Long Guo, Zhao Lihua, Tang Biao, Zhou Ledu, Mi Xingyu, Su Wenxin, Xiao Liang
Department of General Surgery, Xiangya Hospital, Central South University, Changsha, China.
Department of Translational Medicine, Genecast Biotechnology Co., Ltd, Wuxi City, Jiangsu, China.
Heliyon. 2023 Aug 25;9(9):e19434. doi: 10.1016/j.heliyon.2023.e19434. eCollection 2023 Sep.
Altered gene methylation precedes altered gene expression and the onset of disease. This study aimed to develop a potential model for predicting recurrence of early to mid-stage hepatocellular carcinoma (HCC) using methylation loci.
We used data from early to mid-stage HCC patients (TNM I-II) in the TCGA-LIHC dataset and lasso-cox regression model to identify an 18-DNA methylation site panel from which to calculate the riskScore of patients. The correlation of high/low riskScore with recurrence-free survival (RFS) and immune microenvironment in HCC patients was analyzed by bioinformatics. It was also validated in the GSE56588 dataset and the final dynamic nomogram was constructed.
The results showed that riskScore was significantly correlated with RFS in HCC patients. The differential mutated genes between the two groups of HCC patients with high/low riskScore were mainly enriched in the TP53 signaling pathway. The immune microenvironment was better in HCC patients in the low-riskScore group compared to the high-riskScore group. This was validated in the GSE56588 dataset. Based on the subgroup stratification analysis of the relationship between high/low riskScore and RFS, as well as univariate and multivariate cox analyses, the riskScore was found to be independent of clinical indicators. We found that riskScore, vascular invasion and cirrhosis status could effectively differentiate RFS in HCC patients, and we also constructed prediction model based on these three factors. The model we constructed were validated in the TCGA-LIHC database and a web calculator was built for clinical use.
The methylation riskScore is a predictor of RFS independent of clinical factors and can be used as a marker to predict recurrence in HCC patients.
基因甲基化改变先于基因表达改变和疾病发生。本研究旨在开发一种利用甲基化位点预测早期至中期肝细胞癌(HCC)复发的潜在模型。
我们使用了TCGA-LIHC数据集中早期至中期HCC患者(TNM I-II)的数据以及套索-考克斯回归模型,以识别一个由18个DNA甲基化位点组成的面板,据此计算患者的风险评分。通过生物信息学分析高/低风险评分与HCC患者无复发生存期(RFS)和免疫微环境的相关性。该模型在GSE56588数据集中进行了验证,并构建了最终的动态列线图。
结果显示,风险评分与HCC患者的RFS显著相关。高/低风险评分的两组HCC患者之间的差异突变基因主要富集在TP53信号通路中。与高风险评分组相比,低风险评分组HCC患者的免疫微环境更好。这在GSE56588数据集中得到了验证。基于高/低风险评分与RFS关系的亚组分层分析以及单因素和多因素考克斯分析,发现风险评分独立于临床指标。我们发现风险评分、血管侵犯和肝硬化状态可以有效区分HCC患者的RFS,并且我们还基于这三个因素构建了预测模型。我们构建的模型在TCGA-LIHC数据库中进行了验证,并构建了一个网络计算器供临床使用。
甲基化风险评分是独立于临床因素的RFS预测指标,可作为预测HCC患者复发的标志物。