A 6-gene risk score system constructed for predicting the clinical prognosis of pancreatic adenocarcinoma patients

Abstract

Pancreatic adenocarcinoma (PAC) is the most common type of pancreatic cancer, which commonly has an unfavorable prognosis. The present study aimed to develop a novel prognostic prediction strategy for PAC patients. mRNA sequencing data of PAC (the training dataset) were extracted from The Cancer Genome Atlas database, and the validation datasets (GSE62452 and GSE79668) were acquired from the Gene Expression Omnibus database. The differentially expressed genes (DEGs) between good and poor prognosis groups were analyzed by limma package, and then prognosis-associated genes were screened using Cox regression analysis. Subsequently, the risk score system was constructed and confirmed using Kaplan-Meier (KM) survival analysis. After the survival associated-clinical factors were screened using Cox regression analysis, they were performed with stratified analysis. Using DAVID tool, the DEGs correlated with risk scores were conducted with enrichment analysis. The results revealed that there were a total of 242 DEGs between the poor and good prognosis groups. Afterwards, a risk score system was constructed based on 6 prognosis-associated genes ( CXCL11, FSTL4, SEZ6L, SPRR1B, SSTR2 and TINAG), which was confirmed in both the training and validation datasets. Cox regression analysis showed that risk score, targeted molecular therapy, and new tumor (the new tumor event days after the initial treatment according to the TCGA database) were significantly related to clinical prognosis. Under the same clinical condition, 6 clinical factors (age, history of chronic pancreatitis, alcohol consumption, radiation therapy, targeted molecular therapy and new tumor (event days) had significant associations with clinical prognosis. Under the same risk condition, only targeted molecular therapy was significantly correlated with clinical prognosis. In conclusion, the 6-gene risk score system may be a promising strategy for predicting the outcome of PAC patients.