International Journal of Gynecology and Obstetrics (2005) 89, 251 — 257

www.elsevier.com/locate/ijgo

CLINICAL ARTICLE

Prediction of preeclampsia with maternal mid-trimester placental growth factor, activin A, fibronectin and uterine artery Doppler velocimetry R. MadazliT, B. Kuseyrioglu, H. Uzun, S. Uludag, V. Ocak Department of Obstetrics and Gynecology, Cerrahpasa Medical Faculty, University of Istanbul, 7-8. KVsVm, ¨y, Istanbul, Turkey L1-D, D:30, 34750 Atako Received 5 September 2004; accepted 24 February 2005

KEYWORDS Preeclampsia; Placental growth factor; Activin A; Fibronectin; Uterine artery Doppler

Abstract Objective: To assess whether alterations in the serum levels of placental growth factor, activin A and fibronectin could be detected in patients before they become preeclamptic. And to evaluate and compare the clinical utility of these markers and uterine artery Doppler velocimetry in predicting preeclampsia. Methods: One hundred and twenty-two normotensive women were prospectively included in the study. Peripheral venous blood samples were obtained and Doppler examination of the uterine arteries was performed between 21 and 26 weeks’ gestation. Serum levels of placental growth factor, activin A and fibronectin were measured by enzyme-linked immunoassay and radial immundiffusion technic. Results: Mid-trimester maternal serum activin A and fibronectin levels and average S / D ratios were significantly higher whereas placental growth factor levels were significantly lower in women who subsequently developed preeclampsia than remained normotensive (p b 0.001). The best cut-off values for predicting preeclampsia of placental growth factor, activin A and fibronectin based on ROC curve analysis were 90 pg/ml, 14 ng/ml and 370 mg/l respectively. The areas under the curve equal to 0.993, 0.972, 0.872 and 0.813 for placental growth factor, activin A, fibronectin and uterine artery Doppler respectively were determined for the prediction of preeclampsia. Conclusion: Placental growth factor, activin A, fibronectin and uterine artery Doppler are all potentially useful as predictors of preeclampsia. Maternal serum midtrimester PGF has the highest

T Corresponding author. Tel.: +90 212 559 22 11. E-mail address: [email protected] (R. Madazli). 0020-7292/$ - see front matter D 2005 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved. doi:10.1016/j.ijgo.2005.02.008

252

R. Madazli et al. predictive value and activin A with a compatible accuracy for early identification of preeclampsia. D 2005 International Federation of Gynecology and Obstetrics. Published by Elsevier Ireland Ltd. All rights reserved.

1. Introduction Preeclampsia is one of the leading causes of maternal and perinatal morbidity and mortality. Although the exact pathogenesis of preeclampsia still remains to be unravelled and is most likely multifactorial; it is increasingly clear that pathological processes at the interface of the fetal and maternal circulation leading to generalized endothelial cell dysfunction contribute to the spectrum of the disease [1]. The signs and symptoms of the disease are well recognised and they characteristically manifest in the second to third trimester, although the underlying pathology is present at earlier stages of pregnancy. The identification of reliable screening markers that could predict the subsequent onset of preeclampsia before maternal clinical manifestations become apparent would permit major improvements in obstetric care through better targeting of antepartum surveillance. The association between abnormal placentation and preeclampsia is well known and is thought to involve inadequate trophoblast invasion of maternal spiral arteries during early gestation [2,3]. The uteroplacental circulation can be assessed by Doppler ultrasound of the uterine arteries. Over the last 25 years a number of Doppler ultrasound studies of the uteroplacental circulation have confirmed that increased impedance to flow of the uterine arteries is associated with an increased risk for subsequent development of preeclampsia and/or fetal growth restriction [4,5]. Evidence of placental derangement may be reflected in maternal circulation by alterations in the concentration of biochemical markers involved in the process of normal placental development. Placental growth factor (PGF), member of the platelet-derived growth factor family, is an important local mediator of angiogenesis in the human placenta [6]. Recent evidence has shown that the presence and function of PGF is critical during cytotrophoblastic invasion and plays a central role within the placentation process [7]. Several studies have shown that the concentration of PGF is dramatically decreased in the plasma of preeclamptic women [8,9]. Activin A is a dimeric protein produced primarily by the placental troph-

