Diagnostic and Prognostic Value of B4GALT1 Hypermethylation and Its Clinical Significance as a Novel Circulating Cell-Free DNA Biomarker in Colorectal Cancer.
Cancers (Basel)
; 11(10)2019 Oct 19.
Article
en En
| MEDLINE
| ID: mdl-31635093
ABSTRACT
Epigenetic modifications of glyco-genes have been documented in different types of cancer and are tightly linked to proliferation, invasiveness, metastasis, and drug resistance. This study aims to investigate the diagnostic, prognostic, and therapy-response predictive value of the glyco-gene B4GALT1 in colorectal cancer (CRC) patients. A Kaplan-Meier analysis was conducted in 1418 CRC patients (GEO and TCGA datasets) to assess the prognostic and therapy-response predictive values of the aberrant expression and methylation status of B4GALT1. Quantitative methylation-specific PCR (QMSP) and droplet digital quantitative methylation-specific PCR (dd-QMSP) were respectively used to detect hypermethylated B4GALT1 in metastasis and plasma in four cohorts of metastatic CRC cases (mCRC). Both the downregulated expression and promoter hypermethylation of B4GALT1 have a negative prognostic impact on CRC. Interestingly a low expression level of B4GALT1 was significantly associated with poor cetuximab response (progression-free survival (PFS) p = 0.01) particularly in wild-type (WT)-KRAS patients (p = 0.03). B4GALT1 promoter was aberrantly methylated in liver and lung metastases. The detection of hypermethylated B4GALT1 in plasma of mCRC patients showed a highly discriminative receiver operating characteristic (ROC) curve profile (area under curve (AUC) value 0.750; 95% CI 0.592-0.908, p = 0.008), clearly distinguishing mCRC patients from healthy controls. Based on an optimal cut-off value defined by the ROC analysis, B4GALT1 yield a 100% specificity and a 50% sensitivity. These data support the potential value of B4GALT1 as an additional novel biomarker for the prediction of cetuximab response, and as a specific and sensitive diagnostic circulating biomarker that can be detected in CRC.
Texto completo:
1
Banco de datos:
MEDLINE
Tipo de estudio:
Diagnostic_studies
/
Prognostic_studies
Idioma:
En
Revista:
Cancers (Basel)
Año:
2019
Tipo del documento:
Article
País de afiliación:
Italia