Preterm birth continues to be a major public health problem with lasting family and societal repercussions.1 Despite tremendous research effort, prevention strategies have failed, and the prevalence of preterm birth in the United States reached 12.3% in 2003.2
The etiology of most preterm births remains elusive and is likely multifactorial, with many pathological-physiological pathways being involved, such as excessive stretching, oxidative stress, decidual hemorrhage, and infection. To wit, the past efforts to reduce prematurity may have failed because each causative pathway requires a distinct therapeutic intervention.3 Of the multiple mechanisms implicated, it was suggested that the fetus may be in control of the timing of birth via activation of its own hypothalamic-pituitary-adrenal axis.4
Current thinking is that placental corticotropin-releasing hormone (CRH) promotes activation of the fetal hypothalamic-pituitary-adrenal axis, which in turn stimulates the production of cortisol by the fetal adrenal gland, followed by activation of a cascade of events that suppress the mechanisms responsible for uterine quiescence.5 This paradigm is supported by the observational evidence that, at term, before the onset of labor, the weight of the fetal adrenal gland is the same as that of the adult adrenal gland.6 Further, the observation that the fetal adrenal gland increases in size in pregnancies complicated by preterm birth makes it extremely likely that the fetus is indeed an active participant in the process of premature parturition.7 The usefulness of fetal adrenal gland volume as a disease marker that can aid development of new strategies for the prediction of preterm birth has not been previously investigated.8
Most investigators agree that ultrasonography remains an accurate way to define the size of the normal fetal adrenal gland during gestation.9–11 Although 2-dimensional (2D) ultrasonography remains the main imaging tool for assessing fetal anatomical structures in utero, refinements and the development of new software packages have allowed 3-dimensional (3D) ultrasonography to be more widely used in routine clinical practice12 and to add diagnostic value by allowing evaluation of organ structures and growth in volume, including that of the fetal adrenal gland.11,13 A clinical challenge in perinatology that can be met with quantification of fetal adrenal gland volume is prediction of preterm birth. In this study we examined the clinical use of 3D ultrasound imaging of the fetal adrenal gland volume as a marker of preterm birth.
MATERIALS AND METHODS
We prospectively assessed the relationship between fetal adrenal gland volume measurement by 3D ultrasound and preterm birth (birth<37 weeks) in 150 women evaluated at Yale New Haven Hospital from August 2005 to July 2006 (Fig. 1). Twenty-four patients were excluded from the final analysis: in 22 women (14.5%) our evaluation of the fetal adrenal gland volume was unsatisfactory because of poor data acquisition, and 2 patients were lost to follow-up. The fetal adrenal gland volume was successfully examined in 126 women. Of these, 53 consecutive women presented to the labor and birth unit or the antepartum high- and low-risk units with symptoms of preterm labor or preterm premature rupture of membranes (PROM). These women made up the preterm labor/preterm PROM group, while 73 asymptomatic healthy women undergoing ultrasonography as part of their routine prenatal care served as control patients. We enrolled our control patients consecutively based on the availability of two of the investigators (O.M.T., S.T.). For preterm labor/preterm PROM participants, inclusion criteria consisted of a singleton fetus, uterine contractions refractory to tocolysis, and advanced cervical dilatation (3 cm or more) or preterm PROM. The adrenal gland volume measurements in preterm labor/preterm PROM subjects were performed within 24 hours after admission. Exclusion criteria included suspected fetal growth restriction (sonographic fetal weight less than the 10th percentile for gestational age), maternal medical complications (ie, hypertension, preeclampsia, diabetes, thyroid or adrenal diseases), and presence of fetal heart rate abnormalities at enrollment (bradycardia or prolonged variable decelerations). The Human Investigation Committee of Yale University approved our study protocol, and written informed consent was obtained from all participants before enrollment.
