Spatial metabolomics identifies distinct tumor-specific subtypes in gastric cancer patients.

PURPOSE: Current systems of gastric cancer (GC) molecular classification include genomic, molecular, and morphological features. GC classification based on tissue metabolomics remains lacking. This study aimed to define metabolically distinct GC subtypes and identify their clinicopathological and molecular characteristics. EXPERIMENTAL DESIGN: Spatial metabolomics by high mass resolution imaging mass spectrometry was performed in 362 GC patients. K-means clustering was used to define tumor and stroma-related subtypes based on tissue metabolites. The identified subtypes were linked with clinicopathological characteristics, molecular features, and metabolic signatures. Responses to trastuzumab treatment were investigated across the subtypes by introducing an independent patient cohort with HER2-positive GC from a multicenter observational study. RESULTS: Three tumor- and three stroma-specific subtypes with distinct tissue metabolite patterns were identified. Tumor-specific subtype T1(HER2+MIB+CD3+) positively correlated with HER2, MIB1, DEFA-1, CD3, CD8, FOXP3, but negatively correlated with MMR. Tumor-specific subtype T2(HER2-MIB-CD3-) negatively correlated with HER2, MIB1, CD3, FOXP3, but positively correlated with MMR. Tumor-specific subtype T3(pEGFR+) positively correlated with pEGFR. Patients with tumor subtype T1(HER2+MIB+CD3+) had elevated nucleotide levels, enhanced DNA metabolism, and a better prognosis than T2(HER2-MIB-CD3-) and T3(pEGFR+). An independent validation cohort confirmed that the T1 subtype benefited from trastuzumab therapy. Stroma-specific subtypes had no association with clinicopathological characteristics, however linked to distinct metabolic pathways and molecular features. CONCLUSIONS: Patient subtypes derived by tissue-based spatial metabolomics are a valuable addition to existing GC molecular classification systems. Metabolic differences between the subtypes and their associations with molecular features could provide a valuable tool to aid in selecting specific treatment approaches