The rate of preterm birth has steadily increased over the last two decades, to recently reach a high of 12.7%.1 Several mechanisms have been implicated in triggering preterm birth. Of these, genetic predisposition, excessive stretching, decidual hemorrhage, fetal stress, and inflammation and infection have received increased attention as pathophysiologic pathways, ultimately converging in a cascade of events leading to premature onset of myometrial contractions, rupture of membranes, and cervical ripening.2
In most animal species, the fetal hypothalamic-pituitary-adrenal axis plays a central role in the initiation of parturition.3 Still, the threshold and factors responsible for inducing fetal stress in human pregnancy remain elusive. One paradigm proposes that intraamniotic inflammation advances an outpouring of proinflammatory cytokines (interleukins-1, -2, -6, tumor necrosis factor-alpha), which, in turn, stimulate the release of placental corticotrophin-releasing hormone (CRH).4 It is believed that placental CRH activates the fetal pituitary-adrenocortical axis, which increases production of cortisol and dehydroepiandrosterone sulfate (DHEAS). In turn, these hormones exert a coordinated upregulation of key enzymes responsible for prostaglandin and estrogen synthesis, leading to activation of myometrial contractility and premature birth.5,6
Findings that link immune and neuroendocrine function may provide an explanation for the importance of the activation of the fetal hypothalamic pituitary adrenal gland axis in the response to inflammation.7 The general view that the reaction of the fetal adrenal gland to inflammation is aimed to avert end-organ damage is supported by evidence that demonstrates that cytokine activation leads to alterations in brain function ranging from behavioral changes to neuronal destruction.8,9 This damage can be prevented through the action of cortisol.7 That adrenal gland hormones have protective roles is also established by the data demonstrating that DHEAS and cortisol have opposing immuno-modulating effects: cortisol suppresses, while DHEAS enhances, the function of the immune system.10,11 Therefore, the ratio of cortisol to DHEAS (stress ratio) would be expected to reflect the ability of the fetus to engage adaptive mechanisms to limit the damaging effects of the inflammatory attack.12,13
Tools for molecular discoveries and fetal surveillance in utero are evolving rapidly to address old challenges and open new opportunities. We have previously shown that proteomic profiling of the amniotic fluid has all the attributes necessary to provide a rapid and accurate diagnosis of intraamniotic inflammation.14,15 We further presented evidence that evaluation of the fetal adrenal gland volume by three-dimensional ultrasonography can identify women at risk for impending preterm birth.16 However, whether the assessment of the fetal adrenal gland volume can provide information with regard to the extent of adrenocortical activation in the context of intraamniotic inflammation remained unexplored. The objective of this study was to estimate the impact of intraamniotic inflammation on fetal adrenal gland volume and fetal stress hormones cortisol and DHEAS in pregnancies complicated by preterm birth.
MATERIALS AND METHODS
We conducted a prospective study in 51 consecutive fetuses of mothers who had a clinically indicated amniocentesis to rule out infection or inflammation. The study period ranged from August 2004 to July 2007. The Yale University Human Investigation Research Board approved the study protocol. We obtained written consent from all the participants before the procedure.
All women presented at Yale New Haven Hospital with advanced cervical dilatation (more than 3 cm), preterm labor, or preterm premature rupture of the membranes (PROM). Eligible women had a singleton fetus at less than 34 weeks of gestation. Exclusion criteria included gestational age greater than 34 weeks, anhydramnios, viral infections (human immunodeficiency virus [HIV], hepatitis), and presence of fetal heart rate abnormalities (fetal bradycardia, recurrent, late, severe, and prolonged variable deceleration) that required immediate delivery.
