[Interpreting the Tumor Microenvironment via Spatial Transcriptomics - Pathology Imaging and Cell-Cell Interactions]

In recent years, spatial transcriptomics has matured as a technology that simultaneously acquires positional information on tissue sections and gene expression, enabling a re-interpretation of histomorphology (histopathology) in molecular terms. In practice, it is useful to (ⅰ) group regions by clustering, (ⅱ) overlay representative markers onto section coordinates, (ⅲ) align results with the pathology image, and (ⅳ) assess reproducibility across samples and individuals. In addition, by performing ligand-receptor-based cell-cell interaction analyses and observing, along the morphological sequence of normal →dysplasia→tumor, the emergence, increase, decrease, or loss of specific interaction axes, one can connect a spatial snapshot to information that reflects temporal change. Methodologically, sequence-based platforms, suited to surveying the global landscape, and in situ hybridization-based platforms, suited to precise reading of boundaries and neighborhood relations, may be selected and combined according to purpose; with continued advances, this integrated approach may support patient stratification and treatment selection. Taken together, viewing the tumor microenvironment through both spatial and temporal perspectives using spatial transcriptomics provides a practical framework to understand the TME as a dynamically evolving environment under continuous construction.