Integral assessment of markers of the local infectious and inflammatory process in women with preterm birth in multiple pregnancies

Full Text

Introduction

The problem of preterm birth (PB) in multiple pregnancies has become a significant field in modern obstetrics. This priority issue is primarily due to the global trend of an increase in the frequency of multiple pregnancies. Because of the prevalence in the category of complications and significance of consequences, PB is considered the leading factor in adverse outcomes [1, 2]. Premature infants account for 60–70% of cases of early neonatal mortality and 50% of all established neurological diseases. Moreover, one-third share of PBs is associated with the induction of labor caused by complications in the mother and/or fetus, whereas the rest occur spontaneously [3].

Spontaneous PB represents a “major obstetric syndrome” characterized by a multifactorial nature. Its development is known to be contributed mainly by etiological factors such as a progesterone receptor decrease, cervical factors, vascular disorders, impaired maternal and fetal tolerance, and placental insufficiency. In case of multiple pregnancies, the most significant causes of PB include overdistension of the uterus, specific complications of multiple pregnancy, such as feto-fetal syndrome, and selective retardation in the development of one fetus [4–7].

The task of determining the leading cause of PB is difficult, and even impossible in some cases. In most studies, it was revealed that it can be infectious and inflammatory processes of the lower organs of the genital tract (40%) in singleton pregnancy [7–9]. At the same time, in multiple pregnancies, there are insufficient data on the contribution of inflammation to the polyetiological process of PB implementation. The pathways of infection spreading (predominance of the ascending or hematogenous pathway), pathogens that initiate the cascade of immune responses (pathogenic and opportunistic microorganisms), and mechanisms leading to PB have also been studied insufficiently.

Today, inflammatory processes play a significant role in the development of labor, namely, remodeling of the cervix during maturation, rupture of membranes, and occurrence of regular uterine contractions. A large amount of data presented that by the onset of labor in the amnion, chorion, and decidual membranes, the number of pro-inflammatory cytokines and prostaglandins increases. Subpopulations of leukocytes, macrophages, and neutrophils increase in the cervix. A change in the ratio of pro-inflammatory and anti-inflammatory cytokines toward a decrease in the latter was developed evolutionarily to activate the inflammatory process necessary for the physiological act of delivery [7, 9].

During pregnancy, the inflammatory response to the permeation of infectious agents can reproduce this model of reactions leading to the preterm onset of labor. Microbial invasion, initiating an inflammatory reaction from the amniotic epithelium of the fetal membranes and umbilical cord, can promote the production of pro-inflammatory cytokines, chemokines, and prostaglandins, which in turn trigger contractile activity of the uterus, shortening of the cervix, rupture of membranes and, as a result, PB [5–7].

In multifetal pregnancy, which in most cases occurs because of the use of assisted reproductive technology (70%) and is often accompanied by a complicated course, in the process of pregravid preparation and antenatal monitoring, local infectious and inflammatory processes are controlled and corrected especially carefully. Despite this, the persisting high rate of PB in multifetal pregnancy (about 54%) indicates the ambiguity of the idea of the suppressive role of the infectious factor in the induction of PB in multifetal pregnancy. Studying the degree of contribution of inflammatory processes to multifactorial mechanisms of implementation is an important aspect not only from the standpoint of the possibility of their identification as predictors and targets for exposure but also a general understanding of the genesis of PB in multifetal pregnancy.

Materials and methods

This prospective cohort study included 30 women with multiple pregnancies, who were admitted at the D.O. Ott Scientific Research Institute of Obstetrics, Gynecology, and Reproductology with a diagnosis of threatened preterm labor. The inclusion criteria were pregnancy with dichorionic diamniotic twins and presence of clinical manifestations of threatening PB at a gestational age of 22–36 weeks (29.4 ± 2.1).

A comprehensive assessment of the qualitative and quantitative compositions of microorganisms of the discharge of the lower segments of the genital tract was performed to all pregnant women using microscopic, bacteriological, molecular, and biological studies (Femoflor-16 test, DNA-Technology, Russia), and an indicator of local inflammatory status was identified by the ratio of profiles of immune response gene expression using an ImmunoQuantex test system (DNA-Technology).

