CERVICOVAGINAL FLUID CYTOKINES' LEVELS IN PRETERM BIRTH HIGH-RISK WOMEN COULD BE USED AS EARLY PREDICTORS FOR RECURRENT PRETERM BIRTH

 CERVICOVAGINAL FLUID CYTOKINES' LEVELS IN PRETERM BIRTH HIGH-RISK WOMEN COULD BE USED AS EARLY PREDICTORS FOR RECURRENT PRETERM BIRTH

By

Sameh Said, El-Sayed El-Desouky and Mahmoud Hashish*

Departments of Obstetrics & Gynecology and Clinical Pathology *,

Faculty of Medicine, Al-Azhar University, Cairo

Corresponding authors: drsameh50@yahoo.com

elsayedeldesouky@yahoo.com

Dr_hashish68@yahoo.com

ABSTRACT

Background: Differences in cytokine distribution across different biological fluids; maternal plasma, cervicovaginal fluid (CVF), and amniotic fluid are evident throughout normal pregnancy, but profiles of numerous cytokines varied across trimesters in women delivering term versus preterm in both CVF and serum.

Objective: To evaluate cytokine profile in cervicovaginal fluid (CVF) of women had previous preterm birth (PTB) and the effect imposed by cerclage on these levels.

Patients and Methods: Seventy study women with previous PTB and currently had cercical length (CL) of <25 mm and 20 control women with no history of PTB and had CL of >25 mm were studied. At the 24^th gestational week (GW; T1), all women had CL estimation and CVF sampling. At time of labor or removal of the suture (T2), study women had CL re-estimation and CVF re-sampling. CVF levels of monocyte chemoattractant protein-1 (MCP-1), tumor necrosis factor-α (TNF-α), interleukin (IL)-1β and IL-6 were ELISA estimated. Shirodkar cervical cerclage (CC) was performed within 4 days after T1 sampling. Outcomes included difference in T1-CVF cytokines' levels between study and control women, effect of cerclage on CL and CVF cytokines' levels and the value of T1-CVF cytokines' levels as predictors of pregnancy duration.

Results: T1-CVF levels were significantly higher in study than control women and to their T2- levels. T1- and T2-CL in study women was significantly shorter than T1-CL in controls with significant difference between T1- and T2-CL. Percentage of T2-CL decrease was negatively correlated with the percentage of decrease of T2-CVF levels. Duration of pregnancy of study women was negatively correlated with percentage of CL shortening, while was positively correlated with percentage of decrease of T2-CVF levels of MCP-1 and TNF-α. Statistical analyses defined high T1-CVF levels of MCP-1, IL-β and TNF-α as significant early predictor for PTB.

Conclusion: High CVF cytokines' levels were associated with increased risk of PTB. Cerclage worked beyond its mechanical action through reduction of CVF cytokines' levels. High CVF cytokines' levels at 24th GW, especially high levels of MCP-1 and TNF-α, may help to predict PTB.

Keywords: Preterm birth, Cervicovaginal fluid, Cytokines, Pregnancy duration

INTRODUCTION

     Cervical insufficiency (CI) is defined as premature, progressive dilation and shortening of the cervix during pregnancy (Monsanto et al., 2017). Mid-trimester CI has deleterious maternal and fetal effects secondary to development of intra-amniotic infection/inflammation that occurs in up to 50% of women who developed CI (Mönckeberg et al., 2019). CI if untreated it can lead to prolapse and rupture of amniotic membrane resulting in mid-trimester pregnancy loss or preterm birth (PTB), which leads to severe newborn complications (Raja et al., 2019).

     Midtrimester ultrasound is a valuable method for identifying asymptomatic women at risk for spontaneous preterm delivery (Kiefer et al., 2016). Cervical cerclage, vaginal progesterone and pessary were used in multiple studies and showed similar effectiveness for management of women with short cervix or previous PTB (Alfirevic et al., 2013). However, the response to various treatments was variable in the clinical setting (Kiefer et al., 2016).

