1 Department of Obstetrics and Gynaecology, The Chinese University of Hong Kong, Hong Kong, 2 The Mothers and Babies Research Centre, University of Newcastle, John Hunter Hospital, Australia, 3 Department of Chemical Pathology, 4 Centre for Clinical Trials and Epidemiological Research, The Chinese University of Hong Kong, Hong Kong, 5 Discipline of Reproductive Medicine and Department of Obstetrics and Gynaecology, University of Newcastle, John Hunter Hospital, Australia
![]() |
Abstract |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key words: alpha-fetoprotein/corticotrophin releasing hormone/pre-eclampsia/prediction
![]() |
Introduction |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Corticotrophin-releasing hormone (CRH) is synthesized by the placenta during human pregnancy and released into both maternal and fetal circulations. Abnormal elevation of CRH concentrations has been reported in pregnancies complicated by pregnancy induced hypertension (PIH) and pre-eclampsia (PE) (Wolfe et al., 1988; Laatikainen et al., 1991
; Emanuel et al., 1994
; Perkins et al., 1995
; Liapi et al., 1996
). A rise in CRH concentrations has also been demonstrated 11 weeks before the development of signs and symptoms of the disease (Wolfe et al., 1988
) although Petraglia et al. (1996) showed that changes in CRH and CRH binding protein concentrations did not precede the onset of clinical disease. In most of these studies, the `hypertensive' group consisted of a combination of pregnant women with PIH (without proteinuria) and PE (with proteinuria) (Wolfe et al., 1988
; Laatikainen et al., 1991
; Emanuel et al., 1994
; Liapi et al., 1996
; Petraglia et al., 1996
). While PE is a more stringent diagnosis specific to pregnancy, PIH may include a group of women with transient hypertension and its inclusion may lead to a heterogeneous cohort (Roberts and Redman, 1993
). In the current study, data were collected on CRH concentrations from over 1500 Chinese pregnant subjects between 1542 weeks of gestation. Firstly the difference in CRH concentrations in the third trimester between normal pregnant subjects and those with established PE was studied. Then the CRH data from the mid-trimester samples were analysed to test if concentrations were elevated in those who subsequently developed PE. The clinical utility of plasma CRH for use in the prediction of PE was then assessed.
Alpha-fetoprotein (AFP) is a valued screening test for both neural tube defects and biochemical screening for Down's syndrome. Pre-eclampsia has also been associated with an elevation of AFP in maternal serum (Moore and Redman, 1983; Walters et al., 1985
). Using a threshold value of two multiples of the median (MoM), elevated AFP in the mid-trimester has been shown to be associated with a 2.3- to 3.8-fold increased risk of developing pre-eclampsia (Milunsky et al., 1989
; Williams et al., 1992
). The mid-trimester AFP concentrations for over 1000 subjects in this cohort were available for analysis. The performance using AFP alone or in combination with CRH for prediction of pre-eclampsia was also examined.
![]() |
Materials and methods |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Sampling details
Ten ml of venous blood were collected by venepuncture of the antecubital vein of each subject between 15 and 42 weeks gestation and placed immediately into ice-cold lithium heparin tubes. After centrifugation at 3000 g for 20 min at 4°C, the plasma was collected in duplicate aliquots and stored at 70°C until assay for CRH. A total of 1021 subjects recruited between 1522 weeks also had AFP concentrations assayed.
CRH assay
Frozen plasma samples were transported in dry ice to The Mothers and Babies Research Centre, University of Newcastle, Australia for measurement of CRH by radioimmunoassay as previously described (Leung et al., 1999). The mean extraction recovery for the radioimmunoassay was 103% and the intra- and interassay coefficients of variation were 7% and 10% respectively. The threshold for detection was 1.35 pmol/l. The investigators responsible for the laboratory assay were blinded to the obstetric outcome.
AFP assay
AFP concentrations were measured by microparticle enzyme immunoassay (Imx analyzer®; Abbott Laboratories, Abbott Park, IL, USA) in the Department of Chemical Pathology, Prince of Wales Hospital, Shatin, Hong Kong.
Obstetric outcome
All subjects were followed until delivery. The gestational age at delivery, mode of delivery, obstetric complications if any, and neonatal outcomes were recorded. The study subjects and clinical staff responsible for their care were blinded to the CRH results. For those subjects who subsequently delivered in another hospital, the obstetric information was obtained by telephone from the subject or via contact with staff in the other hospitals.
