Pathophysiology, Predictive Modeling, and Long-Term Cardiovascular Outcomes in Severe Preeclampsia

Preeclampsia (PE) is a pregnancy-specific disorder that is characterized by oxidative stress, angiogenic imbalance, and endothelial dysfunction leading to multiorgan manifestations.The resulting endothelial damage have a life-long impact on cardiovascular health, contributing to an increased risk of cardiovascular disease (CVD), stroke and diabetes.The aims of this thesis were to describe the pathophysiology of severe PE by analyzing clinical risk factors and a subset of biomarkers, representing various signaling pathways described in PE.Another aim was to develop a predictive model for PE with need for intensive care and to evaluate the longterm postpartum impact of severe PE on the cardiovascular system.In Paper I, biomarkers for oxidative stress, hemopexin (Hpx), alpha-1-microglobulin (A1M), and angiogenic imbalance, soluble fms-like tyrosine kinase-1 (sFlt-1), placental growth factor (PlGF), were analyzed in a cohort of intensive care patients with severe PE (n=41) versus PE patients not admitted to intensive care, PE controls (n=40), and normotensive pregnancies (n=40).In Paper III, markers for endothelial dysfunction, sphingosine-1-phosphate (S1P), hyaluronic acid (HA), syndecan-1 (SDC-1), heparan sulfate proteoglycan 2 (HPSG2), were analyzed in the same cohorts.The results showed that patients with severe PE had lower postpartum plasma concentrations of Hpx, A1M and S1P, and higher plasma concentrations of the glycocalyx degradation products SDC-1 and HA, indicating an impairment in the endogenous defense system against the oxidative stress in PE, in combination with signs of increased endothelial damage and vascular dysfunction.Correlation analysis revealed significant associations between analyzed biomarkers and physiological disease severity parameters such as blood pressure.Also, intensive care patients showed a postpartum sFlt-1:PlGF ratio in parity with antenatal levels indicative of adverse pregnancy outcomes previously described in PE.In Paper II, the potential of using machine learning models to predict intensive care need in patients with preeclampsia was explored by incorporating numerical routine clinical and laboratory parameters from the intensive care cohort (n=41) and the PE controls (n=40).The variables that were most indicative of severe disease were aspartate aminotransferase, uric acid and body mass index, with an area under the curve (AUC) of 0.91.The model was validated internally on a test-set yielding a lower AUC of 0.85.Paper IV comprised a prospective study that evaluated cardiovascular status in a selection of 10 patients from the intensive care cohort and 10 normotensive pregnancies, 4-7 years postpartum.Participants were examined with blood pressure measurements, echocardiography and cardiac magnetic resonance (CMR) imaging.Prior PE patients had a significantly higher E:A ratio (peak velocity blood flow during early filling versus peak velocity flow during atrial contraction) obtained by echocardiography.However no patient fullfilled the current clinical criterias for diastolic dysfunction.There was no difference in CMRaquired parameters, indicating a normal heart morphology and function.The normal morphology was likely related to the lack of persistent hypertension in these cases.