Angiogenic growth factor levels in maternal and fetal blood: correlation with Doppler ultrasound parameters in pregnancies complicated by pre-eclampsia and intrauterine growth restriction

Authors


Abstract

Objectives

To correlate levels of angiogenic growth factors with Doppler ultrasound parameters in pregnancies complicated by pre-eclampsia and intrauterine growth restriction (IUGR).

Methods

In 16 women with pre-eclampsia and 15 women with isolated IUGR, pulsatility indices (PI) in the umbilical and uterine arteries were measured by Doppler ultrasonography. At delivery, maternal and fetal blood (umbilical vein and artery separately) was sampled and angiogenic growth factors measured by means of enzyme linked immunosorbent assay (ELISA).

Results

Umbilical artery PI was significantly higher in women with IUGR than in those with pre-eclampsia, whereas uterine artery PI was not statistically significantly different. Maternal soluble fms-like tyrosine kinase-1 (sFlt-1) levels were higher in women with pre-eclampsia than in those with IUGR (P < 0.0001). Umbilical vein basic fibroblast growth factor (bFGF) levels were lower in women with pre-eclampsia than in those with IUGR (P < 0.05). Placental growth factor (PlGF) levels in the umbilical vein were below the detection limit in nearly all samples of IUGR fetuses and lower than in those with pre-eclampsia (P < 0.001). Maternal PlGF levels were inversely correlated with PI values of both vessels. In the umbilical vein sFlt-1 was positively and soluble kinase insert domain receptor (sKDR) negatively correlated with umbilical artery PI. No correlation could be found in the serum of the umbilical artery for all growth factors and for vascular endothelial growth factor (VEGF) in all compartments.

Conclusions

The correlations between maternal and fetal angiogenic growth factor serum levels and Doppler ultrasound indices of uterine and umbilical arteries in pre-eclampsia and IUGR reflect the severity of the disorders especially for the fetus. A combination of both measurements may be useful in future screening for early prediction of pregnancy complications. Copyright © 2007 ISUOG. Published by John Wiley & Sons, Ltd.

Introduction

Adequate placentation and placental development are crucial steps for normal pregnancy, and angiogenesis and vascular transformation are important for normal placental development. Hypertensive disorders in pregnancy (especially pre-eclampsia) and intrauterine growth restriction (IUGR) still remain the leading causes of maternal and fetal morbidity and mortality1, 2. Despite extensive research the exact pathophysiology of both disorders—especially that of pre-eclampsia—remains elusive3, 4. A two-stage disorder model is commonly accepted—reduced placental perfusion and shallow trophoblast invasion leading to abnormal placentation in the first and second trimesters, leading in turn to the clinical disorders in later pregnancy5, 6.

The current etiological and pathophysiological concept for these pregnancy disorders implies an imbalance of angiogenic growth factors. Recent studies have reported that placental growth factor (PlGF) and soluble fms-like tyrosine kinase-1 (sFlt-1) are altered in pre-eclampsia and IUGR. In numerous studies, PlGF has been demonstrated to be diminished in pre-eclamptic serum7–10, most likely owing to binding on sFlt-1, the levels of which are elevated: the higher the sFlt-1 concentration, the lower the PlGF level10. Serum PlGF levels inversely correlate with the severity of the disease10. Additionally, in one study, serum PlGF levels in normotensive women with small-for-gestational-age (SGA) infants were significantly lower than those in gestational-age-matched controls with appropriate-for-gestational-age infants at 33 weeks' gestation11.

A soluble and endogenously secreted form of the vascular endothelial growth factor (VEGF) receptor (sFlt-1) is produced by alternative splicing and contains the extracellular ligand-binding domain, but not the transmembrane and cytoplasmic portions12. sFlt-1 is able to block the effects of VEGF and PlGF by inhibiting interaction with its receptors.

In pre-eclampsia and, to a lesser extent, in IUGR, placental expression and maternal blood levels of sFlt-1 are upregulated7–10, 13, 14. Increased sFlt-1 during pre-eclampsia is associated with decreased free VEGF and free PlGF in maternal blood. Levels of sFlt-1 are positively correlated with the severity of the disease15: sFlt-1 concentrations are higher in women with early onset of the disease10–16, more severe disease7, 10, 15, and SGA or IUGR infants10, 16, 17.

