Immune Cell Profiling Reveals a Common Pattern in Premetastatic Niche Formation Across Various Cancer Types

Background: Metastasis is the major cause of cancer-related mortality. The premetastatic niche is a promising target for its prevention. However, the generality and cellular dynamics in premetastatic niche formation have remained unclear.

Aims: This study aimed to elucidate the generality and cellular dynamics in premetastatic niche formation.

Materials and methods: We performed comprehensive flow cytometric analysis of lung and peripheral immune cells at three time points (early premetastatic, late premetastatic, and micrometastatic phases) for mice with subcutaneous implants of three types of cancer cells (breast cancer, lung cancer, or melanoma cells). The immuno-cell profiles were then used to predict the metastatic phase by machine learning.

Results: We found a common pattern of changes in both lung and peripheral immune cell profiles across the three cancer types, including a decrease in the proportion of eosinophils in the early premetastatic phase, an increase in that of regulatory T cells in the late premetastatic phase, and an increase in that of polymorphonuclear myeloid-derived suppressor cells and a decrease in that of B cells in the micrometastatic phase. Machine learning using immune cell profiles could predict the metastatic phase with approximately 75% accuracy.

Discussion: Validation of our findings in humans will require data on the presence or absence of micrometastases in patients and the accumulation of comprehensive and temporal information on immune cells. In addition, blood proteins, extracellular vesicles, DNA, RNA, or metabolites may be useful for more accurate prediction.

Conclusion: The discovery of generalities in premetastatic niche formation allow prediction of metastatic phase and provide a basis for the development of methods for early detection and prevention of cancer metastasis in a cancer type-independent manner.