Diagnostic significance of autoantibody combination in fetal growth restriction with early and late manifestation

Full Text

BACKGROUND

Fetal growth restriction (FGR) is a common placenta-associated complication of pregnancy that increases the risk of preterm delivery and premature birth, antepartum and intrapartum death of the fetus and newborn, and other adverse outcomes [1]. FGR is characterized by impaired fetoplacental blood flow, fetal weight 5th percentile, frequent oligohydramnios, and signs of hypoxia according to cardiotocography (CTG).

Pregnancy complications caused by placental dysfunction negatively affect the growth and functional status of the fetus and further development of the individual throughout postnatal life, which enabled us to describe “adverse fetal programming”. Thus, prediction and early diagnostics of FGR, which involves the optimization of the obstetric approach, are significant in improving long-term pediatric and therapeutic outcomes.

A population-based study (n=480,448) conducted in France in 2017 has demonstrated that the costs of antenatal and postnatal healthcare for mothers and newborns with FGR were much higher than those for children corresponding to a gestational age. It is estimated that maternal and neonatal hospital care costs associated with FGR account for 23% of total maternal healthcare costs in France [2].

The etiology of FGR remains unclear. However, FGR is believed to be caused by an interaction between environmental and genetic factors of embryonal, placental, or maternal origin [3]. Understanding the etiopathogenesis of FGR is crucial for timely diagnosis and optimization of the obstetric approach. FGR usually occurs because of disturbances in the formation and development of the placental complex. Impairment of wave 2 of trophoblast invasion and vascular remodeling leads to the development of ischemic placental disease, which underlies obstetric syndromes such as preeclampsia (PE), placental abruption, and FGR [4]. The more pronounced the disturbance in the remodeling of spiral arteries, damage to the chorionic villi, and decrease in oxygen supply to the intervillous space, the earlier the FGR manifests, reaching a critical condition of the fetus, and the higher the risk of developing adverse perinatal outcomes and long-term consequences.

There are two phenotypes of fetal growth restriction: FGR with early manifestation until week 32 of pregnancy and FGR with late manifestation after week 32 of pregnancy. In clinical practice, to distinguish between early and late phenotypes of FGR, the criteria proposed by the international Delphi consensus are used [6].

Currently, despite a wide variety of studies on the search for informative diagnostic and prognostic markers, FGR is considered a complex obstetric and perinatal problem.

Natural autoantibodies (AAB) are an integral part of the immune system of healthy people. Scientists continue to study the frequency of deviations in the level of AAB in the blood serum of pregnant women to predict various somatic diseases, reproductive disorders, and pregnancy complications [6].

This study aimed to evaluate the diagnostic significance of isolated and combined changes in the level of autoantibodies in the prognosis and diagnostics of FGR with early and late manifestation.

MATERIALS AND METHODS

The main study group consisted of pregnant women with FGR (n = 90) hospitalized at the Maternity Hospital of the S.S. Yudin City Clinical Hospital, and the control group included women with a normal pregnancy (n=27) registered in the Consultative and Diagnostic Department of the same hospital. The study was conducted from September 2020 to September 2021. This study was approved by the Local Ethics Committee of the I.M. Sechenov First Moscow State Medical University (dated February 19, 2020; extracted from protocol No. 03-20). All pregnant women provided informed voluntary consent to participate in the study. The study groups included patients aged >18 years with a gestation period of ≥28 weeks. The study did not include patients aged <18 years or those with multifetal pregnancies. The exclusion criteria in all groups were diabetes mellitus (gestational, types I and II), intake of immunosuppressants, antiphospholipid syndrome, and acute infectious process.

The diagnosis of FGR was made based on clinical recommendations [5], and then in the hospital, venous blood was collected from pregnant women in the study group to determine the serum AAB level. Pregnant women from the control group underwent a standard blood sampling procedure for a clinical study at weeks 28–38.

