Maternal plasma syndecan-1: a biomarker for fetal growth restriction

Abstract Objective The identification of fetal growth disorders is an important clinical priority given that they increase the risk of perinatal morbidity and mortality as well as long-term diseases. A subset of small-for-gestational-age (SGA) infants are growth-restricted, and this condition is often attributed to placental insufficiency. Syndecan-1, a product of the degradation of the endothelial glycocalyx, has been proposed as a biomarker of endothelial damage in different pathologies. During pregnancy, a “specialized” form of the glycocalyx—the “syncytiotrophoblast glycocalyx”—covers the placental villi. The purpose of this study was to determine whether the concentration of maternal plasma syndecan-1 can be proposed as a biomarker for fetal growth restriction. Study design A cross-sectional study was designed to include women with normal pregnancy (n = 130) and pregnant women who delivered an SGA neonate (n = 50). Doppler velocimetry of the uterine and umbilical arteries was performed in women with an SGA fetus at the time of diagnosis. Venipuncture was performed within 48 h of Doppler velocimetry and plasma concentrations of syndecan-1 were determined by a specific and sensitive immunoassay. Results (1) Plasma syndecan-1 concentration followed a nonlinear increase with gestational age in uncomplicated pregnancies (R2 = 0.27, p < .001); (2) women with a pregnancy complicated with an SGA fetus had a significantly lower mean plasma concentration of syndecan-1 than those with an appropriate-for-gestational-age fetus (p = .0001); (3) this difference can be attributed to fetal growth restriction, as the mean plasma syndecan-1 concentration was significantly lower only in the group of women with an SGA fetus who had abnormal umbilical and uterine artery Doppler velocimetry compared to controls (p = .00071; adjusted p = .0028). A trend toward lower syndecan-1 concentrations was also noted for SGA with abnormal uterine but normal umbilical artery Doppler velocimetry (p = .0505; adjusted p = .067); 4) among women with an SGA fetus, those with abnormal umbilical and uterine artery Doppler findings had a lower mean plasma syndecan-1 concentration than women with normal Doppler velocimetry (p = .02; adjusted p = .04); 5) an inverse relationship was found between the maternal plasma syndecan-1 concentration and the umbilical artery pulsatility index (r = −0.5; p = .003); and 6) a plasma syndecan-1 concentration ≤ 850 ng/mL had a positive likelihood ratio of 4.4 and a negative likelihood ratio of 0.24 for the identification of a mother with an SGA fetus who had abnormal umbilical artery Doppler velocimetry (area under the ROC curve 0.83; p < .001). Conclusion Low maternal plasma syndecan-1 may reflect placental diseases and this protein could be a biomarker for fetal growth restriction. However, as a sole biomarker for this condition, its accuracy is low.

The glycocalyx, a multifunctional surface layer of glycans, covers the surface of epithelial and other types of cells [30]. A specialized glycocalyx lines the endothelium where it acts as a mechanosensory receptor; contributes to the regulation of vascular permeability and the maintenance of anticoagulation; and regulates the interaction between circulating blood cells and the endothelium itself [31][32][33][34][35]. In addition, the glycocalyx can sequester cytokines, angiogenic factors, and enzymes; therefore, it modulates their effects [31,36,37]. In pregnancy, there is a unique glycocalyx-the "syncytiotrophoblast glycocalyx"-that covers the villous tree of the placenta, which is thought to have an important role in the maintenance of blood fluidity in the intervillous space and in the cell-to-cell interaction between maternal white blood cells and the syncytiotrophoblasts [38][39][40][41][42].
The current study was conducted to determine whether pregnant women with an SGA fetus have a difference in the concentration of plasma syndecan-1 as a function of umbilical and uterine artery Doppler velocimetry. We performed a cross-sectional study of pregnancies with an SGA fetus, with and without umbilical and uterine artery Doppler abnormalities, and measured the maternal plasma concentrations of syndecan-1. We tested the hypothesis that maternal plasma syndecan-1 changes when mothers have an SGA fetus with evidence of placental disease reflected by abnormal umbilical artery Doppler velocimetry.

Methods
A retrospective cross-sectional study was conducted by searching the clinical database and bank of biological specimens of the Perinatology Research Branch, NICHD/ NIH/DHHS. All patients were enrolled at Hutzel Women's Hospital at the Detroit Medical Center (Detroit, Michigan, USA) between March 2000 and December 2002. This study included pregnant women between 20 weeks and 42 weeks of gestation who were allocated to the following groups: women with an uncomplicated pregnancy and appropriate-for-gestational-age (AGA) fetus (n ¼ 130) and women diagnosed to have an SGA fetus (n ¼ 50). Exclusion criteria included known fetal chromosomal or major structural anomalies, multiple gestation, and delivery at an outside institution.