The soluble form of the kinase insert domain receptor (KDR) has recently been detected in human plasma18. Although KDR appears to play a more important functional role in mediating VEGF signaling events compared with fms-like tyrosine kinase-1 (Flt-1), little information regarding sKDR in pregnant women with or without pre-eclampsia or IUGR is available19.

Fibroblast growth factor-2 (basic FGF or bFGF) is a member of the FGF family and has pleiotropic effects in different cell and organ systems. It is a potent angiogenic factor, stimulates hematopoiesis, and in addition plays an important role in the differentiation and/or function of the nervous system20, 21. In pregnancies complicated by SGA a trend to lower FGF levels in maternal and fetal blood has been reported22, whereas in pregnancies complicated by hypertensive disorders bFGF levels were found to be higher23.

The severity of both pre-eclampsia and IUGR can be estimated with Doppler ultrasound measurements. Uterine artery evaluation depicts maternal vascular effects, as poor placentation results in impaired remodeling of the uterine spiral arteries5, 6, resulting in an increase in peripheral resistance, which can be shown by typical Doppler ultrasound waveforms24. Fetal well-being can also be estimated by Doppler ultrasound by investigating fetal and fetoplacental vessels, which are known to change in a characteristic pattern24.

The purpose of this study was to examine a potential correlation of angiogenic growth factor levels in maternal and fetal serum with Doppler ultrasound parameters of uterine and umbilical arteries.

Patients and Methods

During a 19-month period women who were admitted to the perinatal and labor ward of the University of Erlangen-Nuremberg and in whom pre-eclampsia or isolated IUGR was diagnosed, were recruited. The guidelines of the International Society for the Study of Hypertension in Pregnancy (ISSHP) provided the basis for the classification of pre-eclampsia, which was diagnosed when proteinuria (> 0.3 g/24h) and hypertension (new onset after 20 weeks' gestation, > 140/90 mmHg by repeated measurements) were present.

IUGR was suspected by ultrasound measurements (abdominal circumference and estimated fetal weight < 10th centile according to the formula of Hadlock et al.25) and confirmed postnatally by birth weight below the 10th percentile corrected for gestational age. In order to ensure the inclusion only of fetuses in which IUGR was caused by placental insufficiency and to avoid inclusion of ‘normal’ and otherwise healthy SGA fetuses, the definition of IUGR was expanded to include oligohydramnios and/or pathologic umbilical artery pulsatility indices.

In this period about 59 women with early onset pre-eclampsia and 37 women with isolated IUGR were admitted to and treated in our department. We finally selected 16 women with pre-eclampsia and 15 women with isolated IUGR with no signs of pre-existing or pregnancy-induced hypertension or pre-eclampsia with postnatally confirmed IUGR. Fetuses with chromosomal or structural anomalies were excluded from the study. To avoid a possible hypoxic effect of labor on angiogenic growth factor levels we selected only those women ultimately delivered by Cesarean section without labor. The study was approved by the ethical committee at the University of Erlangen-Nuremberg, and informed written consent was obtained from all study subjects.

Pulsed Doppler recordings were done either on the day of delivery or 1 day prior to delivery; recordings from the umbilical artery were made with a 60-Hz filter, and absent or abnormal end-diastolic blood flow velocity was documented from a minimum of five consecutive waveforms in two free-floating loops of cord. The uterine arteries were examined similarly, special attention being paid to a possible notch phenomenon. The mean reading of both uterine arteries was calculated and used for statistical analysis.

Maternal venous blood samples were collected from the antecubital vein prior to Cesarean section. Fetal blood samples were taken from the umbilical artery and vein separately, immediately after delivery of the placenta. All blood samples were centrifuged immediately after collection at 3000 rpm for 15 min and serum was aliquotted and stored at − 80 °C.

Serum bFGF, VEGF, PlGF, and sFlt-1 levels were measured in all three compartments (maternal, placental (umbilical vein), and fetal (umbilical artery)). Soluble VEGF receptor 2 (sKDR) levels were measured only in maternal and umbilical vein serum. All factors were measured by means of quantitative sandwich immunoassay techniques, following the manufacturer's (R&D Systems GmbH, Wiesbaden, Germany) instructions. According to the manufacturer mean minimal detectable doses are: 9 pg/mL for VEGF, 7 pg/mL for PlGF, 5.01 pg/mL for sFlt-1, 4.6 pg/mL for sKDR, and 0.22 pg/mL for bFGF.

