Background:Human chemokines are linked to tumor survival and perform a critical role in the pancreatic adenocarcinoma (PAAD) microenvironment. It is not known if abnormal expression of chemokines contributes to the prognostic features of PAAD. Therefore, this study explored gene alteration profiles of chemokines in PAAD and constructed a chemokine family-based prognosis signature (CPS).Methods:RNA-Seq and clinical data from 176 PAAD patients were obtained from The Cancer Genome Atlas (TCGA) database. Bioinformatics analysis tools were used to assess the expression and prognostic value of chemokines. A profile affecting patient survival was obtained using unsupervised hierarchical clustering and a prognostic model of chemokine-related genes was constructed to explore the immune landscape.Findings:Twenty-nine chemokines were highly expressed in PAAD tissues, and 10 were significantly associated with poor prognosis. Hierarchical clustering resulted in the classification of the PAAD cohort into 2 clusters. High levels of chemokines indicated a worse prognosis (hazard ratio [HR] 5.4, 95% confidence interval [CI] 1.7-17; P = .004). A CPS containing 7 hub genes was constructed using the least absolute shrinkage and selection operator (LASSO) and stepwise multivariable Cox analysis. This signature separated PAAD patients into high- and low-risk groups and was confirmed as an independent prognostic factor. An ESTIMATE score and TIMER2.0 analysis further reflected the immune profile of PAAD, which had a higher percentage of localized immune cell infiltration with predominantly suppressive cells.Interpretation:This is the first chemokine family-based model for predicting outcomes in PAAD patients. Our work highlights the need to understand the mechanism of chemokine contributions to PAAD occurrence.