Autoimmune antibodies were determined using an enzyme-linked immunosorbent assay using the ELI-P-Test panel, which includes 12 immunoreagents to autoantibodies: AAB markers of anti-hCG syndrome (AAB to hCG, human chorionic gonadotropin); AAB markers of infectious-inflammatory, scar-adhesive, and autoimmune processes (AAB to DNA, AAB to β₂-GP, AT to collagen, AAB to Fc-IgC); AAB markers of changes in the Langerhans islets (AAB to insulin); AAB markers of changes in thyroid tissue (AAB to thyroglobulin); AAB markers of changes in the central and/or peripheral nervous system (AAB to S100); AAB markers of changes in the pelvic organs (AAB to Spr); AAB markers of changes in the vessels and hemostatic system (AAB-TrM, AAB-ANCA); and AAB markers of changes in kidney tissue (AAB-KiMS).

Statistical methods

Statistical analysis was performed using the StatTech v. 2.8.5 program (developer Stattekh, Russia). Quantitative indicators were assessed for compliance with normal distribution using the Shapiro–Wilk test (for <50 participants) or the Kolmogorov–Smirnov test (for >50 participants). Quantitative indicators with a normal distribution were described using arithmetic means (M) and standard deviations (SD) and limits of the 95% confidence interval (95% CI). In the absence of a normal distribution, quantitative data were described using the median (Me) and lower and upper quartiles (Q1–Q3). Categorical data were described using absolute values and percentages. Comparison of three or more groups according to a quantitative indicator, whose distribution differed from normal, was performed using the Kruskal–Wallis test, and post hoc comparisons were performed using Dunn’s test with Holm’s correction.

Comparison of percentages in the analysis of four-field contingency tables was performed using Pearson’s chi-square test (for values of the expected phenomenon >10) and Fisher’s exact test (for values of the expected phenomenon <10).

A prognostic model of the probability of a certain outcome was constructed using the logistic regression method. Nigelkirk’s R² coefficient served as a measure of certainty, indicating that part of the variance can be explained using logistic regression. To assess the diagnostic significance of quantitative characteristics in predicting a certain outcome, the ROC curve analysis method was used. The separating value of a quantitative characteristic at the cut-off point was determined by the highest value of the Youden index.

Differences in indicators were considered statistically significant at p <0.05.

RESULTS

The examined pregnant women of the main group, depending on the period of FGR manifestation, were divided into two subgroups in accordance with the Delphi consensus criteria [6], where subgroup 1 (1A) included 45 patients with early FGR manifestation, and subgroup 2 (1B) included 45 patients with late FGR manifestation. The age of pregnant women in subgroup 1A was 34 (29; 37) years and in subgroup 1B was 33 (25; 35) years. The age of pregnant women in group 2 (control group) ranged from 22 to 28 years and averaged 23 years. When assessing age, statistically significant differences were revealed depending on the groups (p <0.001).

Furthermore, 26 (59.1%) primigravidas among 45 patients (subgroup 1A) had early manifestation of FGR; 22 (48.9%) primigravidas among 45 patients of subgroup 1B had late manifestation of FGR; and 8 (29.6%) primigravidas among 27patients of group 2 (control) had normal pregnancy. No statistical differences by parity was noted (p=0.054).

Current pregnancy occurred spontaneously in all patients with early manifestation of FGR and in 42/45 (93.3%) patients with late manifestation of FGR, and in 3/45 (6.7%) patients of this subgroup, pregnancy occurred because of in vitro fertilization (IVF). In all patients with normal gestation, pregnancy occurred spontaneously.

According to prenatal screening 1, a high risk of FGR was identified in 27/45 (61.4%) pregnant women with early manifestation of FGR and in 14/45 (31.1%) pregnant women with late manifestation of FGR (p=0.004). In 21 of 45 (47.7%) pregnant women with early manifestation of FGR and in 14/45 (31.1%) pregnant women with late manifestation of FGR, a high risk of preeclampsia (PE) was registered according to prenatal screening 1. However, according to prenatal screening 1, patients in the main group with a high risk of FGR and PE did not take acetylsalicylic acid for prophylactic purposes. Five of 27 (26.3%) pregnant women from the control group had a high risk of FGR and PE according to trimester 1 screening conducted using the Astraia program. All patients with a high risk of FGR and PE from the control group took acetylsalicylic acid (150 mg orally) [5, 7].

The complications of current pregnancy and somatic diseases in the examined patients are presented in Table 1. Table 1 shows statistically significant differences in the characteristics of the course of pregnancy between the groups.