     Multiple previous studies found vaginal progesterone reduced the rate of early PTB in women with short cervix with reduction of rate of neonatal morbidity/ mortality in singleton pregnancies, but failed to prevent PTB in twin pregnancies (Rode et al., 2011, Serra et al., 2013 and Schuit et al., 2015).  The use of a pessary proved to prevent PTB in asymptomatic women with mid-trimester short cervical length (CL); however, the precise mechanisms for this effect remain unclear (Mendoza Cobaleda et al., 2019). Cervical cerclage (CC) involves positioning of a stitch around the cervical neck to give the cervix mechanical support to reduce risk of PTB (Alfirevic et al., 2017).

     Differences in cytokine distribution across different biological fluids; maternal plasma, cervicovaginal fluid (CVF), and amniotic fluid are evident throughout normal pregnancy (Ashford et al., 2018a), but profiles of numerous cytokines were varied across trimesters in women delivering term versus preterm in both CVF and serum (Ashford et al., 2018b).

     This study aimed to evaluate the cytokine profile in CVF of women with history of previous PTB and the effect imposed by cerclage on these levels.

PATIENTS AND METHODS

Design: Prospective comparative study.

Setting: Obstetrics & Gynecology Department in association with Clinical Pathology Department, Faculty of Medicine, Al-Azhar University.

Time: Thirty months starting from January 2016 till June 2018.

Patients: The study included women with history of previous PTB and currently were pregnant in singleton fetus, free at time of diagnosis of pregnancy-induced complications and having CL of <25 mm as determined by transvaginal ultrasonography (TVU). Exclusion criteria included multiple pregnancies, fetal anomalies, diabetes mellitus, systemic or local inflammatory disorders and current vaginal infection at time of sample collection. Pregnant women having singleton fetus, had no history of previous PTB, free at time of diagnosis of pregnancy-induced complications, having CL of >25 mm and free of exclusion criteria were included as control women.

     At time of 1^st attendance, baseline data included age, number of previous pregnancies and its outcome, blood pressure and random blood glucose level of pregnant women eligible for evaluation and then all women underwent full obstetric history taking and examination including TVU. Women were asked to attend the Antenatal Care Unit (ACU) at the beginning of the 24^th gestational week (GW; T1 time) for assurance of inclusion and exclusion criteria especially the presence of cervicovaginal infections, TVU estimation of CL (T1 CL) and women with CL <25 mm were considered at a risk for having PTB (Owen et al., 2001). Samples of cervicovaginal fluid (CVF) were assayed for T1 levels of cytokines and were twenty cross-matched age and body mass index pregnant women, free of previous PTB, fulfilling the inclusion and exclusion criteria for the study group and has T1 CL> 25 mm were included as control group.

     All enrolled women were maintained on vaginal toilet to guard against development of infection. Women of the study were asked to attend the ACU biweekly for follow-up for development of PTB, which was defined as any birth before 37 completed weeks of gestation (Meher and Alfirevic, 2014). At time of labor or removal of the suture (T2), all study women undertook estimation of CL (T2 CL) and another CVF sample was obtained to estimate T2 CVF cytokines' levels. For both T2 CL and CVF cytokines' level. The percentage of change between T1 and T2 measures was calculated as T1-T2 measures divided by T1 measures and multiplied by 100 (% of change= ([T1-T2 levels]/[T1 level]) x (100)).

Methods

     CVF sample obtaining and processing: Vaginal speculum was applied, and to obtain high vaginal smear of CVF a Dacron swab was placed in posterior vaginal fornix, maintained in situ for 10 seconds to achieve saturation, and then was transferred into 750 ml of standard phosphate-buffered saline solution mixed with freshly prepared protease inhibitor solution. The swab was then removed, placed in a clean tube, vortexed for 10 sec and centrifuged at 2500 g for 10 minutes, at 4^oC.

