Circulating vascular endothelial growth factor receptor-3, a pro-lymphangiogenic and pro-angiogenic mediator, is decreased in pre-eclampsia.

The demographic and clinical characteristics of women included in this case-control study are displayed in Table 1 . Compared with HP women, PIH patients had similar age, but NP women were older (p<0.05). At study enrollment, body mass index (BMI) was lower in NP, but higher in PIH (both p<0.05). Fasting blood glucose levels were also higher in PIH compared to HP (p=0.046), but did not reach pre-diabetic (100–125 mg/dL) or diabetic (≥126 mg/dL) levels. NP women exhibited similar SBP and higher DBP (p<0.05). As expected, both SBP and DBP levels were increased in PIH (both p<0.05), despite the fact that most of these women were receiving anti-hypertensive therapy. The following anti-hypertensive drugs were prescribed in isolation or combination depending on disease severity, blood pressure attenuation response, and patient tolerance: methyldopa, nifedipine, labetalol, and/or hydralazine.[ 23 ] Moreover, we found lower gestational age at blood sampling (GAS), gestational age at delivery (GAD), newborn weight, and newborn height in PIH than in HP (p=0.031, p<0.001, p<0.001, and p=0.014, respectively). Heart rate, hemoglobin, hematocrit, red blood cell (RBC), mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), neutrophil, lymphocyte, eosinophil and platelet numbers, and placental weight values were not statistically different between the PIH and HP groups. We additionally noted several differences with regard to demographic and clinical characteristics when the PIH group was broken down into the diagnosis of GH (n=61) and PE (n=74), as shown in Table 2 . Compared with GH women, PE patients exhibited increased SBP (p=0.003), DBP (p=0.043), neutrophil number (p=0.002), urea nitrogen (p<0.001), creatinine (p=0.013), and proteinuria (p<0.001), whilst decreased hemoglobin (p=0.037), lymphocyte (p=0.012) and platelet (p=0.016) numbers, GAS (p=0.032), GAD (p<0.001), and newborn weight (p=0.018) and height (p=0.011) values. By classifying the PE group according to different clinical presentations, n=22 (29.7%), n=15 (20.3%), n=19 (25.7%), and n=17 (23.0%) of patients fit into the subgroups of PE with severe features, early-onset PE, preterm PE, and PE with SGA, respectively. The median (25–75% percentile) values for circulating VEGFR-3 concentration in the study groups are depicted in Figures 1 and 2 . Plasma VEGFR-3 was increased in HP [1207.0 (646.2–1993.0) pg/mL] compared to NP [132.8 (65.7–213.9) pg/mL) women; however, PIH [729.2 (309.3–1403.0) pg/mL] patients had lower levels than NP and HP women ( Figure 1A ; all p<0.05). Considering just our pregnant population and setting apart PE from GH in the PIH group, plasma VEGFR-3 remained significantly lower in PE [537.4 (132.9–1100.0) pg/mL] than in HP (p<0.05); however, levels were statistically similar in GH [979.9 (388.6–1615.0) pg/mL] and HP ( Figure 1B ). In addition, plasma VEGFR-3 was decreased in PE compared to GH patients ( Figure 1B ; p<0.05). When PE was categorized into the different subgroups, plasma VEGFR-3 was further decreased in the cases defined as PE with severe features compared to those without severe features [196.7 (89.4–629.0) vs. 665.5 (357.4–1516.0) pg/mL] ( Figure 2A ; p=0.004). Likewise, we found lower plasma VEGFR-3 in PE women who had preterm birth versus term birth [204.9 (91.9–729.2) vs. 589.1 (335.2–1360.0) pg/mL] and in those who delivered SGA babies versus babies with appropriate weight for gestational age [204.9 (115.9–755.5) vs. 592.6 (336.7–1516.0) pg/mL] ( Figure 2C and 2D ; p=0.034 and p=0.028, respectively). We also observed a trend to VEGFR-3 to be decreased in the circulation of early-onset PE versus late-onset PE [397.7 (82.0–1165.0) vs. 589.1 (392.9–1004.0) pg/mL] ( Figure 2B , p=0.072). We then performed correlation analyses of plasma VEGFR-3 concentration with the demographic and clinical characteristics differentially distributed among pregnant groups, including fasting glucose, blood pressure, GAS, GAD, and newborn weight/height, as displayed in Table 3 . While there were significant positive correlations of VEGFR-3 with GAS, GAD, newborn weight, and newborn height (all p<0.05), VEGFR-3 was negatively correlated with SBP and DBP (both p<0.05) in our pregnant population (HP+PIH). When we segregated these groups in an attempt to determine which one was driving