Purpose:
The molecular properties of TLSs in pancreatic cancer are still not well comprehended.
This research delved into the molecular properties of intratumoral TLSs in pancreatic
cancer through the exploration of multi-omics data.
Methods:
Seven key genes were identified through Cox regression analysis and random survival
forest analysis from a total of 5908 genes related to TLSs. These genes were utilized to construct a
prognosis model, which was subsequently validated in two independent cohorts. Additionally, the
study investigated the molecular features of different populations of TLSs from multiple perspectives.
The model’ s forecasting accuracy was verified by analyzing nomogram and decision curves,
taking into account the patients’ clinical traits.
Results:
The analysis of immune cell infiltration showed a notably greater presence of Macrophage
M0 cells in the group at high risk than in the low-risk group. The pathway enrichment analysis
demonstrated the activation among common cancer-related pathways, including ECM receptor interaction,
pathways in cancer, and focal adhesion, in the high-risk group. Additionally, the methylation
study revealed notable disparities in DNA methylation between two TLS groups across four
regions: TSS200, 5’ UTR, 1stExon, and Body. A variety of notably distinct sites were linked with
PVT1. Furthermore, by constructing a competing endogenous RNA network, several mRNAs and
lncRNAs were identified that compete for the binding of hsa-mir-221.
Conclusion:
Overall, this research sheds light on the molecular properties of TLSs across various
pancreatic cancer stages and suggests possible focal points for the treatment of pancreatic cancer.