Integrated Analysis Identifies Molecular Signatures and Specific Prognostic Factors for Different Gastric Cancer Subtypes.

摘要: Abstract BACKGROUND: Gastric cancer (GC) is the fifth leading cause of cancer-related deaths worldwide. As an effective and easily performed method, microscopy-based Lauren classification has been widely accepted by gastrointestinal surgeons pathologists for GC subtyping, but molecular characteristics different subtypes were poorly revealed. METHODS: GSE62254 was used as a derivation cohort, GSE15459 validation cohort. The difference between diffuse intestinal on gene expression level measured. Gene ontology (GO) enrichment analysis both subgroups. Hierarchical clustering heatmap exhibition also performed. Kaplan-Meier plot Cox proportional hazards model to evaluate survival grouped given genes or hierarchical clusters. RESULTS: A total 4598 found differentially expressed GC. Immunity- cell adhesion–related GOs enriched GC, whereas DNA repair– cycle–related We proposed 40-gene signature ( χ 2 =30.71, P =12.11, =.002). FRZB [RR (95% CI)=1.824 (1.115-2.986), =.017] EFEMP1 CI)=1.537 (0.969-2.437), =.067] identified independent prognostic factors only in not patients. KRT23 CI)=1.616 (0.938-2.785), =.083] factor patients Similar results achieved CONCLUSION: that GCs with classifications had FRZB, EFEMP1, subtype-specific