- Email: [email protected]
Second-trimester maternal serum inhibin A concentration as an early marker for preeclampsia
Second-trimester maternal serum inhibin A concentration as an early marker for preeclampsia Joseph Aquilina, MD,a Adrian Barnett, BSc,b Olaleye Thompson, MD,a and Kevin Harrington, MDa London, United Kingdom OBJECTIVE: Maternal serum inhibin A concentration is elevated in established preeclampsia. The aim of this study was to investigate whether this relationship antedates the appearance of the classic signs of preeclampsia. STUDY DESIGN: A retrospective analysis was performed on trisomy 21 screening data from 685 women at between 15 and 19 weeks’ gestation. The main outcome measures were preeclampsia and small for gestational age (<5th percentile) infants. RESULTS: Preeclampsia developed in 35 women (5.5%). Women with inhibin A concentration >2.0 multiples of the median were significantly more likely to acquire preeclampsia (P < .00001) and to be delivered of a small for gestational age infant (<5th percentile, P < .00001) than were women with inhibin A concentration ≤2.0 multiples of the median. The odds ratios were 9.4 (95% confidence interval 4.6-19.3) for development of preeclampsia and 18.2 (95% confidence interval 6.0-54.8) for preeclampsia necessitating delivery at <37 weeks’ gestation. The association remained statistically significant for nulliparous women. CONCLUSION: Elevated maternal inhibin A concentration in the second trimester was strongly associated with a subsequent risk of preeclampsia. The potential role of second trimester inhibin A measurement in a screening strategy for preeclampsia needs to be investigated further. (Am J Obstet Gynecol 1999;181:131-6.)
Key words: Inhibin, preeclampsia, screening
Preeclampsia remains an important cause of maternalfetal morbidity and mortality.1, 2 Although the cause of the condition is not known, placental hypoperfusion and endothelial dysfunction have been cited as the main pathophysiologic mechanisms associated with the disorder.3 The syndrome is unpredictable in onset and progression and is only cured by delivery. Randomized trials of low-dose aspirin prophylaxis started in early pregnancy in low-risk populations4 or in populations identified as being at high risk on the basis of clinical and historical risk factors5 have highlighted the failure of therapeutic trials that have used ineffective screening tests to select patients. Before we can begin to evaluate therapies, we require an effective screening test for the selection of women who are at sufficiently high risk for development of preeclampsia. Ideally a screening test should be simple, inexpensive From The Homerton Hospital National Health Service Trust and St Bartholomew’s and the Royal London School of Medicine and Dentistry, Queen Mary & Westfield College,a and MRC Epidemiology and Medical Care Unit.b Received for publication November 30, 1998; revised January 27, 1999; accepted February 24, 1999. Reprint requests: Kevin Harrington, MD, Academic Department of Obstetrics and Gynaecology, Homerton Hospital, Homerton Row, London, E9 6SR, United Kingdom. Copyright © 1999 by Mosby, Inc. 0002-9378/99 $8.00 + 0 6/1/98098
and reproducible, with a high sensitivity and positive predictive value for the subsequent development of the disease. To ensure maximum patient acceptability it should also be easy to perform and noninvasive. Research has focused on screening tests that have the potential to fulfil these criteria. A number of screening tests have already been proposed,6 but as yet no test fulfils all the criteria that would make it an acceptable screening test in routine obstetric practice. Inhibins are members of the transforming growth factor β superfamily of growth factors.7 They are dimeric glycoprotein hormones consisting of disulfide-link α and βA subunits (inhibin A) or α and βB subunits (inhibin B) that are mainly produced by ovarian granulosa cells in nonpregnant women.8 During pregnancy the placenta is the main source of inhibin,9 which has subsequently been identified as being predominantly of the A form.10, 11 Inhibin has been postulated to play a role in the feedback control of placental human chorionic gonadotropin (hCG).12 Inhibin A is raised during gestations of fetuses with Down syndrome13 and its measurement in conjunction with α-fetoprotein, estriol, and hCG14 levels has been shown to improve screening performance for Down syndrome. The association between inhibin A and preeclampsia was first reported by Muttukrishna et al,15 who found that levels of this analyte are significantly higher in 131
132 Aquilina et al
Fig 1. Histogram of inhibin concentration distribution among women with known outcome (N = 640). Median is 0.98, interquartile range is 0.71. Open bar, Normal; solid bar, preeclampsia.
women with established preeclampsia in the third trimester than among normotensive women. There appears to have been no work reporting a relationship between inhibin A assayed in the second trimester of gestation of chromosomally normal fetuses and subsequent preeclampsia. The null hypothesis for our study was that there was no relationship between second-trimester maternal serum inhibin A concentration and the subsequent development of preeclampsia. Providing that hypothesis was disproved an initial assessment of the potential value of maternal serum inhibin A measurement as an early screening test for preeclampsia would be undertaken. Methods A retrospective analysis was undertaken of consecutive unselected women who had inhibin A measured as part of the serum screening program for trisomy 21 between September 1996 and June 1997 in an inner-city teaching hospital. Exclusion criteria were as follows: multiple gestation, diabetic pregnancy, hypertension diagnosed before 20 weeks’ gestation, and prenatal or postnatal diagnosis of a chromosomal or structural abnormality. All the maternal serum samples were collected between 15 and 19 completed weeks’ gestation, as estimated from the last menstrual period or from early ultrasonographic dates when the menstrual dates differed by >2 SD for the specified gestation. Inhibin A was assayed with a solid-phase sandwich enzyme-linked immunoassay kit (Serotec Ltd, Oxford, United Kingdom). There is a significant inverse relationship between maternal weight and inhibin A concentration.13 Inhibin A concentrations were therefore corrected for weight and expressed as multiples of the median (MoM), as previously described
July 1999 Am J Obstet Gynecol