Clinical definitions
A patient was considered to have an uncomplicated pregnancy if the following criteria were met: (1) no major medical, obstetrical, or surgical complications, (2) absence of labor at the time of venipuncture, and (3) delivery of a term (!37 weeks) infant whose birth weight was between the 10th and 90th percentiles for gestational age. The diagnosis of SGA was based on an ultrasonographic estimated fetal weight (EFW) below the 10th percentile according to the Hadlock 4 estimated fetal weight nomogram [70] and confirmed by a birth weight below the 10th percentile for gestational age, according to the reference range proposed by Alexander et al. [71]. Preeclampsia was defined in the presence of hypertension (systolic blood pressure !140 mmHg or diastolic blood pressure !90 mmHg on at least two occasions, 4 h to 1 week apart, after the 20th week of gestation) and proteinuria (! 300 mg in a 24-h urine collection or one dipstick measurement !1þ) [72].
All women provided written informed consent prior to the collection of plasma samples. The collection of samples and their utilization for research purposes were approved by the Institutional Review Boards of Wayne State University and the NICHD. Many of these samples were previously used in studies of intravascular inflammation, soluble adhesion molecules, and cytokine biology in normal and complicated pregnancies.

Doppler velocimetry of the umbilical and uterine arteries
Color and pulse-wave Doppler ultrasound examinations of the umbilical and uterine arteries were performed by trained personnel in a subset of patients (41 and 40 patients, respectively) at the time of diagnosis of an SGA fetus. The umbilical artery pulsatility index (PI) was measured in a free loop of the umbilical cord. Measurements were taken during periods of fetal inactivity and apnea, with the angle of insonation close to zero. Umbilical artery Doppler velocimetry was defined as abnormal if either the PI was above the 95th percentile for gestational age, or if waveforms were abnormal (absent or reversed end-diastolic velocities) [73,74]. The maternal uterine artery Doppler resistance index (RI) was calculated after sampling the uterine artery within one centimeter of its crossing with the external iliac artery. Measurements were taken bilaterally in duplicate and the average of means of the two sides was the final mean RI. Uterine artery Doppler velocimetry was defined as abnormal if the mean RI (average of right and left) was above the 95th percentile for gestational age and/or if waveform notching was present [75].

Sample collection and human syndecan-1 immunoassay
Blood was collected through venipuncture in ethylenediamine-tetra-acetic acid (EDTA) tubes within 48 h of the diagnosis of SGA and Doppler velocimetry examinations or at a prenatal visit for women in the control group. Plasma was obtained by centrifuging blood at 1300 g for 10 min at 4 C and then stored at À70 C. Concentrations of soluble syndecan-1 were determined by a commercially available human syndecan-1 enzyme-linked immunoassay (ELISA) Kit (Cell Sciences, Canton, MA, USA). The sensitivity of the assay was <2.56 ng/mL, and the coefficients of intra-assay variation and inter-assay variation were 7.6% and 6.8%, respectively.

Statistical analysis
After log (base 2) transformation, syndecan-1 concentration data in normal pregnancy were fit by using a quadratic function of gestational age at venipuncture. Differences between SGA subgroups and controls were assessed using a linear model that included the patient group and linear and quadratic terms of gestational age at sampling. Nulliparity was also considered as a covariate but not retained in the final model. P-values for differences between subgroups and the normal pregnancy group were adjusted for multiple comparisons by using the false discovery rate method to obtain q-values, and the false discovery rate was controlled at the 5% level. Syndecan-1 data in cases and controls were displayed in violin plots and summarized as geometric mean and 95% confidence intervals (CI). Chi-square tests were utilized to compare proportions. The relationship between the plasma syndecan-1 concentration and the umbilical or uterine artery Doppler velocimetry was examined by Spearman's rank correlation tests. A Receiver Operating Characteristic (ROC) curve was generated to assess the diagnostic performance of plasma syndecan-1 concentration for the identification of patients with abnormal umbilical artery Doppler velocimetry. Data were analyzed by using the statistical language and environment version R 3.6.1 and the IBM SPSS version 19.0 (IBM Corporation., Armonk, NY).

