PRENATAL DIAGNOSTIC OF PREECLAMPSIA

Objective: Preeclampsia is a major cause of morbidity and mortality in pregnancy and still lacking an early prediction. Oxylipins, auto- and paracrine signaling molecules derived from polyunsaturated fatty acids (PUFAs), are implicated in the pathogenesis. We performed oxylipin profiling and developed a novel prediction model for preeclampsia. Design and method: We measured eighty-one free circulating oxylipins derived from linoleic-, eicosapentaenoic-, docosahexaenoic- and arachidonic acids by Triple-Quad-Tandem mass spectrometry in serum collected at gestational week (GW) 7-12 in a cohort study (total, 90 cases, 608 matched controls) before clinical onset of maternal syndrome. The later arising preeclampsia phenotype was heterogeneous with predominantly mild, late-onset disease. Using a machine learning regression-decision tree approach, prediction models incorporated clinical data, angiogenic markers and oxylipins, compared with a model including only standard clinical variables in combination with sFlt-1 and PlGF, bio markers used to diagnose PE late in pregnancy. Results: A regression tree approach for predicting preeclampsia in first trimester including X (number) specific oxylipins demonstrated good predictive capacity in early pregnancy with sensitivity 71, specificity 99 and AUC 87. Conclusions: Placentally derived PUFA metabolites (oxylipins) substantially add to the known clinical and biochemical predictive parameters for preeclampsia. A regression-tree based approach including known clinical and biochemical parameters, blood pressure and key oxylipins is a promising tool for predicting preeclampsia already in the first trimester.