Microscopic examination. A microscopic examination of the urogenital tract discharge was performed with an assessment of the microbiocenosis when staining the preparations with methylene blue and using gram staining. In the vaginal discharge, the ratio of polymorphonuclear leukocytes (PMNL) to squamous epithelial cells (SEC) was assessed, as well as lactobacilli in relation to other microorganisms, clue cells, yeast-like fungi, and trichomonas. In addition, in the cervical discharge, columnar epithelial cells, SEC, PMNL, mucus, and intracellular and extracellular gram-negative diplococci were assessed.

Bacteriological study. Microorganisms were isolated using standard nutriculture media for gram-positive and gram-negative bacteria. The microorganisms isolated were identified by mass spectrometry using a matrix-assisted laser desorption and ionization time-of-flight mass spectrometry bacteriological analyzer (Bruker, Germany).

Molecular biological study. Using the Femoflor-16 test system, the discharge of the lower parts of the genital tract was studied using polymerase chain reaction in real time with a quantitative assessment of opportunistic pathogens, Lactobacillus spp., and determination of total bacterial mass. When interpreting the results, the quantitative ratio of Lactobacillus spp. and the total bacterial mass were calculated, and the following conclusions about the microbiocenosis state were made: (1) Lactobacillus spp. > 80% indicated normocenosis characterized by the dominance of the normal flora, (2) Lactobacillus spp. 20–80% indicated moderate dysbiosis, and (3) Lactobacillus spp. < 20% indicated severe dysbiosis.

Using the ImmunoQuantex test system, according to the manufacturer’s instructions, indices of local inflammatory status were determined by the ratio of the expression profiles of the immune response genes. To isolate nucleic acids, the Proba NK kits (DNA-Technology) were used. After isolation of the total pool of DNA and RNA, a reverse transcription reaction was performed to obtain from the template the mRNA complementary to DNA, which was further amplified by the polymerase chain reaction method. After the amplification stage, the mRNA expression level of eight innate immunity genes (IL-1B, IL-10, IL-18, TNFα, TLR4, GATA3, CD68, and B2М) was calculated using the indicator cycle value. Based on the integral assessment of the obtained gene expression levels, it was concluded on the presence or absence of a local inflammatory response according to the value of the inflammation index. An inflammation index value of more than 60% was considered inflammation, less than 50% was considered the absence of an inflammatory response, and 50–60% implied that an inflammatory response could not be ruled out (gray area).

Laboratory studies were conducted after obtaining biomaterials from all pregnant women. (No correction was performed on the basis of the results obtained.) After obtaining data on pregnancy outcomes, the study results were evaluated and analyzed. Comparison groups were formed depending on the gestational age at the time of delivery and type of delivery. Women with spontaneous PB (13) were included in the main group, and those with term delivery (16) constituted the control group. Women with medically induced PB (2) were excluded from the comparative analysis. The results of their testing were described only when characterizing the data.

Statistical analysis

Statistical processing of the results was performed using the STATISTICA ver. 10 program. For analyzing qualitative attributes, the Pearson chi-square test (χ²) was used. If the number of expected attributes was less than 10 in the analysis of fourfold tables, the chi-square test (χ²) with Yates’ correction was used; at frequencies less than 5, two-tailed Fisher’s exact test was used (p). With a normal distribution of quantitative data, the results were presented as arithmetic mean and mean-square deviation (M ± σ), and the significance was calculated using Student’s t-test (t). In the absence of a normal distribution, the median indicating the interquartile range Me (L–H) and Mann–Whitney test (U) were calculated.

A mathematical model of PB probability was developed, based on an integral assessment of markers of the local inflammatory process by the mRNA level of innate immunity genes in epithelial cells of the lower parts of the reproductive tract. A linear discriminant analysis was performed to identify the predictive value of the parameters under study, deduce the formula for calculating the probability, and check the operation of the algorithm. For all types of analysis, p-values of <0.05 were considered statistically significant.