for other analytes such as alpha-fetoprotein, hCG, and estriol.14 An arbitrary cutoff of 2.0 MoM was chosen prospectively. Data collected retrospectively included age, race, previous obstetric history, parity, medical history, smoking habit, and blood pressure at <20 weeks’ gestation. The main outcome measure was development of preeclampsia. Secondary outcome measures were gestational age at delivery, gestational hypertension, delivery of a small for gestational age (SGA) baby (<5th percentile), abruptio placentae, stillbirth, and early neonatal death. The criteria for gestational hypertension and preeclampsia were based on the definitions given by Davey and MacGillivray.16 Gestational hypertension was defined as the occurrence in a previously normotensive and nonproteinuric women of (1) a diastolic blood pressure of ≥90 mm Hg on at least ≥2 consecutive occasions ≥4 hours apart after the 20th week of gestation or (2) a diastolic blood pressure of ≥110 mm Hg on a single occasion after the 20th week of gestation. Preeclampsia was diagnosed when gestational hypertension was associated with significant proteinuria, which was defined as >300 mg proteinuria on 24-hour urinary collection or the appearance of ≥2+ proteinuria on dipstick testing on 2 separate occasions ≥4 hours apart, without evidence of urinary tract infection. Preeclampsia necessitating delivery at <37 completed weeks’ gestation was also analyzed as a subset. Other premature deliveries were defined as deliveries at <37 completed weeks’ gestation in the absence of preeclampsia or pregnancy-induced hypertension. The diagnosis of abruptio placentae was made when a clinical diagnosis of antepartum hemorrhage and abdominal pain was associated with a retroplacental clot on examination of the placenta after delivery. An intrauterine death was referred to as a stillbirth if it occurred after 24 completed weeks’ gestation, as confirmed by ultrasonographic dating if the menstrual dates were uncertain. The term neonatal death was used to refer to any death occurring within the first week after birth. The definition of SGA at birth was based on a fetal weight lower than the 5th percentile according to charts currently in use at our hospital.17 Any complication was the term used to include all cases of preeclampsia, delivery of an SGA (<5th percentile) baby, abruptio placentae, and stillbirth or early neonatal death. During the data collection the investigators were blinded to the inhibin data. Labor ward, neonatal intensive care unit records (where appropriate), computer database records, and notes (individual patient charts) were reviewed by the investigators for all the women recruited. The data were analyzed with the SPSS statistical package version 6.1.3 (SPSS Inc, Chicago, Ill) for windows (Microsoft Corporation, Redmond, Wash). Statistical significance was assessed with the χ2 test or with the Fisher exact test when cell counts were small. The Wilcoxon
Aquilina et al 133
Volume 181, Number 1 Am J Obstet Gynecol
Table I. Comparison of demographic characteristics of women with inhibin A concentration ≤2.0 MoM and women with inhibin A concentration >2.0 MoM Total (N = 640) Characteristic Race White African Caribbean Asian Oriental (Far Eastern origin—Chinese, Japanese) Nulliparous Smoker Maternal weight in first trimester of pregnancy (kg, median ± SD) Previous preeclampsia necessitating early delivery‡ Previous delivery at >37 wk with birth weight <2500 g Previous other premature delivery§
Inhibin A ≤2.0 MoM (n = 568)
Inhibin A >2.0 MoM (n = 72) Statistical significance
No.
%
No.
%
No.
%
341 252 40 5
53.4 39.5 6.3 0.8
306 223 32 5
54.1 39.4 5.7 0.9
35 29 8 0
48.6 40.3 11.1 0.0
313 148
49.0 23.2
281 125 63.0 ± 13.4
49.5 22.0
32 23 64.7 ± 16.7
45.1 31.9
P = .48 P = .06 P = .562*
19
3.0
11
1.9
8
11.1
P = .0005†
10
1.6
8
1.4
2
2.8
P = .31†
11
1.7
10
1.8
1
1.4
P = 1.00†
P = .26
*By Mann-Whitney test. †By 2-sided Fisher exact test. ‡Delivery at <37 weeks’ gestation. §Delivery at <37 weeks’ gestation with no evidence of gestational hypertension or preeclampsia.
rank sum test was used to compare the distribution of inhibin values between women with known outcomes and those unavailable for follow-up. The Mann-Whitney test was used to compare the weights of the women with normal and abnormal inhibin concentrations. Results The records of inhibin A assays of 685 women were available for analysis after the exclusion of diabetic pregnancies, chromosomally and structurally abnormal fetuses, women with hypertension before the 20th week of pregnancy, and multiple gestations. Forty-five women (6.6%) were unavailable for follow-up, leaving 640 women for analysis. The distribution of maternal serum inhibin concentrations among women with known outcomes is shown in Fig 1. Seventy-two women (11.3%) had inhibin A levels >2.0 MoM. There were no significant differences in the demographic characteristics or distribution of inhibin A concentrations between the patients with known outcomes and those unavailable for follow-up (results available on request). The demographic characteristics of the women with normal and abnormal serum inhibin A concentrations were compared; there were no significant differences between them with the exception of women with a history of preeclampsia who required delivery at <37 weeks’ gestation (1.9% of women with normal inhibin A concentration versus 11.1% of women with abnormal inhibin A concentration, P = .0005; Table I). There was also no significant difference between maternal weights recorded in the first trimester between the 2 groups (P value = .562 by Mann-Whitney test).
Preeclampsia developed in 35 women (5.5%), 15 of whom (2.3%) required delivery at <37 weeks’ gestation and 19 of whom (3.1%) acquired gestational hypertension. Seventy-one women (11.1%) were delivered of SGA infants (<5th percentile), 5 (0.8%) had abruptio placentae, 49 (7.7%) had other premature deliveries, and 14 (2.2%) had stillbirths or early neonatal deaths. Women with inhibin A concentration >2.0 MoM were more likely to acquire preeclampsia (P < .00001), to be delivered of an SGA infant (<5th percentile, P < .00001), and to have a stillbirth or neonatal death (P = .002); they were no more likely to acquire pregnancy-induced hypertension (P = .40) or premature delivery (P = .82; Table II). For women with inhibin A concentration >2.0 MoM the odds ratios were 9.4 (95% confidence interval 4.619.3) for development of preeclampsia, 18.2 (95% confidence interval 6.0-54.8) for preeclampsia necessitating delivery at <37 weeks’ gestation, and 5.5 (95% confidence interval 3.2-9.4) for development of any complication. When women with a history of preeclampsia necessitating early delivery were excluded the association between inhibin A >2.0 MoM and preeclampsia remained statistically significant, although as would be expected the odds ratio was slightly reduced (data not shown but available on request). When the analysis was restricted to nulliparous women the results were similar to those of the entire study population, with an odds ratio among women with inhibin A concentration >2.0 MoM for preeclampsia of 11.2 (95% confidence interval 4.3-29.1; Table III). With a cutoff value of >2.0 MoM the sensitivity for all
134 Aquilina et al
July 1999 Am J Obstet Gynecol
Table II. Comparison of outcomes between 640 women with inhibin A concentration ≤2.0 MoM and women with inhibin A concentration >2.0 MoM Total (N = 640)
Inhibin A ≤2.0 MoM (n = 568)
Inhibin A >2.0 MoM (n = 72) Odds ratio
95% Confidence interval
Outcome
No.