the relationships, we found VEGFR-3 remained significantly correlated with GAS in PE and GH (both p<0.05) as well as with newborn weight and newborn height in PE (both p<0.05). However, the correlations of VEGFR-3 with SBP, DBP, and GAD lost significance. In addition, VEGFR-3 was positively correlated with placental weight only in the PE group and negatively correlated with fasting glucose exclusively in the GH group (both p<0.05). Furthermore, as the lymphatic system is a vascular network critical to interstitial fluid homeostasis and immunity, we determined the relationship of plasma VEGFR-3 concentration with hemogram parameters in our pregnant groups ( Table 3 ). In the overall HP+PIH population, there was a significant positive correlation between VEGFR-3 and the number of lymphocytes in blood (p=0.032), which was driven mainly by the PE and GH groups (both p<0.05). We also observed a significant negative correlation between VEGFR-3 and the number of neutrophils only in PE (p=0.015). Curiously, VEGFR-3 was negatively correlated with MCV and MCH exclusively in the PE group (both p<0.05).

This study was approved by the Institutional Review Board (CAAE-37738620.0.0000.5440, October 19, 2020) of the University Hospital of the Ribeirao Preto Medical School (FMRP), University of Sao Paulo (USP, Ribeirao Preto, Brazil). A total of 252 women (PIH, n=135; HP, n=68; NP, n=49) were consecutively enrolled in the outpatient and inpatient clinics of the Department of Gynecology and Obstetrics. We conducted our study in accordance to the principles of the Declaration of Helsinki and all participants provided written informed consent. Following the Brazilian Federation of Gynecology and Obstetrics Associations (FEBRASGO) and the International Society for the Study of Hypertension in Pregnancy (ISSHP) guidelines, GH was defined as systolic blood pressure (SBP) ≥140 mmHg or diastolic blood pressure (DBP) ≥90 mmHg, or both, on two or more measurements at least 4 h apart after 20 weeks of gestation, in a previously normotensive woman. PE was defined as GH along with proteinuria and/or other sign/symptom of target maternal organ damage and uteroplacental dysfunction after 20 weeks of gestation.[ 2 , 23 ] PE was further categorized according to different clinical presentations as following: PE with severe features, patients with one or more of the following features: SBP ≥160 mmHg and/or DBP ≥110 mmHg, thrombocytopenia, hepatic impairment, renal insufficiency, pulmonary edema, and brain or visual disturbances; early-onset PE, patients with onset of symptoms occurring before 34 weeks of gestation; preterm PE, patients delivering before 37 +0 weeks of gestation; and PE with small for gestational age (SGA), patients with a newborn weighting below the 10 th percentile for gestational age.[ 1 , 23 , 24 ] Demographic and clinical data, including results of routine laboratory tests on pregnant women and their respective newborns, were gathered from medical records. Women with chronic hypertension, including those diagnosed with superimposed PE, were excluded from this study. Additional exclusion criteria comprised the clinical evidence of multiple pregnancy, hemostatic abnormalities, (chronic or gestational) diabetes, cancer, and/or autoimmune, cardiovascular, renal and hepatic diseases. Maternal venous blood from NP, HP, and PIH women was collected into sterile tubes containing ethylenediaminetetraacetic acid (EDTA) as anticoagulant at the time of clinic attendance. No sample was collected during delivery. Plasma samples were obtained from centrifugation of whole blood, and stored at −80 °C until used to assess VEGFR-3 concentration by the methodology described below. The concentration of VEGFR-3 in plasma samples was determined using a commercially available Duoset enzyme-linked immunosorbent assay (ELISA) kit (Catalog number: DY349B-05, R&D Systems, Minneapolis, MN, USA), according to the manufacturer’s instructions. Assay range was 93.8 – 6.000 pg/mL. GraphPad Prism 9.5 was used for statistical analysis and preparation of graphs. Data distribution was analyzed by D’Agostino-Pearson normality test. Student’s t -test, Mann-Whitney U test, parametric/nonparametric one-way analysis of variance (ANOVA) followed by post-hoc tests, and chi-square test were used as appropriate to examine the differences in demographic and clinical parameters among NP, HP, and PIH women, considering the HP group as controls. Differences