Results

Microscopic examination of the discharge from the vagina and cervical canal revealed several aspects of the state of the microbiocenosis of the genital tract organs in women with preterm and term birth.

When examining vaginal discharges, in about half of the cases, both during spontaneous and PB, in relation to PMNL to SEC, inflammation was noted (PMNL > SEC). However, there were no statistically significant differences in this indicator in the groups (p > 0.05).

There were also no significant differences between the groups in the lactobacilli count (Lactobacillius spp.) and their ratio with other microorganisms (p > 0.05). Only lactobacilli or lactobacilli prevalence was found in 84.4% of pregnant women. In the remaining 15.6% of women, lactobacilli were either absent or present in the minority. In both groups, pseudomycelium was detected in single samples (4.65%), but clue cells or trichomonas were not found.

When examining the cervical canal discharge in the group of women with preterm labor, an inflammatory reaction was determined significantly more often compared with the control group, characterized by several leukocytes of more than 10 in the field of view of a light microscope with a magnification of × 1000 (53.9% and 31.3%, respectively, p = 0.033) and mucus content (73.9% and 26.1%, p = 0.026).

Bacteriological examination of discharge from the cervical canal found microorganisms such as Lactobacillius spp., Gardnerella vaginalis, Escherichia coli, Streptococcus agalactiae, Streptococcus spp., Staphylococcus spp., Enterococcus faecalis, Corynebacterium spp., Klebsiella spp., and Candida spp.

In the group of PB women, lactobacilli were isolated less frequently, but more often, opportunistic microorganisms were revealed. In the main group, Lactobacillius spp. was identified in 15.4% of cases. In the group of women whose pregnancy ended in term delivery, lactobacilli were isolated in 32.5%, but the differences were not significant (p > 0.05).

In the main group, the isolation frequency of E. faecalis was two times higher than in the control group, but the differences were not statistically significant (23.1% and 12.5%, respectively; p > 0.05).

According to the frequency of isolated E. coli, S. agalactiae, Streptococcus spp., Staphylococcus spp., Klebsiella spp., and Candida spp., the groups were comparable and did not differ significantly (p > 0.05). Corynebacterium spp. were not identified by microbiological examination but were detected by molecular biological examination (polymerase chain reaction in real time), which necessitates the application of methods for identifying genital infections in aggregate.

In a molecular biological study using the Femoflor-16 test system (assessment of the characteristics of the state of the vaginal microbiocenosis in the groups being studied through multiple analysis), it was revealed that in the PB group, the frequency of dysbiotic disorders and degree of their severity were higher than in the control group (p = 0.029).

In women whose pregnancy resulted in PB, the frequency of physiological microbiocenosis of the vagina was insignificantly lower (76.9%) than in those who had term delivery (87.5%). When assessing vaginal dysbiosis, the control group patients were characterized by a milder form of microbiocenosis disorders, namely, moderate anaerobic dysbiosis (12.5%), whereas in patients whose pregnancy resulted in spontaneous preterm labor, moderate dysbiosis was not registered. In the PB group, more pronounced disorders of the vaginal microbiocenosis were noted, including pronounced anaerobic dysbiosis (15.38%) and pronounced aerobic dysbiosis (7.69%). There were no pronounced dysbiotic disorders during term delivery (Figure).

 

Characteristics of vaginal microbiocenosis in the study groups (* p < 0.05)

Характеристика состояния микробиоценоза влагалища у беременных в исследуемых группах (* p < 0,05)

 

In the control group, the amount of Lactobacillius spp. included in the vaginal microflora during pregnancy corresponded to the physiological indicator in 100% (16) of cases, whereas in the main group, physiological microbiocenosis was registered in 69% (9) of cases. The median of the number of lactobacilli in the main group was significantly lower and amounted to 5.0 GE/sample, whereas at term delivery, it was 6.8 GE/sample (p = 0.0019). There were no intergroup differences in the frequency of optionally anaerobic microorganisms (bacteria of the Enterobacteriaceae family, Streptococcus spp., and Staphylococcus spp.; p > 0.05). In the main group, obligate anaerobic microorganisms (Lachnobacterium spp./Clostridium spp. and Peptostreptococcus spp.) were noted with a frequency of 7.69% (1), whereas in the control group, these microorganisms were not identified. The group of microorganisms Snethia spp./Leptotrichia spp./Fusobacterium spp. was not identified in any group under study. The frequency of detection of Ureaplasma (urealyticum + parvum) did not differ significantly in the main and control groups (46.15% and 31.3%, respectively; p > 0.05; Table 1).

