Colorectal cancer is one of the most common gastrointestinal cancers and the second leading cause of cancer-related death. Although colonoscopy screening has greatly improved the early diagnosis of colorectal cancer, its recurrence and metastasis are still significant problems. Tumour cells usually have the hallmark of metabolic reprogramming, while fatty acids play important roles in energy storage, cell membrane synthesis, and signal transduction. Many pathways of fatty acid metabolism (FAM) are involved in the occurrence and development of colon cancer, and the complex molecular interaction network contains a variety of genes encoding key enzymes and related products.
Clinical information and RNA sequencing data were collected from TCGA and GEO databases. The prognosis model of colon cancer was constructed by LASSO-Cox regression analysis among the selected fatty acid metabolism genes with differential expression. Nomogram for the prognosis model was also constructed in order to analyze its value in evaluating the survival and clinical stage of the colon cancer patients. The differential expression of the selected genes was verified by qPCR and immunohistochemistry. GSEA and GSVA were used to analyze the enrichment pathways for high- and low-risk groups. CIBERSORT was used to analyze the immune microenvironment of colon cancer and to compare the infiltration of immune cells in the high- and low-risk groups. The “circlize” package was used to explore the correlation between the risk score signature and immunotherapy for colon cancer.
We analysed the differential expression of 704 FAM-related genes between colon tumour and normal tissue and screened 10 genes with prognostic value. Subsequently, we constructed a prognostic model for colon cancer based on eight optimal FAM genes through LASSO Cox regression analysis in the TCGA-COAD dataset, and its practicality was validated in the GSE39582 dataset. Moreover, the risk score calculated based on the prognostic model was validated as an independent prognostic factor for colon cancer patients. We further constructed a nomogram composed of the risk score signature, age and American Joint Committee on Cancer (AJCC) stage for clinical application. The colon cancer cohort was divided into high- and low-risk groups according to the optimal cut-off value, and different enrichment pathways and immune microenvironments were depicted in the groups.
Since the risk score signature was significantly correlated with the expression of immune checkpoint molecules, the prognostic model might be able to predict the immunotherapy response of colon cancer patients. In summary, our findings expand the prognostic value of FAM-related genes in colon cancer and provide evidence for their application in guiding immunotherapy.