     The resulting fluid was collected and added to the fluid in the original tube, well-mixed and centrifuged for a further 10 minutes to remove cell debris. Cell-free supernatants were collected and divided into aliquots and stored at -80^oC until being ELISA assayed.

Investigations: CVF levels of monocyte chemo attractant protein-1 (MCP-1), tumor necrosis factor-α (TNF-α), interleukin (IL)-1β and IL-6 were measured using ELISA kits according to the manufacturer's instructions and were read using a 96 well micro plate ELISA reader (Dynatech, MR 7000):

1. Monocyte chemoattractant protein-1 was measured with the enzyme linked immunoassay (ELISA) kit (catalogue no. ab181421, abcam Inc., Cambridge, USA) by quantitative sandwich enzyme immunoassay technique (Deshmane et al., 2009).

2. Human TNF-α was measured with the enzyme linked immunoassay (ELISA) kit (catalogue no. ab179886, abcam Inc., Cambridge, USA) by quantitative sandwich enzyme immunoassay technique (Coughlan et al., 2001).

3. IL-1β was measured with the enzyme linked immunoassay (ELISA) kit (catalogue no. ab46052, abcam Inc., Cambridge, USA) by quantitative sandwich enzyme immunoassay technique (Heng et al., 2014).

4. IL-6 using the Eagle Biosciences Human IL-6 ELISA Assay Kit (Eagle Bioscience Inc., USA; Catalogue No IL631-K01) which employs the quantitative sandwich enzyme immunoassay technique (Gaines-Das and Poole, 1993).

Cervical cerclage: Cervical cerclage (CC) was performed using the Shirodkar procedure with a non-absorbable suture within 4 days after T1 CVF sample obtaining. After performing CC, women were asked to avoid any sexual activity, use of tampons or douching, prolonged standing for >4 h, heavy physical work, lifting heavy weights, straining or any activity that brings on symptoms of pelvic pressure or discomfort.

Study Outcomes:

1. Primary outcome: The difference in T1-CVF cytokines' levels between women with PTB (Study group) and without history of PTB (Control group).

2. Secondary outcomes:

a.     The frequency of PTB among study women.

b.    The effect of cerclage on CL and CVF cytokines' levels defined as the calculated percentage of change.

c.     The value of T1-CVF cytokines' levels as predictors for extent of shortening of CL as measured on T2 time and for duration of gestation.

Statistical analysis:

     Obtained data were presented as mean±SD, numbers and percentages. Results were analyzed using paired t-test, One-way ANOVA Test followed by post – HOC test (Bonferroni procedure) and Chi-square test (X2 test). Possible relationships were investigated using Pearson linear regression analysis. Regression analysis (Stepwise method) was used for stratification of T1 estimated cytokines' levels as specific predictors. T1 estimated CVF levels as predictors for duration of pregnancy were evaluated using the receiver operating characteristic (ROC) curve analysis judged by the area under the curve (AUC) that was compared versus null hypothesis that AUC=0.5. Statistical analysis was conducted using the IBM SPSS (Version 23, 2015) for Windows statistical package. P value <0.05 was considered statistically significant.

RESULTS

     During 30 month study duration, 70 women at high-risk of PTB and fulfilling the inclusion criteria were enrolled as study group and 20 women free of risk for PTB and were age and BMI-matching women of study group were also enrolled in the study as control group. There were non-significant differences between study and control women regarding inclusion criteria as shown in table 1 apart from the presence of history of previous PTB (Table 1).

Table (1): Enrolment data of women included in both groups

Data were presented as mean±SD, numbers & percentages; BMI: Body mass index; PTB: Preterm birth

Mean T1-CVF cytokines' levels were significantly higher in women of study group than in women of control group, while mean T2-CVF cytokines' levels were significantly lower in study women compared to their corresponding T-1 levels. On comparison to T1 control levels, mean T2-CVF levels of MCP-1 were non-significantly higher than T1 levels, while mean T2-CVF of TNF-α, IL-1β and IL-6 were still significantly higher than T1-levels of control women. The percentages of decreased T2 cytokines' levels were 57±19.7%, 36.9±3.8%, 20.1±8% and 38.9±10%, for MCP-1, TNF-α, IL-1β and IL-6, respectively (Table 2).