in circulating VEGFR-3 concentrations were analyzed by Kruskal-Wallis test followed by Dunn’s multiple comparison test when NP, HP, and PIH (GH and PE) groups were compared or by Mann Whitney test when the PE group was broken down into the subgroups of PE with severe features versus without severe features, early-onset versus late-onset PE, preterm versus term PE, or PE with SGA versus without SGA. The relationships of plasma VEGFR-3 concentration with clinical parameters in mothers and newborns were analyzed using Pearson’s or Spearman’s correlation tests (r and p-values) as appropriate. A value of p <0.05 was considered the level of statistical significance.

The main findings reported here are that circulating VEGFR-3 is increased in HP compared with NP, but decreased in PIH when compared with HP women. In addition, plasma VEGFR-3 was lower in PE than in HP, and further reduced levels were detected in cases identified as PE with severe features, preterm PE, and PE with SGA. Plasma VEGFR-3 was also decreased in PE compared with GH, whereas its levels were similar between GH and HP women. Moreover, increasing plasma VEGFR-3 levels were associated with decreasing SBP and DBP as well as with increasing blood lymphocyte number, GAS, GAD, newborn weight and height in our overall pregnant groups. In PE specifically, plasma VEGFR-3 was significantly correlated with GAS, newborn weight and height, placental weight, and different hemogram parameters (MCV, MCH, and neutrophil and lymphocyte numbers). The lymph vessels are important for drainage and return of interstitial fluids to the bloodstream as well as for transport of immune cells, dietary lipids, and fat-soluble vitamins. As an important component of the lymph endothelial cells, VEGFR-3 along with its ligand, vascular endothelial growth factor (VEGF)-C facilitate not only these functions of the lymphatic system in the adult life, but also mediate angiogenesis and lymphangiogenesis during early gestation.[ 25 ] Indeed, increased plasma VEGFR-3 was detected in HP compared with NP. However, circulating VEGFR-3 was not associated with any of the considered maternal and neonate demographic and clinical parameters in the HP group. Although we are unaware of prior publications comparing circulating levels of VEGFR-3 between healthy pregnant and healthy non-pregnant women, Kim and colleagues (2011) demonstrated lower endometrial mRNA expression of VEGFR-3 in pregnant than in non-pregnant women; however, patients in both of these groups were diagnosed with endometriosis and had their tissues collected as part of in vitro fertilization procedures.[ 26 ] Yet, previous studies have shown that increases in VEGFR-3 mRNA expression may not be paralleled by increases in VEGFR-3 protein expression in placental tissue from PIH patients.[ 19 ] Our findings of decreased circulating VEGFR-3 in PE compared with HP are in alignment with earlier studies describing reduced VEGFR-3 protein levels in placental and decidual tissues from PE patients.[ 18 – 21 ] Only Lely and collaborators (2013) had previously assessed VEGFR-3 in the circulation of PIH patients, and they likewise reported decreased plasma VEGFR-3 levels in PE compared with normotensive pregnancy. However, mean concentrations in the PE and GH groups were very similar, being plasma VEGFR-3 in GH also significantly decreased compared with normotensive pregnancy.[ 22 ] In view of the association between VEGFR-3 and GAS noted here, we speculate that inconsistent results between Lely et al . (2013) study and ours regarding the comparison of circulating VEGFR-3 in GH versus PE might be explained by differences in the gestational age of blood collection. Although Lely and colleagues do not specify the mean GAS in each study group, they mentioned that samples were collected pre-delivery within 2 weeks of diagnosis of GH or PE. Thus, blood collection was performed any time after 20 weeks and before 35.4 and 34.1 weeks – the mean GAD in their GH and PE groups, respectively.[ 22 ] Discrepancies in clinical practice related to recommendations for expectant management and anti-hypertensive therapy might also contribute to conflicting results among studies. At the time Lely et al. study was published,[ 22 ] academic medical institutions in the USA used to follow the 2002’s American College of Obstetricians and Gynecologists (ACOG) practice bulletin for the management of PE,[ 27 ] which was developed based on the 2000’s Report of the National High Blood Pressure Education Program Working Group on High Blood Pressure in Pregnancy.