JMIR Publications

ABSTRACT

Background:

Breast Cancer (BRCA) is the leading cause of death in females worldwide. Although progress in mammography-based screening, the diagnosis of BRCA remains a challenge. However, the sensitivity decreases with high breast density of mammography and a high level of heterogeneity with different prognoses are necessary to improve prognosis in BRCA patients. The valuable molecular targets and therapeutic biomarkers improve the prognoses of BRCA patients, leading to a lower incidence of recurrences.

Objective:

This report is conductive to provide new ideas for the clinical potential diagnosis and treatment of BRCA.

Methods:

The transcriptome and methylation genes were downloaded from The Cancer Genome Atlas (TCGA) database. Methyl Mix algorithm was performed to obtain methylation-driven genes. Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analysis were applied for identification of methylation-driven genes with functional enrichment. After screening clinical data, a risk score model was established based on univariate and multivariate cox regression analyses. The risk evaluation model was performed using the Kaplan-Meier (K-M) methods, Receiver Operating Curves (ROC) and Area Under Curve (AUS). Overall survival analysis of methylation-driven genes was used survival R package. The analysis of gene methylation and gene expression was conducted to explore the relationships in BRCA.

Results:

A total of 1213 samples from TCGA database. 153 differentially expressed methylation-driven genes were obtained. 6 methylation-driven genes (GYPC, KCNH8, USP44, ZNF502, ZNF677, and ZSCAN23) were significant negative correlations between methylation and gene expression levels. Functional analysis suggested that methylation-driven genes enriched in mammary gland development and prolactin signaling pathway. The expression of 9 methylation genes (ADHFE1, KNSTRN, BUB1B, ABCB1, QKI, NDST4, CKMT1B, GALNT8, and GLT1D1) was differently profiled from two groups related to clinical information. The ROC (AUC=0.703) indicated that risk score (P＜0.001) showed accuracy for predicting models. The univariate and multivariate cox regression analysis demonstrated age (P＜0.001) was an independent prognostic factor for patients with BRCA. Methylation and gene expression analysis were identified that ALDH1L1-AS2-1, PTPRZ1-1, and UBE2T-1(P＜0.05) independently predicting prognosis in BRCA patients.

Conclusions:

Our study suggested that ALDH1L1-AS2-1, PTPRZ1-1, and UBE2T-1 might as predictors of great clinical potential biomarkers in BRCA. These results will provide new ideas for the clinical treatment of BRCA patients.