Maternal Serum EG-VEGF as A First Trimester Predictor of Hypertensive Disorders of Pregnancy: A Prospective Cohort Study

This study aims to explore whether plasma endocrine gland-derived vascular endothelial growth factor (EG-VEGF) in the rst trimester can be used as a predictor of hypertensive disorders of pregnancy (HDP), and compare it with placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1) to evaluate its prediction of HDP value. This is a prospective cohort study that records the medical history of the pregnant women included in the study at 11–13 weeks’ gestation, and analyzes serum biochemical markers including EG-VEGF, PIGF, sFlt-1 and sFlt-1/PIGF. The predictive values of these tests were determined. We used the receiver operating characteristic (ROC) curve to nd the optimal cut-off value for each biomarker and compare the operating characteristics (sensitivity, specicity). Logistic regression analysis was used to create a prediction model for HDP based on maternal characteristics and maternal biochemistry. sFlt-1. Combining maternal clinical characteristics and biochemical tests at 11–13 weeks can effectively identify women at high risk of HDP. arterial pressure, EG-VEGF endocrine gland-derived vascular endothelial growth factor


Introduction
Hypertensive disorders of pregnancy (HDP) are pregnancy-related diseases including gestational hypertension (GH) and preeclampsia (PE) that constitute one of the leading causes of maternal and perinatal mortality worldwide [1]. In Latin America and the Caribbean, hypertensive disorders account for nearly 26% of maternal deaths, whereas in Africa and Asia they are responsible for 9% of deaths [2].
Currently, the only treatment for HDP is delivery. However, this approach usually increases the incidence of preterm birth and the risk of neonatal morbidity.
Previous studies have shown that if clinicians can predict the risk of HDP in advance and take corresponding preventive measures, the resulting mortality can be greatly reduced [3,4]. Therefore, in recent years, considering its epidemiological and clinical importance, people have tried to nd speci c and practical biomarkers to predict and evaluate HDP. Recent studies have shown that a large number of serum circulating factors including soluble endoglin, placental growth factor (PlGF), soluble fms-like tyrosine kinase 1 (sFlt-1) are related to the occurrence of HDP [5]. However, there is still no indicator that can effectively predict HDP.
Endocrine gland derived vascular endothelial growth factor (EG-VEGF), also known as Prokineticin 1 (PROK1), belongs to a novel family of angiogenic mitogens [6]. It is mainly expressed in the endocrine organs, including the placenta, and the main target is endothelial cells (EC) derived from these endocrine tissues [6]. The biological activity of EG-VEGF is mediated by two G protein-coupled prokineticin receptors(PROKR), termed PROKR1 and PROKR2. Its receptors, PROKR1 and PROKR2 are highly expressed in trophoblasts and placental endothelial cells [7]. EG-VEGF is not only a uterine receptor marker and key protein during embryo implantation, but also a key endocrine factor that controls placental development after pregnancy [8]. The development of the placenta in normal early pregnancy is mainly controlled by two key steps: rst, angiogenesis period(8-10 weeks): trophoblast proliferation and differentiation, and angiogenesis; second. vascular recasting period (11-20 weeks): The cytotrophoblast differentiates into extravillous trophoblast cells (EVTs) with high invasion ability. The moderate in ltration of EVTs makes the maternal spiral arteries change from high-resistance, low-discharge to lowresistance, high-discharge vessels, and the uterine spiral arteries are recast to establish the placental vascular bed. The expression of EG-VEGF can be upregulated by hypoxia and βHCG, both of which are factors that are highly related to the development of HDP [9]. Although higher levels of EG-VEGF has been demonstrated during preeclampsia [10], there is still no investigation of this cytokine before the disease onset and its predictive potential for this disorder.
Therefore, This study aims to explore the concentration of EG-VEGF in the plasma of pregnant women before HDP onset, and use the maternal characteristics of early pregnancy, MAP and EG-VEGF and other biomarkers alone or in combination for effective prediction of HDP. In order to verify if EG-VEGF is a potential biomarker for HDP prediction.

