Predicting prognosis through the discovery of specific biomarkers according to colorectal cancer lymph node metastasis.

ABSTRACT

Colorectal cancer (CRC) is a prevalent cancer worldwide, ranking as the third most common cancer and the second leading cause of cancer-related deaths. The presence or absence of lymph node metastases is one of the representative markers for predicting CRC prognosis, but often yields heterogeneous results. In this study, we conducted an integrative molecular analysis of CRC using publicly available data from The Cancer Genome Atlas database and NCBI's Gene Expression Omnibus. Through our analysis, we identified 372 upregulated genes that were differentially expressed in CRC patients. Additionally, Kyoto Encyclopedia of Genes and Genomes analysis revealed five significant pathways, including Hippo, FC-gamma, and forkhead box O signaling pathways, which are known to be associated with cancer. Survival analysis of 28 genes involved in these pathways led to the identification of 13 genes with prognostic significance (P < 0.05). To validate our findings, logistic regression models were generated and tested in multiple cohorts, demonstrating significant accuracy. Moreover, we identified six genes (BNIP3, CD63, RDX, RGCC, WASF1, and WASF3) whose combination predicted the best prognosis based on survival analysis. This predictive model holds promise as a potential biomarker for prognosis, survival, and treatment efficacy. In conclusion, our study provides valuable insights into the molecular characteristics of CRC and identifies prognostic biomarkers. The combination of differentially expressed genes and their involvement in cancer-related pathways enhances our understanding of CRC pathogenesis and opens avenues for personalized treatment approaches and improved patient outcomes.