NEW YORK – A team led by investigators in China has used single-cell gene expression data to characterize brain metastases linked to several primary cancer types, defining five distinct "ecotypes."
"This work defines hallmarks of [brain metastases] biology across cancer types and suggests that shared dependencies may exist, which may be exploited clinically," co-senior and co-corresponding author Fan Bai, a researcher with Peking University's Biomedical Pioneering Innovation Center, and colleagues wrote in a paper published in Cancer Cell on Thursday.
For their study, Bai and colleagues combined 10x Genomics single-cell RNA sequencing data and published single-cell RNA-seq data to characterize more than 684,200 individual cells from 111 pan-cancer primary tumor and 108 brain metastasis samples covering melanoma, sarcoma, esophageal squamous cell carcinoma, hepatocellular carcinoma, lung cancer, breast cancer, and colorectal cancer.
Based on cell clusters, cell states, and pathway activity identified, the team was able to untangle the cellular composition of the tumors and their microenvironment, as well as molecular features expected to inform treatment selection and response.
"These data will serve as a valuable resource for advancing the understanding of the overarching features of [brain metastases]," the authors wrote, "and provide critical clinical implications for the treatment of [brain metastases]."
To understand the dynamic changes and remodeling events that take place as primary tumors progress to brain metastases, the team developed a "remodeling index" (RI), which compares the cellular composition between primary tumors and brain metastases.
Future studies should incorporate more features that can shape tumor microenvironment (TME), they wrote, "which will further help us characterize the remodeling of tumor metastasis."
The team's analyses revealed an overall uptick in chromosomal instability in malignant cells from metastatic samples, along with expression signatures associated with processes such as proliferation, angiogenesis, and neural-like cell state features.
On the TME side, the investigators uncovered features backed up by subsequent immunofluorescence and proximal ligation assay experiments, including immunosuppressive tumor-infiltrating myeloid immune cell subclusters coinciding with poorer-than-usual patient outcomes and diminished immunotherapy response.
"Together, myeloid cells demonstrated diversity in the TME and [brain metastases] across various cancer types and subtypes," the researchers reported, "with diverse lineages, transcriptional states, and altered cell compositions transitioning from immune-stimulating to immunosuppressive states as [brain metastases] progressed."
The team also identified five ecological groups, or "ecotypes," marked by a range of tumor and TME features, dubbed the "desert type" (D-type), "lymphoid-enriched type" (L-type), "myeloid-enriched type" (M-type), "stromal-enriched type" (S-type), and "balanced type" (B-type).
Tumor cells were prevalent in the D-type group, while other metastatic tumor ecotypes were marked by enrichment for lymphoid, myeloid, or stromal cells. A more balanced set of immune cells and subtypes characterized the B-type group.
The prevalence of the brain metastasis ecotypes varied depending on factors such as primary cancer type, tumor histology, cancer subtype, biological sex, and treatment history. This pointed to the possibility of bringing in ecotype data to inform the treatment and management of cancer cases involving metastasis to the brain.
"[W]e propose five distinct ecotypes that encapsulate compositional differences across cancer lineages during tumor progression," the authors reported, adding that their broader findings "underscore the importance of drug selection based on different ecotypes for [brain metastasis] therapy."