Clinical characteristics of the study population
The clinical and demographic characteristics of the study population are displayed in Table 1. Eighty percent (40/50) of patients with an SGA fetus delivered a neonate whose birth weight was below the 5th percentile for gestational age. Women with an SGA neonate were more often nulliparous than those with an uncomplicated pregnancy (52% vs. 27%; p ¼ .001). There were no significant differences in the median gestational age at venipuncture, maternal age, and frequency of tobacco and illicit drug usage between the two groups (all, p > 0.05). As expected, patients with SGA had a lower median gestational age at delivery (35.7 weeks vs. 39.5 weeks; p < .001) and a higher rate of preexisting medical diseases (15% vs. 0%; p ¼ .001) than women in the control group (Table 1). A total of seven patients presented preexisting medical conditions: four, essential hypertension; one, diabetes mellitus; one, chronic hypertension with diabetes mellitus; and one, chronic hypertension with left nonfunctioning kidney. Four of these patients had abnormal umbilical artery Doppler velocimetry and three had abnormal uterine artery Doppler velocimetry. None of the patients in the SGA group met the criteria to diagnose preeclampsia at the time of diagnosis of SGA. However, one patient developed preeclampsia 10 days later, prior to delivery.
Plasma syndecan-1 concentration followed a nonlinear increase with gestational age in uncomplicated pregnancies Maternal soluble plasma syndecan-1 concentrations (ng/mL) showed a nonlinear increase with gestational age in the group of uncomplicated pregnancies (R 2 ¼ 0.27, p < .001) (Figure 1).  There was an inverse relationship between the plasma syndecan-1 concentration and the umbilical artery PI (r ¼ À0.5; p ¼ .003) as well as the mean uterine artery RI (r ¼ À0.4; p ¼ .04).
Among women with an SGA fetus, the group with abnormal umbilical and uterine artery Doppler findings had the lower mean plasma syndecan-1 concentration When pregnancies complicated with an SGA fetus were considered, the group of women with

Principal findings
(1) Plasma syndecan-1 concentration followed a nonlinear increase with gestational age in uncomplicated pregnancies (R 2 ¼ 0.27, p < .001); (2) women with pregnancies complicated with an SGA fetus had a significantly lower mean plasma concentration of syndecan-1 than those with an AGA fetus (p ¼ .005); (3) this difference can be attributed to FGR, as the mean plasma syndecan-1 concentration was significantly lower only in the group of women with an SGA fetus who had abnormal umbilical and uterine artery Doppler velocimetry compared to controls (p ¼ .00071; q ¼ 0.0028). A non-significant trend toward lower syndecan-1 concentrations was also noted for SGA with abnormal uterine but normal umbilical artery Doppler velocimetry (p ¼ .0505; q ¼ 0.067); 4) among women with an SGA fetus, those with abnormal umbilical and uterine artery Doppler findings had a significantly lower mean plasma syndecan-1 concentration than women with normal Doppler velocimetry (p ¼ .02; q ¼ 0.04); 5) an inverse relationship was found between the mean maternal plasma syndecan-1 concentration and the umbilical artery PI (r ¼ À0.5; p ¼ .003); and 6) a plasma syndecan-1 concentration 850 ng/mL had a positive likelihood ratio of 4.4 and a negative likelihood ratio of 0.24 for the identification of a mother with an SGA fetus who had abnormal umbilical artery Doppler velocimetry (AUC 0.83; p < .001).

Results in the context of what is known
Syndecan-1 in normal pregnancy A growing body of evidence suggests that placental syndecan-1 plays an important role in normal pregnancy where it seems to be involved in implantation [76], cell migration and proliferation [39], and fetal-maternal inter-communication [40]. Syndecan-1 also modulates the effects of multiple growth factors, e.g. angiogenic factors such as vascular endothelial growth factor (VEGF) [76][77][78][79][80]. Plasma syndecan-1 concentrations have been reported to increase throughout pregnancy [38,66]. Gandley et al. evaluated maternal plasma syndecan-1 concentrations across uncomplicated pregnancies and found that plasma syndecan-1 starts to increase in early pregnancy, reaches a peak at term, and decreases within 24-48 h postpartum [81]. The rapid drop in concentration after delivery has been interpreted as indicating that the placenta is a major source of syndecan-1 in maternal plasma [38,62,81]. Studies investigating syndecan-1 immunostaining of placentas at different gestational ages showed no differences in the expression of syndecan-1 on the villous surface throughout gestation [38,41,63,82]. Syndecan-1 mRNA expression has been reported to be increased in BeWo cells exposed to forskolin, a culture model that simulates the process of trophoblast differentiation. This model provides evidence for a possible role of syndecan-1 in syncytialization of cytotrophoblasts [38]. A possible explanation for the increase in plasma syndecan-1 concentrations during pregnancy, given the absence of changes in syndecan-1 expression on the surface of the villi as reported from immunostaining findings, is the increase in placental size, thus the release of syndecan-1 into the maternal circulation.