 

Table 1. Features of the course of the present pregnancy and somatic diseases in the examined patients, abs. (%)

Indicator	Subgroup 1A (FGR with early manifestation, n=45)	Subgroup 1B (FGR with late manifestation, n=45)	Group 2, control (n = 27)	р
Somatic anamnesis
Chronic arterial hypertension	18 (40,0%)	8 (17,8%)	1 (3,7%)	р1А–1В–2=0,001* р1А–1В =0,04* р1А–2 =0,02*
Mitral valve prolapse stage I	4 (8,9%)	4 (8,9%)	1 (3,7%)	р1А–1В–2=0,395
Chronic pyelonephritis	9 (20,0%)	5 (11,1%)	–	р1А–1В–2 =0,040* р1А–2=0,039*
Myopia of varying degrees	31 (70,5%)	33 (73,3%)	4 (14,8%)	р1А–1В–2 <0,001* р1А–2 <0,001* p1В–2 <0,001*
Varicose veins of the lower extremities	9 (20%)	13 (28,9%)	1 (3,7%)	р1А–1В–2 =0,034* p1В–2 =0,027*
Complications of pregnancy
Moderate preeclampsia	23 (51,1)	14 (31,1)	2 (7,4)	р1А–1В–2 <0,001* p1А–2 <0,001* p1В–2 =0,038*
ARVI with fever	18 (40,9)	19 (42,2)	2 (7,4)	р1А–1В–2 =0,004* р1А–2 =0,005* p1В–2=0,005*
Chronic cystitis and exacerbation	9 (20,5%)	6 (13,3%)	2 (7,4%)	р1А–1В–2 =0,304
Anemia during pregnancy	24 (54,5)	27 (60,0)	3 (15,8)	р1А–1В–2 =0,004* р1А–2=0,009* p1В–2 =0,004*
Threatened miscarriage with the formation of a retrochorial hematoma in trimester I	8 (17,8%)	1 (2,2%)	–	р1А–1В–2 =0,005* p1А–1В =0,042* p1А–2 =0,042*
Preterm placental abruption	–	2 (4,4%)	–	р1А–1В–2 =0,196

* The differences are statistically significant (p <0.05); p_1A–1B–2 ― comparison of three groups: 1A, 1B and the 2nd group (control); p_1A–1B ― comparison of subgroups 1A and 1B; p_1A–2 ― comparison of subgroup 1A and control group 2; p_1B–2 ― comparison of subgroup 1B and control group 2.

 

Moderate preeclampsia and chronic arterial hypertension were significantly more common in the subgroup with early manifestation of FGR than in the subgroup with late manifestation and in the control group; the study results [8] confirm this conclusion.

Acute respiratory viral infection (ARVI) during pregnancy occurred in subgroups 1A and 1B with the same incidence, but statistically significantly more often in the main group than in the control group (p=0.004).

The threat of miscarriage with the formation of retrochorial hematoma in trimester I in patients in the subgroup with early manifestations of FGR was significantly more frequent (8/45; 17.8%). A study conducted in 2016 [9] revealed a correlation between an increased concentration of autoantibo dies to hCG and the formation of retrochorial hematomas in trimester I of pregnancy.

Anemia in pregnant women was diagnosed statistically significantly more often in the main group; when comparing the incidence in the subgroup with early and late manifestations of FGR, no statistically significant differences were revealed. Pregnant women with anemia have a high risk of obstetric complications, such as preterm delivery, FGR, and maternal postpartum infections, which indicates the need to pay attention to the diagnosis and treatment of anemia for its prevention [10]. Statistically significant differences were revealed when assessing myopia of varying degrees and varicose veins of the lower extremities; their incidence in the main group was significantly higher than that in the group with normal pregnancy, which may indicate connective tissue dysplasia.

Based on the abnormal levels of autoimmune antibodies in the blood serum in pregnant women of the main and control groups, statistically significant differences were obtained, namely, AABs to the hCG antigen (p <0.001), AABs to DNA (p <0.001), AABs to β₂-GP (p=0.035), AABs to collagen (p <0.001), AABs to Fc-IgG (p=0.001), AABs to insulin (p <0.001), AABs to S100 (p <0.001), AABs to Spr (p <0.001), AABs to TrM (p <0.001), AABs to ANCA (p <0.001), and AABs to KiMS (p <0.001) (Fig. 1).

