Abstract C042: Discovery of a core inflammatory gene network associated with poor prognosis and characterized immune infiltration pattern in pancreatic ductal adenocarcinoma

Abstract

Abstract Introduction: Inflammation has a significant impact on the tumor microenvironment (TME) of pancreatic ductal adenocarcinoma (PDAC) and is closely associated with disease prognosis. However, the inflammation features and immune infiltration profiles vary greatly among patients. It is still unclear how inflammation is associated with PDAC prognosis, and contradictory findings have been reported frequently. Therefore, it is essential to explore the core inflammatory genes and the related immune infiltration pattern in PDAC. Method: We analyzed bulk RNA sequencing data of 183 PDAC patients from the Cancer Genome Altas by combining 104 inflammatory gene sets (IGSs, 3427 genes in total) from the Molecular Signatures Database. Firstly, we calculated the inflammatory gene expression matrix using the gene set variation analysis (GSVA). Secondly, consensus clustering was executed to evaluate the stability of clustering for the above matrix. Furthermore, the overall survival (OS) and differential expression analyses were used to screen for core inflammatory genes. Finally, immune infiltration pattern among groups were profiled with CIBERSORT. Results: The 183 PDAC patients were initially clustered into two distinct groups according to the consensus clustering of our formed GSVA inflammatory gene expression matrix. Group 1 (94 patients) exhibited evident inflammation features, while Group 2 (89 patients) displayed limited inflammation features. However, no significant difference in OS was observed (p=0.33). We hypothesized that there might be a core inflammatory gene network (CIGN) which is potently related to poor prognosis in PDAC, rather than the overall inflammation features reflected by the collection of all IGSs. Therefore, Group 1 was further clustered into two subgroups: Group 1A (61 patients, high inflammation features) and Group 1B (33 patients, medium). Interestingly, a strong association with poor prognosis was found for Group 1B (medium) (p=0.045), while patients from Group 1A (high) and Group 2 (limited) presented no significant difference in OS (p=0.98). Finally, we discovered a CIGN consisting of two inflammatory genes, PLAU and DCBLD2, which were upregulated in Group 1B and positively correlated with poor prognosis. This association was also validated with two other PDAC datasets. The built CIGN defined groups showed the greatest association with poor prognosis (p=0.0045). As for the immune infiltration pattern, CIGN group showed a higher accumulation of M0 macrophages and neutrophils compared to non-CIGN group, while CD8+ T cells infiltration in CIGN group was much lowerer, implying a more immune suppressive TME in CIGN patients. Conclusion: Despite the lack of association between overall inflammation features and prognosis in PDAC, a CIGN associated with poor prognosis was discovered in this work with a characterized immune infiltration pattern. The CIGN may serve as diagnosis markers and potential therapeutic targets for PDAC toward precision medicine. Fundings: Hong Kong Theme-based Scheme (T12-201/20-R) Citation Format: Liu Yang, Zheng Chen, Yalan Sheng, Debajyoti Chowdhury, Hiu Fung Yip, Shuangying Qiao, Meiheng Sun, Aiping Lu, Fangfei Li. Discovery of a core inflammatory gene network associated with poor prognosis and characterized immune infiltration pattern in pancreatic ductal adenocarcinoma [abstract]. In: Proceedings of the AACR Special Conference in Cancer Research: Pancreatic Cancer; 2023 Sep 27-30; Boston, Massachusetts. Philadelphia (PA): AACR; Cancer Res 2024;84(2 Suppl):Abstract nr C042.