A signature serum biomarker for non-invasive diagnosis of preeclampsia patients.

Abstract

Purpose: This study seeks to identify a non-invasive biomarker for preeclampsia (PE), given its considerable influence on both maternal and fetal health.

Methods: The identification of differentially expressed genes (DEGs) in PE serum was conducted utilizing GSE192902. Weighted gene co-expression network analysis (WGCNA) was employed to identify functional modules, which were subsequently evaluated for their biological functions. Binary logistic regression was employed to evaluate genes derived from the intersection of DEGs and the most correlated module, with the aim of developing a biomarker model. The analysis of placental gene expression profiles was conducted utilizing GSE234729, and the model underwent validation in GSE149437.

Results: Over 1500 DEGs were identified in the serum of PE patients, with 63% exhibiting downregulation. Co-expression analysis revealed that the expression patterns of PE are structured into 13 distinct modules, with the dark-red module, comprising 55 genes, demonstrating the most significant correlation to the onset of PE. Following this, eight genes from the 26 differentially expressed genes (ADRB1, ARX, C2orf72, FOXB2, HIC1, IRX4, MEX3D, and MIR6724-4) were employed to construct a biomarker model, which attained an area under the curve of 76% (95% CI: 69-83%) in the training cohort and 74% (95% CI: 61-87%) in the validation cohort. Six DEGs were identified from the intersection of results pertaining to serum, placenta, and the dark-red module. However, only two, C2orf72 and RASGEF1C, exhibited consistent downregulation in both placenta and blood.

Conclusion: This comprehensive analysis reveals a promising biomarker model that may facilitate early detection of PE.