oblast during pregnancy [10]. It has been demonstrated that activin A is abnormally elevated in preeclamptic patients [11,12]. Fibronectin is a glycoprotein synthesized in the endothelial cell and its elevated level may be an indicator of endothelial damage. The association of elevated plasma fibronectin in preeclampsia have been documented in previous studies indicating an endothelial cell dysfunction [13,14]. This study was performed to assess whether alterations in the serum levels of PGF, activin A and fibronectin could be detected in patients before they become preeclamptic. And to evaluate and compare the clinical utility of these markers and uterine artery Doppler velocimetry in predicting preeclampsia.

2. Material and methods 2.1. Patient selection Informed consent for the investigations described herein was obtained from all women. Approval for the study was given by the Ethics Committee of our hospital. A total of 122 normotensive women, seen for their antenatal care at our department were prospectively included in the study. Blood pressure was measured in the sitting position. Diastolic pressure was defined as the pressure at which a Korotkoff phase V was obtained. Blood pressure was monitored monthly throughout pregnancy. Normal pregnancy was defined as a unifetal pregnancy that occurred in a normotensive woman without proteinuria, who remained normotensive and delivered after 36 weeks of pregnancy, a healthy neonate with a weight adequate for gestational age. Preeclampsia was defined as the onset of hypertension (blood pressure of 140 / 90 mm Hg or greater, 6 h or more apart) and consistent proteinuria (300 mg/day or more) during the latter half of the pregnancy with both remitting remotely after delivery. The diagnostic criteria for severe preeclampsia were: (1) blood pressure z 160 mm Hg systolic or z 110 mm Hg diastolic on two occasions greater than 6 h apart; (2) significant proteinuria (protein excretion z 3 + on a dipstick random sample); and (3) oliguria (urinary output b 400

Preeclampsia prediction using markers and uterine artery doppler velocimetry ml / 24 h). In addition, any patient with cerebral or visual disturbances, pulmonary oedema or cyanosis, impaired liver function, or thrombocytopenia (platelet count less than 100,000/ml) was included in severe preeclampsia.

2.2. Doppler study All Doppler examinations of the uterine arteries were performed by the first author (R.M.) between 21 and 26 weeks’ gestation. An Ultramark 9 (Advanced Technologies Laboratory, Bothell, WA, USA) with a 3.5-MHz color-pulsed Doppler, was used. By placing the transducer in the lower lateral quadrant of the uterus, the hypercrossover of the external iliac artery and the main uterine artery was identified. High-pass filter was set at 100 Hz. Doppler waveform was obtained from each uterine artery. When three consecutive waveforms of a satisfactory quality were obtained, the S / D ratio (S, peak systolic velocity; D, least diastolic velocity) was measured for each waveform. The mean value of three waveforms was used for subsequent analysis. Abnormal uterine velocimetry was defined as an average of (left and right) S / D ratio N 2.6 [15].

2.3. Blood samples and PGF, activin A and fibronectin measurements Single peripheral venous blood samples, obtained between 21 and 26 weeks’ gestation, were collected into heparinized vacutainer tubes. Plasma samples were isolated by 2500 g centrifugation at 4 8C for 10 min. The samples were stored at 70 8C until the analyses were performed. For measurement of plasma PGF and activin A commercially available enzyme immunoassays were used. Placental growth factor (Research and Diagnostics Systems, Minneapolis, USA) and activin A (Serotec, Oxford, UK). Analyses of all samples, standards and controls were run in duplicate. The minimal detectable levels were for PGF 7 pg/ml and activin A 0.14 ng/ml. Plasma levels of fibronectin were detected by radial immundiffusion technic using LC Partigen Fibronectin (Behringwerke AG, Marburg, Germany).

2.4. Statistical analysis The data were tested for normal distribution and Student’s t-test or chi square test was used for comparison of two groups where appropriate. A receiver—operator characteristic (ROC) curve was constructed and the area under the curve (AUC) was used to compare the performance of various

253

prediction tests. The optimal cut-off point of testing values was selected as the point on the curve that had the shortest distance to the top-left corner (sensitivity = 1 and false positive = 0) of the graph. The sensitivity, specificity, positive predictive value, negative predictive value and likelihood ratio with 95% confidence intervals in the prediction of preeclampsia were calculated.

