastasis samples from lung cancer patients.
A global team led by Dr. Takahiro Tsuji at Nagoya University tracked seed cells traveling from the lung to the brain to understand their behavior at the critical entry moment. The window during which microglia can act lasts about 12 days; by the time symptoms appear or imaging detects trouble, tumors may already be established. Using two-photon microscopy to visualize these interactions in live mice, the researchers observed a spectrum: some microglia actively destroyed cancer cells, while others appeared to aid cancer survival and growth. To probe this difference, they applied an innovative “opto-omics” approach, which selectively labeled the microglia in contact with tumor cells with light, enabling isolated analysis of those specific cells.
What allows brain cancer cells to slip past microglia? The culprit signals are don’t-eat-me molecules CD24 and CD47. When researchers removed these signals, microglia digested the seed cells, and brain tumors were notably reduced. This suggests a potential clinical angle: therapies that block CD24 and CD47 could disable the cancer’s early escape mechanism and empower microglia to clear seeds before tumors take hold.
As senior author Professor Hiroaki Wake notes, some microglia receive signals that reprogram them to support tumor growth by forming blood vessels and creating a protective niche that shields cancer from immune attack. Understanding why microglia polarize toward destruction in some cases and support in others at the very first encounter is pivotal for designing better prevention strategies.
The proposed path forward involves developing treatments that remove CD24 and CD47 with antibodies or other methods to eradicate cancer seed cells during the early window when microglia can phagocytose them. The goal is not just to treat established brain tumors but to stop metastasis at its earliest stage by leveraging the brain’s own immune system.
Looking ahead, the researchers plan to refine therapies that target CD24 and CD47, potentially in combination with existing treatments. They’re also working on biomarkers to identify patients most likely to benefit from early intervention and expanding their imaging toolkit — combining two-photon microscopy with opto-omics — to study metastasis in other cancers and organs.
Reference: Tsuji T, Hirose H, Sugiyama D, et al. Microglia display heterogeneous initial responses to disseminated tumor cells. Cancer Res. 2025. doi:10.1158/0008-5472.CAN-25-3425
