Differential expression and bioinformatics analysis of exosome circRNAs in pancreatic ductal adenocarcinoma

Background: Pancreatic ductal adenocarcinoma (PDAC) is a fatal malignant tumor with an unfavorable prognosis. Increasing evidence indicated circRNAs were associated with the pathogenesis and progression of tumors, but data on the expression of serum exosomal circRNAs in PDAC are scarce. This study attempted to explore the prognostic value and function of serum exosomes in PDAC patients. Methods: Microarray-based circRNA expression was determined in PDAC and paired with normal serum samples, and the intersection of differentially expressed circRNAs (DECs) in serum exosomal samples and GSE79634 tissue samples was conducted. A specific CircRNA database was applied to investigate DECs binding miRNAs. Target genes were predicted using the R package multiMiR. Cox regression analyses were applied for constructing a prognostic model. The immunological characteristics analysis was carried out through the TIMER, QUANTISEQ, XCELL, EPIC, and ssGSEA algorithms. Results: 15 DECs were finally identified, and a circRNA-miRNA-mRNA network was established. A prognostic risk model was developed to categorize patients according to the risk scores. Furthermore, the association between risk score and immune checkpoint genes including CD80, TNFSF9, CD276, CD274, LGALS9, and CD44 were significantly elevated in the high-risk group, while ICOSLG and ADORA2A were upregulated in the low-risk group. Conclusions: Our results may provide new clues for the prognosis and treatment of PDAC.