RNA Predicts Colorectal Cancer Prognosis

Spatial transcriptomics applies color coding to digital histology images to create a map of active and inactive genes on RNA molecules on the tumor landscape. Combined with AI, the technique is increasingly being harnessed for cancer research.

Researchers in New Zealand have been looking to RNA molecules to predict the likelihood of bowel cancer returning later in life. By examining long non-coding RNA (lncRNA) expression within distinct tumor regions, researchers uncovered three lncRNAs that show potential as biomarkers for assessing the risk of metastatic disease, particularly in early-stage colorectal cancer (CRC) patients.

In the study, the researchers analyzed CRC tissue samples from various stages, including stage II and stage IV tumors, using spatial transcriptomics to map lncRNA expression at near single-cell resolution. This method enabled a detailed investigation of malignant tissue, distinguishing it from adjacent normal tissue, and identified 301 lncRNAs prevalent in cancerous regions. Following validation with public datasets, three specific lncRNAs emerged as key markers due to their elevated expression in metastatic tissues, versus non-metastatic, and their detectability in stage II tumors.

The three lncRNAs were validated with RNA-FISH, revealing that they are minimally expressed in normal colon tissue but significantly upregulated in malignant and metastatic regions. Analysis of these markers in the TCGA-COAD cohort further demonstrated their association with patient survival, with high expression levels correlating with poorer outcomes. These findings suggest that incorporating these lncRNAs into routine clinical assessments could help identify high-risk CRC patients, potentially guiding more targeted treatments and monitoring.

Overall, the study suggests that spatial transcriptomics is a powerful tool for identifying prognostic biomarkers in CRC, laying the groundwork for further research into these lncRNAs as part of early-stage CRC diagnosis and risk assessment.