Objective
Elevated interleukin-6 (IL-6) level in midtrimester amniotic fluid is associated with preterm delivery. We hypothesized that, in patients with elevated IL-6, vitamin C and alpha-fetoprotein may provide protection from spontaneous preterm delivery through antioxidant functions.
Study Design
Antioxidant potential of alpha-fetoprotein was assessed in vitro. Amniotic fluid was collected from a prospective cohort of patients who underwent midtrimester amniocentesis. In patients with IL-6 >600 pg/mL, alpha-fetoprotein, vitamin C, tumor necrosis factor–alpha, tumor necrosis factor receptors, and antioxidant capacity were compared between subjects with spontaneous preterm and term deliveries.
Results
Alpha-fetoprotein demonstrated 75% the antioxidant capacity of albumin in vitro. Of 388 subjects, 73 women had elevated IL-6 levels. Among these subjects, alpha-fetoprotein, but not vitamin C, was significantly lower in 9 women with preterm birth. Antioxidant capacity correlated with vitamin C and tumor necrosis factor receptors, but not with alpha-fetoprotein or pregnancy outcome.
Conclusion
Amniotic fluid alpha-fetoprotein, but not vitamin C, may protect against preterm birth in patients with elevated midtrimester IL-6 levels.
Preterm birth continues to be a serious and costly problem because it occurs in >12% of pregnancies or 1 in 8 births in the United States. In addition, prematurity accounts for most perinatal morbidity and neonatal death. Intrauterine infection and inflammation are associated strongly with preterm birth. Inflammatory cytokines such as interleukin-6 (IL-6) and tumor necrosis factor –alpha (TNF-α) are elevated significantly in amniotic fluid samples that are obtained at the time of preterm labor. Evidence also indicates that subclinical inflammation may start as early as 15-20 weeks gestation, as reflected by elevated levels of IL-6 in amniotic fluid samples from asymptomatic patients who subsequently delivered at <34 weeks of gestation.
Preterm premature rupture of membranes (PPROM) occurs in 2-4% of pregnancies but is associated with up to 40% of admissions of premature babies to the neonatal intensive care unit. Recognized risk factors for PPROM include intrauterine infection, midtrimester bleeding, smoking, and cocaine abuse. Substantial evidence indicates that reactive oxygen species are generated through these same processes. Specifically, free radicals are released by macrophages and neutrophils during microbial killing and by tissue ischemia followed by reperfusion because of transient vasoconstriction that is related to smoking or cocaine use. In addition, iron that is released from subchorionic hemorrhage is able to catalyze the conversion of superoxide to the hydroxyl radical, which is a highly active reactive oxygen species.
Vitamins C and E scavenge reactive oxygen species and may protect the chorioamnion from reactive oxygen species–induced damage during pregnancy. Vitamin C, a water soluble vitamin, protects collagen from enzymes (metalloproteinases) that degrade collagen. Vitamin C also stimulates fibroblasts in the chorioamnion to produce collagen and stabilizes the cross-linkages in the collagen triple helix that provide strength to the collagen at body temperature. Several clinical studies have shown a correlation between low plasma vitamin C levels and increased risk of PPROM. In addition, low vitamin C levels in the amniotic fluid also have been associated with PPROM.
To date, there are no studies that evaluate amniotic fluid vitamin C levels in the mid trimester in humans. In term, uncomplicated cases before the onset of labor, amniotic fluid vitamin C levels are significantly higher than vitamin C levels in maternal or fetal plasma. Studies in guinea pigs have revealed decreasing levels of amniotic fluid vitamin C with advancing gestational age. If vitamin C plays an antioxidant role in the intrauterine environment, higher levels in pregnancies that are at risk for spontaneous preterm delivery may be protective.
