DataSheet2.docx

Abstract

Background: Pancreatic ductal adenocarcinoma (PDAC) is one of the most aggressive tumors with an extremely low 5-year survival rate. Accumulating evidence has unveiled that inflammatory response promotes tumor progression, enhances angiogenesis, and causes local immunosuppression. Herein, we aim to develop an inflammatory related prognostic signature, and found it could be used to predict gemcitabine response in PDAC. Methods: PDAC cohorts with mRNA expression profiles and clinical information were systematically collected from the four public databases. An inflammatory response related genes (IRRGs) prognostic signature was constructed by LASSO regression analysis. Kaplan–Meier survival analysis, receiver operating characteristic analysis, principal component analysis, univariate and multivariate Cox analyses were carried out to evaluate effectiveness and reliability of the signature. The correlation between gemcitabine response and risk score was evaluated in the TCGA-PAAD cohort. The GDSC database, pRRophetic algorithm and Connectivity map analysis were used to predict gemcitabine sensitivity and identify potential drugs for the treatment of PDAC. Finally, we analyzed differences in frequencies of gene mutations, infiltration of immune cells, as well as biological functions between different subgroups divided by the prognostic signature. Results: We established a seven IRRGs (ADM, DCBLD2, EREG, ITGA5, MIF, TREM1 and BTG2) signature which divided the PDAC patients into low- and high-risk groups. The prognostic value of the signature was validated in 11 PDAC cohorts consisted of 1,337 PDAC patients from six countries. A nomogram that integrated the IRRGs signature and clinicopathologic factors of PDAC patients were constructed. The risk score showed positive correlation with gemcitabine resistance. Two drugs (BMS-536924 and dasatinib) might have potential therapeutic implications in high-risk PDAC patients. We found that high-risk group had higher frequencies of KRAS, TP53 and CDKN2A mutations, increased infiltration of macrophages M0, neutrophils and macrophages M2 cells, as well as upregulated hypoxia and glycolysis pathways, while low-risk group had increased infiltration of CD8+ T, naïve B, plasma and macrophages M1 cells. Conclusions: We constructed and validated an IRRGs signature that could be used to predict the prognosis and gemcitabine response of patients with PDAC, as well as two drugs (BMS-536924 and dasatinib) may contribute to PDAC treatment.