Transformative insights from transcriptome analysis of colorectal cancer patient tissues: identification of four key prognostic genes

Colorectal cancer (CRC) is a leading cause of cancer-related deaths worldwide, necessitating accurate and robust predictive approaches to assist oncologists with prognosis prediction and therapeutic decision-making in clinical practice. Here, we aimed to identify key genes involved in colorectal cancer pathology and develop a model for prognosis prediction and guide therapeutic decisions in CRC patients. We profiled 49 matched tumour and normal formalin-fixed paraffin-embedded (FFPE) samples using Affymetrix HGU133-X3P arrays and identified 845 differentially expressed genes (FDR ≤ 0.001, fold change ≥2), predominantly enriched in the extracellular matrix (ECM)-receptor interaction pathway. The integrative analysis of our data with publicly available mRNA and miRNA datasets, including their differentially expressed gene analyses, identified four overexpressed genes in the ECM-receptor interaction pathway as key regulators of human CRC development and progression. These four genes were independently validated for their differential expression and association with prognosis in a newly collected CRC cohort and publicly available datasets. A prognostic risk score was developed using these genes, with patient stages weighted by multivariate Cox regression coefficients to stratify patients into low-risk and high-risk groups, showing significantly poorer overall survival (OS) in the high-risk group. In conclusion, our risk assessment model exhibits strong potential for predicting poor survival and unfavorable clinicopathological features in CRC patients, offering valuable insights for personalised management strategies.