Abstract
Background: Adrenomedullin, secreted by decidua and trophoblast cells, is considered to participate in regulating uterine and placental blood flow, leading to control of placental hormonal secretion. Furthermore, adrenomedullin has an antimicrobial activity. The objective of this study was to determine whether adrenomedullin concentrations in midtrimester amniotic fluid can be used as a predictor of preterm delivery. Patients and Methods: Amniotic fluid samples were collected in a retrospective cross-matched study that included 362 women with singleton pregnancies who presented for genetic amniocentesis. Adrenomedullin concentrations were determined by ELISA in amniotic fluid taken from women with spontaneous preterm delivery (n=41) and maternal age-matched controls who had normal pregnancy at term (n=41). Results: No difference was found in adrenomedullin concentrations between women with spontaneous preterm delivery (median: 1.33 ng/ml, range: 0.36-8.53 ng/ml) and controls (median: 1.32 ng/ml, range: 0.33-4.07 ng/ml), nor between a subset of cases of preterm premature rupture of membranes (n=19) and their controls (n=19). Conclusion: Adrenomedullin concentration in amniotic fluid cannot serve as a predictor of preterm delivery.
Adrenomedullin is a member of the calcitonin family of peptides with sequence homology to calcitonin gene-related peptide and islet amyloid polypeptide (1). It is a peptide, a secretory product of vascular endothelium, smooth muscle cells, cardiomyocytes and cardiac fibroblasts (2, 3). Adrenomedullin is highly expressed in the uterus, ovary (4), decidua and trophoblast cells (5). More specifically, immunohistochemical analysis, Western blot analysis, (RT-PCR) and in situ (RT-PCR) showed that the fetal amniotic membranes are the source of adrenomedullin in the amniotic fluid (6-8). Di Iorio et al. were the first to demonstrate that both the amnion and chorio-decidua secrete adrenomedullin (9). Furthermore, it was shown that adrenomedullin is localized in placental syncytiotrophoblast and fetal amniotic membranes in both the first and third trimester (10).
Adrenomedullin is a hypotensive and natriuretic peptide (5) acting as a potential uterine smooth muscle and blood vessel relaxant (11-14), implicated in maternal systemic vasodilation, placental vessel relaxation and reduced fetoplacental vascular resistance (15). Therefore, it is important for the regulation of placentation, angiogenesis and an adequate blood supply to the fetus (16). It has been shown that its levels increase during spontaneous and preterm delivery (15).
Preterm delivery complicates 7-12.8% of deliveries (17, 18) and leads to increased neonatal mortality and morbidity (e.g. blindness, deafness, developmental delay, cerebral palsy and chronic lung disease) (19-21). The etiology of preterm delivery remains unknown (22). A major cause of preterm delivery is intrauterine infection (23-25). Microbial invasion of the amniotic cavity has been previously reported to range between 4% and 38% of preterm deliveries (26, 27). However, intra-amniotic infection is usually subclinical (26, 27) and pathological events activating the cytokine signaling system can lead to preterm delivery. More specifically, bacterial endotoxins activate the decidua and the fetal membranes to produce tumor necrosis factor (TNF-α), interleukin (IL)-1ra, (IL-6), (IL-8), (ITAC) and ADAM metallopeptidase domain 8 (ADAM-8) (28-32). The latter stimulate prostaglandin synthesis and release which lead to uterine contraction. In parallel, activated metalloproteases are responsible for rupture of the membranes (28). It is assumed that adrenomedullin might regulate fetomaternal and vascular function during chorioamnionitis in an autocrine/paracrine manner (33). On the other hand, adrenomedullin, stimulated by IL-1, TNF-α and prostaglandin E1, participates in cervical ripening by inducing NO synthase (34-36).
Taking into account the properties of adrenomedullin, we hypothesized that this relatively new cytokine might be implicated in the preterm delivery pathway. Therefore, we aimed to determine amniotic fluid concentrations of adrenomedullin in the second trimester of pregnancy and correlate them with preterm delivery and preterm delivery with preterm premature rupture of membranes.
Patients and Methods
During the period September 2005 - December 2006, a retrospective cohort study was performed through collaboration of the Second Department of Obstetrics and Gynecology University of Athens Aretaieion Hospital and the Department of Fetal Medicine of Lito Maternity Hospital. The study population consisted of Greek women with singleton pregnancies who presented for genetic amniocentesis. Women with twin pregnancies, known history of uterine abnormalities, cone biopsies, significant vaginal bleeding, or fetal malformations were excluded from the study.