All statistical analyses were conducted using the statistical software program GraphPad PRISM 4.03 for Windows (GraphPad Software, San Diego, CA, USA). Test results are expressed as mean ± standard deviation. Data were tested for normal distribution and demographic and clinical data were compared by means of independent-samples t-test, Mann–Whitney U-test or Fisher's exact test if appropriate. To analyze correlations between angiogenic growth factor levels and Doppler parameters, Pearson's or Spearman's correlation coefficient was calculated. A two-sided P < 0.05 was considered statistically significant.

Results

The characteristics of the study patients are shown in Table 1. There were no statistically significant differences between the groups in maternal age, body mass index and gestational age at delivery, primigravidity and primiparity, as well as smoking habits. Fetal birth weight was not statistically different but, owing to the study design, in the IUGR group the birth-weight centiles were significantly lower than in the pre-eclampsia group, in which only 50% of the fetuses were growth restricted.

Table 1. Demographic characteristics of the study population
 Pre-eclampsia group (n = 16)IUGR group (n = 15)P
  • Values are given as mean ± SD or n (%).

  • *

    According to the International Society for the Study of Hypertension in Pregnancy (i.e. < 0.3 g/24 h). IUGR, intrauterine growth restriction; NS, not significant.

Age (years)31.81 ± 4.9730.07 ± 5.87NS
Body mass index (kg/m2)29.99 ± 7.3025.94 ± 3.72NS
Primigravida8 (50.0)7 (46.7)NS
Primipara12 (75.0)9 (60.0)NS
Gestational age at delivery (days)226 ± 34236 ± 26NS
Smoker2 (12.5)4 (26.7)NS
Systolic blood pressure (mmHg)185 ± 24125 ± 22< 0.0001
Diastolic blood pressure (mmHg)115 ± 1080 ± 16< 0.0001
Mean arterial blood pressure (mmHg)138 ± 1495 ± 17< 0.0001
Proteinuria (g/dL)6.64 ± 7.05Negative* 
Fetal birth weight (g)1568 ± 8531350 ± 657NS
Birth weight centile16 ± 163 ± 2< 0.01
IUGR
 ≥ 5th and < 10th centile1 (12.5)1 (6.7) 
 < 5th centile7 (87.5)14 (93.3) 

Umbilical artery pulsatility indices (PI) were significantly higher in women with IUGR than in those in the pre-eclamptic group (P < 0.05), indicating the severity of the alterations of fetoplacental circulation, whereas PI of the uterine arteries (mean of both uterine arteries) were not statistically significantly different (Table 2).

Table 2. Doppler parameters of the study population (mean ± SD)
VesselPre-eclampsia groupIUGR groupPReference PI (32–35 weeks)*30
  • *

    Mean (10th and 90th centiles).

  • Mean of both uterine arteries.

  • Absent or reversed end-diastolic flow was set to a value of 2. IUGR, intrauterine growth restriction; NS, not significant; PI, pulsatility index.

Uterine artery PI1.23 ± 0.72(n = 14)1.68 ± 0.78(n = 13)NS0.45 (0.25–0.74)
Umbilical artery PI1.29 ± 0.35(n = 15)1.76 ± 0.55(n = 15)< 0.051.00 (0.71–1.32)

The levels of angiogenic growth factors are shown in Table 3. In maternal serum, levels of sFlt-1 were significantly higher in women with pre-eclampsia (11 529.97 ± 4289.39 pg/mL) than in those in the IUGR group (4479.17 ± 2633.21 pg/mL) (P < 0.0001). Umbilical vein PlGF levels were below the manufacturer's minimal detectable dose in nearly all samples of IUGR fetuses (1.90 ± 3.37 pg/mL) and lower than in the women with pre-eclampsia (8.81 ± 5.80 pg/mL) (P < 0.001). PlGF levels in umbilical artery serum were lower in women with pre-eclampsia (11.73 ± 6.18) compared to in IUGR (17.45 ± 40.58) (P < 0.05). All other factors were not statistically significantly different in either of the compartments (Table 3).

