Comprehensive prognostic and immune analysis of a glycosylation related risk model in pancreatic cancer

Abstract

Background

Pancreatic cancer (PC) is a malignant tumor with extremely poor prognosis, exhibiting resistance to chemotherapy and immunotherapy. Nowadays, it is ranked as the third leading cause of cancer-related mortality. Glycation is a common epigenetic modification that occurs during the tumor transformation. Many studies have demonstrated a strong correlation between glycation modification and tumor progression. However, the expression status of glycosylation-related genes (GRGs) in PC and their potential roles in PC microenvironment have not been extensively investigated.

Method

We systematically integrated RNA sequencing data and clinicopathological parameters of PC patients from TCGA and GTEx databases. A GRGs risk model based on glycosylation related genes was constructed and validated in 60 patients from Pancreatic biobank via RT-PCR. R packages were used to analyze the relationships between GRGs risk scores and overall survival (OS), tumor microenvironment, immune checkpoint, chemotherapy drug sensitivity and tumor mutational load in PC patients. Panoramic analysis was performed on PC tissues. The function of B3GNT8 in PC was detected via in vitro experiments.

Results

In this study, we found close correlations between GRGs risk model and PC patients’ overall survival and tumor microenvironment. Multifaceted predictions demonstrated the low-risk cohort exhibits superior OS compared to high-risk counterparts. Meanwhile, the low-risk group was characterized by high immune infiltration and may be more sensitive to immunotherapy or chemotherapy. Panoramic analysis was further confirmed a significant relationship between the GRGs risk score and both the distribution of PC tumor cells as well as CD8 + T cell infiltration. In addition, we also identified a unique glycosylation gene B3GNT8, which could suppress PC progression in vitro and in vivo.

Conclusion

We established a GRGs risk model, which could predict prognosis and immune infiltration in PC patients. This risk model may provide a new tool for PC precision treatment.