Gestational age was established based on an ultrasonographic examination before 20 weeks in all instances. We defined preterm labor as the presence of regular uterine contractions in the context of cervical effacement, advanced cervical dilatation, or both. We confirmed a diagnosis of preterm PROM by visualization of amniotic fluid “pooling” through the cervical os during speculum examination, “Nitrazine,” “ferning,” or amniocentesis-dye–positive tests. Management of the patients was left to the discretion of the clinical team. In the absence of signs or symptoms of clinical chorioamnionitis (fever over 38.0°C, abdominal tenderness, fetal tachycardia), and/or abnormalities of fetal heart rate (variable or late decelerations), and/or abruption, preterm PROM was managed expectantly. In preterm PROM patients, digital examinations were not permitted. In the preterm labor/preterm PROM group, 80% of patients received corticosteroids for lung maturity after adrenal gland volume measurement if at less than 32 weeks of gestational age. Women with preterm PROM also received antibiotic therapy (ampicillin/erythromycin or clindamycin) if at less than 34 weeks of gestational age. Women were monitored by cardiotocography at least twice daily for the presence of fetal heart abnormalities, uterine contractions, or both. After adrenal gland volume data acquisition, each woman was followed prospectively to the point of delivery. The delivery interval was defined as the time passed (days or hours) from adrenal gland volume acquisition to delivery of the fetus. The adrenal gland volume results were not used for clinical management.
We performed all our scans with the Voluson 730 system (Voluson Expert, General Electric Medical Systems, Milwaukee, WI), equipped with a 4–8 MHz curved array transducer. 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. 2). Harmonic imaging was used as needed to provide a clear contrast between tissue structures. Subsequently, we optimized the sector angle and depth of the penetration. We set the volume sweep angle to between 55° and 80°, depending on gestational age, to allow both adrenal gland and kidney volumes to be acquired within the same block. The duration of the scan was adjusted to obtain the best image resolution. We obtained at least 3 volume blocks in each fetus, and after each data recording, one examiner (O.M.T.) verified image quality by means of multiplanar evaluation (transverse, frontal, and longitudinal). For each woman we performed fetal weight estimations by combining biparietal diameter, abdominal circumference, and femur length.14
Calculation of the adrenal gland volumes was performed in all instances by the same observer (O.M.T.) using VOCAL (Virtual Organ Computer-Aided Analysis, 4D View; General Electric Medical System, Milwaukee) software. For the multiplanar technique, we used the transverse plan as a reference, and for a better visualization, we chose to analyze the volume of the fetal adrenal gland closest to the probe (79% of the cases on the right side). For best visualization, we rotated the adrenal and kidney organs in the X- and Y-axes by keeping the Z-axis in a fixed position. The adrenal gland and kidney volumes were assessed separately. To calculate the volumes, we used the cursor to manually outline six sections of the adrenal gland and kidney contour. We first set the upper and lower anatomical edges of the adrenals and kidneys as the first and last images delineating the volume for each organ. By design, the software keeps track of the distance between the first and last slices of the analysis. We also measured the length, width, and depth of each kidney and adrenal gland by using conventional 2D ultrasound views. The intraobserver and interobserver coefficients of variation for calculation of adrenal gland volume were 2.5% and 3.5%, respectively.
The Kolmogorov-Smirnov test was used for data normality testing. Comparisons between the two groups were performed with Student t tests or Mann-Whitney rank sum tests. Proportions were compared with Fisher exact or χ2 tests (SPSS 11; SPSS Inc, Chicago, IL; and MedCalc; MedCalc Software, Mariakerke, Belgium). Linear and nonlinear function curve fitting was performed with TableCurve 2D 5.01 (Systat Software, Point Richmond, CA). We examined the test accuracy, sensitivity and specificity values, and positive and negative likelihood ratios by examining the distribution of the adrenal gland volume on receiver operating characteristic curves after correction for estimated fetal weight. Stepwise logistic and linear regressions were used for multivariable analysis with dependent and independent variables as specified, with P<.05 used for variable entry and P≥.1 for variable removal as performed with SigmaStat 2.0 (Systat Software). We judged P<.05 as indicating statistical significance. For sample size calculation, we used the accuracy to predict delivery within 5 days. Assuming that a valuable ultrasound parameter will be able to correctly predict the outcome in 80% of cases (compared with 50 accuracy for chance alone), we estimated a minimum required sample size of 45, given 80% power and a confidence coefficient of 95% (one-sample test of hypothesis for proportion; PASS 2005, NCSS Statistical Software, Kaysville, UT).15
In Table 1 we present the clinical characteristics of the 126 women included in our analysis. As expected, a history of preterm birth was more frequently identified in women with preterm labor/preterm PROM. Women with preterm labor/preterm PROM delivered at an earlier gestational age than did control patients, with almost half of them being delivered within 5 days from the time of enrollment. The fetal adrenal grand volume was successfully examined in 86.5% of the cases. Table 2 summarizes the ultrasonographic findings of the fetuses. We found that fetuses of the preterm labor/preterm PROM women had significantly larger adrenal gland volume than asymptomatic controls, despite lack of differences in ultrasonographic estimated fetal weight and abdominal circumferences.
