Received: 5 September 2022 | Revised: 18 January 2023 | Accepted: 21 January 2023 DOI: 10.1111/aogs.14524 ORIGINAL RESEARCH ARTICLE Reference values for interleukin-­6 in the amniotic fluid of asymptomatic pregnant women Ester del Barco1 | Clara Franco-­Jarava2 | Mireia Vargas1 | Nerea Maíz1 | Silvia Arevalo | Maria Ángeles Sánchez | Maria Teresa Avilés | Carlota Rodó1 | 1 1 1 Manel Mendoza1 | Maria Goya1,3 | Manuel Hernández-­González2 | Elena Carreras1,3 1 Maternal-­Fetal Medicine Unit, Department of Obstetrics, Hospital Abstract Universitari Vall d'Hebron, Universitat Introduction: Nowadays, proinflammatory factors are considered to play an impor- Autònoma de Barcelona, Barcelona, Spain 2 Immunology Department, Hospital tant role in the pathophysiology of threatened preterm labor or chorioamnionitis. The Universitari Vall d'Hebron, Barcelona, aim of this study was to establish the normal reference range for interleukin-­6 (IL-­6) Spain 3 levels in the amniotic fluid and to identify factors which may alter this value. Maternal and Child Health and Development Network (SAMID), Instituto Material and methods: Prospective study in a tertiary-­level center including asymp- Salud Carlos III, Madrid, Spain tomatic pregnant women undergoing amniocentesis for genetic studies from October Correspondence 2016 to September 2019. IL-­6 measurements in amniotic fluid were performed using Maria Goya, Maternal-­Fetal Medicine, a fluorescence immunoassay with microfluidic technology (ELLA Proteinsimple, Bio Dept. Obstetrics and Gynecology, Hospital Universitari Vall d'Hebron, Techne). Maternal history and pregnancy data were also recorded. Universitat Autònoma de Barcelona, Results: This study included 140 pregnant women. Of those, women who underwent Passeig de la Vall d'Hebron, 119-­129 08035 Barcelona, Spain. termination of pregnancy were excluded. Therefore, a total of 98 pregnancies were Email: maria.goya@vallhebron.cat included in the final statistical analysis. The mean gestational age was 21.86 weeks (range: 15–­38.7) at the time of amniocentesis, and 38.6 weeks (range: 30.9–­41.4) at delivery. No cases of chorioamnionitis were reported. The log10 IL-­6 values follow a normal distribution (W = 0.990, p = 0.692). The median, and the 5th, 10th, 90th, and 95th percentiles for IL-­6 levels were 573, 105, 130, 1645, and 2260 pg/mL, respectively. The log10 IL-­6 values were not affected by gestational age (p = 0.395), maternal age (p = 0.376), body mass index (p = 0.551), ethnicity (p = 0.467), smoking status (p = 0.933), parity (p = 0.557), method of concep- tion (p = 0.322), or diabetes mellitus (p = 0.381). Conclusions: The log10 IL-­6 values follow a normal distribution. IL-­6 values are inde- pendent of gestational age, maternal age, body mass index, ethnicity, smoking status, parity and method of conception. Our study provides a normal reference range for IL-­6 levels in the amniotic fluid that can be used in future studies. We also observed that normal IL-­6 values were higher in the amniotic fluid than in serum. Abbreviations: BMI, body mass index; ELISA, enzyme-­linked immunosorbent assays; IL-­6 , interleukin-­6; PPROM, preterm premature rupture of membranes. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. © 2023 The Authors. Acta Obstetricia et Gynecologica Scandinavica published by John Wiley & Sons Ltd on behalf of Nordic Federation of Societies of Obstetrics and Gynecology (NFOG). 480 |  wileyonlinelibrary.com/journal/aogs Acta Obstet Gynecol Scand. 2023;102:480–485. del BARCO et al. | 481 KEYWORDS amniocentesis, amniotic fluid, chorioamnionitis, interleukin-­6, intra-­amniotic infection, intra-­ amniotic inflammation, preterm labor, spontaneous preterm delivery 1 | I NTRO D U C TI O N Key message Preterm birth is defined as delivery before the 37th gestational week and is the most important cause of neonatal mortality Interleukin-­6 has been linked to the pathophysiology of and morbidity, being intra-­a mniotic infection and inflammation preterm labor and chorioamnionitis. Interleukin-­6 levels in some of its causes. Pathogenesis of preterm labor is not well amniotic fluid remain fairly constant during the pregnancy understood, but is thought to be a syndrome initiated by mul- and are not affected by maternal factors such as ethnicity, tiple mechanisms, including infection, inflammation, ischemia, age, BMI, smoking status, parity, or method of conception. uterine overdistension