Table (2): Mean levels of cytokines estimated in CVF of women included in both groups

Data were presented as mean±SD; T1: at 24th GW; T2: at time of spontaneous labor; MCP-1: Monocyte chemoattractant protein-1, TNF-α: Tumor necrosis factor-α, IL: Interleukin; P1 indicates significance versus control levels; P2 indicates significance versus T1 levels.

     Mean T1-CL (23.6±2.1 mm) and T2-CL (18.9±1.7 mm) in study women was significantly shorter compared to T1-CL of control women (26.8±1.6 mm) with significantly shorter T2-CL than the corresponding T1-CL in study women. The percentage of T2-CL decrease in relation to T1-CL was negatively correlated with the percentage of decrease of T2-CVF levels of studied cytokines in relation to T1-CVF levels (Table 3).

     Throughout duration of pregnancy, 13 study women (18.6%) had PTB (<37GW) for a mean duration of pregnancy of 35±0.8 GW, while the remaining 57 women (81.4%) had mean duration of pregnancy of 38.2±1.3 GW.

     There were negatively significant correlations between duration of pregnancy and the percentage of CL shortening, while the correlations were positively significant between duration of pregnancy and the percentage of decrease of T2-CVF levels of MCP-1 and TNF-α in relation to T1 levels. However, duration of pregnancy showed positive non-significant correlation with percentage of decrease of T2-CVF levels of IL-6 and IL-1β (Table 3).

Table (3): Pearson's correlation between percentage of decrease T2 CL and CVF levels of studied cytokines and duration of pregnancy

     Regression analysis of high T1-CVF cytokines' levels as predictors for short pregnancy duration excluded high T1-CVF level of IL-6 and TNF-α as predictors, but defined high T1-CVF levels of MCP-1 (β=-0.486, <0.001) and IL-β (β=-0.345, p=0.001) as significant early predictors for short duration of pregnancy and development of PTB. On other hand, ROC curve analysis defined high T1-CVF level of MCP-1 and TNF-α as the significant (p=0.006 & 0.004) early predictor for PTB with AUC= 0.872 (95% CI: 0.744-1.001) and 0.885 (95% CI: 0.728-1.041), respectively.

DISCUSSION

     Women had history of previous PTB were at high-risk for redeveloping PTB, throughout follow-up during the current study, PTB occurred by a rate of 18.6%, among studied high-risk women, which is higher than that reported by Blencowe et al. (2013) who reported that worldwide an estimated 11.1% of all live births in 2010 were born preterm, with rates increasing in most countries with reliable trend data. Also, the reported PTB rate was higher than that reported by Hernandez-Andrade et al. (2014) who found the spontaneous PTB at < 37 GW among women with internal cervical os low strain values on US elastography was 11%. On the other hand, the reported figure for PTB among studied high-risk women was in accordance with Chan et al. (2015) who reported PTB in 17.4% of women who had history-indicated cerclage, but was lower than that reported by Sun et al. (2016) and Jung et al. (2016) who reported spontaneous PTB rate of 35.6% and 49%, respectively, among women who had significantly more advanced cervical dilatation at presentation

     For estimation of cytokines' concentrations, the current study used cervicovaginal fluid (CVF) sample that was collected as high as possible from the vagina; depending on previous findings that levels of proinflammatory cytokines were elevated in CVF not in serum in patients with cervical insufficiency than in control women (Monsanto et al., 2017). Also, Short et al. (2018) detected highest concentrations of IL-1β, IL-8 and IL-6 in ranking order of cervical fluid> CVF > plasma, and documented that CVF collection was simpler, provided the largest volume of sample with the potential for undiluted usage, and allowed for self-insertion.

