Integrating Tertiary Lymphoid Structure-Associated Genes into Computational Models to Evaluate Prognostication and Immune Infiltration in Pancreatic Cancer

Abstract Pancreatic ductal adenocarcinoma (PDAC) is characterized by poor response to all therapeutic modalities and dismal prognosis. The presence of tertiary lymphoid structures (TLSs) in various solid cancers is of crucial prognostic significance, highlighting the intricate interplay between the tumor microenvironment and immune cells aggregation. However, the extent to which TLS and immune status affect PDAC prognosis remains incompletely understood. Here, we sought to unveil the unique properties of TLS in PDAC by leveraging both single-cell and bulk transcriptomics, and culminating in a risk model that predicts clinical outcomes. We used TLS score based on 12 genes (CCL2, CCL3, CCL4, CCL5, CCL8, CCL18, CCL19, CCL21, CXCL9, CXCL10, CXCL11 and CXCL13) and 9 genes (PTGDS, RBP5, EIF1AY, CETP, SKAP1, LAT, CCR6, CD1D and CD79B) signature, respectively, and examined their distribution in cell clusters of single-cell data from PDAC samples. The markers involved in these clusters were selected to develop a prognostic model using The Cancer Genome Atlas Program (TCGA) database as the training cohort and Gene Expression Omnibus (GEO) database as the validation cohort. Further we compared the immune infiltration, drug sensitivity, enriched and differentially expressed genes between the high-risk and low-risk groups in our model. Therefore, we established a risk model that has significant implications for the prognostic assessment of PADC patients with remarkable differences in immune infiltration and chemo-sensitivity between the low-risk and high-risk groups. And this paradigm established by TLS-related cell marker genes provides a prognostic prediction and a panel of novel therapeutic targets for exploring potential immunotherapy.