and other immunologically mediated processes.1 Previous studies have shown that inflammatory mediators and cytokines such as interleukin-­6 (IL-­6) may be involved in the patho- 2.2 | Amniocentesis physiology of preterm labor and premature rupture of membranes, indicating that infection and subclinical inflammation play a crucial Amniocentesis was performed by an experienced obstetrician with role.1–­3 Currently, systemic inflammation is considered to be less the patient in the supine position. Skin asepsis of the abdominal wall important than increased inflammation at the maternal-­fetal in- was performed using chlorhexidine or povidone-­iodine. The punc- terface. In addition, intra-­amniotic inflammation is associated with ture site was selected by ultrasound, based on placental location, adverse perinatal outcomes, even in the absence of intra-­amniotic fetal position and movements, and area with the largest pool of am- microbes.4 niotic fluid. Amniotic fluid was collected using a 22 gauge needle. Many maternal and fetal characteristics, including markers such When possible, transplacental puncture was avoided. as IL-­6, have been associated with preterm birth. However, to date, the number of studies measuring IL-­6 in asymptomatic patients is limited.3,5–­7 In this study, we aimed to establish a normal reference range for IL-­6 values in amniotic fluid and to investigage if IL-­6 values are affected by maternal or fetal factors. TA B L E 1 Demographic characteristics Demographic characteristics (n = 98) 2 | M ATE R I A L A N D M E TH O DS Age (years), mean (SD) 33.84 (4.55) 2 BMI (kg/m ), mean (SD) 24.27 (4.41) 2.1 | Study design Ethnicity, n (%) This prospective study included pregnant women who under- Arab 14 (14.3%) went amniocentesis for genetic studies at Vall d'Hebron Hospital, Asian 1 (1.0%) Barcelona, between October 2016 and September 2019. Women Caucasian 74 (75.5%) with clinical signs of chorioamnionitis, preterm premature rupture Indian 1 (1%) of membranes, threatened preterm labor and age below 18 years old South American 8 (8.2%) were excluded. Smoking status, n (%) Measurements of IL-­6 levels in the amniotic fluid were per- Smoker 8 (8.2%) formed using a fluorescence immunoassay with microfluidic tech- Non-­smoker 90 (91.8%) nology (ELLA Proteinsimple, Bio Techne). Samples were stored at Assisted reproductive techniques, n (%) −20°C until their processing. Samples were run in triplicate with a Yes 14 (14.3%) 1:5 dilution using the buffer provided in the kit. IVF with oocyte donor 4 (4.0%) Maternal history and pregnancy data were recorded. Maternal IVF 9 (9.2%) and pregnancy factors included age, body mass index (BMI), parity, ethnicity, smoking status and method of conception. Delivery vari- AI 1 (1.0%) ables included gestational age at delivery, birthweight, type of birth No 84 (85.7%) (vaginal or cesarean section), perinatal complications and newborn Abbreviations: AI, artificial insemination; BMI, body mass index; IVF, in gender. vitro fertilization; SD, standard deviation. 482 | del BARCO et al. F I G U R E 1 Distributions of (A) IL-­6 values and (B) log10-­transformed IL-­6 values. 2.3 | Statistical analyses TA B L E 2 IL-­6 levels Gestational age at inclusion (weeks), mean (SD) 22.49 (5.21) All statistical analyses were performed using the computing envi- Gestational age at delivery (weeks), mean (SD) 38.06 (2.33) ronment R (version 4.0.0).8 The Shapiro test was used to assess nor- IL-­6 levels (pg/mL), mean (SD) 791.28 mality of distribution. The log10 -­transformed IL-­6 values followed a (782.93) Gaussian distribution. IL-­6 levels (log10 -­transformed), mean (SD) 2.71 (0.43) The association of log10 IL-­6 values with continuous variables IL-­6 levels, percentiles was assessed using the Pearson's correlation coefficient or the 5th 105 Spearman's rank correlation coefficient, as appropriate, whereas 10th 130 association of log10 IL-­6 values with categorical variables was 90th 1645 assessed using the Wilcoxon test for variables with two catego- 95th 2260 ries and the Kruskal Wallis test for variables with more than two Log10 -­transformed IL-­6 levels, percentiles categories. Results are reported as the median, mean and 5th, 10th, 90th 5th 2.022 and 95th percentile. 