 

Table 1 / Таблица 1

Vaginal microbiota composition in the study groups

Состав микробиоты влагалища у женщин в основной и контрольной группах

Microbiota composition	Preterm delivery (n = 13), % (n)	Term birth (n = 16),% (n)	p-value
Normal flora
Lactobacillius spp.	69.2 (9)	100 (16)	0.0019
Optionally anaerobic microorganisms
Enterobacteriaceae	30.8(4)	18.8 (3)	0.68
Streptococcus spp.	30.8 (4)	18.8 (3)	0.46
Staphylococcus spp.	53.9 (7)	43.8 (7)	0.96
Obligate anaerobic microorganisms
Gardnerella vaginalis/Prevotella bivia/Porphyromonas spp.	38.5 (5)	37.5 (6)	0.84
Eubacterium spp.	30.8 (4)	50.0 (8)	0.54
Megasphaera spp./Veillonella spp./Dialister spp.	15.4 (2)	18.8 (3)	0.84
Lachnobacterium spp./Clostridium spp.	7.7 (1)	0	0.29
Mobillincus spp./Corynebacterium spp.	7.7 (1)	18.8 (3)	0.49
Peptostreptococcus spp.	7.69 (1)	0	0.29
Atopobium vaginae	15.38 (2)	25.0 (4)	0.55
Yeast-like fungi
Candida spp.	0	12.5 (2)	0.21
Mycoplasma
Mycoplasma hominis	38.5 (5)	0	0.0085
Ureaplasma (urealyticum + parvum)	46.2 (6)	31.3 (5)	0.84

Note. Significant values at p < 0.05 are presented in bold.

 

There were significant differences in the groups of women for detecting Mycoplasma hominis. In the main group, M. hominis was revealed in 38.5% of cases (5) with a median number of 0.92 GE/sample, whereas in the control group, these microorganisms were not detected in any case (p = 0.0085).

Assessment of the state of the local immune status in women with dichorionic twins. To assess the presence and severity of local inflammation in pregnant women with dichorionic twins, the ImmunoQuantex test was performed, based on determining the mRNA expression profile of eight innate immunity genes in the epithelial cells of the lower reproductive tract and four indicators of their ratios (Table 2).

 

Table 2 / Таблица 2

Mean values of local immune status parameters in the study cohort

Средние значения показателей локального иммунного статуса у женщин в исследуемой когорте

Indicator,lg GE/mL	M ± σ	Min	Max
IL-1B	5.0 ± 1.2	2.1	7.0
IL-10	2.3 ± 1.3	0	5.1
IL-18	4.6 ± 1.3	1.4	6.3
TNFα	3.1 ± 1.3	0	5.1
TLR4	2.5 ± 1.5	0	4.6
GATA3	3.7 ± 1.5	0	5.7
CD68	4.3 ± 1.6	0	6.1
B2M	4.9 ± 1.5	1.3	6.9
TLR4/GATA3	0.9 ± 2.1	0	9.2
TNFα/IL-18	0.6 ± 2.3	0	13
IL-10/IL-18	238.9 ± 1162.8	0	6498
IL-1B/CD68	13.5 ± 24.7	0	90.5
Inflammation index, %	33.5 ± 44.0	0	100

 

Among all the patients examined, the median expression level of the gene that encodes interleukin (IL)-1b, which is an important mediator of the inflammatory response, was 5.2 lg GE/mL. An inhibitor of cellular immunity is IL-10, and the expression value of this interleukin gene was 2.4 lg GE/mL. The level of IL-18, which has pleiotropic effects in the immune response, was 4.9 lg GE/mL.