Table 3. Maternal and fetal serum levels of angiogenic growth factors in women with pre-eclampsia and isolated intrauterine growth restriction (IUGR)
ParameterPre-eclampsia (n = 16)IUGR (n = 15)P
  1. Values are given as mean ± SD. bFGF, basic fibroblast growth factor; PlGF, placental growth factor; sFlt-1, soluble fms-like tyrosine kinase-1; sKDR, soluble kinase insert domain receptor; NS, not significant; UA, umbilical artery serum; UV, umbilical vein serum; VEGF, vascular endothelial growth factor.

VEGF (pg/mL)
 MaternalNot detectable 
 UV490.98 ± 524.32448.49 ± 504.02NS
 UA413.47 ± 480.66672.63 ± 615.23NS
PlGF (pg/mL)
 Maternal62.92 ± 31.4048.44 ± 41.63NS
 UV8.81 ± 5.801.90 ± 3.37< 0.001
 UA11.73 ± 6.1817.45 ± 40.58< 0.05
sFlt-1 (pg/mL)
 Maternal11 529.97 ± 4289.394479.17 ± 2633.21< 0.0001
 UV278.14 ± 476.34624.71 ± 804.08NS
 UA238.16 ± 172.43546.49 ± 844.78NS
sKDR (pg/mL)
 Maternal3962.25 ± 1098.864726.47 ± 1338.12NS
 UV5602.94 ± 1416.254666.07 ± 1273.55NS
 UANot measured 
bFGF (pg/mL)
 Maternal7.17 ± 8.8812.14 ± 11.92NS
 UV10.86 ± 4.6824.20 ± 22.69NS
 UA29.64 ± 19.1845.04 ± 64.16NS

Comparing pre-eclamptic pregnancies with associated IUGR (n = 8) with pre-eclamptic pregnancies without IUGR (n = 8), no statistically significant differences could be detected (Table 4), although some factors did show decreasing or increasing trends in serum levels from pre-eclampsia without IUGR to pre-eclampsia with IUGR to isolated IUGR (Figure 1).

Figure 1.

Trends (mean ± SD) in angiogenic growth factors in different serum samples in women with isolated intrauterine growth restriction (IUGR), pre-eclamptic women with associated IUGR (PE with IUGR) and pre-eclamptic women without IUGR (PE without IUGR). (a) Maternal serum (▴, sFlt-1 (soluble fms-like tyrosine kinase-1); ▪, PlGF (placental growth factor)); (b) umbilical artery serum (▴, VEGF (vascular endothelial growth factor); ▪, bFGF (basic fibroblast growth factor)); (c) umbilical vein serum (▴, sKDR (soluble kinase insert domain receptor); ▪, sFlt-1).

Table 4. Maternal and fetal serum levels of angiogenic growth factors in women with pre-eclampsia (PE) with and without intrauterine growth restriction (IUGR)
ParameterPE with IUGR (n = 8)PE without IUGR (n = 8)P
  1. Values are given as mean ± SD. bFGF, basic fibroblast growth factor; PlGF, placental growth factor; sFlt-1, soluble fms-like tyrosine kinase-1; sKDR, soluble kinase insert domain receptor; NS, not significant; UA, umbilical artery serum; UV, umbilical vein serum; VEGF, vascular endothelial growth factor.

VEGF (pg/mL)
 MaternalNot detectable 
 UV625.96 ± 706.47356.00 ± 219.86NS
 UA535.43 ± 614.05291.52 ± 289.87NS
PlGF (pg/mL)
 Maternal58.14 ± 32.9367.70 ± 31.23NS
 UV8.61 ± 8.409.01 ± 1.25NS
 UA11.06 ± 7.7612.40 ± 4.65NS
sFlt-1 (pg/mL)
 Maternal11 359.69 ± 4293.7711 700.25 ± 4574.22NS
 UV409.70 ± 664.99146.58 ± 66.58NS
 UA254.33 ± 154.06224.02 ± 196.55NS
sKDR (pg/mL)
 Maternal3663.13 ± 1375.424261.38 ± 700.86NS
 UV5063.63 ± 795.546142.25 ± 1732.16NS
 UANot measured 
bFGF (pg/mL)
 Maternal4.66 ± 4.869.71 ± 11.44NS
 UV9.93 ± 4.7811.79 ± 4.72NS
 UA34.58 ± 20.8024.70 ± 17.31NS