Alpha-fetoprotein is the main serum-binding globulin during fetal life. Fetal alpha-fetoprotein is synthesized primarily in the liver, although the gastrointestinal tract and yolk sac are responsible for some production as well. Despite the fact that hepatic production of alpha-fetoprotein remains relatively constant in the fetus throughout gestation, amniotic fluid alpha-fetoprotein peaks at 20 μg/mL at 12-14 weeks, which corresponds to the peak in fetal serum alpha-fetoprotein, and falls through gestation after this point. Presumed to be similar to albumin in its role, alpha-fetoprotein appears to have the antioxidant properties that albumin has been shown to exhibit, which includes the ability to protect tumor cells from the cytotoxic effects of TNF. It has recently become apparent that alpha-fetoprotein also is capable of modulating and inhibiting growth, which may explain the reason that macrosomia is associated with very low maternal serum alpha-fetoprotein. The interactions between alpha-fetoprotein and cell regulation are not well-characterized and are extremely complex, but it is clear that alpha-fetoprotein has significant regulatory effects on cell growth through apoptosis, signal transduction, and cytostatic or cytotoxic mechanisms.
Unexplained elevated levels of alpha-fetoprotein in maternal serum (ie, those not associated with neural tube or ventral wall defects) are associated with poor pregnancy outcome and include preterm birth, PPROM, fetal growth restriction, and preeclampsia. The association between “unexplained” elevated amniotic fluid alpha-fetoprotein and pregnancy outcome is less conclusive; 1 study shows an increase in the incidence of preeclampsia and preterm delivery, and other studies show no predictive value for subsequent pregnancy complications. If alpha-fetoprotein does have antioxidant properties, elevated levels in the amniotic fluid actually may provide protection from oxidative damage and prevent spontaneous preterm birth (SPTB).
We hypothesized that alpha-fetoprotein possesses similar antioxidant properties to albumin and that, in pregnancies that are at risk for preterm delivery because of midtrimester elevations in amniotic fluid IL-6, higher levels of vitamin C and alpha-fetoprotein in amniotic fluid may protect against SPTB. This could explain the reason that, even though elevated midtrimester amniotic fluid IL-6 is associated with SPTB, not all patients with elevated IL-6 levels deliver preterm. Furthermore, we hypothesize that the total antioxidant capacity of midtrimester amniotic fluid will correlate with vitamin C and alpha-fetoprotein levels as well as pregnancy outcome.
Materials and Methods
The study consisted of an in vitro study of antioxidant potential of human alpha-fetoprotein and albumin and a nested case-control study in a prospective cohort of patients who underwent genetic amniocentesis. For the in vitro study, samples of human alpha-fetoprotein (Lee Biosolutions, St. Louis, MO) and human albumin (Sigma-Aldrich, St. Louis, MO) were assayed for total antioxidant capacity in triplicate with the use of a commercially available kit (Randox, Crumlin, UK) at concentrations of 0, 0.1, 0.2, and 0.3 mmol/L. These concentrations are equivalent to 0-207,000 μg/mL of alpha-fetoprotein and would correspond to 0, 0.84, 1.7, and 2.5 multiples of the median (MoM) in amniotic fluid at 18 weeks gestation. In the assay, chromogen-linked metmyoglobin reacts with hydrogen peroxide to form a radical cation species (ABTS+). ABTS+ has a blue-green color that can be measured by spectroscopy at 734 nm. Added antioxidants inhibit the generation of ABTS+ and suppress color generation. Each assay was repeated 5 times. With simple linear regression on the optical absorption of the reagent dilutions, slopes and standard errors were calculated. Pairwise differences in slopes were compared with the use of t -tests.