Pre-eclampsia was defined on the basis of a diastolic blood pressure recording greater than 110 mm Hg on one occasion or greater than 90 mm Hg on two or more occasions at least 4 h apart with the presence of significant proteinuria in subjects with no history of hypertension. Significant proteinuria was defined as proteinuria greater than 0.3 g/day or 2+ on dipstix testing in two clean-catch mid-stream urine specimens collected at least 4 h apart.
Statistics
Statistical analyses were performed using the Statistical Package for the Social Sciences Version 9.0 (SPSS Inc., IL, USA). As both CRH and AFP concentrations vary with gestational age, their results were corrected for gestational age by expression as multiples of the median (MoM) of the normal pregnancies at that gestational age to allow legitimate comparison between normal and PE groups. Median concentrations of AFP at different gestational ages for the local population were used for the adjustment of AFP concentrations. The median concentrations of the CRH concentrations at different gestational ages were calculated from this cohort of subjects since such values for the local population were not otherwise available. The difference in the CRH and AFP concentrations in subjects who had PE and normal uncomplicated term pregnancies were examined using the MannWhitney U test. A value of P < 0.05 was considered statistically significant. The ability of mid-trimester CRH or AFP concentrations to predict PE was assessed by receiver operating characteristic (ROC) curves. The sensitivity, specificity, positive predictive value, negative predictive value, and the positive likelihood ratio for a threshold value of 2.0 MoM were then calculated and discussed.
The test performance of combining the mid-trimester CRH and AFP concentrations was also assessed using ROC curves. Using logistic regression, a probability P for developing PE was formulated based on CRH and AFP concentrations {P = 1/[1 + exp(5.81670.3981xCRH1.1019xAFP)]}. The P value for each individual subject was calculated and a ROC curve was then plotted according to the P value in predicting PE. The optimal sensitivity, specificity, positive predictive value, negative predictive value, and the positive likelihood ratio were then reported.
![]() |
Results |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
A summary of the characteristics of the normal and PE groups is shown in Table II. There was no difference in the maternal age between the two groups but a significantly larger proportion of subjects in the PE group were nulliparous. The mean gestational age at delivery was significantly lower for the PE group due to a higher incidence of preterm delivery because of the disease. Babies born to the PE group were significantly lighter and a significantly larger proportion were `small-for-dates', which was defined as birth weight below the 10th centile for the respective gestational age at delivery.
|
|
|
![]() |
Discussion |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The reason for elevation of maternal CRH concentrations in subjects with pre-eclampsia is unknown. In human pregnancy, CRH is of placental origin and is secreted into both maternal and fetal circulations (Majzoub et al., 1999). CRH has been shown to exhibit a vasodilatory effect on the human feto-placental circulation and therefore may be a regulator of placental vascular tone (Clifton et al., 1994
). Peripheral administration of CRH to human subjects has also been shown to cause vasodilatation and drop of blood pressure (Hermus et al., 1987
). In the presence of hypoperfusion of the fetoplacental unit as in the case of PE, the elevation of CRH concentrations could be a protective response to placental ischaemia. Alternatively, the placental CRH may act on the fetal hypothalamic pituitaryadrenal axis to cause increased production of cortisol, which is important for fetal maturation. A positive feedback loop between the fetal hypothalamicpituitaryadrenal axis and placental CRH has been suggested by Robinson et al. (1988). It has been shown that the umbilical-cord plasma cortisol and CRH concentrations were both significantly elevated in neonates born to mothers with pre-eclampsia (Goland et al., 1995
). It has also been reported that pulmonary maturation was more advanced in neonates born to women with pregnancy induced hypertension (Bent et al., 1982
). The current data have shown that subjects with higher mid-trimester CRH concentrations in the PE group tended to require earlier delivery. This was mostly related to earlier onset of disease. It is possible that placental CRH production is elevated early in pregnancies with pre-eclampsia as a protective mechanism better to prepare the fetus for extrauterine life. However, we cound not demonstrate any correlation between maternal CRH concentrations and the severity of hypertension.
Of greater clinical relevance is the utility of the measurement for identification of those at high risk for PE. Although the data in this study showed that CRH measured in the mid-trimester could be used for prediction of PE, the performance was far from satisfactory. When a threshold of 2.0 MoM was used, the sensitivity, specificity, positive predictive value, negative predictive value and positive likelihood ratio were 38.1, 79.9, 3.8, 98.4%, and 1.9 respectively when the background prevalence was 2.1%. The performance of the CRH test in the mid-trimester was worse than when it was used for prediction of spontaneous preterm delivery (Leung et al., 1999).