 

Fig. 1. The frequency of detection of deviations in the level of autoantibodies depending on the groups of subjects and forms of fetal growth retardation, %.

 

However, when comparing the incidence of abnormalities in the level of autoantibodies in early and late manifestation of FGR, an isolated increase in AABs to the hCG antigen, AABs to TrM, AABs to ANCA, and AABs to KiMS was noted with early manifestation of FGR.

With the late manifestation of FGR, the isolated increase in the level of antibodies to DNA and insulin is statistically significantly higher. Abnormal levels of AABs in the S100 protein occur with the same frequency in both early and late manifestations of FGR.

We conducted binary logistic regression to study the probability of developing FGR with early and late manifestation depending on the level of autoimmune AABs determined by the ELI-P test.

We developed a prognostic model to determine the probability of early FGR manifestation depending on the level of AABs to hCG, S100 protein, ANCA, and KiMS using binary logistic regression.

The dependence observed is described by the following equation:

P = 1/(1+e^-z) × 100%

z = 0.605 – 0.092X_{AABs to hCG antigen} + + 0.147X_{AABs to S100} + 0.113X_{AABs to ANCA} + 0.192X_{AABs to KiMS},

where P is the probability of FGR with early manifestation, e is the base of natural logarithms and has a value of 2.71828182845904, and z is the standard regression equation.

The resulting regression model is statistically significant (p <0.001). Based on the value of the Nigelkirk coefficient of determination, the model explains 59.4% of the observed group variance (Table 2).

 

Table 2. Characteristics of the association between autoantibodies and the probability of detecting fetal growth retardation with early manifestation

Autoantibodies	COR (95% CI)	p	AOR (95% CI)	p
AAB to HCG antigen	0,957 (0,913–1,002)	0,060	0,912 (0,837–0,995)	0,037*
AAB to S100 protein	1,049 (0,994–1,107)	0,082	1,158 (1,035–1,296)	0,011*
AAB to ANCA	1,078 (1,016–1,145)	0,012*	1,119 (1,010–1,240)	0,032*
AAB to KiMS	1,107 (1,038–1,182)	0,002*	1,212 (1,075–1,366)	0,002*

*Differences between subgroup 1A with delayed fetal development with early manifestation and the control group are statistically significant (p <0.05); COR (crude odds ratio) unadjusted odds ratio; AOR (adjusted odds ratio) ― adjusted odds ratio; AT to ANCA ― cytoplasmic antigen of vascular endothelial cells; AT to KiMS ― membrane antigen of kidney glomerular cells.

 

The area under the ROC curve was 0.924±0.044 (95% CI, 0.838–1.000). The resulting model was statistically significant (p <0.001).

The threshold value of the logistic function P at the cut-off point, which corresponded to the highest value of the Youden index, was 0.405. FGR with early manifestation was predicted when the value of the logistic function P was higher than or equal to this value. The sensitivity and specificity of the model were 94.7 and 87.8%, respectively (Fig. 2).

 

Fig. 2. ROC curve characterizing the dependence of the probability of fetal growth retardation with early manifestation on a combination of autoantibodies.

 

A prognostic model was developed to determine the probability of developing FGR with late manifestation depending on autoantibodies to DNA, collagen, and S100 protein using binary logistic regression.

The noted dependence is described by the following equation:

P = 1/(1+e^-z) × 100%

z = 0.290 + 0.077X_{AAB to DNA} + 0.102X_{AAB to collagen} + + 0.072X_{AAB to S100},

where P is the probability of FGR with late manifestation, e is the base of natural logarithms and has a value of 2.71828182845904, and z is the standard regression equation.

The resulting regression model is statistically significant (p <0.001). Based on the value of the Nigelkirk coefficient of determination, the model explains 42.6% of the observed group variance (Table 3).

