Aim: To detect significant clusters of co-expressed genes associated with tumorigenesis that might help to predict stomach adenocarcinoma (SA) prognosis.
Methods: The Cancer Genome Atlas database was used to obtain RNA sequences as well as complete clinical data of SA and adjacent normal tissues from patients. Weighted gene co-expression network analysis (WGCNA) was used to investigate the meaningful module along with hub genes. Expression of hub genes was analyzed in 362 paraffin-embedded SA biopsy tissues by immunohistochemical staining. Patients were classified into two groups (according to expression of hub genes): Weak expression and over-expression groups. Correlation of biomarkers with clinicopathological factors indicated patient survival.
Results: Whole genome expression level screening identified 6,231 differentially expressed genes. Twenty-four co-expressed gene modules were identified using WGCNA. Pearson's correlation analysis showed that the tan module was the most relevant to tumor stage (r = 0.24, P = 7 × 10-6). In addition, we detected sorting nexin (SNX)10 as the hub gene of the tan module. SNX10 expression was linked to T category (P = 0.042, χ2 = 8.708), N category (P = 0.000, χ2 = 18.778), TNM stage (P = 0.001, χ2 = 16.744) as well as tumor differentiation (P = 0.000, χ2 = 251.930). Patients with high SNX10 expression tended to have longer disease-free survival (DFS; 44.97 mo vs 33.85 mo, P = 0.000) as well as overall survival (OS; 49.95 vs 40.84 mo, P = 0.000) in univariate analysis. Multivariate analysis showed that dismal prognosis could be precisely predicted clinicopathologically using SNX10 [DFS: P = 0.014, hazard ratio (HR) = 0.698, 95% confidence interval (CI): 0.524-0.930, OS: P = 0.017, HR = 0.704, 95%CI: 0.528-0.940].
Conclusion: This study provides a new technique for screening prognostic biomarkers of SA. Weak expression of SNX10 is linked to poor prognosis, and is a suitable prognostic biomarker of SA.
Keywords: Clinicopathological predictors; Disease-free survival; Overall survival; Sorting nexin 10; Stomach adenocarcinoma; The Cancer Genome Atlas; Weighted gene co-expression network analysis.