Elucidating epigenetic landscape of gastric premalignant lesions through genome‐wide mapping of 5‐hydroxymethylcytosines: A 12‐year median follow‐up study

作者：Zhongguang Luo, Wenshuai Li, Wanwei Zheng, Yi‐Xiang Shi, Maolin Ye, Xiangyu Guo, Kaiyi Fu, Changsheng Yan, Bowen Wang, Bin Lv, Shaocong Mo, Hongyang Zhang, Jun Zhang, Chuan He, Feifei Luo, Wei Zhang, Jie Liu · 发表于：Clinical and Translational Medicine · 年份：2024 · DOI：10.1002/ctm2.70114 · 被引用次数：6 · 研究领域：Epigenetics and DNA Methylation、Ferroptosis and cancer prognosis、Genetic Associations and Epidemiology

BACKGROUND: Epigenetic modifications are crucial in tumourigenesis, yet the changes in novel epigenetic regulators like 5-hydroxymethylcytosines (5hmC) during the evolution of gastric premalignant lesions remain poorly understood. This study aims to investigate the implications of 5hmC in the progression from gastric premalignant lesions to gastric adenocarcinoma (GAC). METHODS: To our knowledge, we conducted the largest and longest longitudinal study of a Chinese population with gastric precursor lesions, involving 29,176 patients with gastritis who underwent gastroscopy and biopsy between 2001 and 2015, with follow-up until 1 August, 2022. The median follow-up time was 12.2 years, and the overall GAC incidence rate was 0.82%. Genome-wide mapping of 5hmC in gastric premalignant lesions from a subset of individuals was performed using the 5hmC-Seal assay, including 21 samples that progressed to GAC during follow-up and 48 non-progressed age- and sex-matched controls. RESULTS: We identified 213 differentially modified gene bodies, primarily concentrated in pathways related to cell division, cell cycle, energy metabolism, inflammation and tumourigenesis. An exploratory study was conducted to summarize a 5hmC-based epigenetic model for predicting cancer progression using multivariable logistic regression and machine learning. The nine-gene model showed an area under the curve of 87.5% (95% confidence interval: 72%-100%) in the validation samples (one of three), which were set as...