Molecular characterization and clinical relevance of metabolic signature subtypes in gastric cancer.

Abstract

e16040 Background: Metabolic reprogramming dictates the tumor molecular attributes and therapeutic potentials. However, the comprehensive metabolic characteristics in gastric cancer (GC) remain obscure. Methods: We comprehensively assessed the metabolic process alterations and identified robust metabolite transcriptional patterns based on 456 metabolite genes-related signatures. Three distinct metabolic patterns were identified by unsupervised consensus clustering and validated across independent datasets, including the SDPH in-house dataset. Results: Here, the metabolic signature-based clustering analysis identifies three subtypes with distinct molecular and clinical features: MSC1 showed better prognosis and upregulation of TCA cycle and lipid metabolism, combined with frequent TP53 and ROHA mutation; MSC2 had moderate prognosis and elevated nucleotide and amino acid metabolism, enriched by intestinal histology and dMMR; MSC3 exhibited poor prognosis and enhanced glycan and energy metabolism, accompanied with with diffuse histology and frequent CDH1 mutation. The molecular features and clinicopathologic characteristics of these three metabolite clusters have distinct metabolic gene expression, pathway enrichment, genomic variation, and survival outcomes. The SDPH in-house dataset with matched transcriptomic, metabolomic, and spatial-metabolomic also validated these findings. We further construct the metabolic subtype-related prognosis genes (MSPG) scoring model to quantify the activity of individual tumor and found a positive correlation with cuproptosis signaling. Conclusions: In conclusion, comprehensive recognizing of the metabolite signature can enhance the understanding of diversity and heterogeneity in GC.