%
No.
%
No.
%
Statistical significance
Preeclampsia Preeclampsia necessitating delivery at <37 wk Gestational hypertension† Small for gestational age (<5th percentile)‡ Other premature delivery§ Abruptio placentae‡ Stillbirth or neonatal death‡ Any complicationll
35 5
5.5 2.3
18 5
3.2 0.9
17 10
23.6 13.9
P < .00001* P < .00001*
9.4 18.2
4.6-19.3 6.0-54.8
19 71
3.1 11.1
16 47
2.9 8.3
3 24
5.5 33.3
P = .40* P < .00001
1.9 5.6
0.54-6.8 3.1-9.9
49 5 14 95
7.7 0.8 2.2 14.8
43 4 8 65
7.6 0.7 1.4 11.4
6 1 6 30
8.3 1.4 8.3 41.7
P = .82 P = .45* P = .002* P < .00001
1.1 0.51 6.4 5.5
0.46- 2.7 0.06-4.6 2.1-18.9 3.2-9.4
*By 2-sided Fisher exact test. †Excluding women with preeclampsia. ‡Including women with preeclampsia or gestational hypertension. §Delivery at <37 weeks’ gestation with no evidence of gestational hypertension or preeclampsia. llPreeclampsia, SGA, abruptio placentae, stillbirth, or neonatal death.
Table III. Comparison of outcomes between 313 nulliparous women with inhibin A concentration ≤2.0 MoM and women with inhibin A concentration >2.0 MoM Total (N = 313) Outcome Preeclampsia Preeclampsia necessitating delivery at <37 wk Gestational hypertension† SGA (<5th percentile)‡ Other premature delivery§ Any complicationll
Inhibin A ≤2.0 MoM (n = 281)
Inhibin A >2.0 MoM (n = 32)
No.
%
No.
%
No.
%
Statistical significance
21 8
6.7 2.6
11 2
3.9 0.7
10 6
31.1 18.8
P < .00001* P = .00002*
10 43 17 55
3.4 13.7 5.4 17.6
8 31 16 40
3.0 11.0 5.7 14.2
2 12 1 15
9.1 37.5 3.1 46.9
P = .17* P = .0003* P = 1.00* P < .00001
Odds ratio
95% Confidence interval
11.2 32.2
4.3-29.1 6.2-167.6
3.3 4.8 0.53 5.3
0.65-16.5 2.2-10.8 0.07- 4.2 2.5-11.5
*By 2-sided Fisher exact test. †Excluding women with preeclampsia. ‡Including women with preeclampsia and gestational hypertension. §Delivery at <37 weeks’ gestation with no evidence of gestational hypertension or preeclampsia. llPreeclampsia, SGA, abruptio placentae, stillbirth, or neonatal death.
preeclampsia was 47%, the specificity was 91%, the positive predictive value was 24%, and the negative predictive value was 97%. At the same cutoff in the subgroup of women with preeclampsia necessitating preterm delivery the sensitivity was 67%, the specificity was 90%, the positive predictive value was 14%, and the negative predictive value was 99%. Among nulliparous women the sensitivity, specificity, positive predictive value, and negative predictive value were 48%, 93%, 31%, and 96%, respectively, for preeclampsia and 75%, 92%, 19%, and 99%, respectively, for preeclampsia necessitating early delivery. The sensitivity for any complication among all women in the analysis was 31%, the specificity was 92%, the positive
predictive value was 42%, and the negative predictive value was 89%. Comment This study indicates a significant association between elevated second-trimester maternal serum levels of inhibin A and the subsequent development of preeclampsia. The higher odds ratio for preeclampsia necessitating delivery at <37 weeks’ gestation suggests that inhibin A is an even more sensitive marker for the more severe forms of preeclampsia necessitating early delivery. This association remained significant when analysis was restricted to nulliparous women. Our findings also suggest that in-
Aquilina et al 135
Volume 181, Number 1 Am J Obstet Gynecol
hibin A is not a marker for gestational hypertension, supporting previous observations that isolated gestational hypertension is likely to be a different disease from preeclampsia.18 The inverse relationship between serum levels of inhibin A and maternal weight makes it unlikely that the association between raised inhibin levels and preeclampsia can be attributable to weight. In any case, there is no reported association between maternal weight in the second trimester (used in our study) and preeclampsia. In addition, the difference in weight at booking between women with high and low inhibin levels in our study was not significant. In preeclampsia there is partial or complete failure of trophoblastic invasion of the myometrial segments of the spiral arteries,3 a process that normally has taken place by 20 weeks’ gestation. The failure of trophoblastic invasion is associated with ischemic damage to the syncytiotrophoblast and increased proliferation of the cytotrophoblast,19 causing functional alteration of the surface layer of the syncytiotrophoblast. Inhibin A is predominantly localized within the syncytiotrophoblast, as demonstrated by immunohistochemical studies.20 It is therefore postulated that in preeclampsia and other disorders associated with placental vasculopathy, such as intrauterine growth restriction, there are 2 possible mechanisms accounting for the increased inhibin levels. The increase could stem either from the reactive hyperplasia of cytotrophoblastic cells, resulting in increased production of inhibin, or the functional alteration of the syncytiotrophoblast, giving rise to increased leakage of placental proteins localized within this layer into the maternal circulation. A statistically significant association between high levels of hCG in chromosomally normal fetuses and subsequent development of preeclampsia is well documented,21 but the correlation is not robust enough to translate it into an effective screening test. Inhibin A has already been shown to be a more sensitive marker than hCG in established preeclampsia.15 Our findings suggest that inhibin A might also be a more sensitive early marker than hCG and that it seems to perform better as an early screening test. In vitro hCG stimulates trophoblastic inhibin A secretion and reciprocally inhibin A suppresses hCG secretion.22 The in vivo relationship in the second trimester between these 2 hormones in women who eventually develop preeclampsia has not yet been investigated. Such a study might shed more light on the pathogenesis of preeclampsia at the level of the syncytiotrophoblast microvillous membranes. A clinically useful screening test must have high sensitivity, a clinically relevant positive predictive value, and an acceptable false-positive rate (high specificity). Among all the women, including nulliparous women, elevated second-trimester maternal inhibin A levels found 1 in 2