Maternal plasma AFP concentrations were analysed to see if a combination of AFP and CRH concentrations would improve detection of high-risk subjects. The current data confirmed that mid-trimester maternal AFP concentrations were significantly elevated in subjects who would develop PE in the subsequent course of pregnancies. However, when used alone for prediction, mid-trimester maternal AFP concentrations were not clinically useful. Most of the studies on AFP and PE were case-controlled studies (Moore and Redman, 1983; Walters et al., 1985
; Williams et al., 1992
; Yaron et al., 1999
) and the association of high mid-trimester maternal AFP concentrations with a higher risk of developing PE was well established. However, the clinical utility of the test for the purpose of prediction of PE can only be reflected in a prospective study. Milunsky et al. (1989) reported the AFP concentrations in 13 486 singleton pregnancies and showed that although high AFP was associated with a 2.3-fold increase of developing PE, the performance of the test for predicting PE was poor. The sensitivity, specificity, positive predictive value and negative predictive value were 8.9, 96.0, 3.6 and 98.4% respectively when the background prevalence was 1.7% (Milunsky et al., 1989
). The results of the current study were in agreement with theirs in that the sensitivity and positive predictive value were very low.
The combination of AFP and CRH, however, offered increased detection of at-risk subjects. At the same specificity of 79.9% as when CRH 2.0 MoM was used, the sensitivity was increased to 52.4% from 38.1%. The likelihood ratio was also increased to 2.6 from 1.9. However, this small increase in the likelihood ratio does not suggest the combination would be of great clinical value.
There is recent scientific evidence to suggest that poor placentation is not the only underlying pathology leading to pre-eclampsia (Ness and Roberts, 1996). Rather, the clinical syndrome is suggested to represent the end result of heterogeneous causes which lead to endothelial dysfunction or exaggerated maternal inflammatory response (Ness and Roberts, 1996
; Redman et al., 1999
). Although this remains hypothesis, it could explain why there is so far no early predictor that can accurately detect pre-eclampsia. The current findings are of no exception in this regard.
In summary, significantly elevated maternal CRH concentrations were demonstrated in subjects suffering from PE in the third trimester of pregnancy. This elevation of maternal CRH concentrations could be detected as early as in the mid-trimester before clinical signs of the disease have appeared. However, when used for prediction of pre-eclampsia, neither the CRH nor AFP concentrations alone in the mid-trimester had strong predictive value. Although a combination of both tests improved the predictive performance, the small improvement did not render the testing clinically useful.
![]() |
Acknowledgments |
---|
![]() |
Notes |
---|
![]() |
Reference |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Clifton, V.L., Read, M.A., Leitch, I.M. et al. (1994) Corticotrophin-releasing hormone-induced vasodilation in the human fetal placental circulation. J. Clin. Endocrinol. Metab., 79, 666669.[Abstract]
Emanuel, R.L., Robinson, B.G., Seely, E.W. et al. (1994) Corticotrophin releasing hormone concentrations in human plasma and amniotic fluid during gestation. Clin. Endocrinol., 40, 257262.[ISI][Medline]
Goland, R.S., Tropper, P.J., Warren, W.B. et al. (1995) Concentrations of corticotrophin-releasing hormone in the umbilical-cord blood of pregnancies complicated by pre-eclampsia. Reprod. Fertil. Dev., 7, 12271230.[ISI][Medline]
Hermus, A.R., Pieters, G.F., Willemsen, J.J. et al. (1987) Hypertensive effects of ovine and human CRFs in man. Eur. J. Clin. Pharmacol., 31, 531534.[ISI][Medline]
Hong Kong College of Obstetricians and Gynaecologists (1996) Territory-wide Audit in Obstetrics and Gynaecology. Hong Kong College of Obstetricians and Gynaecologists, Hong Kong.