Syndecan-1 in complications of pregnancy
Altered placental syndecan-1 expression has been reported in pregnancies complicated by preeclampsia [38,63,[81][82][83][84] and HELLP syndrome [38] and in a subset of patients with FGR [84,85]. The heterogeneity of the reported results can be attributed to the differences in the applied methodologies as well as to the heterogeneity of these syndromes [86][87][88].
Circulating syndecan-1 has been evaluated in patients with preeclampsia. Several studies showed that plasma syndecan-1 concentrations are lower in patients with preeclampsia compared to normal pregnancies [38,81,[89][90][91]. Szabo et al. [38] proposed this may be due to a defective syncytiotrophoblast transport of syndecan-1, possibly associated with an altered network of the syncytiotrophoblast's cytoskeletal proteins reported in this syndrome [92]. The finding that plasma syndecan-1 concentrations are lower in women with preeclampsia has been therefore attributed to a placental dysfunction rather than to endothelial damage. Recently, pregnant women with a growth-restricted fetus have also been reported to have lower plasma syndecan-1 concentrations than controls [68,69]. Collectively, the available evidence seems to support a role for syndecan-1 in various diseases of pregnancy, and some authors proposed that placental syndecan-1 in early gestation may be predictive of pregnancy outcomes [93].
Plasma syndecan-1 concentration followed a nonlinear increase with gestational age in uncomplicated pregnancies In this study, we found that maternal plasma syndecan-1 concentrations increased in a nonlinear fashion with gestational age in the group of uncomplicated pregnancies (R 2 ¼ 0.27, p < .001) (Figure 1). This finding is consistent with the results of previous studies that showed a similar trend in plasma syndecan-1 concentrations throughout pregnancy [38,66,81].
Plasma syndecan-1 is lower in pregnant women with SGA fetuses than in women with AGA fetuses In the present study, plasma syndecan-1 concentrations were lower in patients with an SGA fetus than in normal pregnant women (Figure 2). This finding is in agreement with the results reported by two previous studies published by Tong et al. in which maternal plasma syndecan-1 concentrations were lower in women with an SGA fetus at the time of diagnosis and, importantly, prior to the diagnosis, compared to women with normal pregnancy outcomes [68,69]. Moreover, circulating syndecan-1 concentrations were significantly correlated with birth weight centiles and placental weight [68]. Interestingly, Garcha et al. [68] reported that neither protein nor mRNA expression of syndecan-1 was altered in the placenta of patients who delivered an SGA neonate. However, the study by Chui et al. found that the expression of syndecan-1 mRNA and protein was significantly lower in the placenta of mothers with FGR fetuses than in those of controls [85]. Relevant to these findings, in vitro experiments showed that exposure of primary cytotrophoblasts to hypoxic conditions (1% oxygen) decreased the secretion of syndecan-1 into the culture medium, an effect thought to be regulated by matrix metalloproteinases, which have been implicated in the cleavage of syndecan-1 from the syncytiotrophoblast. Thus, the low plasma syndecan-1 concentrations observed in mothers with an SGA fetus might be a consequence of hypoxic stress on the trophoblasts, interfering with the cleavage process of syndecan-1. Lastly, the same investigators reported that silencing syndecan-1 expression with siRNA determined a decrease in trophoblast proliferation [68].
Maternal plasma syndecan-1 represents a biomarker for the identification of growth-restricted fetuses The present study is the first to evaluate maternal plasma concentrations of syndecan-1 in pregnancies with an SGA fetus and to examine the relationship between syndecan-1 and Doppler velocimetry findings in umbilical and uterine arteries. Doppler velocimetry of the umbilical artery examines the impedance to blood flow in the placenta, and it has been used to assess placental disease [94][95][96][97]. Deterioration of umbilical artery blood flow is associated with an abnormal biophysical profile [98,99], umbilical acidemia in cordocentesis [99][100][101][102][103], and fetal vascular lesions of malperfusion in placental pathology [94,95,[104][105][106][107][108][109][110]. Abnormal Doppler velocimetry of the uterine artery is thought to reflect high impedance to uterine blood flow, as a consequence of failure of physiologic transformation of spiral arteries, which can lead to reduced uteroplacental perfusion [105,106,[111][112][113][114]. In 2016, a Delphi consensus of experts identified the alterations in umbilical or uterine artery Doppler velocimetry as contributory criteria to be applied in the differential diagnosis of SGA fetuses with EFW <10th percentile from fetal growth-restricted fetuses [22]. The importance of distinguishing between an SGA fetus and an FGR fetus relies on the higher risk of abnormal perinatal outcome associated with FGR compared to SGA, which may include a large number of constitutionally small but healthy fetuses [19][20][21].
