MON-129 - Multi-Omic Biomarkers to Improve Sflt-1/Plgf Ratio-Based Pre-Eclampsia Prediction

Pre-eclampsia (PET) is a heterogeneous disorder affecting 3-5% of pregnancies, which involves metabolic, inflammatory, cardiovascular, and endocrine dysfunction. It is a leading cause of maternal and neonatal morbidity and mortality worldwide. Current clinical prediction of PET among symptomatic patients uses the ratio of two angiogenic markers released from the placenta: soluble fms-like tyrosine kinase-1 (sFlt-1) and placental growth factor (PlGF), offering an acceptable ability to rule out but suboptimal sensitivity to accurately rule in PET. Identification of additional serum biomarkers that enhance the performance of the sFlt-1/PlGF ratio is therefore necessary to improve risk stratification and consequently patient outcomes.

To address this, we performed targeted proteomic profiling of established PET-related pathways in retrospective serum samples from symptomatic patients with suspected PET (n=88), with sFlt-1/PlGF ratios measured by NICE-approved assays and pregnancy outcomes. This approach used Olink proximity extension assay panels to quantify 92 cardiometabolic and 92 inflammatory proteins in these samples.

Differential expression analysis identified 23 proteins significantly altered in individuals later diagnosed with PET, most notably the extracellular matrix remodelling protein PCOLCE, angiogenic protein Tie1 and vascular inflammation marker VCAM1. Decision tree models incorporating PCOLCE alongside the sFlt-1/PlGF ratio demonstrated improved sensitivity (84.9% vs. 71.9%), specificity (85.7% vs. 80.4%), and overall predictive accuracy (85.2% vs. 77.2%) compared with the current sFlt-1/PlGF-only model. Functional enrichment analysis and protein-protein interactions networks also revealed dysregulation of interconnected cardiometabolic and inflammatory processes in PET.

Proteomic profiles were integrated with previously generated NMR-based metabolomic profiles from the same serum samples, resulting in paired metabolomic-proteomic datasets. This enabled the development of multi-omic predictive models for PET and provides insight into biological processes underpinning disease heterogeneity. From the multi-omic dataset, cross-validated Elastic Net regression identified an 8-marker signature of PET, including the sFlt-1/PlGF ratio, with high sensitivity (98%), specificity (71.2%), PPV (85.7%), NPV (95.8%) and overall accuracy (88.5%). In a subset of samples (n=30) with a high proportion (50%) of misclassifications by the sFlt-1/PlGF ratio, untargeted mass-spectrometry further identified potential biological mechanisms underlying false-negative and false-positive classifications.

These findings demonstrate that circulating protein biomarkers, particularly PCOLCE, may enhance the predictive performance of the sFlt-1/PlGF ratio and support multi-omic approaches to improve clinical prediction of PET.

*Unless otherwise noted, all abstracts presented at ENDO must not be released to the press or the public until the date and time of presentation. For oral presentations, the abstracts are embargoed until the session begins. The Endocrine Society reserves the right to lift the embargo on specific abstracts that are selected for promotion prior to or during ENDO.*