The expression level of the gene encoding tumor necrosis factor (TNF)α, a protein of the acute phase of inflammation, was 3.2 lg GE/mL. The expression level of TLR4, which determines the first line of local antibacterial defense, corresponded to 2.7 lg GE/mL, and the level of GATA3 transcription factor was 4.1 lg GE/mL. The expression value of the gene encoding CD68, which is significant in the phagocytic activity of cellular macrophages, was 4.7 lg GE/mL. The expression level of the gene encoding B2M associated with an increase in the activity of the immune system was 5.5 lg GE/mL.

The parameter ratio was 0.04 lg GE/mL for TLR4/GATA3, 0.02 lg GE/mL for TNFα/IL-18, 2.1 lg GE/mL for IL-10/IL-18, and 2.8 lg GE/mL for IL-1B/CD68. The inflammation index was calculated from the ratio of mRNA expression levels of pro-inflammatory and anti-inflammatory cytokine genes; and the median was 3.3%. With an inflammation index of less than 50%, the test was regarded as negative, and a conclusion was made about the absence of inflammation in 58% of cases. With an index of 50–60% (“gray zone”), a reliable assessment of the local immune status was impossible in 3% of cases, and with an index of more than 60%, the test was determined as positive with the presence of local inflammatory processes stated in 26%.

It was not possible to estimate the inflammation index in 13% of cases because of the level of B2M expression below the critical value (not less than 4 lg GE/mL), which determines the test validity. It was noted that all patients with “ambiguous” test results gave birth prematurely. Further analysis showed that spontaneous PB occurred in 28.57% (8) of patients, which level of B2M expression in the sample was less than 4 lg GE/mL. The expression level of the B2M-encoding gene differed significantly in the PB group, and the median was significantly lower (4.5 lg GE/mL) compared with term delivery (6.0 lg GE/mL; p = 0.035).

A significant decrease in the expression of most genes, including IL-1B, TNFα, TLR4, and GATA3, was also found in the PB group compared with the term delivery group (Table 3).

 

Table 3 / Таблица 3

ImmunoQuantex test values in the study groups

Показатели теста «ИммуноКвантэкс» у пациенток в сравниваемых группах

Indicator,lg GE/mL	Preterm delivery (n = 13)		Term delivery (n = 16)		p-value
	M ± σ	min–max	M ± σ	min–max
IL-1B	4.4 ± 1.3	2.1–6.2	5.5 ± 1.0	3.9–7.0	0.034
IL-10	2.2 ± 1.5	0–5.1	2.2 ± 1.2	0–3.7	>0.05
IL-18	4.4 ± 1.5	1.4–6.3	4.9 ± 0.6	2.3 ± 6.1	>0.05
TNFα	2.5 ± 1.5	0–4.3	3.8 ± 0.9	2.1–5.1	0.024
TLR4	1.7 ± 1.6	0–4.1	3.2 ± 1.0	1.9–4.6	0.048
GATA3	3.1 ± 1.7	0–5.2	4.5 ± 1.0	1.9–5.7	0.023
CD68	3.6 ± 1.9	0–5.9	5.0 ± 0.84	3.2–6.1	0.077
B2M	4.3 ± 1.7	1.3–6.7	5.7 ± 0.79	4.2–6.9	0.035
TLR4/GATA3	0.9 ± 2.0	0–7.5	1.0 ± 2.6	0.001–9.2	>0.05
TNFα/IL-18	1.0 ± 3.3	0–13.0	0.2 ± 0.26	0.009–0.81	>0.05
IL-10/IL-18	84.2 ± 169.1	0–649.8	14.2 ± 35.8	0–125	>0.05
IL-1B/CD68	10.5 ± 19.3	0–64	13.5 ± 24.5	0.25–78.8	>0.05
Inflammation index, %	31.6 ± 44.3	0–100	37.7 ± 45.8	0–100	>0.05

Note. Significant values at p < 0.05 are presented in bold.

