SAT-124 Identification of Metabolic Biomarkers Complementing the sFlt-1/PlGF Ratio in Pre-eclampsia

Abstract Disclosure: H.M. Hooker: None. N. Khovanova: None. D. Grammatopoulos: None. Pre-eclampsia (PE) is a serious complication of 3-5% of pregnancies and a leading cause of maternal and neonatal morbidity and mortality worldwide. Accurate prediction of symptomatic patients that will develop PE is vital to improve patient outcomes. The pathogenesis of the condition originates in the placenta, where dysfunction results in the release of proinflammatory and antiangiogenic molecules into the maternal bloodstream, causing maternal endothelial dysfunction. The ratio between two of these markers, soluble fms-like tyrosine kinase-1 (sFlt-1) and placental growth factor (PlGF), has recently been adopted into clinical practice for PE prediction and diagnosis, offering acceptable ability to rule-out, but sub-optimal sensitivity to accurately rule-in, PE. Investigation into additional serum biomarkers to improve the predictive performance of the sFlt-1/PlGF ratio is therefore urgently required. The aetiology of PE is multifactorial and includes metabolic dysregulation, making the metabolome an important source of potential biomarkers. Therefore, this study aimed to use broad metabolic profiling to identify biomarkers that improve performance of the current clinical biomarkers, sFlt-1 and PlGF, for improved PE prediction and greater understanding into disease phenotype. 250 metabolite biomarkers were measured using an NMR-based metabolomics platform in retrospective serum samples from 124 symptomatic patients with suspected PE, investigated by measurements of sFlt-1/PlGF ratio by NICE-approved assays and defined pregnancy outcomes. Univariate analysis using modified Z-scores identified 38 significantly altered metabolite biomarkers in those diagnosed with PE, with glycerol exhibiting the most significant increase (2.8-fold, p<0.001). Glycerol also significantly correlated with the sFlt-1/PlGF ratio (ρ=0.359, p<0.001) and showed significant differences in patients stratified by clinical sFlt-1/PlGF cut-offs (p<0.001). Importantly, Decision Tree regression models demonstrated improved sensitivity, specificity and overall predictive accuracy using sFlt-1/PlGF ratio alongside glycerol, compared to current models using ratio alone.Combining the univariate analysis with dimensionality reduction techniques and multivariate regression analysis resulted in the identification of a 24-biomaker metabolite signature of PE. These biomarkers cover multiple metabolic pathways including amino acid metabolism, fatty acid metabolism, glycolysis, lipid and cholesterol pathways, capturing the complex, multi-factorial nature of PE. Identification of new unique biomarker signatures associated with both the sFlt-1/PlGF ratio and PE diagnosis broadens our insight into metabolic disruptions in this condition but also demonstrates the potential utility of metabolomic biomarkers to refine clinical prediction models for pre-eclampsia. Presentation: Saturday, July 12, 2025