The integrated landscape of fatty acid metabolism subtypes reveals with prognostic and therapeutic relevance in pancreatic cancer

Abstract

Background Pancreatic Cancer (PAAD) is one of the most commonly diagnosed malignancies and the leading cause of cancer-related death worldwide. Aberrantly expressed long noncoding RNAs (lncRNAs) are involved in tumourigenesis of PAAD, and associated with the overall survival and tumor fatty acid metabolism in PAAD patients. Methods The data on gene expression and corresponding clinical characteristics of PAAD patients in TCGA-PAAD (N=177) and GSE62452 (N=65) are taken from The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases. Consensus cluster analysis to identify distinct fatty acid metabolism subtypes in PAAD based on 62 fatty acid metabolism gene. The single sample GSEA (ssGSEA) algorithm was developed for evaluation of tumor infiltrating immune cells between fatty acid metabolism subtypes. As well, the R package “pRRophetic” was used to predict chemotherapeutic response in PAAD patients. Tumor Immune Dysfunction and Exclusion (TIDE) was used to predict immunotherapy response in PAAD patients. Univariate and multivariate Cox analysis were utilized to calculate the prognostic-related lncRNAs. Results Totally, three fatty acid metabolism subtypes were obtained in PAAD based on 62 fatty acid metabolism gene. Kaplan-Meier (K-M) analysis showed that the overall survival rate of cluster3 group was significantly higher than the other two groups. Significant differences were seen between the three subtypes in immune cell infiltration characteristics and the immunotherapy response indicators, including Tumor mutational burden (TMB), immunophenoscore (IPS), and immune checkpoint molecules. The cluster1 group and cluster3 group were speculated to have the higher response to immunotherapy patients in cluster2 gains more benefit from chemotherapy than other groups. A 4-lncRNA signature was constructed based on the value of gene expression and regression coefficients which stratified patients into two risk groups. Patients in the higher-risk group had lower survival probabilities than those in the lower-risk group, based on the Kaplan-Meier analysis and Cox regression analysis. Receiver operating characteristic (ROC) curve analysis confirmed the predictive capability. In GO and KEGG analysis, genes in the high-risk group were linked to PAAD development. Conclusions We constructed a signature that could predict prognosis of PAAD and provide certain theory guidance for novel therapeutic approaches of PAAD.