We documented gestational age based on last menstrual period and early ultrasound evaluations (less than 20 weeks of gestation).17 We defined preterm labor as presence of regular uterine contractions associated with advanced cervical dilatation or effacement at less than 37 weeks of gestation. Rupture of the membranes was confirmed either by pooling on speculum examination, positive nitrazine and fern tests, or by a positive amino-dye test. Clinical chorioamnionitis was diagnosed in the presence of maternal fever (body temperature greater than 37.8ºC [100ºF]), maternal leukocytosis (more than 15,000 cells/mm3), uterine tenderness, foul-smelling amniotic fluid or visualization of pus at the time of the speculum examination, and maternal or fetal tachycardia.18 Women in preterm labor received steroids if they were at less than 34 weeks of gestation. Those with preterm PROM received steroids for enhancement of fetal maturity if at 24 or more weeks but less than 32 weeks of gestation. Antibiotic therapy was recommended as clinically indicated.19 In women with preterm PROM a nonstress test was indicated at least twice daily. A biophysical profile followed in the presence of a nonreassuring fetal heart rate. Induction of labor or a surgical delivery (indicated delivery) was recommended for clinical indications, such as amniotic fluid laboratory results suggestive of intraamniotic infection, nonreassuring fetal heart rate, breech presentation, and in the context of preterm PROM for prolapsed umbilical cord, fetal lung maturity, and/or gestational age of 34 weeks or more.20 The amniocentesis procedure and the decision to deliver the fetus were recommended by the attending physician, independently of our study protocol. We collected data prospectively on whether delivery of the fetus occurred subsequent to spontaneous onset of uterine contractions or secondary to indicated delivery (gestational age more than 34 weeks in the context of preterm PROM, evidence of intraamniotic infection, nonreassuring fetal heart rate). Amniotic fluid was retrieved by ultrasound-guided amniocentesis under sterile conditions. The clinical laboratory tests performed for the purpose of diagnosing inflammation or infection were glucose, lactate dehydrogenase activity, Gram stain, white blood cell count, and microbial cultures for aerobic and anaerobic bacteria, Ureaplasma, and Mycoplasma species. These results were available for clinical management. Remaining amniotic fluid was spun at 3,000g at 4ºC for 20 minutes, aliquoted, and stored at –80ºC for research purposes.
This ultrasonographic technique for measuring the fetal adrenal gland volume has been previously described.16 Briefly, scans were performed with a Voluson 730 or E8 system (Voluson Expert, General Electric Medical Systems, Milwaukee, WI) equipped with a 4–8 MHz three-dimensional probe. We first visualized the fetal abdomen in a cross-sectional way to allow identification of the fetal spine at the 10 or 2 o'clock positions (Fig. 1). Subsequently, we optimized the sector angle and depth of penetration. We set the volume sweep angle between 55º and 80º to allow acquisition of the adrenal volume within one block. At least three volume blocks were obtained for each fetus. Ultrasonographic weight estimations were performed for each fetus at the same time.17 For the purpose of this analysis one examiner who was unaware of clinical presentation or outcome (O.M.T.) verified image quality by multi-planar evaluation (transverse, frontal, and longitudinal) and calculated all fetal adrenal volumes using VOCAL software (Virtual Organ Computer-aided AnaLysis, 4D View; General Electric Medical System). The intraobserver coefficient of variation for calculation of adrenal gland volume was less than 2.5%. As previously described, the adrenal gland volume corrected for estimated fetal weight yields a sonographic variable (corrected adrenal gland volume) that is independent of gestational age.16,17
To confirm or exclude intraamniotic inflammation, we used SELDI-TOF (surface-enhanced laser desorption ionization time of flight) mass spectrometry (PBSIIC, Ciphergen Biosystems, Fremont, CA). The criteria for assessing intraamniotic inflammation were previously reported.14 Briefly, immediately after amniocentesis we searched for the presence of for proteomic biomarkers (peaks corresponding to neutrophil defensin-1, neutrophil defensin-2, calgranulin C, and calgranulin A) making up the Mass Restricted score. A value of 1 is assigned if a particular biomarker peak is present and 0 if absent. The Mass Restricted score ranges thus from 0 to 4, depending upon the absence or presence of each of the four protein biomarkers. A score of 3–4 indicates the presence of “severe” intraamniotic inflammation.14 All SELDI tracings were scored “blindly” by investigators unaware of either clinical presentation or outcome.
At Yale New Haven Hospital, placental pathologic study is a routine part of the evaluation of a pregnancy complicated by preterm labor or preterm PROM. The diagnosis of histologic chorioamnionitis was based on infiltration by the neutrophils of the chorionic plate and amniochorionic membranes. We used well-recognized histologic stages and grading systems to establish the severity of histologic inflammation.21,22 The perinatal pathologist who examined all the placentas was unaware of the results of adrenal measurements, amniotic fluid, and umbilical cord blood analysis.