10th 2.113 90th 3.216 95th 3.351 2.4 | Ethics statement Abbreviations: IL-­6, interleukin-­6; SD, standard deviation. This study was approved by the Ethics Committee of Vall d'Hebron Research Institute (registration number PR(AMI)541-­2016) on September seroconversion of toxoplasma in the third trimester; there was a 1, 2016. All participants provided their written informed consent. positive PCR in amniotic fluid for toxoplasma. No cases of chorioam- nionitis were reported. Raw values of IL-­6 levels in the amniotic fluid did not follow a 3 | R E S U LT S normal distribution (W = 0.789, p < 0.001), however, log10-­IL-­6 val- ues did (W = 0.990, p = 0.692) (Figure 1). A total of 140 patients were recruited. Of those, 42 underwent ter- Median IL-­6, and 5th, 10th, 90th, and 95th percentiles were 573, mination of pregnancy for genetic anomalies and were excluded 105, 130, 1645, and 2260 pg/mL, respectively (Table 2). from the analysis. Therefore, 98 pregnancies were included in the Log10 IL-­6 levels were independent of gestational age final statistical analysis (Table 1). In one of these pregnancies, a tri- (rho = 0.086, p = 0.395), maternal age (rho = 0.090, p = 0.376), somy 18 case was found; the patient decided to carry on with the BMI (rho = −0.064, p = 0.551), ethnicity (p = 0.467), smoking sta- pregnancy. tus (p = 0.933), parity (p = 0.557), method of conception (p = 0.322) Mean gestational age was 21.8 weeks (range 15.0–­38.7) at the or diabetes mellitus (p = 0.381) (Figure 2). Log10 IL-­6 levels did not time of amniocentesis and 38.6 weeks (range 30.9–­41.4) at delivery. show an association with gestational age at delivery (rho = −0.192, Late amniocentesis at 38 gestational weeks was performed to study p = 0.089). del BARCO et al. | 483 (A) (B) (C) (D) (E) (F) (G) (H) F I G U R E 2 IL-­6 levels according to clinical data: Gestational age (A), maternal age (B), gestational age at birth (C), body mass index (D), ethnicity (E), smoking status (F), multiparity (G) and assisted reproductive techniques (ART) (H). 484 | del BARCO et al. 4 | DISCUSSION Of note, in our study we observed increased IL-­6 levels in the amniotic fluid after 15 gestational weeks as compared to mean nor- Our results show that there is no association between IL-­6 levels and mal IL-­6 values in serum throughout the duration of pregnancy. maternal factors such as gestational age, maternal age, BMI, ethnic- Fu et al. measured serum IL-­6 in healthy non-­pregnant women ity, smoking status, parity and method of conception. Moreover, we and healthy pregnant women and established the following ref- observed that IL-­6 levels in amniotic fluid are substantially higher erence intervals for serum IL-­6: <1.518 pg/mL for healthy non-­ than in serum (0–­20 pg/mL). pregnant women, <3.52 pg/mL for healthy pregnant women during Many studies have examined inflammatory markers in symp- the first trimester, and <4.40 pg/mL for healthy pregnant women tomatic patients, either with premature rupture of membranes or during the second and third trimesters.15 Our study offers the ad- threatened preterm labor.3,6,9,10 These studies have shown increased ditional benefit of including asymptomatic women in the second inflammatory markers in patients with these pathologies and an as- and third trimesters of pregnancy. Additionally, we studied several sociation between intra-­amniotic inflammation and adverse perina- maternal and fetal factors that may alter IL-­6 values, and confirmed tal outcomes, even in the absence of intra-­amniotic microbes. Lee that there was no association between these factors and IL-­6 values. et al. found that women with preterm premature rupture of mem- A study has recently been published showing that immuno- branes (PPROM) and intra-­amniotic infection have lower IL-­6 levels assays using the Cobas system (Roche) and the traditional ELISA than women with intra-­amniotic infection but intact membranes.11 method are comparable for measuring IL-­6 levels in amniotic fluid.16 In our study, patients with PPROM or suspected I intra-­amniotic in- Clinically, the Cobas system has advantages over the ELISA method, fection were excluded. such as being more robust and rapid, plus totally automated. The For this reason, prediction models have been designed to pre- ELISA system used in our study (ELLA) gives results comparable to dict spontaneous preterm delivery or microbial invasion of the those obtained with the Cobas system (data not shown). Therefore, amniotic cavity based on levels of inflammatory markers, such as the reference values reported in our study can also be applied to 4,9,12–­14 