 

Table 3. Characteristics of the association of autoantibodies with the probability of detecting delayed fetal development with late manifestation

Autoantibodies	COR (95% CI)	p	AOR (95% CI)	p
AAB to DNA	1,079 (1,022–1,138)	0,006*	1,080 (1,013–1,153)	0,019*
AAB to collagen	1,114 (1,022–1,214)	0,013*	1,108 (1,003–1,224)	0,044*
AAB to S100	0,911 (0,856–0,970)	0,004*	0,931 (0,868–0,998)	0,045*

*Statistically significant differences between subgroup 1B and the control group (p <0.05); COR (crude odds ratio) ― unadjusted odds ratio; AOR (adjusted odds ratio) ― adjusted odds ratio; AT to DNA ― autoantibodies to DNA.

 

The area under the ROC curve was 0.866±0.057 (95% CI, 0.754–0.978). The resulting model was statistically significant (p <0.001).

The threshold value of the logistic function P at the cut-off point, which corresponded to the highest value of the Youden index, was 0.338. The late FGR was predicted when the logistic function P value was higher than or equal to this value. The model sensitivity and specificity were 100.0% and 81.6%, respectively (Fig. 3).

 

Fig. 3. ROC curve characterizing the dependence of the probability of fetal growth retardation with late manifestation on a combination of autoantibodies.

 

CONCLUSION

Thus, when studying the probability of identifying FGR with early and late manifestation based on the determination of autoimmune autoantibodies using an enzyme-linked immunosorbent assay using the ELI-P-Test panel, diagnostically significant combinations were identified. For early manifestation of FGR, the combination was AABs to hCG + AABs to S100 + AABs to ANCA + AABs to KiMS. For late manifestation of FGR, the combination AABs to DNA + AABs to collagen + AABs to protein S100 showed diagnostic significance.

The results obtained have practical significance; the implementation of these results in practical activities will allow for preventive measures, which will help reduce the incidence of adverse perinatal outcomes.

ADDITIONAL INFO

Author contribution. All authors made a substantial contribution to the conception of the work, acquisition, analysis, interpretation of data for the work, drafting and revising the work, final approval of the version to be published and agree to be accountable for all aspects of the work. The concept of the study ― I.V. Ignatko, D.I. Yakubova; data collection and processing of the material ― D.I. Yakubova, A.D. Megrabyan, T.M. Silaeva, I.M. Bogomazova; statistical processing of the material ― D.I. Yakubova; writing text ― D.I. Yakubova, I.V. Ignatko; editing ― D.I. Yakubova, I.M. Bogomazova.

Funding source. This study was not supported by any external sources of funding.

Competing interests. The authors declares that there are no obvious and potential conflicts of interest associated with the publication of this article.

About the authors

Diana I. Yakubova

I.M. Sechenov First Moscow State Medical University (Sechenov University)

Author for correspondence.
Email: diana_28_03_94@mail.ru
ORCID iD: 0000-0002-7561-0706
SPIN-code: 4541-8640

post-graduate student

Russian Federation, Moscow

Irina V. Ignatko

I.M. Sechenov First Moscow State Medical University (Sechenov University)

Email: ignatko_i_v@staff.sechenov.ru
ORCID iD: 0000-0002-9945-3848
SPIN-code: 8073-1817

MD, Dr. Sci. (Med.), Professor, Сorresponding member of the Russian Academy of Sciences, Head of the the Department

Russian Federation, Moscow

Aren D. Megrabyan

I.M. Sechenov First Moscow State Medical University (Sechenov University)

Email: arentek@mail.ru
ORCID iD: 0000-0002-8542-7630

post-graduate student

Russian Federation, Moscow

Tatyana M. Silaeva

I.M. Sechenov First Moscow State Medical University (Sechenov University)

Email: Tatyana.s1laeva@yandex.ru
ORCID iD: 0000-0003-3330-6367
SPIN-code: 8712-0487

Assistant Lecturer

Russian Federation, Moscow

Irina M. Bogomazova

I.M. Sechenov First Moscow State Medical University (Sechenov University)

Email: bogomazova_i_m@staff.sechenov.ru
ORCID iD: 0000-0003-1156-7726
SPIN-code: 9414-1218

MD, Cand. Sci. (Med.), Assistant Professor

Russian Federation, Moscow

References

Supplementary files