Glycosylation (Glyc) is prevalently related to gastric cancer (GC) pathophysiology. However, studies on the relationship between glycosylation regulators and tumor microenvironment (TME) and immunotherapy of GC remain scarce. We extracted expression data of 1,956 patients with GC from eight cohorts and systematically characterized the glycosylation patterns of six marker genes into phenotype clusters using the unsupervised clustering method. Next, we constructed a Glyc. score to quantify the glycosylation index of each patient with GC. Finally, we analyzed the relationship between Glyc. score and clinical traits including molecular subtype, TME, and immunotherapy of GC. On the basis of prognostic glycosylation-related differentially expressed genes, we constructed the Glyc. score and divided the samples into the high- and low-Glyc. score groups. The high-Glyc. score group showed a poor prognosis and was validated in multiple cohorts. Functional enrichment analysis revealed that the high-Glyc. score group was enriched in metabolism-related pathways. Furthermore, the high-Glyc. score group was associated with the infiltration of immune cells. Importantly, the established Glyc. score would contribute to predicting the response to anti-PD-1/L1 immunotherapy. In conclusion, the Glyc. score is a potentially useful tool to predict the prognosis of GC. Comprehensive analysis of glycosylation may provide novel insights into the epigenetics of GC and improve treatment strategies.
Keywords: gastric cancer; glycosylation; immunotherapy; prognosis; scoring system.
Copyright © 2022 Li, Huang, Pan, Xiao, Zheng and Zhong.