Preterm delivery was defined as labor before 37 weeks of gestation with regular uterine contractions (at least two uterine contractions/10 minutes during 30 minutes) in combination with the characteristic cervical changes (23, 37-40). Preterm premature rupture of the fetal membranes was defined as the rupture of the amniotic membranes with release of the amniotic fluid more than one hour before the onset of preterm labor (41). The gestational age was calculated from the last menstruation and was confirmed during routine ultrasound in the second trimester (16-19 weeks of gestation). Microbial invasion of the amniotic fluid was defined as a positive (PCR) for Mycoplasma hominis and Chlamydia trachomatis and/or growth of any bacteria (aerobic or anaerobic) in the amniotic fluid cultures except for coagulase-negative Staphylococcus, which was considered to be a skin contamination. All patients were followed until delivery for the occurrence of pregnancy complications. Maternal and perinatal data were entered into a database. The Ethics Committee of the Aretaieion Hospital, University of Athens approved the study. Each woman gave informed consent before enrolment in the study and completed a questionnaire including personal and family data.
Genetic amniocentesis for advanced maternal age, and/or increased risk for aneuploidy based on nuchal translucency ultrasound was performed under aseptic conditions with a 21-gauge needle in 362 women. The first 0.5 ml of amniotic fluid was discarded to avoid maternal contamination. Thereafter, 20 ml of amniotic fluid were aspirated from each woman with a 20-ml syringe. Of this, 15 ml were used for genetic diagnosis; 1 ml of the uncentrifuged amniotic fluid was immediately transported to the microbiological laboratory and was cultured for aerobic and anaerobic bacteria; 1 ml of the uncentrifuged amniotic fluid was tested by (PCR) for M. hominis and C. trachomatis detection. The remaining 3 ml of amniotic fluid were immediately placed in a refrigerator (4°C) and were centrifuged within the next 6 hours at 3000 xg and at 4°C for 10 minutes. The supernatant was stored in polypropylene tubes at −80°C until analysis.
Adrenomedullin concentrations in amniotic fluid were determined with an enzyme-linked immunosorbent assay (ELISA) [adrenomedullin (1-52) (human), EIA by Phoenix Pharmaceuticals INC (Burlingame, CA, USA)]. The amniotic fluid samples ran in duplicates. The intra- and interassay coefficients of variation (CV%) were <5% and <14%, respectively. The detection limit was 0.17 ng/ml. Laboratory personnel were blinded to the clinical history of the involved women.
The antibodies used in the adrenomedullin assay had a 100% crossreactivity for adrenomedullin (1-52) (human), adrenomedullin (1-50) (rat, mouse), and adrenomedullin (13-52) (human) and <1% for adrenomedullin-gly (human). No cross-reactivity existed for amylin (human), (cGRP) (human), cGRP 2 (human), endothelin-1 (human, porcine, rat, mouse, canine, bovine), α-atrial natriuretic peptide (α-ANP) (1-28) (human, canine), β-natriuretic peptide 32 (BNP-32) (human), nor natriuretic peptide precursor C (CNP-22) (human, porcine, rat).
Statistical analysis. The data, except for birth weight, were not normally distributed (Kolmogorov-Smirnov test) thus, nonparametric procedures were applied in the analysis. Independent samples t-test was used to detect differences between groups where continuous variables were normally distributed, otherwise, the Mann-Whitney U-test was applied. Pearson's Chi square test was used to detect differences between categorical variables. Conditional logistic regression analysis was applied to examine the possible associations of human adrenomedullin with preterm delivery. Women who gave a spontaneous preterm delivery were defined as cases (N=41) while for each case, a woman matched for age with normal pregnancy served as control (N=41). Furthermore, subgroup analysis was conducted in order to examine any possible association of human adrenomedullin levels with the incidence of preterm delivery with premature rupture of membranes. Nineteen women with preterm labor and premature rupture of membranes were defined as cases, while for every case a woman matched for maternal age delivering at term served as control (N=19). Cases and controls delivered either spontaneously or by caesarean section. STATA v8.2 (STATA Corp, Texas, USA) and SPSS v11.5 (SPSS Inc, Chicago, IL, USA) were used for the analysis. A p-value of <0.05 was considered to be statistically significant.
Results
Out of 362 women who were included in the study, 41 had spontaneous preterm delivery (incidence: 11.3%) and of these, 19 delivered preterm with premature rupture of membranes (incidence: 5.2%). Ten were excluded after amniocentesis indicated the presence of fetal chromosomal abnormalities (two with trisomy 18, two with trisomy 21, one with Turner syndrome, one with Klinefelter syndrome and four others with less common pathological karyotypes). Eight women were lost to follow-up. Four infants were delivered by cesarean section before the onset of labor for maternal (severe preeclampsia) or fetal reasons (compromised fetal growth or umbilical Doppler flow abnormalities). Two women who delivered within 30 days following amniocentesis were excluded from the study as their delivery was considered to be related to the procedure of amniocentesis (42-46).