     Mean T1-CVF cytokines' levels were significantly higher in study than in control women; this finding spots light on a possible role for disturbed local cytokine milieu for development of PTB. In support of this assumption, there was negative significant correlation between duration of pregnancy and T1-CVF cytokines' levels and such correlation was highly significant with T1-CVF levels of TNF-α and MCP-1.

     In line with these findings and assumption, Monsanto et al. (2017) reported that women with cervical insufficiency have higher CVF levels of proinflammatory cytokines in comparison to normal women. Also, Ashford et al. (2018b) found CVF values of IL-6, 8 and 10, TNF-α, and CRP was significantly higher in those who delivered preterm than those who delivered full-term. Moreover, Caritis et al. (2018) detected significantly elevated baseline cervical fluid concentration of IL-6 and 10, TNF-α among women whose earliest prior delivery occurred between 16 and 23 GW compared to those with earliest delivery occurred between 32 and 36 GW and these levels were unaffected by 17-hydroxyprogesterone caproate. Recently, Buxton et al. (2019) detected high concentrations of CVF cytokines measured in a monthly sample collected from the 2nd-9th month of term pregnancies, but these levels were largely stable over time and documented that this finding indicated that pregnancy is associated with an active inflammatory state.

     At time of labor or removal of stitch (T2), mean CVF levels were significantly lower than T1-CVF levels, but were non-significantly higher compared to T1- levels of control women. Similarly, Monsanto et al. (2017) detected significant decline in CVF proinflammatory cytokines after cerclage and suggested that cerclage may help reduce local inflammation in CI. Also, Mönckeberg et al. (2019) found patients free of intra-amniotic infection/inflammation who underwent cerclage had longer admission-to-delivery interval, higher gestational age at delivery and had higher neonatal survival rate than those with intra-amniotic infection/inflammation.

     These results illustrate the beneficial effect of cerclage on local immune milieu and indicated an effect of cerclage more than its well-known mechanical effect as a preventive measure against PTB. This effect was manifested as decreased percentage of CL shortening and its positive significant correlation with the percentage of decrease in T2-CVF levels of examined cytokines. In accordance with these data, Steenhaut et al. (2019) detected significantly thicker chorion overlying the cervix in women had cerclage and significantly higher expression of 15-hydroxyprostaglandin dehydrogenase and toll-like receptor-2 expression with significantly decreased senescence than in chorion located far from the cervix, and attributed prevention of triggering of parturition to these membrane changes.

     Duration of pregnancy showed positive significant correlation with the percentage of decrease of T2-CVF levels of MCP-1 and TNF-α and Regression analysis of high T1-CVF cytokines' levels defined high T1-CVF levels of MCP-1 and IL-β as significant early predictors for short pregnancy duration. ROC curve analysis defined high T1-CVF level of MCP-1 and TNF-α as the significant early predictor for PTB. In line with the obtained data, Pandey et al. (2017) found increased maternal levels of IL-1β and IL-6 and low levels of IL-10 were associated with PTB. Sayaril et al. (2018) found the levels of IL-1α, IL-6 and TNF-α can be used as predicative markers for PTB. Lappas (2018) reported that deficiency of mothers against decapentaplegic homolog (SMADs) in myometrial cells significantly increased IL-1β, 1A, 6 and 8, MCP-1, adhesion molecules and COX-2 mRNA expression with subsequent decreased PGF2α release.

CONCLUSION

     Preterm birth in women who had history of previous PTB occurs frequently. High CVF cytokines' levels are associated with increased risk of PTB. Cerclage works beyond its mechanical action through reduction of CVF cytokines' levels. High CVF levels at 24th GW is predictive of PTB, especially high levels of MCP-1 and TNF-α. Wider scale studies are mandatory to define certain cutoff points for these cytokines. 