3. Results In Table 1, the clinical characteristics of the patient populations are shown. Of the 122 pregnancies, 14 (11.4%) developed preeclampsia during follow-up, of which 5 were severe preeclampsia. No difference was found in the average maternal age, number of primigravidae and gestational age at sampling between the groups. The highest diastolic blood pressure, the average gestational age at delivery and birthweight were found to be significantly different between the preeclampsia and normal pregnancy groups ( p b 0.001). The mean gestational age at delivery and birth weight of severe preeclamptic patients were 33.2 F 4.2 weeks and 1586 F 844 g respectively. Maternal serum midtrimester PGF, activin A and fibronectin levels and an average (left and right) S / D ratio of uterine arteries of women who subsequently developed preeclampsia and remained normotensive are shown in Table 2. Maternal serum activin A and fibronectin levels and average S / D ratios were significantly higher whereas PGF levels were significantly lower in women who subsequently developed preeclampsia than remained normotensive ( p b 0.001). The raw data, including individual factor concentrations Table 1 Clinical characteristics of the patient population of pregnant women (data are presented as mean F standard deviation) Normal pregnancy

Preeclampsia

N 108 14 Age (years) 28.5 F 5.1 29.1 F 4.5 Number of primigravidae 66 9 Gestational age at sampling 23.4 F 1.7 23.9 F 1.5 (weeks) Diastolic blood pressure at 67.1 F 9.7 70.3 F 11.3 sampling (mm Hg) Highest diastolic blood pressure 71.7 F 10.1 103.1 F 13.5T during pregnancy (mm Hg) Gestational age at delivery 38.4 F 1.4 34.6 F 3.1T (weeks) Birth weight (g) 3327 F 482 2085 F 821T T p b 0.001 compared with the normal pregnancy group.

254

R. Madazli et al.

Normal pregnancy (N = 108)

Preeclampsia (N = 14)

Placental growth factor 142.51 F 30.82 69.43 F 27.31T (pg/ml) Activin A (ng/ml) 8.56 F 3.86 26.50 F 10.37T Fibronectin (mg/l) 306.59 F 50.5 380.56 F 43.75T Average (left and right) S / D 2.41 F 0.83 3.35 F 0.93T ratio of uterine arteries T p b 0.001 compared with the normal pregnancy group.

blast and endothelial cells [18]. Accordingly, PGF has been found to directly affect trophoblast proliferation, some aspects of differentiation, and to protect the cells from apoptosis [7]. These

a 220 200 180 160

PGF(pg/ml)

Table 2 Maternal serum placental growth factor, activin A and fibronectin levels and average S / D ratios of uterine arteries of women who developed preeclampsia and remained normotensive (data are presented as mean F standard deviation)

140 120 100 80 60

Preeclamptic Normotensive

40 20

21

22

23

24

25

26

27

Weeks of gestation

b 50

40

Activin A (ng/ml)

and gestational age at the time each sample was obtained, are shown for maternal serum placental growth factor, activin A and fibronectin (Fig. 1a,b,c). The sensitivities and false-positive rates (receiver operating characteristic curve—ROC) for the detection of preeclampsia by measurement of maternal serum PGF, activin A and fibronectin concentrations and abnormal uterine artery velocimetry are shown in Fig. 2. The areas under the curve equal to 0.993, 0.972, 0.872 and 0.813 for PGF, activin A, fibronectin and uterine artery Doppler respectively were determined for the prediction of preeclampsia. The best cut-off values for predicting preeclampsia of PGF, activin A and fibronectin based on ROC curve analysis were 90 pg/ml, 14 ng/ml and 370 mg/l respectively. Screening characteristics for preeclampsia and severe preeclampsia for PGF, activin A, fibronectin and uterine artery Doppler are shown in Table 3.