Materials And Methods
This prospective observational study was conducted in the Department of Obstetrics and Gynecology, The Second Xiangya Hospital of Central South University between June 2019 to June 2020. The study protocol was approved by the ethical board of The Second Xiangya Hospital of Central South University.
All participants gave written informed consent. Women with singleton pregnancy at 11-13 + 6 weeks' gestational were invited into the study. Gestational age(GA) was calculated from the rst day of their last menstrual period and con rmed by ultrasound scanning of the crown-lump length (CRL) in the rst trimester. The exclusion criteria were chromosomal aneuploidy, fetal malformation, miscarriage before 24 weeks of pregnancy, termination of pregnancy, and fetal death. All operations (measurement, data recording) are performed according to routine clinical practice. Regular follow-up of all eligible women, record their demographic characteristics, delivery methods, and perinatal outcomes were recorded.
De nition of pregnancy outcome: Gestational hypertension is de ned as a systolic blood pressure of 140 mm Hg or more or a diastolic blood pressure of 90 mm Hg or more, or both, in previously normotensive women on two occasions at least 4 hours apart after 20 weeks of gestation. Preeclampsia was diagnosed according to the guidelines of the American College of Obstetricians and Gynecologists, based on systolic blood pressure ≥ 140 mmHg and/or diastolic blood pressure ≥ 90 mmHg on two recordings at least 4 hours apart, with the presence of proteinuria of ≥ 0.3g/24h or two readings of at least + + on dipstick analysis or other maternal end-organ damage [2].
Statistical Analysis SPSS 26.0 version(IBM, Armonk, NY, USA)was used for statistical analysis.The suitability of the quantitative data for normal distribution was tested by the Kolmogorov-Smirnov test and the Shapiro-Wilk test, The two-sample t-test was used for the comparison of two groups of quantitative data showing normal distribution, and the Mann-Whitney U test was used for the comparison of two groups of data with abnormal distribution. The Pearson chi-square test or Fisher test was used to compare qualitative data. We tried to use the receiver operating characteristic (ROC) curve to nd the optimal cutoff value for each biomarker and compare the operating characteristics (sensitivity, speci city). Logistic regression analysis was used to create a prediction model for HDP based on maternal characteristics and maternal biochemistry. ROC curve was obtained to assess the performance of this model.