 

The median expression of the IL-1B gene, a pro-inflammatory cytokine that is an important mediator of the inflammatory response inducing prostaglandin synthesis, neutrophil activation, and antibody production, was 4.6 lg GE/mL in PB and 5.8 lg GE/mL in term delivery (p = 0.064).

A relative decrease in the expression of the TNFα gene, which is a multifunctional pro-inflammatory cytokine synthesized mainly by monocytes and macrophages, has a significant effect on the endothelium functioning, production stimulation of IL-1, IL-6, IL-8, and interferon-gamma, and leukocyte activation. The median level of TNFα expression was 2.8 lg GE/mL in patients with PB and 4.35 lg GE/mL in women with term delivery (p = 0.024).

Similar data were obtained when analyzing the expression of receptor genes that determine the first line of natural local defense when permeating pathogenic microorganisms (TLR4) so that the median was significantly lower in PB (1.7 lg GE/mL) than in term delivery (3.7 lg GE/mL; p = 0.048).

The expression of the gene for the transcription factor GATA3, which regulates the expression of the spectrum of genes involved in the development of inflammatory and allergic reactions, was also significantly lower in PB than in term delivery (3.8 and 4.9 lg GE/mL, respectively; p = 0.023).

In PB, there was also a tendency to a decrease in the CD68 level, a glycoprotein that is significant in the phagocytic activity of tissue macrophages, both in intracellular lysosomal metabolism and in extracellular cell–cell and cell–pathogen interactions. The expression level of CD68 was 4.2 and 5.4 lg GE/mL in patients with preterm and term deliveries, respectively (p = 0.077), which is inconsistent with the literature data on the prevalence of pro-inflammatory factors in PB.

There were no statistically significant differences in the levels of gene expression of such regulatory pro-inflammatory and anti-inflammatory cytokines as IL-18 and IL-10, which perform multiple pleiotropic functions in immunoregulation and inflammation. The expression levels of genes IL-18 and IL-10 were 4.9 and 2.3 lg GE/mL, respectively, in PB, and 5.5 and 3.1 lg GE/mL, respectively, in term delivery (p > 0.05). Using various statistical methods, there were also no significant differences in the groups in the parameters of the ratio TLR4/GATA3, TNFα/IL-18, IL-10/IL-18, and IL-1B/CD68 (p > 0.05).

The compared groups did not differ in terms of the inflammation index calculated from the difference in these ratios (p > 0.05). When interpreting the test as an “inflammatory,” the samples differed significantly from the “noninflammatory” ones in the relative decrease in the expression level of IL-18 genes (4.0 and 5.2 lg GE/mL, respectively; p = 0.015), the tendency toward a decrease in the GATA3 expression level (3.3 and 4.4 lg GE/mL, respectively; p = 0.055), and increase in TLR4 (3.5 and 2.5 lg GE/mL, respectively; p = 0.055).

In terms of the ratio of factors, positive tests that determined the presence of inflammation differed significantly from negative ones with relatively higher values of TLR4/GATA3 (3.11 and 0.04; p = 0.0001), TNFα/IL-18 (2.13 and 0.05; p = 0.0002), and IL-1B/CD68 (39.8 and 4.9; p = 0.0004).

Predictive model of probability of PB in multifetal pregnancy

Based on the data obtained, a prognostic model of PB probability in multifetal pregnancy was created based on the mRNA level of innate immunity genes in epithelial cells of the lower parts of the reproductive tract.

Linear discriminant analysis was used to reveal the prognostic value of the parameters under study, deduce the formula for calculating the probability of PB, and check the operation of the algorithm for predicting it.

The values of the discriminant functions were calculated using the following equations:

F = –151.299 + 45.199 ∙ X – 11.166 ∙ Y –– 45.49 ∙ Z – 64.028 ∙ M + 10.267 ∙ K –– 71.962 ∙ L + 129.403 ∙ S + 5.985 ∙ R;

F₁ = –153.151 + 45.155 ∙ X – 11.123 ∙ Y –– 45.375 ∙ Z – 63.212 ∙ M + 11.983 ∙ K –– 72.564 ∙ L + 129.102 ∙ S + 5.209 ∙ R,

where F is the probability of PB; F₁ is the probability of term delivery; X is the expression level of IL-1B gene; Y is the expression level of IL-10 gene; Z is the expression level of TNFα gene; M is the expression level of TLR4 gene; K is the expression level of GATA3 gene; L is the expression level of CD68 gene; S is the expression level of B2M gene; R is the expression level of IL-18 gene. With F < F₁, PB is predicted, and with F < F₁, term delivery is predicted (Table 4).

