阶段二结直肠癌的主要挑战是鉴定具有复发风险增加的患者。目前还缺乏能够区分预后不良患者与无复发患者的生物标志物。本研究旨在开发一种强大的DNA甲基化分类器，以预测阶段二结直肠癌患者的复发和化疗效果。我们对243个阶段二结直肠癌样本进行了全基因组DNA甲基化捕获测序，并确定了由八个CpG位点组成的特异的复发相关DNA甲基化标志。本研究纳入了243名阶段二结直肠癌患者。为了在复发和非复发阶段二结直肠癌样本中选择差异甲基化位点，我们使用Agilent SureSelectXT Human Methyl-Seq进行了DNA甲基化分析，这是一种全面的靶向富集系统，可用于分析CpG甲基化。在243名患者中，我们采用荧光定量测序来量化候选DNA甲基化位点的甲基化水平。我们采用最小绝对收缩和选择算子（LASSO）方法来建立疾病复发预测分类器。我们通过DNA甲基化捕获测序确定了由八个CpG位点组成的复发相关DNA甲基化标志。该分类器的预测准确性显著高于任何临床病理风险因素。Kaplan-Meier生存曲线显示高风险评分与不良预后相关。在多元分析中，该标志是最重要的预后因素，具有HR为2.80（95％CI，1.71-4.58，P<0.001）。该标志可以鉴定适合接受辅助化疗的患者。八个CpG DNA甲基化标志是阶段二结直肠癌患者复发的可靠预测工具。©2023年。作者（专有许可人：Springer Nature Limited）。

A major challenge in stage II colorectal carcinoma is to identify patients with increased risk of recurrence. Biomarkers that distinguish patients with poor prognosis from patients without recurrence are currently lacking. This study aims to develop a robust DNA methylation classifier that allows the prediction of recurrence and chemotherapy benefit in patients with stage II colorectal cancer. We performed a genome-wide DNA methylation capture sequencing in 243 stage II colorectal carcinoma samples and identified a relapse-specific DNA methylation signature consisting of eight CpG sites.Two hundred and forty-three patients with stage II CRC were enrolled in this study. In order to select differential methylation sites among recurrence and non-recurrence stage II CRC samples, DNA methylation profiles of 62 tumour samples including 31 recurrence and 31 nonrecurrence samples were analysed using the Agilent SureSelectXT Human Methyl-Seq, a comprehensive target enrichment system to analyse CpG methylation. Pyrosequencing was applied to quantify the methylation level of candidate DNA methylation sites in 243 patients. Least absolute shrinkage and selection operator (LASSO) method was employed to build the disease recurrence prediction classifier.We identified a relapse-related DNA methylation signature consisting of eight CpG sites in stage II CRC by DNA methylation capture sequencing. The classifier showed significantly higher prognostic accuracy than any clinicopathological risk factors. The Kaplan-Meier survival curve showed an association of high-risk score with poor prognosis. In multivariate analysis, the signature was the most significant prognosis factor, with an HR of 2.80 (95% CI, 1.71-4.58, P < 0.001). The signature could identify patients who are suitable candidates for adjuvant chemotherapy.An eight-CpG DNA methylation signature is a reliable prognostic and predictive tool for disease recurrence in patients with stage II CRC.© 2023. The Author(s), under exclusive licence to Springer Nature Limited.