National Institutes of Health CPTAC researchers have uncovered new insights into tumor therapeutic resistance, heterogeneity, and immune evasion. Their work, published in Nature, leveraged several cutting-edge technologies, including spatial transcriptomics, single-nucleus RNA sequencing, and multiplexed imaging, to map the tumor microenvironment in unprecedented detail to understand how cells interact and how their location affects their behavior.
The study identified distinct “tumor microregions” — organized clusters of cancer cells separated by stromal components. These microregions varied significantly in size, depth, composition, and genetic makeup across samples. In particular, metastatic tumors exhibited microregions that were larger, penetrated deeper, and were more densely packed with cancer cells (relative to stromal or immune cells) compared to primary tumors.
By grouping microregions with shared genetic alterations into “spatial subclones,” researchers investigated how copy number variations (CVNs) and mutations might drive regional differences in tumor behavior. For example, a colorectal cancer liver metastasis sample with 12 microregions was found to have two unique spatial subclones, separated by fibrosis. Analysis indicated the two subclones were closely related (sharing clonal genetic events) but had diverged, acquiring significantly different CVNs and mutation frequencies.
The study also highlighted how metabolic and immune activity appear to exhibit spatial specificity. Metabolic activity was found to be the highest at the center of microregions, while the edges displayed increased antigen presentation and higher concentrations of macrophages.
By analyzing serial sections of tumor tissue, the team was able to create a virtual representation of the tumor’s structure in three dimensions. These models provided insights into the structural organization of tumors and, with the help of deep-learning algorithms, allowed the researchers to identify immune “hot” and “cold” neighborhoods in the 3D space surrounding subclones.
This research showcases the limitations of traditional approaches that imagine tumors as being uniform and suggests the possibility of developing treatments to target specific microenvironments within a given tumor.