First-Trimester Serum Acylcarnitine Levels to Predict Preeclampsia: A Metabolomics Approach. (4th June 2015)

• Record Type: Journal Article
• Authors: Koster, Maria P. H.; Vreeken, Rob J.; Harms, Amy C.; Dane, Adrie D.; Kuc, Sylwia; Schielen, Peter C. J. I.; Hankemeier, Thomas; Berger, Ruud; Visser, Gerard H. A.; Pennings, Jeroen L. A.
• Other Names: Sapin Vincent Academic Editor.
• Is Part Of: Disease markers. Volume 2015(2015)
• Publication Date: 2015-06-04
• Subjects: Diagnosis -- Periodicals; Biochemical markers -- Periodicals; Pathology -- Periodicals; 616
• DOI: 10.1155/2015/857108
• Languages: English
• ISSNs: 0278-0240

Abstract

Abstract : Objective . To expand the search for preeclampsia (PE) metabolomics biomarkers through the analysis of acylcarnitines in first-trimester maternal serum. Methods . This was a nested case-control study using serum from pregnant women, drawn between 8 and 14 weeks of gestational age. Metabolites were measured using an UPLC-MS/MS based method. Concentrations were compared between controls (n = 500 ) and early-onset- (EO-) PE (n = 68 ) or late-onset- (LO-) PE (n = 99 ) women. Metabolites with a false discovery rate <10% for both EO-PE and LO-PE were selected and added to prediction models based on maternal characteristics (MC), mean arterial pressure (MAP), and previously established biomarkers (PAPPA, PLGF, and taurine). Results . Twelve metabolites were significantly different between EO-PE women and controls, with effect levels between −18% and 29%. For LO-PE, 11 metabolites were significantly different with effect sizes between −8% and 24%. Nine metabolites were significantly different for both comparisons. The best prediction model for EO-PE consisted of MC, MAP, PAPPA, PLGF, taurine, and stearoylcarnitine (AUC = 0.784). The best prediction model for LO-PE consisted of MC, MAP, PAPPA, PLGF, and stearoylcarnitine (AUC = 0.700). Conclusion . This study identified stearoylcarnitine as a novel metabolomics biomarker for EO-PE and LO-PE. Nevertheless, metabolomics-based assays for predicting PE are not yet suitable for clinical implementation.