Department of Critical Care Medicine, Hubei Province Hospital of Traditonal Chinese Medicine, 856 Luoyu Street, Wuhan, Hubei 430061, China.
Hubei Province Academy of Traditonal Chinese Medicine, 856 Luoyu Street, Wuhan, Hubei 430061, China.
Biomed Res Int. 2022 Aug 6;2022:3690893. doi: 10.1155/2022/3690893. eCollection 2022.
Septic cardiomyopathy is widespread during sepsis and has adverse effects on mortality. Diagnosis of septic cardiomyopathy now mainly depends on transthoracic echocardiogram. Although some laboratory tests such as troponin T and atrial brain natriuretic peptide play a role in the diagnosis, specific blood biochemistry biomarkers are still lacking. In our study, we sought to find potential biological markers from genes and pathways that are covariant in the blood and myocardium of septic patients. Bioinformatics and machine learning methods were applied to achieve our goal. Datasets of myocardium and peripheral blood of patients with sepsis were obtained from the Gene Expression Omnibus (GEO) database. Differentially expressed genes (DEGs) were selected and received functional enrichment analysis. Unsupervised hierarchical clustering analysis was performed to identify the subtypes of sepsis. Random forest, lasso regression, and logistic regression were used for variable screening and model construction. Internal and external validation sets were applied to verify the efficiency of the model in classifying disease and predicting mortality.
By defining significance for genes using Student's -test, we obtained 1,049 genes commonly changed in both myocardium and blood of patients with sepsis. The upregulated genes (LogFC >0) were related to inflammation pathways, and downregulated (LogFC <0) genes were related to mitochondrial and aerobic metabolism. We divided 468 sepsis patients into two groups with different clinical result based on the mortality-related commonly changed genes (104 genes), using unsupervised hierarchical clustering analysis. In our validation datasets, a six-gene model (, , , , , and ) was obtained and proven to perform well in classifying groups and predicting mortality.
We have identified genes that have the potential to become biomarkers for septic cardiomyopathy. Additionally, the pathophysiological changes in the myocardium of patients with sepsis were also reflected in peripheral blood to some extent. The co-occurring pathological processes can affect the prognosis of sepsis.
脓毒症性心肌病在脓毒症中广泛存在，对死亡率有不良影响。脓毒性心肌病的诊断现在主要依赖于经胸超声心动图。虽然肌钙蛋白 T 和心房脑利钠肽等一些实验室检查在诊断中起作用，但仍缺乏特异性血液生化标志物。在本研究中，我们试图从脓毒症患者血液和心肌中共同变化的基因和途径中寻找潜在的生物学标志物。应用生物信息学和机器学习方法来实现我们的目标。从基因表达综合数据库（GEO）数据库中获得脓毒症患者心肌和外周血的数据集。选择差异表达基因（DEGs）并进行功能富集分析。进行无监督层次聚类分析以识别脓毒症的亚型。随机森林、套索回归和逻辑回归用于变量筛选和模型构建。内部和外部验证集用于验证模型在疾病分类和预测死亡率方面的效率。
通过使用 Student's t 检验定义基因的显著性，我们获得了在脓毒症患者心肌和血液中共同变化的 1049 个基因。上调基因（LogFC >0）与炎症途径有关，下调基因（LogFC <0）与线粒体和有氧代谢有关。我们根据与死亡率相关的共同变化基因（104 个），使用无监督层次聚类分析将 468 名脓毒症患者分为两组，具有不同的临床结果。在我们的验证数据集中，获得了一个六基因模型（、、、、、和），并证明其在分组分类和预测死亡率方面表现良好。
我们已经确定了一些有潜力成为脓毒性心肌病生物标志物的基因。此外，脓毒症患者心肌的病理生理变化在某种程度上也反映在外周血中。共同发生的病理过程会影响脓毒症的预后。