IL-­6. evaluate results obtained in studies using the Cobas system. Studies with asymptomatic patients have also been conducted, Our study had a number of limitations, such as the small sam- although the literature is limited. Melekoglu et al. measured IL-­6 ple size and single-­center population. Furthermore, measurements levels in 22 asymptomatic pregnant women between 16 and 24 of IL-­6 in the maternal serum were not performed. First, this would gestational weeks. They excluded patients with fetuses having any have allowed the direct comparison of levels in both fluids, and sec- congenital or chromosomal abnormality, pre-­gestational diabetes, ond to establish whether there is a correlation between IL-­6 levels hypertensive disease, obstetric cholestasis, chronic maternal dis- in serum and IL-­6 levels in amniotic fluid. However, the number of ease, and use of tobacco or alcohol. Mean IL-­6 values in the amni- patients participating was sufficient to measure IL-­6 levels in the otic fluid of these asymptomatic patients were 95.8 ± 16.4 pg/mL. amniotic fluid and to evaluate the factors that may affect these val- Following the same inclusion and exclusion criteria as in the previ- ues. To date, there were no normal reference ranges for IL-­6 levels in ous study, the same group measured IL-­6 levels in 20 patients diag- amniotic fluid. Moreover, we have investigated this parameter in as- nosed with gestational diabetes and found increased IL-­6 levels in ymptomatic pregnant women and in a wide range of gestational ages. these patients as compared to pregnant women without gestational This study provides a normal reference range for IL-­6 values in diabetes (mean IL-­6 of 136.2 ± 17.3 pg/mL and 98.3 ± 11.5 pg/mL, the amniotic fluid of asymptomatic pregnant women. In addition, respectively). 3,7 In our study, a total of seven patients were diag- since IL-­6 values are not affected by gestational age or other mater- nosed with gestational diabetes and one patient had type 2 dia- nal factors, IL-­6 values can be used without having to convert them betes. None of the patients had type 1 diabetes. There were no to MoMs or without performing other forms of standardization. differences between patients with diabetes and patients without diabetes. Gervasi et al. measured IL-­6 values between 15.6 and 16.9 5 | CO N C LU S I O N gestational weeks in 652 term pregnancies without obstetric com- plications (i.e., placental abruption, small for gestational age, fetal In asymptomatic women, IL-­6 levels remain fairly constant through- death, gestational hypertension and pre-­eclampsia), and their results out the duration of pregnancy, and are not affected by maternal fac- agree with our results (the 95th percentile had mean IL-­6 values of tors such as ethnicity, age, body mass index, smoking status, parity, 5 2935 pg/mL). In Gervasi's studies, IL-­6 levels in the amniotic fluid or method of conception. Our study provides a normal reference were measured with commercially available enzyme-­linked immuno- range for IL-­6 values in the amniotic fluid of asymptomatic pregnant sorbent assays (ELISA), as in our study. women. These normal reference range for IL-­6 values can be used as Although previous studies have reported increased IL-­6 levels a reference in future studies. during pregnancy in asymptomatic women, measurements were performed only during the second trimester. In addition, these stud- AU T H O R C O N T R I B U T I O N S ies did not investigate whether IL-­6 values were affected by mater- EB was the principal investigator at Hospital Vall d'Hebron, con- nal factors. tributed to the study design, patient follow-­u p and data collection, del BARCO et al. | 485 inflammation and associations with spontaneous early (32 weeks) and wrote the first draft of both the protocol and manuscript. EB preterm delivery. J Perinat Med. 2012;40:329-­3 43. is the guarantor for the manuscript. MG was the study coordina- 6. Thomakos N, Daskalakis G, Papapanagiotou A, Papantoniou N, tor, contributed to the study design, patient follow-­u p and data Mesogitis S, Antsaklis A. Amniotic fluid interleukin-­6 and tumor necrosis factor-­α at mid-­trimester genetic amniocentesis: relation- collection, and wrote the protocol and manuscript with EB. MV, ship to intra-­amniotic microbial invasion and preterm delivery. Eur J MM and CF-­J contributed to the study design and data collection. Obstet Gynecol Reprod Biol. 