Cost-effectiveness analysis of a first-trimester screening test for preterm preeclampsia in the Netherlands.

OBJECTIVES: The risk of preterm preeclampsia (PT PE) can significantly be reduced by starting acetylsalicylic acid ≤ 16 weeks of gestational age. First trimester predictive models based on maternal risk factors to effectively start this therapy lacked sufficient power, but recent studies showed that these models can be improved by including test results of biochemical and/or -physical markers. To investigate whether testing a biochemical marker in the first trimester is cost-effective in the Netherlands, a cost-effectiveness analysis was performed in this study. STUDY DESIGN: The outcome of this study was expressed as an incremental cost-effectiveness ratio (ICER) with as effect prevented PT PE cases. To evaluate the impact of each model parameter and to determine model uncertainties, both univariate and probabilistic sensitivity analyses were performed. RESULTS: When compared to the baseline strategy, the test strategy is estimated to save almost 4 million euros per year on a national scale and at the same time this would prevent an additional 228 PT PE cases. The sensitivity analyses showed that the major drivers of the result are the costs to monitor a high-risk pregnancy and the specificity and that most of the model simulations were in the southeast quadrant: cost saving and more prevented complications. CONCLUSIONS: This study showed that a first-trimester test strategy to screen for PT PE in the first trimester is potentially cost-effective in the Dutch healthcare setting. The fact that the specificity is a major driver of the ICER indicates the importance for a (new) screening model to correctly classify low-risk pregnancies.

Early pregnancy biomarker discovery study for spontaneous preterm birth.

(1) OBJECTIVE: discover new candidate biomarkers for spontaneous preterm birth in early pregnancy samples. When fully clinically validated, early pregnancy biomarkers for sPTB give the possibility to intervene or monitor high-risk pregnancies more intensively through, as example, pelvic exams, ultrasound or sonographic cervical length surveillance. (2) STUDY DESIGN: Early pregnancy serum samples of eight spontaneous extreme and very preterm birth cases (<32 weeks of gestational age) without any symptoms of preeclampsia and fetal growth restriction and eight uncomplicated pregnancies were analyzed by liquid chromatography mass spectrometry (LC-MS). Thirteen proteins, which were differentially expressed according to the LC-MS data, were subsequently selected for confirmation by enzyme-linked immunosorbent assay (ELISA). (3) RESULTS: Differential expression of four candidate biomarkers was confirmed by ELISA with decreased early pregnancy levels of gelsolin and fibulin-1 and increased levels of c-reactive protein and complement C5 in the preterm birth group. (4) CONCLUSIONS: The confirmed candidate biomarkers are all to some extent related to inflammatory pathways and/or the complement system. This supports the hypothesis that both play a role in extreme and very preterm birth without any symptoms of preeclampsia and fetal growth restriction. The predictive value of complement C5, c-reactive protein, fibulin-1 and gelsolin should, therefore, be validated in another cohort with early pregnancy samples.

First trimester serum biomarker discovery study for early onset, preterm onset and preeclampsia at term.

INTRODUCTION: Preeclampsia (PE) is a heterogeneous syndrome during pregnancy and postpartum and it is subdivided in this study into early onset (<34 weeks), preterm onset (34-37 weeks) and PE at term (>37 weeks). First trimester models currently lack a sufficient power to predict PE, but inclusion of biochemical markers shows an improvement of their predictive power. The aim of this study was to perform a biomarker discovery study in order to find possible novel first trimester biomarkers for each PE subtype. Further, our findings were related to available literature and the possible role of the proteins in the development of preeclampsia was discussed. METHODS: In this study, 9 early onset (<34 weeks), 8 preterm onset (34-37 weeks), 6 PE at term (>37 weeks) and 23 control samples were drawn between 11 and 14 weeks gestational age. Serum samples were prepared for liquid chromatography mass spectrometry analysis and protein data were exported for statistical analyses. All differentially expressed proteins were further evaluated by searching literature in MEDLINE, Embase and Web of science and differential expression of two proteins, which were not yet associated with PE, was verified through enzyme-linked immunosorbent assay (ELISA). RESULTS: After statistical analysis, six, four and eight proteins were differently expressed in early onset, preterm onset and PE at term, respectively. After exclusion of antibody fragments, only nine proteins remained. Seven out of these nine proteins were already in literature associated with preeclampsia and only three of them were described as differentially expressed in the first trimester or early second trimester of preeclamptic pregnancies. Differential expression of Apolipoprotein D (ApoD), which was not yet associated with PE, was confirmed by ELISA in both early and preterm onset PE in the first trimester. DISCUSSION: In this study, two main observations were made. First, some of the differentially expressed proteins have a role in the same biological pathway, such as the acute phase response or endometrium receptivity, and their differential expression was observed in all three PE subtypes. This observation supports the hypothesis that classification of PE could be more accurate when subtyping is based on the etiology and/or phenotype instead of the arbitrary parameter gestational age at onset or delivery. Second, seven differential expressed proteins were already associated in literature with preeclampsia, but this association was for only three of them observed in the first trimester. In addition, ApoD was not yet associated with PE in other studies and, moreover, its differential expression was confirmed by ELISA. Therefore the predictive power of these proteins in the first trimester is worth evaluating in a larger and more heterogeneous cohort.