Table 5 summarizes the correlation of angiogenic growth factors in maternal or fetal blood with Doppler parameters; no correlations could be found in umbilical artery serum for all growth factors and for VEGF in all compartments. We found a significant inverse correlation of mean uterine artery PI with maternal PlGF levels (Pearson's correlation coefficient, r = − 0.544, P = 0.003) and umbilical vein sKDR levels (r = − 0.385, P < 0.05). In the maternal umbilical artery PI was inversely correlated with PlGF only (r = − 0.657, P < 0.0001), whereas umbilical artery PI was significantly correlated with all growth factors in the umbilical vein serum except VEGF (Table 5, Figures 2 and 3).

Figure 2.

Correlation of placental growth factor (PlGF) in maternal serum with pulsatility indices (PIs) in the uterine (a, r2 linear = 0.296) and umbilical (b, r2 linear = 0.436) arteries. Horizontal rule represents the manufacturer's detection limit of PlGF (9 pg/mL).

Figure 3.

Correlation of soluble vascular endothelial growth factor receptors in umbilical vein serum with umbilical artery pulsatility index (PI). (a) Positive correlation of sFlt-1 (Spearman r = 0.500); (b) negative correlation of sKDR (r2 linear = 0.314). Horizontal rule represents the manufacturer's detection limit of sFlt-1 (5.01 pg/mL); sFlt-1, soluble fms-like tyrosine kinase-1; sKDR, soluble kinase insert domain receptor.

Table 5. Correlation of Doppler parameters with angiogenic growth factors (Pearson's correlation coefficient)
Growth factorUmbilical artery pulsatility indexUterine artery pulsatility index
MaternalUVMaternalUV
  • *

    Spearman correlation. bFGF, basic fibroblast growth factor; PlGF, placental growth factor; sFlt-1, soluble fms-like tyrosine kinase-1; sKDR, soluble kinase insert domain receptor; NS, not significant; UA, umbilical artery serum; UV, umbilical vein serum; VEGF, vascular endothelial growth factor.

VEGFNSNSNSNS
PlGFr = − 0.657, P < 0.0001r = − 0.466, P < 0.01*r = − 0.544, P = 0.003NS*
sFlt-1NS*r = 0.50, P < 0.005*NS*NS*
sKDRNSr = − 0.552, P < 0.005NSr = − 0.385, P < 0.05
bFGFNSr = 0.425, P < 0.05NSNS

Discussion

Angiogenic growth factors play a crucial role in the normal development of the placenta and are altered in pregnancy disorders such as pre-eclampsia and IUGR7–11, 13–17, 23. The activity of these molecules must be carefully orchestrated in the developing placenta in order to form a functioning vascular network. We observed significant correlations between the growth factor levels and Doppler ultrasound parameters indicating the severity of the disorders.

Currently, the severity of both disorders is mainly estimated by time of onset of the disease (early- or late-onset) or by maternal clinical parameters such as blood pressure or estimated fetal weight or weight centile. Definitions of severity vary within the literature depending on the focus of the study as well the fact that pathophysiologic mechanisms in early- and late-onset cases of the diseases may be different. Maternal symptoms such as blood pressure and biochemical markers reflect the severity of the disease for the mother, but their use for estimating fetal well-being is limited. In IUGR additional parameters such as umbilical artery Doppler ultrasonography to measure fetoplacental circulation are used to estimate the severity of the disease in the fetus. This is especially important in early-onset complications. In these cases, delivery of the fetus results in severe prematurity with its associated problems. The current goal is to prolong complicated pregnancies without any harm to mother and fetus.

Levels of angiogenic growth factors have been reported to be correlated with various clinical markers10, 15, 26, but to our knowledge, this is the first study to correlate maternal and fetal levels of angiogenic growth factors with uterine and umbilical artery Doppler PIs.

