Breast cancer is not one disease, and two women with the same subtype can still have very different outcomes. A study in PLOS One, published June 29, asks whether the genes that hold cells together can explain some of that gap.
The team, led by Ishii H and colleagues, focused on cell-cell junctions, the structures that keep the lining of tissue intact. Adhesion molecules have long been tied to how breast cancer progresses, but nobody had systematically checked what junction-related gene activity means inside each intrinsic subtype. So the researchers took 179 genes listed under the Gene Ontology term “cell-cell junction organization” and tested them subtype by subtype.
The data
They drew on two large public datasets, METABRIC and The Cancer Genome Atlas. METABRIC holds 2,509 samples, and 1,479 of them had the survival, gene expression and subtype information the analysis needed. Using Cox proportional hazards models for disease-specific survival, the team picked out the junction genes that predicted outcome in each subtype and combined them into expression signatures.
They then checked the signatures against an independent group, the SCAN-B cohort, where 3,052 samples qualified.
What they found
Higher signature scores went with poorer survival across subtypes. The effect was strongest in Luminal B and Basal-like tumors. In those aggressive subtypes, high scores also predicted more recurrences in the first five years after diagnosis.
Tumors with high scores showed more activity in pathways tied to extracellular matrix organization and the matrisome. Single-cell RNA sequencing added detail on where the signal comes from. In Luminal B, genes such as PARD6B and CDH3 were expressed mainly in tumor epithelial cells. The Basal signature drew on both epithelial cells (MARVELD2, for example) and endothelial cells (RAMP2).
The authors say the signatures sort patients by prognosis and reflect subtype-specific features of the tumor and its surroundings.
Where it fits
Existing multigene tests such as Oncotype DX, MammaPrint and EndoPredict have improved risk assessment in hormone receptor-positive disease. They use fixed gene panels and were built mainly for early-stage luminal cancer treated with endocrine therapy such as tamoxifen, which leaves some of the variation inside subtypes unexplained. The new signatures are meant to complement them, not replace them.
Nothing in the paper describes a clinical test. The work is an analysis of existing datasets.
Funding came from several Japanese sources, among them the Tokyo University of Science, JST Moonshot R&D and Grants-in-Aid for Scientific Research. The authors declared no competing interests.
Sources
- Cell–cell junction gene signatures as subtype-specific prognostic biomarkers in breast cancer
- Cell-cell junction gene signatures as subtype-specific prognostic biomarkers in breast cancer.