After amniocentesis, each woman was followed prospectively up to delivery. For research purposes, immediately after delivery, umbilical cord blood specimens were retrieved in a plain tube, spun at 3,000g, 4ºC for 10 minutes., aliquoted, and serum stored at –80ºC.
Enzyme-linked immunosorbent assay (ELISA) assays were performed in duplicate according to manufacturers' instructions. Interleukin-6 (IL-6) levels (Pierce-Endogen, Rockford, IL) were measured in both amniotic fluid and umbilical cord blood. Levels of cortisol (R&D Systems, Minneapolis, MN) and DHEAS (Alpha Diagnostic, International, San Antonio, TX) were performed only in cord blood. The minimal detectable concentrations for IL-6, cortisol, and DHEAS were 1 pg/mL, 0.071 ng/mL, and 0.005 mcg/mL, respectively, and the inter- and intraassay coefficients of variation were below 10, 7, and 11.5%, respectively. The results of the amniotic fluid analysis, adrenal gland volume, and histologic analysis of the placenta for several of these women and their neonates were previously used in studies aimed to identify presence in the amniotic fluid of biomarkers characteristic of intraamniotic inflammation or evaluate the role of fetal adrenal gland volume in identifying women at risk for preterm birth.14,16
The Kolmogorov-Smirnov method was used for data normality testing. Data are presented as average and standard deviation if normally distributed or, otherwise, as median and range. Statistical analyses were performed with Sigma Stat 2.03 (SPSS Inc., Chicago, IL), MedCalc (MedCalc Software, Broekstraat, Belgium), and StatsDirect (Stats Direct Ltd., Cheshire, UK) statistical software programs. Comparisons between the groups were performed using Student t tests or Mann-Whitney rank sum tests, as appropriate. Proportions were compared with Fisher exact or χ2 tests. Statistical analysis was completed after logarithmic transformation of the cytokine levels before correlation analyses using Pearson product moment correlation coefficient. Multiple stepwise linear or logistic regression analyses were used to adjust P values for potential influences of gestational age or other parameters. In our previous study, we found that the corrected fetal adrenal gland volume was inversely correlated with duration from scan to delivery (r=–0.445, P<.001), a biologically significant outcome.16 Therefore, in a two-sided sample size calculation, we estimated that a minimum of 39 cases are required to detect a similar level of correlation, using 80% power and an alpha of 5%. A sample size of 51 total cases would allow statistical significance in Pearson correlation analysis for an r=0.386. Throughout the study we judged P<.05 as indicating statistical significance.
In Table 1 we present the demographic, clinical, and outcome characteristics of our study population. Women with intraamniotic inflammation were of lower gestational age at both amniocentesis and delivery. The prevalence of a spontaneous delivery was 65% (33 of 51). Indicated preterm delivery was more frequently used in women with intraamniotic inflammation, and therefore, this group had a shorter amniocentesis-to-delivery interval. Neonates delivered by women with intraamniotic inflammation were of lower birth weight and had lower 1-minute Apgar scores. However, these differences were rendered nonsignificant when correcting for gestational age at delivery in multivariable regression analysis. Placental pathology of cases with intraamniotic inflammation more frequently confirmed presence of histologic chorioamnionitis (P<.001), independently of gestational age at birth.
We present the results of the amniotic fluid analysis in Table 2. Intraamniotic inflammation was characterized by a lower glucose concentration, higher lactate dehydrogenase activity, and higher white blood cell count. The overall prevalence of a positive microbial culture result in our study population was 38% (13 of 51), with a higher incidence in women with intraamniotic inflammation. The amniotic fluid IL-6 levels in this group were significantly higher compared with the noninflammation group.
We observed that the cord serum concentrations of IL-6 were significantly higher in fetuses delivered by women with intraamniotic inflammation (Table 2), confirming the higher inflammatory state of fetuses in this group. In contrast, the median concentrations of cortisol and DHEAS were not significantly different in fetuses delivered by women with and without intraamniotic inflammation. However, fetuses delivered in the context of intraamniotic inflammation had a significantly lower stress ratio (P=.034) (Table 2 and Fig. 2A). These results were maintained after adjusting for gestational age, presence of uterine contractions, indicated delivery, amniocentesis-to-delivery interval, race, and steroid exposure in multivariable regression analysis.