Healthy and preeclamptic pregnancies show differences in Guanylate-Binding Protein-1 plasma levels.

The large interferon-inducible anti-angiogenic pro-inflammatory GTPase Guanylate Binding Protein-1 (GBP-1) is produced and secreted by activated endothelial cells and is highly induced by inflammatory cytokines and inhibited by angiogenic growth factors. During pregnancy a generalized mild inflammatory response is observed. During preeclampsia this generalized inflammatory response is even further activated and activation of the endothelium occurs. We hypothesized that GBP-1 is increased in healthy pregnancy and will be even further increased during preeclampsia. In the first experiment, plasma and placentas were collected from healthy and preeclamptic pregnancies. Plasma was also collected from non-pregnant women. For the second experiment longitudinal blood samples from women with a healthy or preeclamptic pregnancy were collected from the end of the first trimester until birth and one sample postpartum. The plasma GBP-1 levels were measured by ELISA and GBP-1 mRNA and protein levels in the placenta were tested by qPCR and immunohistochemistry. During pregnancy higher plasma concentrations of GBP-1 compared with non-pregnant women were observed. Surprisingly, during preeclampsia, plasma GBP-1 levels were lower than in control pregnancies and similar to the level of non-pregnant controls. Placental GBP-1 mRNA levels were not different between healthy and preeclamptic pregnancies and GBP-1 protein was virtually undetectable in the trophoblast by immunohistochemistry in placental tissue. Evaluation of longitudinal samples showed that plasma GBP-1 concentrations increased towards the end of pregnancy in healthy pregnancies, but not in preeclampsia. In line with our hypothesis, we found higher GBP-1 plasma levels during healthy pregnancy. However, plasma GBP-1 did not further increase during preeclampsia, but was stable. Further studies are needed to evaluate why GBP-1 does not increase during preeclampsia.

First trimester secreted Frizzled-Related Protein 4 and other adipokine serum concentrations in women developing gestational diabetes mellitus.

BACKGROUND: The aim of this study was to evaluate whether soluble frizzled-related protein 4 (sFRP4) concentration in the first trimester of pregnancy is individually, or in combination with Leptin, Chemerin and/or Adiponectin, associated with the development of gestational diabetes (GDM). METHODS: In a nested case-control study, 50 women with GDM who spontaneously conceived and delivered a live-born infant were matched with a total of 100 uncomplicated singleton control pregnancies based on body mass index (± 2 kg/m2), gestational age at sampling (exact day) and maternal age (± 2 years). In serum samples, obtained between 70-90 days gestational age, sFRP4, Chemerin, Leptin and Adiponectin concentrations were determined by ELISA. Statistical comparisons were performed using univariate and multi-variate logistic regression analysis after logarithmic transformation of the concentrations. Discrimination of the models was assessed by the area under the curve (AUC). RESULTS: First trimester sFRP4 concentrations were significantly increased in GDM cases (2.04 vs 1.93 ng/ml; p<0.05), just as Chemerin (3.19 vs 3.15 ng/ml; p<0.05) and Leptin (1.44 vs 1.32 ng/ml; p<0.01). Adiponectin concentrations were significantly decreased (2.83 vs 2.94 ng/ml; p<0.01) in GDM cases. Further analysis only showed a weak, though significant, correlation of sFRP4 with Chemerin (R2 = 0.124; p<0.001) and Leptin (R2 = 0.145; p<0.001), and Chemerin with Leptin (R2 = 0.282; p<0.001) in the control group. In a multivariate logistic regression model of these four markers, only Adiponectin showed to be significantly associated with GDM (odds ratio 0.12, 95%CI 0.02-0.68). The AUC of this model was 0.699 (95%CI 0.605-0.793). CONCLUSION: In the first trimester of pregnancy, a multi-marker model with sFRP4, Leptin, Chemerin and Adiponectin is associated with the development of GDM. Therefore, this panel seems to be an interesting candidate to further evaluate for prediction of GDM in a prospective study.