REFERENCES

1. Alfirevic Z, Owen J, Carreras Moratonas E, Sharp AN, Szychowski JM and Goya M (2013): Vaginal progesterone, cerclage or cervical pessary for preventing preterm birth in asymptomatic singleton pregnant women with a history of preterm birth and a sonographic short cervix. Ultrasound Obstet Gynecol., 41(2):146-51.
2. Alfirevic Z, Stampalija T and Medley N (2017): Cervical stitch (cerclage) for preventing preterm birth in singleton pregnancy. Cochrane Database Syst Rev. 6:CD008991.
3. Ashford KB, Chavan N, Ebersole JL, Wiggins AT, Sharma S, McCubbin A, Barnett J and O'Brien J (2018a): Patterns of systemic and cervicovaginal fluid inflammatory cytokines throughout pregnancy. Am J Perinatol., 35(5):455-462.
4. Ashford K, Chavan NR, Wiggins AT, Sayre MM, McCubbin A, Critchfield AS and O'Brien J (2018b): Comparison of serum and cervical cytokine levels throughout pregnancy between preterm and term births. AJP Rep., 8(2):e113-e120.
5. Blencowe H, Cousens S, Chou D, Oestergaard M, Say L, Moller AB, Kinney M and Lawn J (2013): Born Too Soon Preterm Birth Action Group: Born too soon: the global epidemiology of 15 million preterm births. Reprod Health, 10 Suppl 1:1-14.
6. Buxton MA, Meraz-Cruz N, Sánchez BN, Foxman B, Gronlund CJ, Beltran-Montoya J, Castillo-Castrejon M, O'Neill MS and Vadillo-Ortega F (2019): Repeated Measures of Cervicovaginal Cytokines during Healthy Pregnancy: Understanding "Normal" Inflammation to Inform Future Screening. Am J Perinatol., 12: 1-8.
7. Caritis SN, Hankins G, Hebert M, Haas DM, Ahmed M, Simhan H, Haneline LA, Harris J, Chang J, Famy AS, Yorio P, Ren Z, D'Alton ME and Venkataramanan R (2018): Impact of Pregnancy History and 17-Hydroxyprogesterone Caproate on Cervical Cytokines and Matrix Metalloproteinases. Am J Perinatol.,  35(5):470-480.
8. Chan LL, Leung TW, Lo TK, Lau WL and Leung WC (2015): Indications for and pregnancy outcomes of cervical cerclage: 11-year comparison of patients undergoing history-indicated, ultrasound-indicated, or rescue cerclage. Hong Kong Med J.,  21(4):310-7.
9. Coughlan MT, Oliva K, Georgiou HM, Permezel JMH and Rice GE (2001): Glucose-induced release of tumor necrosis factor-alpha from human placental and adipose tissues in gestational diabetes mellitus. Diabet Med., 18:921–7.
10. Deshmane SL, Kremlev S, Amini S and Sawaya BE (2009): Monocyte chemoattractant protein-1 (MCP-1): an overview. J Interferon Cytokine Res., 29:313–326.
11. Gaines-Das RE and Poole S (1993): The international standard for interleukin-6—evaluation in an international collaborative study. J Immunol Methods., 160: 147–53.
12. Heng YJ, Liong S, Permezel M, Rice GE, Di Quinzio MK and Georgiou HM (2014): The interplay of the interleukin 1 system in pregnancy and labor. Reprod Sci.,  21(1):122-30.
13. Hernandez-Andrade E, Romero R, Korzeniewski SJ, Ahn H, Aurioles-Garibay A, Garcia M, Schwartz AG, Yeo L, Chaiworapongsa T and Hassan SS (2014): Cervical strain determined by ultrasound elastography and its association with spontaneous preterm delivery. J Perinat Med., 42(2):159-69.
14. Jung EY, Park KH, Lee SY, Ryu A, Joo JK and Park JW (2016): Predicting outcomes of emergency cerclage in women with cervical insufficiency using inflammatory markers in maternal blood and amniotic fluid. Int J Gynaecol Obstet., 132(2):165-9.
15. Kiefer DG, Peltier MR, Keeler SM, Rust O, Ananth CV, Vintzileos AM and Hanna N (2016): Efficacy of midtrimester short cervix interventions is conditional on intraamniotic inflammation. Am J Obstet Gynecol., 214(2):276.e1-e6.