Demographic data of women with preterm (cases) (N=41) and term delivery (controls) (N=41).
Demographic data of women with preterm labor and premature rupture of membranes (cases) (N=19) and women delivering at term (controls) (N=19).
The demographic data of the study population are presented in Tables I and II. No statistical significant differences were found in mean maternal age, gestational age at amniotic fluid sampling, or parity between the groups of women delivering preterm or at term. Amniotic fluid cultures for common bacteria in the participating mothers were negative, while M. hominis and C. trachomatis were identified in 2/362 and 2/362, respectively. However, one of the two women with Mycoplasma and one of the women with Chlamydia delivered preterm. No correlation was found between (PCR) detection of M. hominis and/or C. trachomatis with preterm labor incidence.
Adrenomedullin was detected in all amniotic fluid samples ranging from 0.33 ng/ml to 8.53 ng/ml. No significant differences were observed in the adrenomedullin concentrations between women with spontaneous preterm delivery (median: 1.33 ng/ml, range: 0.36-8.53 ng/ml) and those with normal (ful-term) pregnancy (median: 1.32 ng/ml, range: 0.33-4.07 ng/ml) (Figure 1A). Conditional regression analysis failed to demonstrate significant associations between adrenomedullin levels and preterm delivery. Furthermore, no significant association of adrenomedullin levels and preterm labor with preterm premature rupture of membranes was detected. Amniotic fluid levels of adrenomedullin in women with preterm premature rupture of membranes (median: 1.51 ng/ml, range: 0.65-2.92 ng/ml) did not differ from those found in women delivering at term (median: 1.22 ng/ml, range: 0.33-2.61 ng/ml) (Figure 1B).
A, Box and whiskers plots of the concentration of amniotic fluid adrenomedullin from women with spontaneous preterm delivery (cases) (median: 1.33 ng/ml, range: 0.36-8.53 ng/ml) and women with normal (fullterm) pregnancies (controls) median: 1.32 ng/ml, range: 0.33-4.07 ng/ml). Each box represents the median concentration with the interquartile range (25th and 75th percentiles). *Represent extreme values. B, Box and whiskers plots of the concentrations of amniotic fluid adrenomedullin from women with preterm labor with premature rupture of membranes (cases) (median: 1.51 ng/ml, range: 0.65-2.92 ng/ml) and women with normal (ful-term) pregnancies (controls) (median: 1.22 ng/ml, range: 0.33-2.61 ng/ml) Each box represents the median concentration with the interquartile range (25th and 75th percentiles).
Discussion
It is known that elevated plasma levels of adrenomedullin are positively correlated with plasma concentrations of inflammatory mediators such as TNF-α, IL-6, and IL-8 (47, 48). Furthermore, it has been stated that adrenomedullin production is increased by a wide range of inflammatory cytokines and growth factors (29-32). It is further known that the fetoplacental unit secretes large amounts of adrenomedullin in the amniotic fluid (6).
Previous studies have shown that amniotic fluid concentrations, on the one hand decreased as pregnancy advanced from the first to the second trimester (49), and on the other, increased from 21-28 weeks up to over 37 weeks of gestation (10, 49). Furthermore, amniotic fluid concentrations have been found to be significantly higher than those in the maternal plasma throughout gestation (10).
In addition, adrenomedullin concentrations in the amniotic fluid have been reported to be higher in complicated pregnancies (e.g. preeclampsia, (IUGR)) versus uncomplicated ones (7), because adrenomedullin is involved in blood flow regulation and growth stimulation (6, 50). Moreover, adrenomedullin has been shown to have antimicrobial activity and be elevated in cases of chorioamnionitis (7). Our study, with no documented infection, showed no statistically significant difference in adrenomedullin concentrations between women delivering preterm and at term. Yamashiro et al. measured adrenomedullin concentrations in amniotic fluid during genetic amniocentesis in 70 term and 3 preterm deliveries (51). In contrast to our results, they found that adrenomedullin levels in women with spontaneous preterm delivery were significantly higher than those in women who delivered at term. However, the number of patients included in the present study (N=41) exceed by far those published in that of Yamashiro et al. (n=3).
In conclusion, our study did not find a correlation between adrenomedullin concentration in second trimester amniotic fluid and incidence of either spontaneous preterm delivery or preterm labor with premature rupture of membranes. Therefore, we consider that adrenomedullin, determined in second trimester amniotic fluid cannot serve as a predictor of preterm delivery.
- Received June 20, 2009.
- Revision received October 1, 2009.
- Accepted October 5, 2009.
- Copyright © 2009 The Author(s). Published by the International Institute of Anticancer Research.