of all women with eventual preeclampsia and 2 of 3 of the women who required early delivery as a result of preeclampsia. Although the positive predictive value might be considered disappointing (31% for nulliparous women and 24% for all the women), it still compares favorably with many other proposed early screening tests6 for preeclampsia. These results suggest that second trimester inhibin A may form the basis of a practical, reproducible, and effective early screening test for the subsequent development of preeclampsia. Although the results are exciting and encouraging, the conclusions from this study are limited by the relatively small numbers investigated in a multiethnic inner-city population in which the prevalences of preeclampsia and SGA (<5th percentile) are higher than the national average. Before the assay can be considered as part of our antenatal screening strategy for preeclampsia and other uteroplacental complications, it needs to be evaluated in larger prospective studies in different populations. The role of maternal serum inhibin A concentration during the second trimester as part of an early screening test in combination with other indicators of placental hypoperfusion23 should also be concurrently investigated. We thank Professor Nicholas Wald and Mr Wayne Huttley (Wolfson Institute of Preventative Medicine, Charterhouse Square, London) for providing us with the serum inhibin A data corrected for maternal weight. REFERENCES
1. Report on confidential enquiries into maternal deaths in the United Kingdom 1994-1996. London: Her Majesty’s Stationery Office; 1997. 2. Eskenazi B, Fenster L, Sidney S, Elkin EP. Fetal growth retardation in infants of multiparous and nulliparous women with preeclampsia. Am J Obstet Gynecol 1993;169:1112-6. 3. Meekins JW, Pijnenborg R, Hanssens M, McFayden IR, Van Asshe A. A study of placental bed spiral arteries and trophoblastic invasion in normal and severe pre-eclamptic pregnancies. Br J Obstet Gynecol 1994;101:669-74. 4. Caritis S, Sibai B, Hauth J, Lindheimer MD, Klebanoff M, Thom E, et al. Low-dose aspirin to prevent preeclampsia in women at high risk. National Institute of Child Health and Human Development Network of Maternal-Fetal Medicine Units. N Engl J Med 1998;338:701-5. 5. Sibai BM, Caritis SN, Thom E, Klebanoff M, McNellis D, Rocco L, et al. Prevention of preeclampsia with low-dose aspirin in healthy, nulliparous pregnant women. The National Institute of Child Health and Human Development Network of MaternalFetal Medicine Units. N Engl J Med 1993:329:1213-8. 6. Conde-Agudelo A, Lede R, Belizan J. Evaluation of methods used in the prediction of hypertensive disorders of pregnancy. Obstet Gynecol Surv 1994;49:210-22. 7. Messague J. The transforming growth factor beta family. Ann Rev Cell Biol 1990;6:597-641. 8. Illingworth PJ, Reddi K, Wickings EJ, Baird DT. The source of inhibin secretion during the human menstrual cycle. J Clin Endocrinol Metabol 1991;73:667-73. 9. Qu J, Thomas K. Inhibin and activin production in the human placenta. Endocr Rev 1995;16:485-507. 10. Illingworth PJ, Groome NP, Duncan WC, Grant V, Tovanabutra S, Baird DT, et al. Measurement of circulating inhibin forms
136 Aquilina et al
11.
12.
13.
14.
15.
16.
17.
during establishment of pregnancy. J Clin Endocrinol Metabol 1996;81:1471-5. Muttukrishna S, George L, Fowler PA, Groome NP, Knight PG. Measurement of serum concentrations of inhibin-A (α−βA dimer) during human pregnancy. Clin Endocrinol 1995;2:391-7. Healy DL, Polson D, Yohkachiya T, de Kretser DM. Inhibin and related peptides in pregnancy. Baillieres Clin Endocrinol Metab 1990;4:233-47. Wald NJ, Densem JW, George L, Muttukrishna S, Knight PG. Prenatal screening for Down’s syndrome using inhibin-A as a serum marker. Prenat Diagn 1996;16:143-5. Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Royston P, Chard T, et al. Maternal serum screening for Down’s syndrome in early pregnancy [published erratum appears in BMJ 1988;297:1029]. BMJ 1988;297:883-7. Muttukrishna S, Knight PG, Groome NP, Redman CW, Ledger WL. Activin-A and inhibin-A as possible endocrine markers for pre-eclampsia. Lancet 1997;349:1285-8. Davey DA, MacGillivray I. The classification and definition of the hypertensive disorders of pregnancy. Am J Obstet Gynecol 1988;158:892-8. Yudkin PL, Aboualfa M, Eyre JA, Redman CW, Wilkinson AR.
July 1999 Am J Obstet Gynecol
18. 19.
20.
21.
22.
23.
New birth weight and head circumference centiles for gestational ages 24-42 weeks. Early Hum Dev 1987;15:45-52. Roberts JM, Redman CW. Pre-eclampsia: more than just pregnancy-induced hypertension. Lancet 1993;341:1447-51. Arnholdt H, Meisel F, Fandrey K, Löhrs U. Proliferation of villous trophoblast of the human placenta in human placenta in normal and abnormal pregnancies. Virchows Arch B Cell Pathol Incl Mol Pathol 1991;60:365-72. McCluggage WG, Ashe P, McBride H, Maxwell P, Sloan JM. Localization of the cellular expression of inhibin in trophoblastic tissue. Histopathology 1998;32:252-6. Muller F, Savey L, Le Fibler F, Bussieres L, Nyadizimba G, Colau JC, Giraudet P. Maternal serum human chorionic gonadotropin level at fifteen weeks is a predictor for preeclampsia. Am J Obstet Gynecol 1996;175:37-40. Petraglia F, Sawchenko P, Lim AT, Rivier J, Vale W. Localization, secretion, and action of inhibin in human placenta. Science 1987;237:187-9. Harrington K, Carpenter RG, Goldfrad C, Campbell S. Transvaginal Doppler ultrasound of the uteroplacental circulation in the early prediction of pre-eclampsia and intrauterine growth retardation. Br J Obstet Gynaecol 1997;104:674-81.
