利用DNA甲基化谱预测银屑病患者发生银屑病关节炎
本研究对银屑病患者全基因组DNA甲基化进行评估,比较后续发生关节炎者(转化者)与未发生者(生物制剂初治、按年龄、性别、银屑病病程和随访时长匹配)的甲基化差异,鉴定出36个高相关甲基化标记,基于这些标记的分类模型识别转化者与非转化者的受试者工作特征曲线下面积为0.9644。
要点
- 在15个基因及若干基因间区鉴定出36个高相关甲基化标记(错误发现率校正P<0.05,甲基化变化最小0.05)。
- 基于这些标记的支持向量机分类模型识别转化者与非转化者的受试者工作特征曲线下面积为0.9644。
- 银屑病疾病早期阶段的DNA甲基化模式可在同一随访期内区分未来发生与不发生银屑病关节炎的患者。
核心结论
- 本研究显示,银屑病疾病早期阶段的DNA甲基化模式可在同一随访期内区分未来发生银屑病关节炎的患者与不发生者。
研究设计与样本
- 研究评估银屑病患者血液样本的全基因组DNA甲基化,比较后续发生关节炎者(转化者)与未发生者(生物制剂初治,按年龄、性别、银屑病病程和随访时长匹配)。
- 采用多变量线性回归模型(纳入年龄、性别、体重指数、吸烟等临床协变量)鉴定转化者与非转化者之间的甲基化差异。
- 通过构建支持向量机分类模型(加与不加临床变量)评估甲基化标记的预测性能。
主要发现
- 在15个基因及若干基因间区鉴定出36个高相关甲基化标记(错误发现率校正P<0.05,甲基化变化最小0.05)。
- 基于这些标记的分类模型识别转化者与非转化者的受试者工作特征曲线下面积为0.9644。
关键数值与推荐
- 甲基化标记筛选标准为错误发现率校正P<0.05且甲基化变化最小0.05。
- 基于36个甲基化标记的分类模型受试者工作特征曲线下面积为0.9644。
临床怎么用
- 原文未据此给出推荐。
局限性
- 原文未报告局限性。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
原文
OBJECTIVE: Psoriatic arthritis (PsA) is an immune-mediated inflammatory arthritis, associated with psoriasis, that significantly increases morbidity and mortality risk. We currently lack the means of predicting which patients with psoriasis will develop PsA, and a large number of patients remain undiagnosed. Regulation of gene expression through DNA methylation can potentially trigger and maintain PsA pathophysiological processes. We aimed to identify DNA methylation markers that can predict which patients with psoriasis will develop PsA prior to the onset of musculoskeletal symptoms. METHODS: Genome-wide DNA methylation was assessed in blood samples from patients with psoriasis who went on to develop arthritis (converters) and patients with psoriasis who did not (biologic naive, matched for age, sex, psoriasis duration, and duration of follow-up). Methylation differences between converters and nonconverters were identified by a multivariate linear regression model including clinical covariates (age, sex, body mass index, smoking). Predictive performance of methylation markers was assessed by developing support vector machine classification models with and without the addition of clinical variables. RESULTS: We identified a set of 36 highly relevant methylation markers (false discovery rate: adjusted P < 0.05 and a minimum change in methylation of 0.05) across 15 genes and several intergenic regions. A classification model relying on these markers identified converters and nonconverters with an area under the receiver operating characteristic curve of 0.9644. CONCLUSION: This study shows that DNA methylation patterns at an early stage of psoriatic disease can distinguish between patients who will develop PsA from those who will not during the same follow-up.
引用
- Cruz-Correa OF, Pollock RA, Machhar R, et al. Prediction of Psoriatic Arthritis in Patients With Psoriasis Using DNA Methylation Profiles. Arthritis Rheumatol. 2023. PMID: 37463128. doi:10.1002/art.42654
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。