[ 28 ] These guidelines encouraged delivery beyond 32–34 weeks of gestation in PE with severe features and did not endorse the administration of anti-hypertensive drugs for women with persistent blood pressure levels <160 mmHg systolic or <110 mmHg diastolic. However, as aforementioned, most of the patients in our study underwent anti-hypertensive treatment once diagnosed with PIH, which likely favored the prolongation of gestation to a mean GAD of 38.9 and 36.5 weeks in our GH and PE groups, respectively. When PE was classified according to different clinical presentations, we found that plasma VEGFR-3 concentrations were decreased in the cases defined as PE with severe features, preterm PE, and PE with SGA compared to respective counterparts. Plasma VEGFR-3 levels also tended to be decreased in early-onset PE compared with late-onset PE. Prior studies have assessed placental and/or decidual VEGFR-3 protein expression in distinct PE phenotypes with contradictory results. Whereas Dubova et al . (2013) failed to demonstrate differences in placental VEGFR-3 immunohistochemistry staining localized in syncytiotrophoblast, endothelial, or mesenchymal cells between moderate or severe PE compared with controls,[ 29 ] Pillay et al . (2020) detected reduced staining in conducting (stem) and exchange (intermediate and terminal) chorionic villi of early-onset PE compared with normotensive pregnancy.[ 20 ] Although no direct comparisons were made between the early- versus late-onset PE subgroups, it seems that VEGFR-3 expression was lower in the conducting villi of early-onset patients compared to late-onset counterparts, while no differences were observed in the exchange villi.[ 20 ] Moreover, Marini et al . (2007) have previously shown decreased placental VEGFR-3 levels in patients with hemolysis, elevated liver enzymes and low platelets (HELLP) syndrome, which is typically considered a complication of PE, compared with both GH and PE.[ 19 ] Furthermore, 20–40% of patients with PE exhibit decidual vasculopathy (DV) or acute atherosis, a pregnancy-specific thrombo-inflammatory lesion in the utero-placental spiral arteries diagnosed by subendothelial accumulation of lipid-laden macrophages and necrosis.[ 30 ] Interestingly, Liu et al . (2015) reported lower decidual VEGFR-3 levels in PE patients with DV than in those without DV.[ 18 ] With regard to the relationship of circulating VEGFR-3 with maternal and neonate parameters in PIH, the association of increasing VEGFR-3 levels with decreasing SBP and DBP and with increasing GAD lost statistical significance after we segregated our pregnancy population into HP, GH, and PE. Although the effect of methyldopa, nifedipine, labetalol, and hydralazine on VEGFR-3 levels is currently unknown, the effect of these anti-hypertensive drugs on attenuating blood pressure and prolonging gestation might have impacted these correlation analyses. Nonetheless, increasing VEGFR-3 levels remained significantly associated with increasing GAS in GH and PE as well as with increasing newborn weight and height and placental weight in PE. These results suggest that circulating VEGFR-3 diminishes as pregnancy progresses and may have negatively influenced fetal growth in PE. Future studies during pregnancy should confirm these findings in distinct populations. PE is linked to impaired adaptative immune response, reduced plasma volume expansion, edema, and anemia.[ 31 – 36 ] Interestingly, VEGFR-3 deficient mouse embryos exhibit abnormal vasculature and severe anemia.[ 37 ] Thus, we also evaluated the relationship between circulating VEGFR-3 and hemogram parameters in our pregnant groups. Increasing VEGFR-3 levels were associated with increasing lymphocyte number in PE and GH. However, a negative correlation with neutrophil number in PE was noted. It is expected that low VEGFR-3 levels lead to an undeveloped lymphatic system and subsequently to decreased recovery of infiltrated leucocytes from tissues like placenta and kidneys into the circulation; conversely, there is compelling evidence that immune cells are involved in the modulation of angiogenesis and lymphangiogenesis, with neutrophils producing both pro- and anti-angiogenic factors.