Availability of Journal back issues As a service to our subscribers, copies of back issues of the American Journal of Obstetrics and Gynecology for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted at a cost of $16 per issue. The following quantity discounts are available: one fourth off on quantities of 12 to 23 and one third off on quantities of 24 or more. Please write to Mosby, Inc, Subscription Services, 11830 Westline Industrial Dr, St Louis, MO 63146-3318, telephone (800)453-4351 or (314)453-4351, for information on availability of particular issues. If back issues are unavailable from the publisher, photocopies of complete issues may be purchased from Bell & Howell Information and Learning, 300 N Zeeb Rd, Ann Arbor, MI 48106-1346. Telephone (734)761-4700.
Key words: Inhibin, preeclampsia, screening
Preeclampsia remains an important cause of maternalfetal morbidity and mortality.1, 2 Although the cause of the condition is not known, placental hypoperfusion and endothelial dysfunction have been cited as the main pathophysiologic mechanisms associated with the disorder.3 The syndrome is unpredictable in onset and progression and is only cured by delivery. Randomized trials of low-dose aspirin prophylaxis started in early pregnancy in low-risk populations4 or in populations identified as being at high risk on the basis of clinical and historical risk factors5 have highlighted the failure of therapeutic trials that have used ineffective screening tests to select patients. Before we can begin to evaluate therapies, we require an effective screening test for the selection of women who are at sufficiently high risk for development of preeclampsia. Ideally a screening test should be simple, inexpensive From The Homerton Hospital National Health Service Trust and St Bartholomew’s and the Royal London School of Medicine and Dentistry, Queen Mary & Westfield College,a and MRC Epidemiology and Medical Care Unit.b Received for publication November 30, 1998; revised January 27, 1999; accepted February 24, 1999. Reprint requests: Kevin Harrington, MD, Academic Department of Obstetrics and Gynaecology, Homerton Hospital, Homerton Row, London, E9 6SR, United Kingdom. Copyright © 1999 by Mosby, Inc. 0002-9378/99 $8.00 + 0 6/1/98098
and reproducible, with a high sensitivity and positive predictive value for the subsequent development of the disease. To ensure maximum patient acceptability it should also be easy to perform and noninvasive. Research has focused on screening tests that have the potential to fulfil these criteria. A number of screening tests have already been proposed,6 but as yet no test fulfils all the criteria that would make it an acceptable screening test in routine obstetric practice. Inhibins are members of the transforming growth factor β superfamily of growth factors.7 They are dimeric glycoprotein hormones consisting of disulfide-link α and βA subunits (inhibin A) or α and βB subunits (inhibin B) that are mainly produced by ovarian granulosa cells in nonpregnant women.8 During pregnancy the placenta is the main source of inhibin,9 which has subsequently been identified as being predominantly of the A form.10, 11 Inhibin has been postulated to play a role in the feedback control of placental human chorionic gonadotropin (hCG).12 Inhibin A is raised during gestations of fetuses with Down syndrome13 and its measurement in conjunction with α-fetoprotein, estriol, and hCG14 levels has been shown to improve screening performance for Down syndrome. The association between inhibin A and preeclampsia was first reported by Muttukrishna et al,15 who found that levels of this analyte are significantly higher in 131
132 Aquilina et al
Fig 1. Histogram of inhibin concentration distribution among women with known outcome (N = 640). Median is 0.98, interquartile range is 0.71. Open bar, Normal; solid bar, preeclampsia.
women with established preeclampsia in the third trimester than among normotensive women. There appears to have been no work reporting a relationship between inhibin A assayed in the second trimester of gestation of chromosomally normal fetuses and subsequent preeclampsia. The null hypothesis for our study was that there was no relationship between second-trimester maternal serum inhibin A concentration and the subsequent development of preeclampsia. Providing that hypothesis was disproved an initial assessment of the potential value of maternal serum inhibin A measurement as an early screening test for preeclampsia would be undertaken. Methods A retrospective analysis was undertaken of consecutive unselected women who had inhibin A measured as part of the serum screening program for trisomy 21 between September 1996 and June 1997 in an inner-city teaching hospital. Exclusion criteria were as follows: multiple gestation, diabetic pregnancy, hypertension diagnosed before 20 weeks’ gestation, and prenatal or postnatal diagnosis of a chromosomal or structural abnormality. All the maternal serum samples were collected between 15 and 19 completed weeks’ gestation, as estimated from the last menstrual period or from early ultrasonographic dates when the menstrual dates differed by >2 SD for the specified gestation. Inhibin A was assayed with a solid-phase sandwich enzyme-linked immunoassay kit (Serotec Ltd, Oxford, United Kingdom). There is a significant inverse relationship between maternal weight and inhibin A concentration.13 Inhibin A concentrations were therefore corrected for weight and expressed as multiples of the median (MoM), as previously described
July 1999 Am J Obstet Gynecol
for other analytes such as alpha-fetoprotein, hCG, and estriol.14 An arbitrary cutoff of 2.0 MoM was chosen prospectively. Data collected retrospectively included age, race, previous obstetric history, parity, medical history, smoking habit, and blood pressure at <20 weeks’ gestation. The main outcome measure was development of preeclampsia. Secondary outcome measures were gestational age at delivery, gestational hypertension, delivery of a small for gestational age (SGA) baby (<5th percentile), abruptio placentae, stillbirth, and early neonatal death. The criteria for gestational hypertension and preeclampsia were based on the definitions given by Davey and MacGillivray.16 Gestational hypertension was defined as the occurrence in a previously normotensive and nonproteinuric women of (1) a diastolic blood pressure of ≥90 mm Hg on at least ≥2 consecutive occasions ≥4 hours apart after the 20th week of gestation or (2) a diastolic blood pressure of ≥110 mm Hg on a single occasion