Development of a streamlined NGS-based TCGA classification scheme for gastric cancer and its implications for personalized therapy.

Background: The Cancer Genome Atlas (TCGA) has identified four distinct molecular subtypes of gastric cancer (GC) with prognostic significance: Epstein-Barr virus (EBV)-positive, microsatellite instability (MSI)-high, genomically stable (GS), and chromosomal instability (CIN). Unfortunately, the complex analysis required for TCGA classification limits its practical use in clinical settings. Our study sought to devise a next-generation sequencing (NGS)-based method to classify GC more efficiently, serving as a promising biomarker for prognosis and immunotherapy efficacy. Methods: This study was a retrospective observation study, and we employed 2 independent GC cohorts. The 3DMed cohort (n=765), comprising data on 733 cancer-related genes along with 4 EBV-encoded genes, was utilized to develop the new NGS classification. Additionally, the secondary Korean cohort (n=55), which includes both genomic data and information on immune checkpoint inhibitor (ICI) treatment, was employed to establish a correlation between NGS subtypes and ICI responsiveness. Results: In the 3DMed cohort, we identified 5.2% EBV, 4.6% MSI, 30.6% GS, and 59.6% CIN subtypes. The MSI subtype exhibited the highest number of mutation events, along with the highest tumor mutational burden (TMB) and strong programmed cell death ligand 1 (PD-L1) expression. CIN tumors showed extensive copy number variations (CNVs) and genomic heterogeneity. The EBV subtype presented recurrent Conclusions: The NGS method successfully maps the mutational landscape of GC, providing a practical TCGA classification surrogate to optimize patient-specific treatment strategies.