A glycosylation signature for predicting the progression and immunotherapeutic response of prostate cancer.

ABSTRACT

BACKGROUND: Glycosylation has been proposed as a new cancer hallmark. However, focusing on specific glycans or glycoproteins may lose much data relevant to glycosylation alterations. The present study aimed to first comprehensively investigate the expression and mutation profiles of glycosylation-related genes (GRgenes) in prostate cancer (PCa) and then develop a glycosylation signature and explore its role in predicting the progression and immunotherapeutic response of PCa. METHODS: Based on The Cancer Genome Atlas database, we comprehensively screened potential prognostic GRgenes and analyzed their expression and mutation profiles in PCa. Through consensus clustering analysis, the study cohort was classified to investigate the effect of glycosylation patterns on the prognosis of PCa. Next, we developed a glycosylation signature (i.e., the glycosylation score [Gly_score]) using the differentially expressed genes between glycosylation pattern groups and evaluated its role in predicting the progression and immunotherapeutic response of PCa. RESULTS: We identified two distinct glycosylation patterns in PCa and found that GRgene expression patterns rather than mutations are associated with the prognosis of PCa. The high Gly_score group had significantly shorter progression-free survival, lower PD-L1 levels, less infiltration of immune cells and lower immunophenoscores than the low Gly_score group. When the patients were grouped according to both the Gly_score and PD-L1 level, patients with a combination of low Gly_score and low PD-L1 expression had the best survival outcomes. CONCLUSIONS: In the present study, for the first time, we developed a glycosylation signature and demonstrated that the proposed glycosylation signature is a promising tool for predicting the prognosis and immunotherapeutic response of PCa.