after the 20th week of gestation. Preeclampsia was diagnosed when gestational hypertension was associated with significant proteinuria, which was defined as >300 mg proteinuria on 24-hour urinary collection or the appearance of ≥2+ proteinuria on dipstick testing on 2 separate occasions ≥4 hours apart, without evidence of urinary tract infection. Preeclampsia necessitating delivery at <37 completed weeks’ gestation was also analyzed as a subset. Other premature deliveries were defined as deliveries at <37 completed weeks’ gestation in the absence of preeclampsia or pregnancy-induced hypertension. The diagnosis of abruptio placentae was made when a clinical diagnosis of antepartum hemorrhage and abdominal pain was associated with a retroplacental clot on examination of the placenta after delivery. An intrauterine death was referred to as a stillbirth if it occurred after 24 completed weeks’ gestation, as confirmed by ultrasonographic dating if the menstrual dates were uncertain. The term neonatal death was used to refer to any death occurring within the first week after birth. The definition of SGA at birth was based on a fetal weight lower than the 5th percentile according to charts currently in use at our hospital.17 Any complication was the term used to include all cases of preeclampsia, delivery of an SGA (<5th percentile) baby, abruptio placentae, and stillbirth or early neonatal death. During the data collection the investigators were blinded to the inhibin data. Labor ward, neonatal intensive care unit records (where appropriate), computer database records, and notes (individual patient charts) were reviewed by the investigators for all the women recruited. The data were analyzed with the SPSS statistical package version 6.1.3 (SPSS Inc, Chicago, Ill) for windows (Microsoft Corporation, Redmond, Wash). Statistical significance was assessed with the χ2 test or with the Fisher exact test when cell counts were small. The Wilcoxon
Aquilina et al 133
Volume 181, Number 1 Am J Obstet Gynecol
Table I. Comparison of demographic characteristics of women with inhibin A concentration ≤2.0 MoM and women with inhibin A concentration >2.0 MoM Total (N = 640) Characteristic Race White African Caribbean Asian Oriental (Far Eastern origin—Chinese, Japanese) Nulliparous Smoker Maternal weight in first trimester of pregnancy (kg, median ± SD) Previous preeclampsia necessitating early delivery‡ Previous delivery at >37 wk with birth weight <2500 g Previous other premature delivery§
Inhibin A ≤2.0 MoM (n = 568)
Inhibin A >2.0 MoM (n = 72) Statistical significance
No.
%
No.
%
No.
%
341 252 40 5
53.4 39.5 6.3 0.8
306 223 32 5
54.1 39.4 5.7 0.9
35 29 8 0
48.6 40.3 11.1 0.0
313 148
49.0 23.2
281 125 63.0 ± 13.4
49.5 22.0
32 23 64.7 ± 16.7
45.1 31.9
P = .48 P = .06 P = .562*
19
3.0
11
1.9
8
11.1
P = .0005†
10
1.6
8
1.4
2
2.8
P = .31†
11
1.7
10
1.8
1
1.4
P = 1.00†
P = .26
*By Mann-Whitney test. †By 2-sided Fisher exact test. ‡Delivery at <37 weeks’ gestation. §Delivery at <37 weeks’ gestation with no evidence of gestational hypertension or preeclampsia.
rank sum test was used to compare the distribution of inhibin values between women with known outcomes and those unavailable for follow-up. The Mann-Whitney test was used to compare the weights of the women with normal and abnormal inhibin concentrations. Results The records of inhibin A assays of 685 women were available for analysis after the exclusion of diabetic pregnancies, chromosomally and structurally abnormal fetuses, women with hypertension before the 20th week of pregnancy, and multiple gestations. Forty-five women (6.6%) were unavailable for follow-up, leaving 640 women for analysis. The distribution of maternal serum inhibin concentrations among women with known outcomes is shown in Fig 1. Seventy-two women (11.3%) had inhibin A levels >2.0 MoM. There were no significant differences in the demographic characteristics or distribution of inhibin A concentrations between the patients with known outcomes and those unavailable for follow-up (results available on request). The demographic characteristics of the women with normal and abnormal serum inhibin A concentrations were compared; there were no significant differences between them with the exception of women with a history of preeclampsia who required delivery at <37 weeks’ gestation (1.9% of women with normal inhibin A concentration versus 11.1% of women with abnormal inhibin A concentration, P = .0005; Table I). There was also no significant difference between maternal weights recorded in the first trimester between the 2 groups (P value = .562 by Mann-Whitney test).
Preeclampsia developed in 35 women (5.5%), 15 of whom (2.3%) required delivery at <37 weeks’ gestation and 19 of whom (3.1%) acquired gestational hypertension. Seventy-one women (11.1%) were delivered of SGA infants (<5th percentile), 5 (0.8%) had abruptio placentae, 49 (7.7%) had other premature deliveries, and 14 (2.2%) had stillbirths or early neonatal deaths. Women with inhibin A concentration >2.0 MoM were more likely to acquire preeclampsia (P < .00001), to be delivered of an SGA infant (<5th percentile, P < .00001), and to have a stillbirth or neonatal death (P = .002); they were no more likely to acquire pregnancy-induced hypertension (P = .40) or premature delivery (P = .82; Table II). For women with inhibin A concentration >2.0 MoM the odds ratios were 9.4 (95% confidence interval 4.619.3) for development of preeclampsia, 18.2 (95% confidence interval 6.0-54.8) for preeclampsia necessitating delivery at <37 weeks’ gestation, and 5.5 (95% confidence interval 3.2-9.4) for development of any complication. When women with a history of preeclampsia necessitating early delivery were excluded the association between inhibin A >2.0 MoM and preeclampsia remained statistically significant, although as would be expected the odds ratio was slightly reduced (data not shown but available on request). When the analysis was restricted to nulliparous women the results were similar to those of the entire study population, with an odds ratio among women with inhibin A concentration >2.0 MoM for preeclampsia of 11.2 (95% confidence interval 4.3-29.1; Table III). With a cutoff value of >2.0 MoM the sensitivity for all
134 Aquilina et al
July 1999 Am J Obstet Gynecol
Table II. Comparison of outcomes between 640 women with inhibin A concentration ≤2.0 MoM and women with inhibin A concentration >2.0 MoM Total (N = 640)
Inhibin A ≤2.0 MoM (n = 568)
Inhibin A >2.0 MoM (n = 72) Odds ratio
95% Confidence interval
Outcome
No.
%
No.
%
No.