 

Table 4 / Таблица 4

Function parameters in the probability model of preterm birth in multiple pregnancies

Параметры функции в модели вероятности реализации преждевременных родов при многоплодии

Indicators	Indicator designation	Coefficients
		Pretermdelivery	Termdelivery
IL-1B	Х	45.199	45.155
IL-10	Y	–11.166	–11.123
TNFα	Z	–45.49	–45.375
TLR4	M	–64.028	–63.212
GATA3	K	10.267	11.983
CD68	L	–71.962	–72.564
B2M	S	129.403	129.102
IL-18	R	5.985	5.209
	Constant	–151.299	–153.151

 

When checking this formula, the program assigned patients to the group with the probable development of PB in 100% of cases. The program identified the probability of term delivery correctly in 86.7% of cases. A false positive result was obtained for PB in 13.3% of cases.

When calculated using the above formula, if the probability by control (term delivery) is less than the probability by PB, then it can be assumed with a probability of 100% that this patient will have spontaneous PB. If the probability by control is greater than that by PB, then it can be assumed with a probability of 86.7% that this patient will have a term delivery.

Discussion

This study was the first to assess the characteristics of vaginal microbiocenosis and local immune status in women with multiple pregnancies and PB. It was noteworthy that, in general, the pregnant women of the study cohort had a low infectious-inflammatory status, but vaginal biotope disorders were registered with a higher frequency in the PB group.

The microscopic presentation of the vagina and cervical canal microbiocenosis in women with PB was characterized by a significantly higher leukocyte count and mucus amount compared with patients with term delivery. A more pronounced degree of leukocyte reaction in the main group may indicate both mucosal immunity activation and cervical remodeling processes (mucous plug discharge) [8].

When determining the spectrum of microorganisms in the vaginal microbiota during bacteriological examination, E. faecalis was isolated in the PB group two times more often than in the control group, but the differences were not significant. As a rule, E. faecalis colonization of the urogenital tract of pregnant women has no significant consequences, but an excessive number of these microorganisms can be associated with pregnancy complications such as PB, chronic placental insufficiency, and small-for-gestational-age fetus [9].

The molecular biological study revealed that vaginal dysbiosis in the PB group is found significantly more often than in the control group, which manifested itself as a small amount or complete absence of lactobacilli in the vaginal microbiocenosis [6, 10, 11]. Lactobacilli are of great importance for maintaining normal vaginal microbiocenosis because of high competition and antagonism with respect to most pathogenic and opportunistic bacteria [11]. In the absence of lactobacilli, favorable conditions are created for activating microorganisms exhibiting pathogenic properties. The resulting inflammation becomes a trigger for premature contractile uterine activity, and an increase in the proteolytic enzyme level can lead to the shortening of the cervix and early rupture of the fetal membranes [12, 13].

In pregnant women with PB, M. hominis was detected in the vaginal microbiota, which was not identified in any case in women who gave birth at full term. M. hominis plays a significant role in the genesis of miscarriage at all stages of gestation [14, 15]. Mycoplasma membrane structures can activate macrophages and monocytes and increase main pro-inflammatory cytokine secretion, especially TNFα, ILs (IL-1, IL-1b, IL-6, IL-8, IL-12, and IL-16), and interferon-gamma, causing local inflammatory processes that contribute significantly to the initiation of PB [16–18].