[ 38 ] Moreover, we found a negative correlation of VEGFR-3 with MCV and MCH in PE. Low MCV and MCH values may indicate iron-deficiency anemia, whereas high MCV and MCH can signal anemia induced by decreased levels of folic acid (vitamin B9) and/or vitamin B12.[ 39 ] Indeed, PE has been previously associated with vitamin B9 and B12 deficiencies.[ 40 , 41 ] Since we have not measured these micronutrients in our population, we are not able to ascertain whether low VEGFR-3 (along with tissue fluid imbalance) and/or decreased vitamin B contributed to increased erythrocyte size (MCV) and hemoglobin quantity (MCH) in our PE group. Nonetheless, these results suggest that low VEGFR-3 levels may have negatively influenced blood cell traffic and oxygen transportation in PE. It is noteworthy some limitations that might impact the interpretation of our results. First, we have measured plasma VEGFR-3 levels after the manifestation of clinical symptoms in PIH patients. Thus, we are unable to establish a causative role for VEGFR-3 in the pathophysiology of PE, neither to assert VEGFR-3 as a predictive biomarker for PE or disease severity/adverse perinatal outcomes in PE. Future studies should measure circulating VEGFR-3 longitudinally in pregnant women, from embryo implantation to the completion of placental-decidual angiogenesis and lymphangiogenesis, and to the appearance of hypertension. In addition, mechanistic studies in cell and animal models could use our interesting correlative observations as a rationale to investigate the role of VEGFR-3 on immune regulation, blood cell and fluid homeostasis, blood pressure control, and fetal growth. Preclinical studies could also evaluate whether pro-lymphangiogenic treatments ameliorates placental pathology, hypertension, and possibly other maternal-fetal adverse outcomes associated with PE. For example, it has been shown that VEGF-C supplementation improves cardiac inflammation, edema, and fibrosis in rodent models of myocardium infarction.[ 42 , 43 ] Furthermore, advanced maternal age, obesity, and diabetes are recognized risk factor for PE.[ 3 – 7 ] Indeed, PIH patients were older, heavier, and presented higher (non-diabetic) fasting blood glucose levels than HP women in the current study. Based on the sample size of our groups (HP, n=68 and PIH, n=61–74) – which is bigger than those in earlier publications (controls, n=10–30 and PIH, n=6–30),[ 19 – 22 , 29 ] with the exception of Liu et al . study (controls, n=20; PE-DV, n=38 and PE+DV, n=52) –[ 18 ] our study was statistically powered (for α=0.05, 1-β=0.935) to address its aims. Moreover, this relatively large population allowed us to perform correlation analyzes between plasma VEGFR-3 levels and demographic/clinical characteristics of pregnant women. As aforementioned, VEGFR-3 was not statistically correlated with age, BMI, and fasting glucose in either HP or PE. These results suggest that low VEGFR-3 levels in the circulation of PE patients compared to HP were not driven by differences in these maternal characteristics between groups. Although plasma VEGFR-3 levels in GH were comparable to those in HP, VEGFR-3 was negatively correlated with fasting glucose in our GH group. While GH may be considered a subtype of PE, the pathophysiological mechanisms leading to maternal hypertension in GH may differ than those in PE, with genetic, epigenetic, behavioral, and environmental factors interacting with placental factors to modify the route to disease.[ 16 ] Previous reports are inconclusive as to whether pregnant women with glucose intolerance have dysregulated VEGFR-3 expression in placental tissue.[ 44 – 46 ] Studies in rodent models are also contradictory regarding the beneficial or detrimental effects of VEGFR-3-mediated lymphangiogenesis on inflammation, renal damage, and wound healing in diabetes.[ 47 – 50 ]

In conclusion, our data indicate reduced circulating VEGFR-3 concentration in patients with PIH, specifically in those diagnosed with PE. In addition, decreased VEGFR-3 was associated with adverse clinical outcomes in PE. We also took a step further than prior studies and provided evidence for significant correlations of plasma VEGFR-3 levels with gestational age, newborn and placental weight, and different hemogram parameters, including the size and hemoglobin quantity of erythrocytes as well as neutrophil and lymphocyte numbers in PE. These findings suggest that diminished VEGFR-3 in the circulation may contribute to impaired immune regulation, blood cell and fluid homeostasis, and fetal growth in PE patients.