%
Statistical significance
Preeclampsia Preeclampsia necessitating delivery at <37 wk Gestational hypertension† Small for gestational age (<5th percentile)‡ Other premature delivery§ Abruptio placentae‡ Stillbirth or neonatal death‡ Any complicationll
35 5
5.5 2.3
18 5
3.2 0.9
17 10
23.6 13.9
P < .00001* P < .00001*
9.4 18.2
4.6-19.3 6.0-54.8
19 71
3.1 11.1
16 47
2.9 8.3
3 24
5.5 33.3
P = .40* P < .00001
1.9 5.6
0.54-6.8 3.1-9.9
49 5 14 95
7.7 0.8 2.2 14.8
43 4 8 65
7.6 0.7 1.4 11.4
6 1 6 30
8.3 1.4 8.3 41.7
P = .82 P = .45* P = .002* P < .00001
1.1 0.51 6.4 5.5
0.46- 2.7 0.06-4.6 2.1-18.9 3.2-9.4
*By 2-sided Fisher exact test. †Excluding women with preeclampsia. ‡Including women with preeclampsia or gestational hypertension. §Delivery at <37 weeks’ gestation with no evidence of gestational hypertension or preeclampsia. llPreeclampsia, SGA, abruptio placentae, stillbirth, or neonatal death.
Table III. Comparison of outcomes between 313 nulliparous women with inhibin A concentration ≤2.0 MoM and women with inhibin A concentration >2.0 MoM Total (N = 313) Outcome Preeclampsia Preeclampsia necessitating delivery at <37 wk Gestational hypertension† SGA (<5th percentile)‡ Other premature delivery§ Any complicationll
Inhibin A ≤2.0 MoM (n = 281)
Inhibin A >2.0 MoM (n = 32)
No.
%
No.
%
No.
%
Statistical significance
21 8
6.7 2.6
11 2
3.9 0.7
10 6
31.1 18.8
P < .00001* P = .00002*
10 43 17 55
3.4 13.7 5.4 17.6
8 31 16 40
3.0 11.0 5.7 14.2
2 12 1 15
9.1 37.5 3.1 46.9
P = .17* P = .0003* P = 1.00* P < .00001
Odds ratio
95% Confidence interval
11.2 32.2
4.3-29.1 6.2-167.6
3.3 4.8 0.53 5.3
0.65-16.5 2.2-10.8 0.07- 4.2 2.5-11.5
*By 2-sided Fisher exact test. †Excluding women with preeclampsia. ‡Including women with preeclampsia and gestational hypertension. §Delivery at <37 weeks’ gestation with no evidence of gestational hypertension or preeclampsia. llPreeclampsia, SGA, abruptio placentae, stillbirth, or neonatal death.
preeclampsia was 47%, the specificity was 91%, the positive predictive value was 24%, and the negative predictive value was 97%. At the same cutoff in the subgroup of women with preeclampsia necessitating preterm delivery the sensitivity was 67%, the specificity was 90%, the positive predictive value was 14%, and the negative predictive value was 99%. Among nulliparous women the sensitivity, specificity, positive predictive value, and negative predictive value were 48%, 93%, 31%, and 96%, respectively, for preeclampsia and 75%, 92%, 19%, and 99%, respectively, for preeclampsia necessitating early delivery. The sensitivity for any complication among all women in the analysis was 31%, the specificity was 92%, the positive
predictive value was 42%, and the negative predictive value was 89%. Comment This study indicates a significant association between elevated second-trimester maternal serum levels of inhibin A and the subsequent development of preeclampsia. The higher odds ratio for preeclampsia necessitating delivery at <37 weeks’ gestation suggests that inhibin A is an even more sensitive marker for the more severe forms of preeclampsia necessitating early delivery. This association remained significant when analysis was restricted to nulliparous women. Our findings also suggest that in-
Aquilina et al 135
Volume 181, Number 1 Am J Obstet Gynecol
hibin A is not a marker for gestational hypertension, supporting previous observations that isolated gestational hypertension is likely to be a different disease from preeclampsia.18 The inverse relationship between serum levels of inhibin A and maternal weight makes it unlikely that the association between raised inhibin levels and preeclampsia can be attributable to weight. In any case, there is no reported association between maternal weight in the second trimester (used in our study) and preeclampsia. In addition, the difference in weight at booking between women with high and low inhibin levels in our study was not significant. In preeclampsia there is partial or complete failure of trophoblastic invasion of the myometrial segments of the spiral arteries,3 a process that normally has taken place by 20 weeks’ gestation. The failure of trophoblastic invasion is associated with ischemic damage to the syncytiotrophoblast and increased proliferation of the cytotrophoblast,19 causing functional alteration of the surface layer of the syncytiotrophoblast. Inhibin A is predominantly localized within the syncytiotrophoblast, as demonstrated by immunohistochemical studies.20 It is therefore postulated that in preeclampsia and other disorders associated with placental vasculopathy, such as intrauterine growth restriction, there are 2 possible mechanisms accounting for the increased inhibin levels. The increase could stem either from the reactive hyperplasia of cytotrophoblastic cells, resulting in increased production of inhibin, or the functional alteration of the syncytiotrophoblast, giving rise to increased leakage of placental proteins localized within this layer into the maternal circulation. A statistically significant association between high levels of hCG in chromosomally normal fetuses and subsequent development of preeclampsia is well documented,21 but the correlation is not robust enough to translate it into an effective screening test. Inhibin A has already been shown to be a more sensitive marker than hCG in established preeclampsia.15 Our findings suggest that inhibin A might also be a more sensitive early marker than hCG and that it seems to perform better as an early screening test. In vitro hCG stimulates trophoblastic inhibin A secretion and reciprocally inhibin A suppresses hCG secretion.22 The in vivo relationship in the second trimester between these 2 hormones in women who eventually develop preeclampsia has not yet been investigated. Such a study might shed more light on the pathogenesis of preeclampsia at the level of the syncytiotrophoblast microvillous membranes. A clinically useful screening test must have high sensitivity, a clinically relevant positive predictive value, and an acceptable false-positive rate (high specificity). Among all the women, including nulliparous women, elevated second-trimester maternal inhibin A levels found 1 in 2