In this study, for the first time, the state of the local immune status in women with dichorionic twins was analyzed by the mRNA expression profile of the IL- IL-1B, IL-10, IL-18, TNFα, TLR4, GATA3, CD68, and B2M genes. During the study, a significant decrease in the expression of the genes IL-1B, TNFα, TLR4, and GATA3 was revealed in women with PB than in the control group. Multivariate factor analysis enabled to characterize the relationship of these parameters with the term of delivery as directly proportional, namely, the lower the level of expression of these markers was, the shorter was the term of delivery.

Researchers who have obtained similar data in singleton pregnancies in their works agree that the most important is not an increase in the expression of these genes but an imbalance toward a decrease in the levels of pro-inflammatory cytokines, which lead to PB [19–22]. Such data are explained by the hyporeactive state of local immunity, which determines favorable conditions for ascending infection, and the development of chorioamnionitis, which is an important pathogenetic link in PB mechanisms.

At the same time, it should be noted that earlier studies were characterized by the isolated determination of the levels of inflammatory markers without a comprehensive assessment of local immunity. In this study, for the first time, a mathematical model was formed on the basis of an integral assessment of the local immune status parameters. The results obtained revealed both the importance of a comprehensive assessment of local inflammatory processes for understanding the multifactorial mechanisms of PB implementation in multiple pregnancies and the possibility of predicting PB using this model with high efficiency (86.7%).

Conclusion

Evaluation of markers of infectious and inflammatory processes of PB is an important aspect not only for identifying these markers as predictors and targets of exposure but also for understanding the pathogenesis of PB in multiple pregnancies.

In accordance with the results obtained, PB markers in multiple pregnancies can be an increase in the level of leukocytes in the cervical canal to more than 10 in the field of view during microscopic examination of cervical discharge, decrease in the count of lactobacilli (vaginal dysbiosis), and presence of genital mycoplasmas in the vaginal microbiocenosis detected using the polymerase chain reaction in real time.

The studied indices of mRNA expression of innate immunity genes (IL-1B, IL-10, IL-18, TNFα, TLR4, GATA3, CD68, and B2M) in the vagina (polymerase chain reaction with reverse transcription) and mathematical model developed based on an integrated assessment of the data obtained (probability 87.6%) demonstrated the prospectivity of their assessment to predict PB in multiple pregnancies.

In turn, the general low inflammatory profile in the cohort under study, including a decrease in the expression of mRNA of the genes IL-1B, TNFα, TLR4, and GATA3 in vaginal discharge, necessitates further research to study the degree of contribution of infectious and inflammatory factors to the mechanisms of PB development in multiple pregnancies and search for markers of other ways of their implementation.

About the authors

Olga V. Kosyakova

The Research Institute of Obstetrics, Gynecology, and Reproductology named after D.O. Ott

Author for correspondence.
Email: for.olga.kosyakova@gmail.com
ORCID iD: 0000-0003-4116-0222
SPIN-code: 1204-3160

MD, Post-Graduate Student. The Department of Pathology of Pregnancy N 1

Russian Federation, Saint Petersburg

Olesya N. Bespalova

The Research Institute of Obstetrics, Gynecology, and Reproductology named after D.O. Ott

Email: shiggerra@mail.ru
ORCID iD: 0000-0002-6542-5953
SPIN-code: 4732-8089

MD, PhD, DSci (Medicine), Deputy Director 

Russian Federation, Saint Petersburg

Olga V. Budilovskaya

The Research Institute of Obstetrics, Gynecology, and Reproductology named after D.O. Ott; Saint Petersburg State Pediatric Medical University

Email: o.budilovskaya@gmail.com
ORCID iD: 0000-0001-7673-6274
SPIN-code: 7603-6982

Researcher. The Laboratory of Microbiology

Russian Federation, Saint Petersburg

Alevtina M. Savicheva

The Research Institute of Obstetrics, Gynecology, and Reproductology named after D.O. Ott; Saint Petersburg State Pediatric Medical University

Email: savitcheva@mail.ru
ORCID iD: 0000-0003-3870-5930
SPIN-code: 8007-2630

D. Sci. (Med.), Prof., Ott Research Institute of Obstetrics, Gynecology and Reproductology, Saint Petersburg State Pediatric Medical University

Russian Federation, Saint Petersburg

References

Supplementary files