Our findings showed that mothers with an SGA neonate with abnormal umbilical and uterine artery Doppler velocimetry had lower plasma concentrations of syndecan-1 compared to those with an AGA fetus ( Figure 3). We also found that there was a significantly lower mean plasma concentration of syndecan-1 in women with an SGA fetus who had abnormal umbilical and uterine artery Doppler velocimetry compared to women with an SGA fetus and normal Doppler findings (p ¼ .02; q ¼ 0.04). Among patients with SGA, those with an abnormal umbilical and uterine artery Doppler velocimetry had the lowest syndecan-1 concentration ( Figure 3). A trend toward a lower mean plasma concentration of syndecan-1 was also found in women with an SGA fetus and abnormal uterine artery Doppler findings only (with normal umbilical Doppler velocimetry) compared to uncomplicated pregnancies, though not statistically significant after adjustment (p ¼ .0505; q ¼ 0.067) (Figure 3). We did not analyze the group of women with abnormal umbilical artery findings only, given that this represents an uncommon clinical scenario in FGR due to placental insufficiency and that only two patients with these characteristics were included in our study. Since uterine artery Doppler velocimetry is thought to reflect the degree of failure of physiologic transformation of the spiral arteries, syndecan-1 concentration in maternal plasma can also be a biomarker of shallow placentation [25]. The relationship between plasma syndecan-1 concentrations and other placental lesions (i.e. chronic inflammatory lesions) observed in a subset of pregnancies with FGR remains to be determined.
Taken together, these results suggest that syndecan-1 may reflect placental disease, thus giving value to its possible role as a biomarker for FGR [68]. In our study, a plasma syndecan-1 concentration 850 ng/ mL enabled the identification of mothers with an SGA fetus who had abnormal umbilical artery Doppler velocimetry among the study population (AUC 0.83; p < .001). The accuracy, however, was low and unlikely to justify the implementation of syndecan-1 as a sole biomarker for this condition.
The diagnosis, surveillance, and time of delivery of fetuses with suspected FGR are major issues in obstetrics [1,3,[115][116][117][118][119][120]. There is a growing interest in finding additional imaging modalities or placental biomarkers able to reflect the presence of placental insufficiency and underperfusion and to improve the detection rate of growth-restricted fetuses [68,[121][122][123][124][125][126]. Proposed candidates are proteins highly expressed in the placenta relative to other tissues such as serine peptidase inhibitor Kunitz type-1 (SPINT1) and growth differentiation factor-15, both previously investigated in the prediction of FGR [69]. The results of this study provide further evidence for a possible role of plasma syndecan-1 in the early identification of fetuses with restricted growth during pregnancy.

Strengths and limitations
This is the first study in which maternal plasma soluble syndecan-1 concentrations have been correlated with umbilical and uterine artery Doppler velocimetry findings in pregnancies complicated with an SGA fetus. Integration of these results with placental findings and neonatal follow-up would add strength to our study. The findings that 80% of patients with an SGA fetus enrolled in this study had an EFW below the 5th percentile for gestational age may reflect our clinical setting and research operation at that time. As a tertiary referral center with a level-three neonatal intensive care unit, our hospital received transfer of care for several patients with early-onset SGA from local and near by regional hospitals. Moreover, patients with a more severe disease were more likely to be sent for ultrasound examinations and diagnosed with SGA due to size discrepancy from dates. We did not perform an ultrasound examination in every patient to screen for SGA. Additionally, the patients who were diagnosed with more severe FGR may have been more inclined to participate in the research study than those with a lesser degree of disease.

Conclusion
Low maternal plasma syndecan-1 may reflect placental diseases and this protein could be a biomarker for fetal growth restriction. However, as a sole biomarker for this condition, its accuracy is low.

Disclosure statement
No potential conflict of interest was reported by the author(s).