Figure 1.

The tumor immune microenvironment in patients with lung squamous cell carcinoma. (A) A representative formalin-fixed paraffin-embedded section (4×) was subjected to hematoxylin and eosin staining to identify the tumor region (the loop with green lines) and the adjacent region. Locations within the tumor region (orange circle) and the adjacent region (red circle) were ablated for imaging mass cytometry analysis. PhenoGraph analysis (B) and t-distributed stochastic neighbor embedding plots (D) were used to divide the cell populations into 30 clusters in the tumor regions according to phenotypic similarity. PhenoGraph analysis (C) and t-distributed stochastic neighbor embedding plots (E) were used to divide the cell populations into 32 clusters in the adjacent regions according to phenotypic similarity. Neighborhood analysis revealed cell-to-cell interactions between different clusters in the tumor regions (F) and the adjacent regions (G), respectively.