Background: The aim of this prospective study was to evaluate the feasibility of predicting GC metastasis using CDH1, GFRA1, P16 and ZNF382 DNA methylation as biomarkers.
Methods: 198 GC patients without metastasis at the time of surgery resection were recruited into the double-blind cohort (NCT02159339). Gene methylation was analysed using MethyLight assays. GC metastasis and survival data were obtained from 178 patients with 94.7% compliance during follow-up.
Results: Twenty six cases of metastasis and 5 cases of recurrence were observed in 178 cases (17.4%) during the follow-up (median, 62.7 months). The GC metastasis rate for GFRA1 methylation-positive patients was significantly reduced compared with GFRA1 methylation-negative patients (odds ratio [OR]: 0.23, 95% confidence interval [CI] 0.08-0.66). Similar results were also observed using ZNF382 methylation as a predictor (OR: 0.17, 95% CI 0.06-0.47). A risk score including methylation of GFRA1 and ZNF382 was generated. The metastasis rate was significantly increased in high-risk GC patients (OR: 4.71, 95% CI: 1.85-12.00). GC patients with high risk had a shorter overall survival, especially for patients with stage I GC (P = 0.024).
Conclusions: The combination of GFRA1 and ZNF382 methylation is a biomarker panel for the prediction of GC metastasis.