Sixty-seven women delivered at term and were considered “normal” pregnant controls. By using regression analysis with linear and nonlinear models in this subgroup we determined that fetal adrenal gland volume and kidney volume changed significantly in a direct linear relationship with estimated fetal weight (fetal adrenal gland volume r=0.88, P<.001; kidney volume r=0.81, P<.001) (Figs. 3A and 3B). In multivariable regression analysis we found that the estimated fetal weight remained the strongest predictor of adrenal gland volume against all other independent variables entered into the model (age, gravidity, parity, gestational age, abdominal circumference, kidney volume). Because of the strong observed association between estimated fetal weight and adrenal gland volume, we calculated a corrected adrenal gland volume index as the ratio of adrenal gland volume to estimated fetal weight (mm3/kg). Women with preterm labor/preterm PROM had significantly higher corrected adrenal gland volume indices than controls (Fig. 4A and Table 2), whereas corrected kidney volume was not significant. By using receiver operating characteristic analysis (Fig. 4B), we determined that a corrected adrenal gland volume index of more than 422 mm3/kg was able to predict preterm birth within 5 days from the date of the scan with the highest accuracy (receiver operating characteristic area: 0.957, 95% confidence interval 0.91–0.99). The sensitivity, specificity, and positive and negative likelihood ratios were 92.0%, 99.0%, 93.5, and 0.08, respectively (Table 3).
In multiple logistic regression analysis, a corrected adrenal gland volume index greater than 422 mm3/kg was the only independent variable predictive of preterm birth within 5 days from the volume evaluation (P<.001), classifying correctly 97.6% of the cases. Variables excluded from the model were age, parity, gravidity, gestational age, kidney volume, and status of the membranes (P>.1). Furthermore, within the group of women presenting with preterm labor/preterm PROM, the median interval to delivery for fetuses with a corrected adrenal gland volume index greater than 422 mm3/kg was 1 day (range 0–5 days), significantly shorter than that of the fetuses with corrected adrenal gland volume index of 422 mm3/kg or less (median 41.5 days, range 0–100 days, P<.001). Conversely, women in the preterm labor/preterm PROM group who delivered at less than 37 weeks of gestation had higher corrected adrenal gland volume index values than did the women who delivered at term (corrected adrenal gland volume index at preterm delivery: 415.6±28.1 mm3/kg versus corrected adrenal gland volume index at term delivery: 301.6±24.1 mm3/kg, P=.039).
Our research was motivated by the fact that the identification of markers that are aimed to aid in the development of new strategies for the prediction, diagnosis, and prevention of preterm birth and its associated complications is critical.16 Our findings provide support for the hypothesis that the fetus itself contributes to initiation of parturition, and by using volume analysis of 3D ultrasonographic images, we provide evidence that an enlarged adrenal gland volume can potentially be used to provide rapid and accurate information about the risk of preterm birth in women with preterm labor.
Traditionally, clinicians have had to rely on the patient’s obstetric history and symptoms to assess her risk of preterm birth. Yet, relying on symptoms is associated with an unacceptably high false-positive rate, which can be as high as 70%.17 Therefore, several maternal and fetal biomarkers have been proposed for the accurate prediction of preterm birth (cytokines, corticotrophin-releasing hormone, C-reactive protein, fetal fibronectin).18–24 Unfortunately, despite tremendous effort, these markers have not proved to be sufficiently sensitive or specific, either alone or in combination, to help avoid the increasing rate of preterm birth.25 Hence, they continue to be poor instruments for clinical decision making.