We found that fetuses of women with intraamniotic inflammation had lower estimated fetal weights but higher corrected adrenal gland volumes compared with those of women without inflammation (Table 3 and Fig. 2B). Fetuses with corrected adrenal volumes greater than 422 mm3/kg×103 had higher median cord blood IL-6 levels (P=.034).
There was a direct relationship between cord blood IL-6 and corrected fetal adrenal gland volume (r=0.372, P=.019), cord blood cortisol (r=0.428, P=.010), and DHEAS (r=0.521, P<.001). We previously demonstrated by receiver operating characteristic analysis that a corrected adrenal gland volume greater than 422 mm3/kg×103 predicts delivery within 5 days from the time of sonographic asessment.16 However, there was no correlation of the corrected adrenal gland volume and either cortisol (r=–0.117, P=.495), DHEAS (r=0.218, P=.218), or the fetal stress ratio (r=–0.200; P=.240). These results were maintained after correcting for the mode of delivery (spontaneous onset of labor or indicated preterm delivery) in stepwise multivariable analysis.
The present study was enabled by our previous findings suggesting that measurement of fetal adrenal gland volume by three-dimensional ultrasonography may be a useful clinical marker for impending preterm birth.16 The long-held view of a complex cross-talk between the innate immune and endocrine systems led us to hypothesize that fetal inflammation is accompanied by alterations in the molar ratio of cortisol/DHEAS (fetal stress ratio). If true, we projected that the level of fetal adrenocortical activation can be recognized noninvasively in utero through an increased corrected adrenal gland volume. Here again, we provided evidence for involvement of cytokines and stress hormones in the fetal adaptive response to inflammation but with particular nuances. Our results indicate that intraamniotic inflammation is associated with increased circulating levels of fetal IL-6, which, in turn, correlated with the fetal adrenal gland volume. We further found that, although the absolute serum levels of cortisol and DHEAS were not significantly different, fetuses exposed to inflammation had lower ratios of these stress hormones. Interestingly, we determined that the corrected fetal adrenal gland volume neither reflects the fetal circulating levels of cortisol, DHEAS, nor the fetal stress ratio, suggesting that these variables are not necessarily interdependent.
The roles of fetal cortisol and DHEAS in the onset of human parturition are less understood than in other animal species.3 Several studies reported increased cortisol levels,23,24 whereas others found no change in relationship to onset of term labor in women.25,26 Controversy also exists related to preterm labor or intrauterine infection.27,28 Our findings are concurrent with the prior observations of Yoon et al27 that fetuses delivered by women with intraamniotic infection do not have higher fetal plasma cortisol or DHEAS concentrations. However, their reported differences in fetal adrenal response based on spontaneous onset of labor and time interval to delivery27 were not confirmed by either our or others' data.28 Possible explanations may be related to differences in study populations or to our focus on intraamniotic inflammation rather than to a positive amniotic fluid culture result as evidence of infection.
Inflammation represents a highly orchestrated process designed to prevent tissue injury.29 In pregnancy, inappropriate control and unsuccessful resolution of an inflammatory process shifts its role from protective to damaging, with grave consequences for the fetus.30 The inflammatory cytokine IL-6 has emerged as an important regulator of this immunological switch through differential control of neutrophil recruitment, activation, and apoptosis.31 Although previous research has demonstrated increased fetal circulating levels of IL-6 in pregnancies complicated by intraamniotic infection,32 to our knowledge the relationship between the intensity of the fetal inflammatory response and adrenal gland volume has not been examined before this study.