Result
This study recruited 226 singleton pregnant women during the rst trimester from June 2019 to June 2020. We excluded a total of 21 cases (during the follow-up period, four participant withdrew consent for the study, nine miscarried after twelve weeks' gestation, three chose termination of pregnancy, three fetuses that later demonstrated morphological anomalies, two with proven aneuploidy). Of the remaining 205 pregnancies, 188 women were normotensive at delivery (91.7%) and 17 developed HDP (8.3%). Of the 17 pregnancies affected by hypertension, 13 developed GH (without PE) and four developed PE; 7.6% and 2.3% of total cohort respectively. Due to the small number of PE-affected pregnancies, GH and PEaffected pregnancies were combined for analysis.
The maternal and pregnancy characteristics of both groups are summarized in Table 1. Women who developed HDP had a signi cantly higher body mass index (BMI) and mean arterial pressure (MAP) in comparison to normotensive pregnancies(P = 0.000). With the exception of preterm birth rate there was no difference in pregnancy outcome between normotensive and hypertensive pregnancies (P < 0.05).
There was no signi cant difference in clinical information such as the age of pregnant women, gestational week of delivery, mode of delivery, and newborn weight. Because of the small number of pregnant women with chronic hypertension, antiphospholipid antibody syndrome, type 1 or type 2 diabetes, there was no statistical difference between the two groups. All analytes exhibited concentration differences between groups as shown in  EG-VEGF endocrine gland-derived vascular endothelial growth factor, PlGF placental growth factor, sFlt-1 soluble fms-like tyrosine kinase-1 EG-VEGF endocrine gland-derived vascular endothelial growth factor, PlGF placental growth factor, sFlt-1 soluble fms-like tyrosine kinase-1 We combined use of maternal clinical characteristics and biomarkers to establish prediction models including: BMI, EG-VEGF, MAP( Table 4). The regression analysis results of the area under the ROC curve (AUC) are shown in (Fig. 1). For a xed false positive rate, the sensitivity, negative value and predictive value of the nal model are shown in Table 5.  Data for sensitivity are given as mean (95% con dence internal). In this prospective cohort study, we found that the BMI and MAP of pregnant women who developed HDP were signi cantly higher than that of normotensive, which indicates that the risk of HDP is greater in women with high BMI and MAP. The research is consistent with Rahman [11]. In addition, the pregnancy method and previous SLE history of pregnant women also suggested that it would affect the occurrence of HDP. Whether other clinical features have an impact on HDP needs further research.
Although the etiology of PE is not fully understood, its development is believed to be mainly related to the super cial invasion of the maternal decidua and spiral artery by extravillous trophoblast(EVT). Hoffmann et al. found that EG-VEGF inhibits the migration and invasion of HTR-8 cells (extravillous trophoblast cell line) and the EVT in the early pregnancy villous explant culture system, and proved that EG-VEGF inhibits HTR-8 cell tissue from entering the tubular structure [10]. The data strongly suggests that EG-VEGF can act as an inhibitor of the differentiation of trophoblasts into an aggressive phenotype. EG-VEGF controls the migration and invasion of trophoblast cells, suggesting that these proteins locally control the process of spiral artery remodeling and fetal-maternal circulation establishment. Our previous studies have shown that during normal pregnancy, EGVEGF peaks in the proliferation and differentiation of placental trophoblasts and early angiogenesis (8-10 + 6 weeks), and then drops to a lower level to maintain until the end of placental recasting (11-20 weeks). This clinical nding suggests that the secretion level of EGVEGF is consistent with its physiological function: high levels of EGVEGF in the rst trimester mainly exercise its pro-angiogenic function, and then effectively withdraw (low levels) to avoid inhibiting the normal invasion of EVT cells. In the normal placenta model, the down-regulation of EG-VEGF expression at around 11 weeks of gestation promotes EVT differentiation. In our study, we found that compared with the normotensive group, the persistently high levels of EG-VEGF that can be observed in the rst trimester of HDP patients. We speculated that the high expression of EG-VEGF at 11-13 weeks' gestation may make EVT invade too shallow and lead to insu cient spiral artery recasting and cause pregnancyinduced hypertension. Sergent et al. have shown that in a mouse model, maintaining EG-VEGF levels more than 11.5 days of pregnancy, which is equivalent to the rst trimester of pregnancy, can lead to the development of the pathogenesis of HDP [12]. EG-VEGF has been described as a new actor in human fertility and plays a major role in the development of the uterus, placenta, and ovaries. During pregnancy, PROK1 can not only promote the development of chorionic villi, but also regulate placental angiogenesis [8,10,[13][14][15]. Its deregulation has been reported to be related to various placental pathologies, such as fetal growth restriction and PE [10,15]. More interestingly, EG-VEGF has recently been con rmed to be involved in the development of tumors in multiple reproductive organs such as testes, prostate [16]. EG-VEGF can promote apoptosis via regulating PI3K/AKT/mTOR pathway [17] and may be involved in the ability of tumor cells to invade other organs [18]. As PROK1 is a well-known actor in cell proliferation and survival [19], one could speculate that PROK1 might then participate with other factors to the development of HDP. However, the exact mechanism by which EG-VEGF causes HDP is unclear, and we still need to make further study.