Additional research is urgently needed to evaluate new methods for early detection of women at risk for preterm birth to allow implementation of the appropriate prevention strategies. Over the years, investigators were mostly interested in providing evidence toward maternal rather than fetal markers of preterm birth. To date, however, the best predictive marker of spontaneous preterm birth is not maternal but rather fetal in its origin—fetal fibronectin.26 Based on the previous findings that the onset of birth is initiated through the fetal genome27 and expressed via fetal endocrine pathways, we provide persuasive evidence that discovery of an enlarged fetal adrenal gland by 3D ultrasonography can be a highly accurate and noninvasive marker of preterm birth. Our finding is clearly in agreement with the results of other studies, which showed that, in primates and humans, activation of the fetal hypothalamic-pituitary-adrenal axis results in increased output of dehydroepiandrosterone, dehydroepiandrosterone sulfate, androstenedione, and cortisol, either at term or preterm.28–30 The specialized ability of cortisol to promote placental CRH production, as well as that of other placenta-derived growth factors, easily explains the increased weight and hypertrophy of the fetal adrenal gland, identified postmortem in cases without an apparent cause of preterm birth.7 This increase in the fetal adrenal gland volume weight apparently occurs independently of kidney or thymic weight, suggesting a possible relationship between the premature onset of labor and overactivity of the fetal adrenal gland. A relationship between an increased fetal adrenal gland weight and a parallel increase in its activity is also sustained by the finding that, in women with preterm labor/preterm PROM, the onset of labor is preceded by an increase in fetal plasma cortisol.31 However, further targeted studies to demonstrate the causal link between fetal cortisol, adrenal volume, and parturition are still needed. It may be possible that the increases in fetal cortisol are a result of, rather than a cause of, or that they occur in parallel with, the process of parturition. Our observation that the adrenal volume increases linearly with the increase in fetal weight may suggest that maturation of the fetus may also be a factor in fetal cortisol production.
The recent development of 3D ultrasound volume data acquisition systems allows imaging and assessment of growth and development of the human fetal adrenal gland throughout gestation.11 With this new imaging technology, we explored in the current study, apparently for the first time, the ability of the fetal adrenal gland volume to identify women at impending preterm birth. Our results are provocative and suggest that examination of the fetal adrenal gland at the time of evaluation for symptoms consistent with preterm labor/preterm PROM may have major beneficial clinical implications. For example, understanding the time when a marker turns positive in relationship to preterm birth is essential for a test with a high diagnostic accuracy. The high accuracy, sensitivity, and specificity of the corrected adrenal gland volume index in predicting preterm birth within 5 days from the time of examination proves that 3D ultrasound evaluation of the fetal adrenal gland volume has the desired test characteristics for defining a population at risk. We certainly recognize the limitations of the current investigation inherent in the novelty of current 3D ultrasound technology, clinical expertise in volume acquisition, additional software and hardware, image resolution, and additional time for rendering. The current study creates the premise for further prospective studies to confirm that 3D ultrasound assessment of the fetal adrenal gland volume can assist clinicians with implementation of better therapeutic and preventive interventions for preterm birth.
In conclusion, our results demonstrate that fetal adrenal gland volume is a reproducible, accurate, and novel noninvasive method of predicting preterm birth. Given the body of evidence that suggests the fetus plays a role in the initiation of parturition, these results are biologically plausible, and fetal adrenal gland volume should be further evaluated as a predictor of preterm birth, either alone or in combination with other useful makers.
1. Marlow N, Wolke D, Bracewell MA, Samara M; EPICure Study Group. Neurologic and developmental disability at six years of age after extremely preterm birth. N Engl J Med 2005;352:9–19.
2. Hamilton BE, Martin JA, Ventura SJ, Sutton PD, Menacker F. Births: preliminary data for 2004. Natl Vital Stat Rep 2005;54:1–17.
3. Lockwood CJ, Kuczynski E. Markers of risk for preterm delivery. J Perinat Med 1999;27:5–20.
4. Norwitz ER, Robinson JN, Challis JR. The control of labor. N Engl J Med 1999;341:660–6.
5. Challis JR. CRH, a placental clock and preterm labour. Nat Med 1995;1:416.
6. Langlois D, Li JY, Saez JM. Development and function of the human fetal adrenal cortex. J Pediatr Endocrinol Metab 2002;15 suppl:1311–22.
7. Anderson AB, Laurence KM, Davies K, Campbell H, Turnbull AC. Fetal adrenal weight and the cause of premature delivery in human pregnancy. J Obstet Gynaecol Br Commonw 1971;78:481–8.
8. Goldenberg RL, Goepfert AR, Ramsey PS. Biochemical markers for the prediction of preterm birth. Am J Obstet Gynecol 2005;192 suppl:S36–46.
9. Jeanty P, Chervenak F, Grannum P, Hobbins JC. Normal ultrasonic size and characteristics of the fetal adrenal glands. Prenat Diagn 1984;4:21–8.
10. Lewis E, Kurtz AB, Dubbins PA, Wapner RJ, Goldberg BB. Real-time ultrasonographic evaluation of normal fetal adrenal glands. J Ultrasound Med 1982;1:265–70.