Our results demonstrate higher maternal sFlt-1 levels in women with pre-eclampsia, indicating a different or an additional pathophysiologic pathway in pre-eclampsia. This is in accord with the results published by Shibata et al., who also found higher sFlt-1 levels in women with pre-eclampsia compared with those with SGA pregnancies16. Owing to the study design—we wanted to investigate maternal and fetal levels—we were not able to include controls matched for gestational age and therefore we do not have sFlt-1 levels from uncomplicated pregnant women. Nevertheless comparing the levels provided in the literature for normal pregnancies at a comparable gestational age10, the levels in our IUGR patients seem to be higher.

Umbilical vein PlGF levels were below the detection limit in nearly all the IUGR fetuses (1.90 ± 3.37 pg/ml) and lower compared to women with pre-eclampsia (8.81 ± 5.80 pg/ml) (P < 0.001). Staff et al. also reported fetal PlGF levels below the detection limit14. As PlGF originates mostly from the placenta, beside the increased sFlt-1 levels, the reduced placental size observed, especially in IUGR, could be responsible for the diminished PlGF levels in maternal and fetal blood.

Additionally in umbilical vein serum, bFGF levels tended to be lower in women suffering from pre-eclampsia compared with isolated IUGR pregnancies. This may also reflect the more pronounced placental pathology in IUGR: bFGF can be secreted via a non-classical pathway such as cell death27, which occurs under hypoxic conditions such as IUGR28.

We did not find any significant difference between pre-eclamptic women with associated IUGR and those without IUGR, therefore the associated IUGR seems not to account for the differences between patients with pre-eclampsia and IUGR. As some factors did show decreasing or increasing trends in serum levels from pre-eclampsia without IUGR to pre-eclampsia with IUGR to isolated IUGR, it may be possible that owing to the small sample size, these differences could have been missed.

We observed significant correlations between maternal and fetal angiogenic growth factor serum levels and Doppler ultrasound indices of the uterine and umbilical arteries in pre-eclampsia and IUGR. These observations reflect the severity of the disorders, especially for the fetus.

In maternal blood we found a significant inverse correlation of PlGF serum levels with umbilical and uterine artery PI. Especially in pre-eclampsia, uterine artery Doppler waveforms are known to show a typical pattern indicative of the alteration of uteroplacental blood flow24. As the failed dilatation of the spiral arteries gets more pronounced, placentation, placental angiogenesis and placental perfusion are further reduced. In addition to an increased secretion of sFlt-1, which may bind free PlGF, reduced uteroplacental ischemia might have down-regulated the PlGF protein expression and production. As the alteration in placentation worsens the blood supply to the fetus, and ultimately fetal weight, is reduced, as can be shown by an increased resistance in the umbilical arteries measured by umbilical artery Doppler ultrasonography.

Levels of the soluble VEGF receptors sFlt-1 and sKDR in umbilical vein serum were also significantly correlated with the PI of the umbilical artery. As sFlt-1 secretion from the placenta into the umbilical vein increases, umbilical artery PI increases, whereas levels of sKDR were inversely correlated with Doppler measurements. The positive correlation for sFlt-1 again reflects the severity of the placental pathology and therefore the severity of the altered blood supply to and nutrition of the fetus. The negative correlation of sKDR with Doppler parameters may be explained by the fact that sKDR is mainly expressed by endothelial cells29. With a more severe placental pathology, beside reduced placental weight, endothelial dysfunction may be more pronounced and therefore less sKDR may be produced.

No differences of angiogenic growth factors and no correlation with Doppler parameters could be found in umbilical artery blood, which drains from the fetus to the placenta, therefore a fetal contribution to the pathophysiology of both disorders can be considered negligible.

Conclusions

The correlation of angiogenic growth factors with Doppler ultrasound parameters, not only parameters indicative of fetal well-being but also parameters found to be useful for screening tools such as uterine artery Doppler, may be interesting as the basis for future screening methods. Early prediction of pregnancy complications is crucial for the effectiveness of any treatment. Estimates of angiogenic factors, especially PlGF and sFlt-1 levels in pregnant women, might be used to screen not only for pre-eclampsia and/or IUGR but also for the most severe form to identify a high-risk group that might benefit from prophylactic treatment. By combining several screening tools such as uterine artery Doppler with biochemical markers, identification of pregnancies destined for early onset problems may be possible.

Acknowledgements

This work was supported and funded by the Wilhelm Sander-Stiftung, Munich, Germany (No. 2000.030.001).

Ancillary