Pregnancy-induced hypertension (PIH) represents a group of obstetric complications characterized by new-onset hypertension after the first half of gestation, with blood pressure returning to normal levels typically within 12 weeks post-partum. Gestational hypertension (GH) is diagnosed when ≥140 mmHg systolic or ≥90 mmHg diastolic blood pressure is the only maternal sign, whereas preeclampsia (PE) is diagnosed when maternal hypertension is accompanied by clinical and/or laboratorial evidence of end-organ damage.[ 1 , 2 ] GH and PE affect 1.8–4.4% and 0.2–9.2% of pregnancies worldwide, respectively.[ 3 , 4 ] However, there are reports indicating that the incidence of PE has increased during the last decades due to the higher prevalence of risk factors (i.e. advanced maternal age, obesity, diabetes, and other pre-existing medical conditions) and to the broad acceptance of these revised diagnostic criteria.[ 5 – 7 ] Importantly, PE has devastating health consequences, contributing to a large portion of the number of maternal and fetal/neonate deaths annually as well as to an increased risk of mother and offspring to develop cardiovascular diseases later in life.[ 3 , 8 – 13 ] Although the underlying molecular mechanism(s) remains unclear, a two-stage disease model has been recognized for PE.[ 14 ] The first stage is marked by placental mal-perfusion, secondary to either 1) shallow implantation and abnormal development of the placental-uterine vasculature in the first trimester of gestation, which portray histological findings typically observed in early-onset PE; or 2) physical constrains on placental growth and compression of the chorionic villi later on gestation, which are frequently associated to late-onset PE. These two pathways converge on a common endpoint of placental pathology: syncytiotrophoblast stress, with the subsequent release of signals into the maternal circulation that prompts widespread endothelial dysfunction. Thus, the heterogeneity in its clinical presentations arises from different pathways leading to the clinical symptoms of PE, influenced by maternal genetics, epigenetics, lifestyle, and environmental factors.[ 15 , 16 ] While angiogenesis, the process of blood vessel formation, is crucial for oxygen and nutrient supply, especially during embryogenesis and placental development; lymphangiogenesis, the establishment of lymph vessels, is essential the for proper absorption of tissue fluids, antigen presentation, and recovery of infiltrated immune cells. Vascular endothelial growth factor receptor (VEGFR)-3 is expressed in lymph endothelial cells as well as in venous endothelial cells during embryogenesis, playing a major role in lymphangiogenesis but also in the regulation of angiogenesis under pathological conditions.[ 17 ] Interestingly, previous studies have found reduced mRNA and protein expression of VEGFR-3 in the placental and decidual tissues of patients with PIH.[ 18 – 21 ] This observation may reflect on decreased circulating VEGFR-3 levels in PIH.[ 22 ] In this study, we aimed to assess the circulating concentration of VEGFR-3 in PIH patients versus those in non-pregnant (NP) and healthy pregnant (HP) women. We further compared VEGFR-3 in plasma from PE patients categorized into different medical presentations, including the subgroups classified as early- or late-onset PE. Finally, we examined the relationship of circulating VEGFR-3 levels with clinical perinatal parameters.