30

20

10 Preeclamptic Normotensive

0 20

21

22

23

24

25

26

27

Weeks of gestation

c

500

Poor placental perfusion and other pathologic changes in the placenta such as defective trophoblast invasion of spiral arteries characterize preeclampsia and form the basis for examination of placental serum markers such as activin A and PGF in that disease. Activins act as both endocrine and paracrine regulators and have a role in the regulation of cell differentiation and tissue remodelling, for instance, during wound repair and in trophoblast invasion during establishment of the placental bed [16]. Activin A is secreted by trophoblast cells of placenta and may have a role in appropriate control of trophoblast invasion and remodelling of the vasculature in the placental bed [17]. PGF is expressed prominently in normal placenta, and its receptors are found on tropho-

Fibronectin (mg/L)

4. Discussion 400

300

Preeclamptic Normotensive

200 20

21

22

23

24

25

26

27

Weeks of gestation

Figure 1 Maternal serum concentrations of placental growth factor (1a), activin A (1b) and fibronectin (1c) collected between 21 and 26 weeks’ gestation. Black circles—preeclamptic patients; Cross—normotensive patients; Solid line—The best cut-off values for each individual marker.

Preeclampsia prediction using markers and uterine artery doppler velocimetry 1,00

Sensitivity

,75

,50

PGF ,25 Activin A Fibronectin 0,00 0,00

UA S/D ,25

,50

,75

1,00

1 - Specificity

Figure 2 Receiver operating characteristic curves showing the prediction performance of placental growth factor, activin A, fibronectin and uterine artery Doppler for preeclampsia.

autocrine effects of PGF imply that it may play an important regulatory role in initiation, growth, and maintenance of the placenta throughout pregnancy. Several studies have confirmed that plasma levels of placenta growth factor are decreased [8,19] and activin A are increased [11,12] in pregnancies complicated by preeclampsia. Alterations in the concentrations of placental markers of poor placentation may precede the clinical onset of preeclampsia. Indeed, as shown in the present study, significantly decreased and increased plasma levels of placenta growth factor and activin A respectively were found in the second trimester (weeks 21—26) in patients who developed preeclampsia. Several other studies have also shown that second trimester plasma levels of placental growth factor are decreased [20,21] and activin A are increased [17,22] in pregnancies destined to become preeclamptic.

255

Fibronectin is synthesized in the endothelial cell and its elevated level may be an indicator of endothelial damage. In the present study increased second trimester plasma levels of fibronectin in patients who later developed preeclampsia has been observed. Fibronectin as a valuable predictor of subsequent preeclampsia has been demonstrated by other investigators [14,23]. Increased plasma fibronectin levels, even before the clinical onset of preeclampsia, confirms the concept of preeclampsia as an endothelial cell dysfunction disease. Doppler ultrasound enables the assessment of blood flow parameters for the adequate and reduced perfusion in vivo. Impaired trophoblastic invasion of the maternal spiral arteries is shown to be associated with increased impedance to flow in the waveforms obtained by Doppler ultrasound examination of the uterine arteries [24,25]. Bilateral high-resistance flow velocity waveforms with early diastolic notches at 22—26 weeks of gestation are associated with subsequent fetal death, IUGR and preeclampsia [5]. High mean uterine artery S / D ratios in pregnancies destined to be preeclamptic has also been demonstrated by the present study. The conversion of spiral arteries into uteroplacental arteries plays a basic role in the establishment of the physiological placental blood supply and the reduced uteroplacental blood flow due to inadequate trophoblastic infiltration of the placental vascular bed can be the cause of a variety of pregnancy complications. Based on ROC curves in the current study, PGF, activin A, fibronectin and uterine artery Doppler were all potentially clinically useful tests for prediction of preeclampsia. Maternal serum midtrimester PGF had the highest predictive value (AUC = 0.993) and activin A with a compatible

Table 3 Screening characteristics for placental growth factor, activin A, fibronectin and mean uterine artery S / D for predicting preeclampsia and severe preeclampsia Cut-off value Placental growth factor Preeclampsia Severe preeclampsia Activin A Preeclampsia Severe preeclampsia Fibronectin Preeclampsia Severe preeclampsia Uterine artery Doppler Preeclampsia Severe preeclampsia

Sensitivity (%)

Specificity (%)

PPV (%)