11. Chang CH, Yu CH, Chang FM, Ko HC, Chen HY. Assessment of fetal adrenal gland volume using three-dimensional ultrasound. Ultrasound Med Biol 2002;28:1383–7.
12. Benacerraf BR, Benson CB, Abuhamad AZ, Copel JA, Abramowicz JS, Devore GR, et al. Three- and 4-dimensional ultrasound in obstetrics and gynecology: proceedings of the American Institute of Ultrasound in Medicine consensus conference. J Ultrasound Med 2005;24:1587–97.
13. Osada H, Iitsuka Y, Masuda K, Sakamoto R, Kaku K, Seki K, et al. Application of lung volume measurement by three-dimensional ultrasonography for clinical assessment of fetal lung development. J Ultrasound Med 2002;21:841–7.
14. Chien PF, Owen P, Khan KS. Validity of ultrasound estimation of fetal weight. Obstet Gynecol 2000;95:856–60.
15. Eng J. Sample size estimation: how many individuals should be studied? Radiology 2003;227:309–13.
16. Green NS, Damus K, Simpson JL, Iams J, Reece EA, Hobel CG, et al. Research agenda for preterm birth: recommendations from the March of Dimes. Am J Obstet Gynecol 2005;193:626–35.
17. King JF, Grant A, Keirse MJ, Chalmers I. Beta-mimetics in preterm labour: an overview of the randomized controlled trials. Br J Obstet Gynaecol 1988;95:211–22.
18. Foulon W, Van Liedekerke D, Demanet C, Decatte L, Dewaele M, Naessens A. Markers of infection and their relationship to preterm delivery. Am J Perinatol 1995;12:208–11.
19. Wenstrom KD, Andrews WW, Hauth JC, Goldenberg RL, DuBard MB, Cliver SP. Elevated second-trimester amniotic fluid interleukin-6 levels predict preterm delivery. Am J Obstet Gynecol 1998;178:546–50.
20. Coleman MA, France JT, Schellenberg JC, Ananiev V, Townend K, Keelan JA, et al. Corticotropin-releasing hormone, corticotropin-releasing hormone-binding protein, and activin A in maternal serum: prediction of preterm delivery and response to glucocorticoids in women with symptoms of preterm labor. Am J Obstet Gynecol 2000;183:643–8.
21. McLean M, Bisits A, Davies J, Walters W, Hackshaw A, De Voss K, et al. Predicting risk of preterm delivery by second-trimester measurement of maternal plasma corticotropin-releasing hormone and alpha-fetoprotein concentrations. Am J Obstet Gynecol 1999;181:207–15.
22. Sibai B, Meis PJ, Klebanoff M, Dombrowski MP, Weiner SJ, Moawad AH, et al. Plasma CRH measurement at 16 to 20 weeks’ gestation does not predict preterm delivery in women at high-risk for preterm delivery. Am J Obstet Gynecol 2005;193:1181–6.
23. Faron G, Boulvain M, Irion O, Bernard PM, Fraser WD. Prediction of preterm delivery by fetal fibronectin: a meta-analysis. Obstet Gynecol 1998;92:153–8.
24. Tsoi E, Akmal S, Geerts L, Jeffery B, Nicolaides KH. Sonographic measurement of cervical length and fetal fibronectin testing in threatened preterm labor. Ultrasound Obstet Gynecol 2006;27:368–72.
25. Arias E, MacDorman MF, Strobino DM, Guyer B. Annual summary of vital statistics–2002. Pediatrics 2003;112:1215–30.
26. Lockwood CJ, Senyei AE, Dische MR, Casal D, Shah KD, Thung SN, et al. Fetal fibronectin in cervical and vaginal secretions as a predictor of preterm delivery. N Engl J Med 1991;325:669–74.
27. Rainey WE, Carr BR, Wang ZN, Parker Jr. CR Gene profiling of human fetal and adult adrenals. J Endocrinol 2001;171:209–15.
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. Rainey WE, Rehman KS, Carr BR. The human fetal adrenal: making adrenal androgens for placental estrogens. Semin Reprod Med 2004;22:327–36.
30. Challis JR, Bloomfield FH, Bocking AD, Casciani V, Chisaka H, Connor K, et al. Fetal signals and parturition. J Obstet Gynaecol Res 2005;31:492–9.
© 2007 by The American College of Obstetricians and Gynecologists.
31. 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.