Laatikainen, T.J., Virtanen, T., Kaaja, R. et al. (1991) Corticotrophin-releasing hormone in maternal and cord plasma in pre-eclampsia. Eur. J. Obstet. Gynecol. Reprod. Biol., 39, 1924.[ISI][Medline]
Leung, T.N., Chung, T.K.H., Madsen, G. et al. (1999) Elevated mid-trimester maternal corticotrophin-releasing hormone concentrations in pregnancies destined to deliver preterm. Br. J. Obstet. Gynaecol., 106, 10411046.[ISI][Medline]
Liapi, C.A., Tsakalia, D.E., Panitsa-Faflia, C.C. et al. (1996) Corticotrophin-releasing hormone concentrations in pregnancy induced hypertension. Eur. J. Obstet. Gynecol., 68, 109114.[ISI][Medline]
Lyail, F. and Greer, J.A. (1994) Pre-eclampsia: a multifaceted vascular disorder of pregnancy. J. Hypertens., 12, 13391345.[ISI][Medline]
Majzoub, J.A., McGregor, J.A., Lockwood, C.J. et al. (1999). A central theory of preterm and term labor: putative role for corticotropin-releasing hormone. Am. J. Obstet. Gynecol., 180, S232241.[ISI][Medline]
Milunsky, A., Jick, S.S., Bruell, C.L. et al. (1989) Predictive values, relative risks, and overall benefits of high and low maternal serum alpha-fetoprotein screening in singleton pregnancies: New epidemiologic data. Am. J. Obstet. Gynecol., 161, 292297.
Moore, M.P. and Redman, C.W.G. (1983) Case-control study of severe pre-eclampsia of early onset. Br. Med. J., 287, 580583.[ISI][Medline]
Ness, R.B. and Roberts, J.M. (1996) Heterogeneous causes constituting the single syndrome of pre-eclampsia: a hypothesis and its implications. Am. J. Obstet. Gynecol., 175, 13651370.[ISI][Medline]
Perkins, A.V., Linton, E.A., Eben, F. et al. (1995) Corticotrophin-releasing hormone and corticotropin-releasing hormone binding protein in normal and pre-eclamptic human pregnancies. Br. J. Obstet. Gynaecol., 102,118122.[ISI][Medline]
Petraglia, F., Florio, P., Benedetto, C. et al. (1996) High concentrations of corticotrophin-releasing factor (CRF) are inversely correlated with low concentrations of maternal CRF-binding protein in pregnant women with pregnancy induced hypertension. J. Clin. Endocrinol. Metab., 81, 852856.[Abstract]
Redman, C.W.G. (1991) Current topic: pre-eclampsia and the placenta. Placenta, 12, 301308.[ISI][Medline]
Redman, C.W.G., Sacks, G.P., Sargent, I.L. (1999) Pre-eclampsia: an excessive maternal inflammatory response to pregnancy. Am. J. Obstet. Gynecol., 180, 499506.[ISI][Medline]
Roberts, J.M. and Redman, C.W.G. (1993) Pre-eclampsia: more than pregnancy-induced hypertension. Lancet, 341, 14471451.[ISI][Medline]
Robertson, W.B., Brosens, L. and Dixon, H.G. (1967) The pathological response of the vessels of the placental bed to hypertensive pregnancy. J. Pathol. Bacteriol., 93, 581592.[ISI][Medline]
Robinson, B.G., Emanuel, R.L., Frim, D.M. et al. (1988) Glucocorticoid stimulates expression of corticotrophin-releasing hormone gene in human placenta. Proc. Natl. Acad. Sci. USA, 85, 52445248.[Abstract]
Walters, B.N.J., Lao, T., Smith, V. et al. (1985) Alpha-fetoprotein elevation and proteinuric pre-eclampsia. Br. J. Obstet. Gynaecol., 92, 341344.[ISI][Medline]
Williams, M.A., Hickok, D.E., Zingheim, R.W. et al. (1992) Elevated maternal serum -fetoprotein concentrations and midtrimester placental abnormalities in relation to subsequent adverse pregnancy outcomes. Am. J. Obstet. Gynecol., 167, 10321037.[ISI][Medline]
Wolfe, C.D.A., Patel, S.P., Linton, E.A. et al. (1988) Plasma corticotrophin-releasing factor in abnormal pregnancy. Br. J. Obstet. Gynaecol., 95, 10031006.[ISI][Medline]
Yaron, Y., Cherry, M., Kramer, R.L. et al. (1999) Second trimester maternal serum marker screening: maternal serum -fetoprotein, ß-human chorionic gonadotrophin, estriol, and their various combinations as predictors of pregnancy outcome. Am. J. Obstet. Gynecol., 181, 968974.[ISI][Medline]
Submitted on November 30, 1999; accepted on May 4, 2000.