NPV (%)

LR

95% CI

92.9 100

94.4 88.1

68.4 26.3

99.1 100

221

24—1983

92.8 100

91.6 85.4

59.1 22.7

99 100

143

16—1221

64.2 80

86.1 82.9

37.5 16.6

94.8 98.9

71.4 100

77.6 74.4

34.5 17.2

94.3 100

90 pg/ml

14 ng/ml

370 mg/l 11.1

3.2—37

11

2.7—43

Mean S / D N 2.6

PPV, positive predictive value; NPV, negative predictive value; LR, likelihood ratio; CI, confidence interval. (The figures are calculated for the 11.4% and 4% incidence of preeclampsia and severe preeclampsia observed in our study group).

256 accuracy (AUC = 0.972) for early identification of women at increased risk for the development of preeclampsia. Fibronectin and uterine artery Doppler was found to be slightly less effective than both mentioned. Best to our knowledge, this is the first study to compare the predictive performance of these tests for preeclampsia. Our study group is a selected high-risk population, not representative of an unselected screened population. The incidence of previous preeclampsia and bad obstetric history in our group was 9% and 8.2% respectively. Therefore the incidence of preeclampsia detected in our group (11.4%) is higher than the 5—7% prevalence of preeclampsia in the normal population [1]. The screening performance of the tests may be high with the high prevalence of preeclampsia in our population. The population selection is, however, unlikely to have affected the relative performance of the tests compared to one another. Midtrimester PGF, activin A, fibronectin and uterine artery Doppler were found to detect women at increased risk of preeclampsia with a sensitivity of 93%, 93%, 64% and 71% and a specificity of 94%, 92%, 86% and 78% respectively. Moreover, all of the women with severe preeclampsia were correctly identified with midtrimester PGF, activin A and uterine artery Doppler with the sensitivities of 100%. From the present study, it has been demonstrated that potentially effective and clinically applicable predicting tests for preeclampsia at least in a high-risk population is possible.

References [1] Roberts JM, Cooper DW. Pathogenesis and genetics of preeclampsia. Lancet 2001;357:53 – 6. [2] Madazli R, Budak E, Calay Z, Aksu MF. Correlation between placental bed biopsy findings, vascular cell adhesion molecule (VCAM-1) and fibronectin levels in preeclampsia. Br J Obstet Gynaecol 2000;107:514 – 8. [3] Meekins JW, Pijnenborg R, Hanssens M, McFayden IR, van Asshe AA. Study of placental bed spiral arteries and trophoblast invasion in normal and severe preeclamptic pregnancies. Br J Obstet Gynaecol 1994;101:669 – 74. [4] Chien PF, Arnott N, Gordon A, Owen P, Khon K. How useful is uterine artery Doppler flow velocimetry in the prediction of pre-eclampsia, intrauterine growth retardation and perinatal death? An overview. Br J Obstet Gynaecol 2000;107: 196 – 208. [5] Papageorghiou AT, Yu CKH, Bindra R, Nicolaides KH. Multicenter screening for pre-eclampsia and fetal growth restriction by transvaginal uterine artery Doppler at 23 weeks of gestation. Ultrasound Obstet Gynecol 2001;18: 441 – 9. [6] Ong S, Lash G, Baker PN. Angiogenesis and placental growth in normal and compromised pregnancies. Bailliere’s Best Pract Res, Clin Obstet Gynaecol 2000;14:969 – 80.