The prospective cohort study was approved by the Institutional Review Board of the University of Rochester. Asymptomatic patients who were scheduled to undergo genetic amniocentesis from January 2003 through December 2007 were offered study enrollment before their procedures, and written consent was obtained. Amniocentesis was performed in the standard fashion by the sterile technique. At the time of the amniocentesis, the first 1-2 mL of amniotic fluid, which normally is discarded, were used for assays of IL-6, vitamin C, TNF-α, TNF receptors I and II, and total antioxidant capacity; amniotic fluid alpha-fetoprotein was determined to be part of standard clinical care. Patients with multiple gestations, chromosomal abnormalities, and fetal anomalies that would affect amniotic fluid alpha-fetoprotein levels (open neural tube defects or ventral wall defects) were excluded. Patients with bloody amniotic fluid samples were also excluded because this has been shown to elevate IL-6 levels markedly. Pregnancy outcome data that were collected from office and hospital records included gestational age at delivery, pregnancy complications, birthweight, and indications for delivery. SPTB was defined as delivery at <37 weeks’ gestation because of preterm labor or PPROM. Patients who delivered at term were included in the control group. All other patients, including those who delivered preterm because of preeclampsia, abruption, abnormal fetal testing, and other maternal or fetal complications were excluded from the current analysis.
Samples for amniotic fluid alpha-fetoprotein were sent to the clinical laboratory of the University of Rochester, and alpha-fetoprotein concentration was determined with the use of Access 2 Immunoassay Systems (Beckman Coulter, Inc, Fullerton, CA). The study amniotic fluid samples for all other assays were centrifuged and divided into 250-μL aliquots. Fluid samples for vitamin C analysis were diluted 1:1 with metaphosphoric acid for stabilization. All samples were frozen continuously at –80°C before batched analysis. Amniotic fluid samples for TNF-α, TNF receptors, and IL-6 analysis were diluted 1:4, 1:100, and 1:2, respectively, in DuoSet assay diluent (1% bovine serum albumin in phosphate buffered saline, pH 7.2-7.4; R&D Systems, Minneapolis, MD) before assay.
Vitamin C concentrations in the amniotic fluid were determined spectrophotometrically by the 2,4-dinitrophenylhydrazine method. Ascorbic acid was converted to dehydroascorbic acid in the presence of thiourea and copper sulfate. Dehydroascorbic acid was then coupled with 2,4-dinitrophenylhydrazine in 9.0 mol/L sulfuric acid to form a bis-2,4-dinitrophenylhydrazine derivative. When treated with 65% sulfuric acid, this derivative yields a stable brownish-red color that was measured with a spectrophotometer at 520 nm.
Amniotic fluid TNF-α, TNF receptors, and IL-6 levels were determined by enzyme-linked immunoabsorbent assay using DuoSet ELISA Development Reagents from R&D Systems. Samples were assayed in duplicate according to the commercial protocol with recombinant human TNF-α, TNF receptors I and II, and IL-6 standards. Tests of IL-6 standard recovery from amniotic fluid samples were also performed. Elevated IL-6 level was defined as levels >600 pg/mL because previous studies have shown an increased risk of adverse pregnancy outcome at these levels, with a sensitivity of 42% and a specificity of 92%.
Total antioxidant capacity of the amniotic fluid samples was measured with Cayman Chemical Antioxidant Assay Kits (Cayman Chemical, Ann Arbor, MI) in samples diluted 1:20 in assay buffer (0.9% sodium chloride, 0.01% glucose, and 5 mmol/L potassium phosphate; pH 7.4). The assay relies on the ability of antioxidants in the sample to inhibit the oxidation of ABTS (2,2’-azino-di-[3-ethylbenzthiazoline sulphonate]) by metmyoglobin. The capacity of the antioxidants in the sample to prevent ABTS oxidation was compared with standard concentrations of trolox, a water-soluble tocopherol analogue, and was quantified as molar trolox equivalents.
Results of these analyses were not provided to patients or providers. Results were compared with the use of medians and the Mann-Whitney U test for nonparametric data, χ 2 and Fisher’s exact testing for categoric variables, and 2-sided t tests that assumed unequal variances for normally distributed data. Logistic regression was used to assess possible interactions. Because of variations in amniotic fluid alpha-fetoprotein and vitamin C by gestational age, both were analyzed as multiples of the median.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
