Abstract
OBJECTIVE: To elucidate the clinical significance of the methylated status of CpG site count of PCDH10 promoter in the survival prediction in gastric cancer (GC). METHODS: In the previous study, we demonstrated that the methylated CpG site count was significantly associated with the overall survival (OS) of GC patients by using the bisulfite genomic sequencing (BGS) with no less than five clones per sample. It was so complex and expensive for patients to undergo the BGS clones. In this study, we detected the different CpG site counts (hypermethylated and hypomethylated) of PCDH10 DNA promoter in GC samples of 471 patients by directly bisulfite genomic sequencing (D-BGS) without any clone. Furthermore, we evaluated the relationships between the methylated status of PCDH10 promoter and OS. RESULTS: Two hundred and fifty-seven of 471 (54.6%) GC patients were identified to present with PCDH10 promoter methylation by D-BGS. Patients who presented with 5 or more methylated CpG site counts of PCDH10 promoter had significantly poorer prognosis than patients who with less than 5 methylated CpG site counts of PCDH10 promoter (p= 0.039). With the multivariate survival analysis, we demonstrated that T stage, N stage and the hypermethylated CpG site counts of PCDH10 DNA promoter were the independent predictors of OS of GC patients. In addition, the hypermethylated CpG site counts of PCDH10 DNA promoter had smaller Akaike information criterion (AIC) and Bayesian information criterion (BIC) values than the other two independent predictors of the OS, indicating the hypermethylated CpG site counts of PCDH10 DNA promoter as the best prognostic predictor of GC. CONCLUSIONS: Our present findings suggested that the hypermethylated CpG site counts of PCDH10 DNA promoter for evaluating the prognosis of GC was reasonable by using the D-BGS.