R. Madazli et al. [7] Torry DS, Mukherjea D, Arroyo J, Torry RJ. Expression and function of placenta growth factor: implications for abnormal placentation. J Soc Gynecol Investig 2003;10:178 – 88. [8] Madazli R, AydVn S, Uludag S, Ocak V, Tolun N. Maternal plasma levels of cytokines in normal and preeclamptic pregnancies and their relationship with diastolic blood pressure and fibronectin levels. Acta Obstet Gynecol Scand 2003;83:1 – 6. [9] Livingston JC, Chin R, Haddad B, McKinney ET, Ahokas R, Sibai BM. Reductions of vascular endothelial growth factor and placental growth factor concentrations in severe preeclampsia. Am J Obstet Gynecol 2000;183:1554 – 7. [10] Qu J, Thomas K. Inhibin and activin production in the human placenta. Endocr Rev 1995;16:485 – 507. [11] Silver HM, Lambert-Messerlian GM, Star JA, Hogan J, Canick JA. Comparison of maternal serum total activin A and inhibin A in normal, preeclamptic, and nonproteinuric gestationally hypertensive pregnancies. Am J Obstet Gynecol 1999;180:1131 – 7. [12] Keelan JA, Taylor R, Schellenberg JC, Groome NP, Mitchell MD, Robyn A, et al. Serum activin A, inhibin A, and follistatin concentrations in preeclampsia or small for gestational age pregnancies. Obstet Gynecol 2002;99:267 – 74. [13] Eriksen HO, Hansen PK, Brocks V, Jensen BA. Plasma fibronectin concentration in normal pregnancy and preeclampsia. Acta Obstet Gynecol Scand 1987;66:25 – 8. [14] Sen C, Madazli R, Kavuzlu C, Ocak V, Tolun N. The value of antithrombin-III and fibronectin in hypertensive disorders of pregnancy. J Perinat Med 1994;22:29 – 38. [15] Aquilina J, Barnett A, Thompson O, Harrington K. Comprehensive analysis of uterine artery flow velocity waveforms for the prediction of pre-eclampsia. Ultrasound Obstet Gynecol 2000;16:163 – 70. [16] Caniggia I, Lye SJ, Cross JC. Activin is a local regulator of human cytotrophoblast cell differentiation. Endocrinology 1997;138:3976 – 86. [17] Davidson EJ, Rileya SC, Roberts SA, Shearing CH, Groomed NP, Martina CW. Maternal serum activin, inhibin, human chorionic gonadotrophin and a-fetoprotein as second trimester predictors of pre-eclampsia. Br J Obstet Gynaecol 2003;110:46 – 52. [18] Shore VH, Wang TH, Wang CL, Torry RJ, Caudle MR, Torry DS. Vascular endothelial growth factor, placenta growth factor and their receptors in isolated human trophoblast. Placenta 1997;18:657 – 65. [19] Reuvekamp A, Velsing-Aarts FV, Poulina IEJ, Capello JJ, Duits AJ. Selective deficit of angiogenic growth factors characterises pregnancies complicated by preeclampsia. Br J Obstet Gynaecol 1999;106:1019 – 22. [20] Polliotti BM, Gordon FA, Saller DN, Mooney RA, Cox C, Miller RK. Second-trimester maternal serum placental growth factor and vascular endothelial growth factor for predicting severe, early-onset preeclampsia developing severe, early-onset preeclampsia. Obstet Gynecol 2003; 101:1266 – 74. [21] Tjoa ML, van Vugt JMG, Mulders MAM, Schutgens RBH, Oudejans CBM, van Wijk IJ. Plasma placenta growth factor levels in midtrimester pregnancies restriction. Obstet Gynecol 2001;98:600 – 7. [22] Muttukrishna S, North RA, Morris J, Schellenberg JC, Taylor RS, Asselin J, et al. Serum inhibin A and activin A are elevated prior to the onset of pre-eclampsia. Hum Reprod 2000;15:1640 – 5. [23] Dreyfus M, Baldauf JJ, Ritter J, van Cauwenberg JR, Hardy A, Foidart JM. The prediction of preeclampsia: reassessment of clinical value of increased plasma levels of fibronectin. Eur J Obstet Gynecol Reprod Biol 1998;78:25 – 8.

Preeclampsia prediction using markers and uterine artery doppler velocimetry [24] Madazli R, Somunkiran A, Calay Z, Ilvan S, Aksu MF. Histomorphology of the placenta and the placental bed of growth restricted foetuses and correlation with the Doppler velocimetries of the uterine and umbilical arteries. Placenta 2003;24:510 – 6.

257

[25] Aardema MW, Oosterhof H, Timmer A, van Rooy I, Aarnoudse JG. Uterine artery Doppler flow and uteroplacental vascular pathology in normal pregnancies and pregnancies complicated by pre-eclampsia and small for gestational age fetuses. Placenta 2001;22:405 – 11.