Since 2012, the focus of biomarker research has been on the prediction accuracy of sFlt-1, PlGF and sFlt-1/PlGF ratio. Previous studies have suggested that in pregnant women who develop HDP the concentration of pro-angiogenic factor PIGF decreases, and the concentration of anti-angiogenic factor sFlt-1 increases [20,21], which is consistent with our research. Although elevated sFlt-1 and low PlGF are described as predictors of GHD, the predictive validity of these parameters is still worth exploring. The detection rate of PIGF in the rst trimester is 41-59% in early-onset PE and 33% in late-onset PE [22]. A recent study of 4,212 singleton pregnancies showed that PlGF cannot improve the screening performance of early-onset preeclampsia [23]. A study by Boucoiran et al. found that sFlt-1 has a speci city of 90% and a sensitivity of 25% [24]. In a multi-center clinical study conducted, it is reported that the DR of the PGF/sFlt-1 ratio is 82%, and the FPR is 5%, which is especially high for EO-PE (the detection rate is 89%) [22]. Therefore, we hope to compare EG-VEGF with PIGF and sFlt-1 to evaluate its predictive value in this study. Our study found that when EG-VEGF ≥ 227.83 pg/ml, the sensitivity, speci city, positive predictive value, and negative predictive value of EG-VEGF for HDP screening were higher than those of PIGF (43% vs. 33%; 93% vs. 70%; 52% vs. 86 %; 51% vs. 62%). In addition, the positive predictive value of EG-VEGF is also higher than that of sFlt-1 (86% vs. 59%). In this study, EG-VEGF is more speci c than PIGF and sFlt-1. This shows that this indicator has a better ability to correctly diagnose it as a non-patient in people who are not actually sick, reducing the waste of medical resources, and alleviating the patients' unprovoked panic and anxiety. The higher positive predictive value allows more people who screen positive to bene t from preventive interventions, reducing unnecessary exposure of the population.
The current guidelines recommend that preventive use of low-dose aspirin before 16 weeks for pregnant women at high risk of preeclampsia can reduce the risk of early-onset PE by more than 60% [25]. The same aspirin intervention after 20 weeks will not be as effective as the rst trimester [26]. Therefore, in order to calculate individualized risks, the current trend in screening involves combining the presence or absence of multiple risk factors at 11-13 + 6 weeks. The prediction model we established in the rst trimester include maternal BMI, MAP, and EG-VEGF, with an AUC of 0.8861 (95%CI: 0.7905-0.9818), which is better than using EG-VEGF alone (AUC: 0.66). This suggests that the combination of maternal characteristics of pregnant women and EG-VEGF can improve the effectiveness of the current prediction model. Our research supplements the existing evidence that during pregnancy, the maternal characteristics routinely measured in clinical practice, combined with the biochemical marker EG-VEGF, can be used to screen the risk of gestational hypertension and preeclampsia in low-risk populations.
This research has the following advantages: In this study, for the rst time, EG-VEGF was included in the prediction model of HDP, and the prediction model showed good predictive performance (AUC: 0.886); all measurements were carried out by trained personnel using recognized standardized methods, enabling the comparison of our ndings with previous reports; both patients and care providers were blinded to HDP risk calculation, so as not to affect further management.
The potential limitations of this study as follows: rst, it is a single-center retrospective study. We hope but have no chance to verify the most suitable model among external populations from more races, different regions, and different geographic regions. There is still a lack of evidence for clinically practice. Therefore, a larger scale multi-center prospective study is still needed to further verify the role of the examined markers in assessing the severity of PE and pregnancy outcome.Second, the combination of the mentioned thresholds of biomarkers as a biomarker test is only a preliminary suggestion. The level of all biomarkers may vary with race,gestational age, and may also depend on parity and smoking status.
Therefore, further research is needed to reduce the bias caused by confounding factors.

Conclusion
To our knowledge, this is the rst time that EG-VEGF has been included in the HDP prediction model. Serum EG-VEGF levels increase signi cantly at 11-13 weeks of gestation. We found that EG-VEGF has better screening performance than PIGF and sFlt-1 in the rst trimester. In addition, by combining with maternal characteristics, EG-VEGF can effectively predict the occurrence of HDP which provide new insights for HDP prevention strategies.

Declarations Data Availability
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.

Funding Information
This study was supported by the Hunan Provincial Key Research and Development Program (Grant no. 2018SK2067).

Author information
A liations