Activation of the hypothalamic-pituitary-adrenal axis represents one of several important responses to critical illnesses. Stress in adults is associated with increased production of adrenal cortisol and DHEAS, which have opposing immunoregulatory effects.33 In vitro, cortisol significantly decreases neutrophil superoxide generation, an effect prevented by co-incubation with DHEAS.12 Healthy young adults respond to stress with a lower cortisol-to-DHEAS ratio, which is a feature of favorable prognosis. In contrast, a high stress ratio is considered a marker of increased mortality in both septic and posttraumatic shock12,13 Although our observation of an increased fetal adrenal gland volume intuitively suggests that activation of the pituitary hypothalamic axis in response to inflammatory stimuli also occurs in the fetus, this concept is not supported by our cortisol or DHEAS data, which did not differ in absolute levels with the presence or extent of inflammation. Several explanations may be pertinent to these apparently contradictory observations. For example, it is possible that the fetal adrenal gland at the time of inflammatory aggression is still functionally immature and, thus, unable to provide the robust endocrinological response seen in adults. This concept is supported by prior studies showing that, despite severe illness, very preterm neonates have relatively low basal cortisol and DHEAS concentrations.34 Therefore, our finding that intraamniotic inflammation is associated with a lower stress ratio suggests that even the premature fetus is able to initiate a stress response aimed at buffering the negative effects of inflammation, albeit at a more subtle level. We propose that the enlarged adrenal glands in the context of a lower stress ratio are features of a complex adaptive stress response mounted in anticipation of birth by premature fetuses that have an overall reduced ability to respond to inflammation.
1. Martin JA, Hamilton BE, Sutton PD, Ventura SJ, Menacker F, Kirmeyer S. Births: final data for 2004. Natl Vital Stat Rep 2006;55:1–101.
2. Lockwood CJ. Testing for risk of preterm delivery. Clin Lab Med 2003;23:345–60.
3. Challis JR. CRH, a placental clock and preterm labour. Nat Med 1995;1:416.
4. Park JS, Park CW, Lockwood CJ, Norwitz ER. Role of cytokines in preterm labor and birth. Minerva Ginecol 2005;57:349–66.
5. Unno N, Wu WX, Wong CH, Bennett PR, Shinozuka N, Nathanielsz PW. Prostaglandin regulation of fetal plasma adrenocorticotropin and cortisol concentrations in late-gestation sheep. Biol Reprod 1998;58:514–9.
6. Dell'Acqua S, Parlati E, Lucisano A, Plotti G, Serri F, Bompiani A. Evaluation of the feto-placental function by means of intraamniotic administration of dehydroepiandrosterone-sulphate. J Perinat Med 1979;7:149–60.
7. Reichlin S. Neuroendocrine-immune interactions. N Engl J Med 1993;329:1246–53.
8. Ma XC, Gottschall PE, Chen LT, Wiranowska M, Phelps CP. Role and mechanisms of interleukin-1 in the modulation of neurotoxicity. Neuroimmunomodulation 2002–2003;10:199–207.
9. Chao CC, Hu S, Sheng WS, Bu D, Bukrinsky MI, Peterson PK. Cytokine-stimulated astrocytes damage human neurons via a nitric oxide mechanism. Glia 1996;16:276–84.
10. Mitchell JB, Dugas JP, McFarlin BK, Nelson MJ. Effect of exercise, heat stress, and hydration on immune cell number and function. Med Sci Sports Exerc 2002;34:1941–50.
11. Khorram O, Vu L, Yen SS. Activation of immune function by dehydroepiandrosterone (DHEA) in age-advanced men. J Gerontol A Biol Sci Med Sci 1997;52:M1–7.
12. Butcher SK, Killampalli V, Lascelles D, Wang K, Alpar EK, Lord JM. Raised cortisol: DHEAS ratios in the elderly after injury: potential impact upon neutrophil function and immunity. Aging Cell 2005;4:319–24.
13. Arlt W, Hammer F, Sanning P, Butcher SK, Lord JM, Allolio B, et al. Dissociation of serum dehydroepiandrosterone and dehydroepiandrosterone sulfate in septic shock. J Clin Endocrinol Metab 2006;91:2548–54.
14. Buhimschi IA, Christner R, Buhimschi CS. Proteomic biomarker analysis of amniotic fluid for identification of intra-amniotic inflammation. BJOG 2005;112:173–81.
15. Buhimschi CS, Bhandari V, Hamar BD, Bahtiyar MO, Zhao G, Sfakianaki AK, et al. Proteomic profiling of the amniotic fluid to detect inflammation, infection, and neonatal sepsis. PLoS Med 2007;4:e18.