of all women with eventual preeclampsia and 2 of 3 of the women who required early delivery as a result of preeclampsia. Although the positive predictive value might be considered disappointing (31% for nulliparous women and 24% for all the women), it still compares favorably with many other proposed early screening tests6 for preeclampsia. These results suggest that second trimester inhibin A may form the basis of a practical, reproducible, and effective early screening test for the subsequent development of preeclampsia. Although the results are exciting and encouraging, the conclusions from this study are limited by the relatively small numbers investigated in a multiethnic inner-city population in which the prevalences of preeclampsia and SGA (<5th percentile) are higher than the national average. Before the assay can be considered as part of our antenatal screening strategy for preeclampsia and other uteroplacental complications, it needs to be evaluated in larger prospective studies in different populations. The role of maternal serum inhibin A concentration during the second trimester as part of an early screening test in combination with other indicators of placental hypoperfusion23 should also be concurrently investigated. We thank Professor Nicholas Wald and Mr Wayne Huttley (Wolfson Institute of Preventative Medicine, Charterhouse Square, London) for providing us with the serum inhibin A data corrected for maternal weight. REFERENCES
1. Report on confidential enquiries into maternal deaths in the United Kingdom 1994-1996. London: Her Majesty’s Stationery Office; 1997. 2. Eskenazi B, Fenster L, Sidney S, Elkin EP. Fetal growth retardation in infants of multiparous and nulliparous women with preeclampsia. Am J Obstet Gynecol 1993;169:1112-6. 3. Meekins JW, Pijnenborg R, Hanssens M, McFayden IR, Van Asshe A. A study of placental bed spiral arteries and trophoblastic invasion in normal and severe pre-eclamptic pregnancies. Br J Obstet Gynecol 1994;101:669-74. 4. Caritis S, Sibai B, Hauth J, Lindheimer MD, Klebanoff M, Thom E, et al. Low-dose aspirin to prevent preeclampsia in women at high risk. National Institute of Child Health and Human Development Network of Maternal-Fetal Medicine Units. N Engl J Med 1998;338:701-5. 5. Sibai BM, Caritis SN, Thom E, Klebanoff M, McNellis D, Rocco L, et al. Prevention of preeclampsia with low-dose aspirin in healthy, nulliparous pregnant women. The National Institute of Child Health and Human Development Network of MaternalFetal Medicine Units. N Engl J Med 1993:329:1213-8. 6. Conde-Agudelo A, Lede R, Belizan J. Evaluation of methods used in the prediction of hypertensive disorders of pregnancy. Obstet Gynecol Surv 1994;49:210-22. 7. Messague J. The transforming growth factor beta family. Ann Rev Cell Biol 1990;6:597-641. 8. Illingworth PJ, Reddi K, Wickings EJ, Baird DT. The source of inhibin secretion during the human menstrual cycle. J Clin Endocrinol Metabol 1991;73:667-73. 9. Qu J, Thomas K. Inhibin and activin production in the human placenta. Endocr Rev 1995;16:485-507. 10. Illingworth PJ, Groome NP, Duncan WC, Grant V, Tovanabutra S, Baird DT, et al. Measurement of circulating inhibin forms
136 Aquilina et al
11.
12.
13.
14.
15.
16.
17.
during establishment of pregnancy. J Clin Endocrinol Metabol 1996;81:1471-5. Muttukrishna S, George L, Fowler PA, Groome NP, Knight PG. Measurement of serum concentrations of inhibin-A (α−βA dimer) during human pregnancy. Clin Endocrinol 1995;2:391-7. Healy DL, Polson D, Yohkachiya T, de Kretser DM. Inhibin and related peptides in pregnancy. Baillieres Clin Endocrinol Metab 1990;4:233-47. Wald NJ, Densem JW, George L, Muttukrishna S, Knight PG. Prenatal screening for Down’s syndrome using inhibin-A as a serum marker. Prenat Diagn 1996;16:143-5. Wald NJ, Cuckle HS, Densem JW, Nanchahal K, Royston P, Chard T, et al. Maternal serum screening for Down’s syndrome in early pregnancy [published erratum appears in BMJ 1988;297:1029]. BMJ 1988;297:883-7. Muttukrishna S, Knight PG, Groome NP, Redman CW, Ledger WL. Activin-A and inhibin-A as possible endocrine markers for pre-eclampsia. Lancet 1997;349:1285-8. Davey DA, MacGillivray I. The classification and definition of the hypertensive disorders of pregnancy. Am J Obstet Gynecol 1988;158:892-8. Yudkin PL, Aboualfa M, Eyre JA, Redman CW, Wilkinson AR.
July 1999 Am J Obstet Gynecol
18. 19.
20.
21.
22.
23.
New birth weight and head circumference centiles for gestational ages 24-42 weeks. Early Hum Dev 1987;15:45-52. Roberts JM, Redman CW. Pre-eclampsia: more than just pregnancy-induced hypertension. Lancet 1993;341:1447-51. Arnholdt H, Meisel F, Fandrey K, Löhrs U. Proliferation of villous trophoblast of the human placenta in human placenta in normal and abnormal pregnancies. Virchows Arch B Cell Pathol Incl Mol Pathol 1991;60:365-72. McCluggage WG, Ashe P, McBride H, Maxwell P, Sloan JM. Localization of the cellular expression of inhibin in trophoblastic tissue. Histopathology 1998;32:252-6. Muller F, Savey L, Le Fibler F, Bussieres L, Nyadizimba G, Colau JC, Giraudet P. Maternal serum human chorionic gonadotropin level at fifteen weeks is a predictor for preeclampsia. Am J Obstet Gynecol 1996;175:37-40. Petraglia F, Sawchenko P, Lim AT, Rivier J, Vale W. Localization, secretion, and action of inhibin in human placenta. Science 1987;237:187-9. Harrington K, Carpenter RG, Goldfrad C, Campbell S. Transvaginal Doppler ultrasound of the uteroplacental circulation in the early prediction of pre-eclampsia and intrauterine growth retardation. Br J Obstet Gynaecol 1997;104:674-81.
Availability of Journal back issues As a service to our subscribers, copies of back issues of the American Journal of Obstetrics and Gynecology for the preceding 5 years are maintained and are available for purchase from Mosby until inventory is depleted at a cost of $16 per issue. The following quantity discounts are available: one fourth off on quantities of 12 to 23 and one third off on quantities of 24 or more. Please write to Mosby, Inc, Subscription Services, 11830 Westline Industrial Dr, St Louis, MO 63146-3318, telephone (800)453-4351 or (314)453-4351, for information on availability of particular issues. If back issues are unavailable from the publisher, photocopies of complete issues may be purchased from Bell & Howell Information and Learning, 300 N Zeeb Rd, Ann Arbor, MI 48106-1346. Telephone (734)761-4700.