2010;148:147-­151. SA, MÁS, MA and CR contributed to data collection. MH-­G and 7. Melekoglu R, Ciftci O, Celik E, Yilmaz E, Bastemur AG. Evaluation EC contributed to the study design. All authors had complete ac- of second trimester amniotic fluid ADAMTS4, ADAMTS5, interleu- cess to the data at the end of the study and the decision to submit kin-­6 and tumor necrosis factor-­α levels in patients with gestational diabetes mellitus. J Obstet Gynaecol Res. 2019;45:824-­829. the manuscript for publication was taken at a team joint meeting 8. R Core Team. R: A Language and Environment for Statistical where everyone reviewed and approved the final version of the Computing. R Foundation for Statistical Computing; 2020 https:// manuscript. This was subsequently submitted by the correspond- www.R-­proje​c t.org/ ing author. 9. Kacerovsky M, Musilova I, Hornychova H, et al. Bedside assess- ment of amniotic fluid interleukin-­6 in preterm prelabor rupture of membranes. Am J Obstet Gynecol. 2014;211:385.e1-­385.e9. AC K N OW L E D G M E N T S 10. Romero R, Miranda J, Chaiworapongsa T, et al. Prevalence and clin- We thank Mar Jiménez Quesada for her help with the English ver- ical significance of sterile intra-­amniotic inflammation in patients sion of this manuscript. Some of the authors are members of the with preterm labor and intact membranes. Am J Reprod Immunol. 2014;72:458-­474. European Reference Network on Rare Congenital Malformations 11. Lee SY, Buhimschi IA, Dulay AT, et al. IL-­6 trans-­signaling sys- and Rare Intellectual Disability ERN-­ITHACA (EU Framework tem in intra-­amniotic inflammation, preterm birth, and preterm Partnership Agreement ID: 3HP-­HP-­FPA ERN-­01-­2016/739516). premature rupture of the membranes. J Immunol Baltim Md. 1950;2011(186):3226-­3236. 12. Cobo T, Aldecoa V, Figueras F, et al. Development and validation C O N FL I C T O F I N T E R E S T of a multivariable prediction model of spontaneous preterm deliv- The authors have no conflicts of interest to declare. ery and microbial invasion of the amniotic cavity in women with preterm labor. Am J Obstet Gynecol. 2020;223:421.e1-­421.e14. ORCID 13. Leaños-­Miranda A, Nolasco-­Leaños AG, Carrillo-­Juárez RI, Molina-­ Pérez CJ, Isordia-­Salas I, Ramírez-­Valenzuela KL. Interleukin-­6 Ester del Barco https://orcid.org/0000-0002-1484-9027 in amniotic fluid: a reliable marker for adverse outcomes in Mireia Vargas https://orcid.org/0000-0002-4980-7652 women in preterm labor and intact membranes. Fetal Diagn Ther. Nerea Maíz https://orcid.org/0000-0001-8261-6383 2021;48:313-­320. Manel Mendoza https://orcid.org/0000-0002-3030-3833 14. Balciuniene G, Gulbiniene V, Dumalakiene I, et al. Prognostic mark- Maria Goya https://orcid.org/0000-0003-4001-2017 ers for chorioamnionitis: IL-­6, TNF-­α , and MMP-­8 in vaginally ob- tained amniotic fluid. J Clin Med. 2021;10:1136. 15. Fu Y, Tang L, Hu M, Xiang Z, Hu Y. Changes of serum interleukin-­6 in REFERENCES healthy pregnant women and establishment of relevant reference 1. Goldenberg RL, Culhane JF, Iams JD, Romero R. Preterm birth 1 intervals. Clin Chim Acta. 2020;502:116-­119. epidemiology and causes of preterm birth. Lancet. 2008;371:10. 16. Cobo T, Aldecoa V, Holeckova M, et al. A rapid amniotic fluid 2. Gilman-­Sachs A, Dambaeva S, Salazar Garcia MD, Hussein Y, Kwak-­ Interleukin-­6 assessment for the identification of intra-­amniotic Kim J, Beaman K. Inflammation induced preterm labor and birth. J inflammation in women with preterm labor and intact membranes. Reprod Immunol. 2018;129:53-­58. Fetal Diagn Ther. 2021;48:327-­332. 3. Melekoglu R, Yilmaz E, Ciftci O, Kafadar YT, Celik E. Associations between second-­trimester amniotic fluid levels of ADAMTS4, ADAMTS5, IL-­6, and TNF-­α and spontaneous preterm delivery in singleton pregnancies. J Perinat Med. 2019;47:304-­310. How to cite this article: del Barco E, Franco-Jarava C, Vargas 4. Combs CA, Gravett M, Garite TJ, et al. Amniotic fluid infection, inflammation, and colonization in preterm labor with intact mem- M, et al. Reference values for interleukin-6 in the amniotic branes. Am J Obstet Gynecol. 2014;210:125.e1-­125.e15. fluid of asymptomatic pregnant women. Acta Obstet Gynecol 5. Gervasi MT, Romero R, Bracalente G, et al. Midtrimester amni- Scand. 2023;102:480-485. doi:10.1111/aogs.14524 otic fluid concentrations of interleukin-­6 and interferon-­g amma-­ inducible protein-­10: evidence for heterogeneity of intra-­amniotic