16. Lappas M (2018): Identification of SMAD3 as a Novel Mediator of Inflammation in Human Myometrium In Vitro. Mediators Inflamm., 27: 3140420.
17. Meher S and Alfirevic Z (2014): Choice of primary outcomes in randomised trials and systematic reviews evaluating interventions for preterm birth prevention: a systematic review. BJOG.  121(10):1188-94; discussion 1195-6.
18. Mendoza Cobaleda M, Ribera I, Maiz N, Goya M and Carreras E (2019): Cervical modifications after pessary placement in singleton pregnancies with maternal short cervical length: 2D and 3D ultrasound evaluation. Acta Obstet Gynecol Scand.,18: 1-8
19. Mönckeberg M, Valdés R, Kusanovic JP, Schepeler M, Nien JK, Pertossi E, Silva P, Silva K, Venegas P, Guajardo U, Romero R and Illanes SE (2019): Patients with acute cervical insufficiency without intra-amniotic infection/inflammation treated with cerclage have a good prognosis. J. Perinat. Med., 47(5): 500–509.
20. Monsanto SP, Daher S, Ono E, Pendeloski KPT, Trainá É, Ḿattar R, and Tayade C (2017): Cervical cerclage placement decreases local levels of proinflamm-atory cytokines in patients with cervical insufficiency. Am J Obstet Gynecol., 217(4):455.e1-455.e8.
21. Owen J, Yost N, Berghella V, Thom E, Swain M, Dildy GA 3rd, Miodovnik M, Langer O, Sibai B and McNellis D (2001): National Institute of Child Health and Human Development, Maternal-Fetal Medicine Units Network: Midtrimester endovaginal sonography in women at high risk for spontaneous preterm birth. JAMA, 286(11):1340-8.
22. Pandey M, Chauhan M and Awasthi S (2017): Interplay of cytokines in preterm birth. Indian J Med Res.,  146(3):316-327.
23. Raja NR, Bakaysa SL, Ghezzi CE, House MD and Kaplan DL (2019): Ex vivo pregnant-like tissue model to assess injectable hydrogel for preterm birth prevention. J Biomed Mater Res., 1–7
24. Rode L, Klein K and Nicolaides KH (2011):  Prevention of preterm delivery in twin gestations (PREDICT): a multicenter, randomized, placebo-controlled trial on the effect of vaginal micronized progesterone. Ultrasound Obstet Gynecol.,  38:272–280.
25. Sayaril N, Bhat RS, Warsy AS, Babay ZA, Addar MH, Shaik J, Al-Marry B and Al-Daihan S (2018): Polymorphisms in proinflammatory cytokine genes, effect on gene expression and association with preterm delivery in Saudi females. Cell Mol Biol (Noisy-le-grand),  64(10):55-60.
26. Schuit E, Stock S and Rode L (2015): Effectiveness of progestogens to improve perinatal outcome in twin pregnancies: an individual participant data meta-analysis. BJOG,  122:27–37.
27. Serra V, Perales A and Meseguer J (2013): Increased doses of vaginal progesterone for the prevention of preterm birth in twin pregnancies: a randomised controlled double-blind multicentre trial. BJOG, 120:50–57
28. Short CS, Quinlan R, Bennett P, Shattock RJ and Taylor GP (2018): Optimising the collection of female genital tract fluid for cytokine analysis in pregnant women. J Immunol Methods, 458:15-20.
29. Steenhaut P, Depoix C, Hubinont C and Debiève F (2019): Changes in fetal membrane histology with cervical insufficiency and transabdominal cerclage. Int J Gynaecol Obstet.,  146(2):223-230.
30. Sun X, Ding XP, Shi CY, Yang HX and Jin YZ (2016): Analysis of clinical effect of McDonald cervical cerclage and the related risk factors. Zhonghua Fu Chan Ke Za Zhi,  51(2):87-91.