16. Turan OM, Turan S, Funai EF, Buhimschi IA, Copel JA, Buhimschi CS. Fetal adrenal gland volume: a novel method to identify women at risk for impending preterm birth. Obstet Gynecol 2007;109:855–62.
17. Hadlock FP, Deter RL, Harrist RB, Park SK. Computer assisted analysis of fetal age in the third trimester using multiple fetal growth parameters. J Clin Ultrasound 1983;11:313–6.
18. Hauth JC, Gilstrap LC 3rd, Hankins GD, Connor KD. Term maternal and neonatal complications of acute chorioamnionitis. Obstet Gynecol 1985;66:59–62.
19. Premature rupture of membranes. ACOG Practice Bulletin No. 80. Clinical management guidelines for obstetrician-gynecologists. American College of Obstetricians and Gynecologists. Obstet Gynecol 2007;109:1007–19.
20. Naef RW 3rd, Allbert JR, Ross EL, Weber BM, Martin RW, Morrison JC. Premature rupture of membranes at 34 to 37 weeks' gestation: aggressive versus conservative management. Am J Obstet Gynecol 1998;178:126–30.
21. Naeye RL. Disorders of the placenta and decidua. In: Disorder of the placenta, fetus and neonate: diagnosis and clinical significance. St. Louis (MO): Mosby; 1992. p. 118–247.
22. Salafia CM, Weigl C, Silberman L. The prevalence and distribution of acute placental inflammation in uncomplicated term pregnancies. Obstet Gynecol 1989;73:383–9.
23. Goldkrand JW, Schulte RL, Messer RH. Maternal and fetal plasma cortisol levels at parturition. Obstet Gynecol 1976;47:41–5.
24. Leong MK, Murphy BE. Cortisol levels in maternal venous and umbilical cord arterial and venous serum at vaginal delivery. Am J Obstet Gynecol 1976;124:471–3.
25. Ohrlander S, Gennser G, Eneroth P. Plasma cortisol levels in human fetus during parturition. Obstet Gynecol 1976;48:381–7.
26. Martinsen K, Peltola J, Tervila L, Virtanen A. Umbilical cord cortisol and arterial pH levels in spontaneous and induced labors. Obstet Gynecol 1982;59:171–5.
27. Yoon BH, Romero R, Jun JK, Maymon E, Gomez R, Mazor M, et al. An increase in fetal plasma cortisol but not dehydroepiandrosterone sulfate is followed by the onset of preterm labor in patients with preterm premature rupture of the membranes. Am J Obstet Gynecol 1998;179:1107–14.
28. Gravett MG, Hitti J, Hess DL, Eschenbach DA. Intrauterine infection and preterm delivery: evidence for activation of the fetal hypothalamic-pituitary-adrenal axis. Am J Obstet Gynecol 2000;182:1404–13.
29. Lawrence T, Gilroy DW. Chronic inflammation: a failure of resolution? Int J Exp Pathol 2007;88:85–94.
30. Murthy V, Kennea NL. Antenatal infection/inflammation and fetal tissue injury. Best Pract Res Clin Obstet Gynaecol 2007;21:479–89.
31. McLoughlin RM, Hurst SM, Nowell MA, Harris DA, Horiuchi S, Morgan LW, et al. Differential regulation of neutrophil-activating chemokines by IL-6 and its soluble receptor isoforms. J Immunol 2004;172:5676–83.
32. Yoon BH, Romero R, Park JS, Kim M, Oh SY, Kim CJ, et al. The relationship among inflammatory lesions of the umbilical cord (funisitis), umbilical cord plasma interleukin 6 concentration, amniotic fluid infection, and neonatal sepsis. Am J Obstet Gynecol 2000;183:1124–9.
33. Cupps TR, Fauci AS. Corticosteroid-mediated immunoregulation in man. Immunol Rev 1982;65:133–55.
34. Kari MA, Raivio KO, Stenman UH, Voutilainen R. Serum cortisol, dehydroepiandrosterone sulfate, and steroid-binding globulins in preterm neonates: effect of gestational age and dexamethasone therapy. Pediatr Res 1996;40:319–24.