Development of machine learning models to predict gestational diabetes risk in the first half of pregnancy.

BACKGROUND: Early prediction of Gestational Diabetes Mellitus (GDM) risk is of particular importance as it may enable more efficacious interventions and reduce cumulative injury to mother and fetus. The aim of this study is to develop machine learning (ML) models, for the early prediction of GDM using widely available variables, facilitating early intervention, and making possible to apply the prediction models in places where there is no access to more complex examinations. METHODS: The dataset used in this study includes registries from 1,611 pregnancies. Twelve different ML models and their hyperparameters were optimized to achieve early and high prediction performance of GDM. A data augmentation method was used in training to improve prediction results. Three methods were used to select the most relevant variables for GDM prediction. After training, the models ranked with the highest Area under the Receiver Operating Characteristic Curve (AUCROC), were assessed on the validation set. Models with the best results were assessed in the test set as a measure of generalization performance. RESULTS: Our method allows identifying many possible models for various levels of sensitivity and specificity. Four models achieved a high sensitivity of 0.82, a specificity in the range 0.72-0.74, accuracy between 0.73-0.75, and AUCROC of 0.81. These models required between 7 and 12 input variables. Another possible choice could be a model with sensitivity of 0.89 that requires just 5 variables reaching an accuracy of 0.65, a specificity of 0.62, and AUCROC of 0.82. CONCLUSIONS: The principal findings of our study are: Early prediction of GDM within early stages of pregnancy using regular examinations/exams; the development and optimization of twelve different ML models and their hyperparameters to achieve the highest prediction performance; a novel data augmentation method is proposed to allow reaching excellent GDM prediction results with various models.

The Role of Long Non-Coding RNAs in Trophoblast Regulation in Preeclampsia and Intrauterine Growth Restriction.

Preeclampsia (PE) and Intrauterine Growth Restriction (IUGR) are two pregnancy-specific placental disorders with high maternal, fetal, and neonatal morbidity and mortality rates worldwide. The identification biomarkers involved in the dysregulation of PE and IUGR are fundamental for developing new strategies for early detection and management of these pregnancy pathologies. Several studies have demonstrated the importance of long non-coding RNAs (lncRNAs) as essential regulators of many biological processes in cells and tissues, and the placenta is not an exception. In this review, we summarize the importance of lncRNAs in the regulation of trophoblasts during the development of PE and IUGR, and other placental disorders.

Supplementation of Omega 3 during Pregnancy and the Risk of Preterm Birth: A Systematic Review and Meta-Analysis.

Preterm birth (PTB) is a major cause of neonatal death and long-term consequences for the newborn. This review aims to update the evidence about the potential benefit of pharmacological supplementation with omega 3 fatty acids during pregnancy on the incidence of PTB. The Medline, Embase, Cochrane Library and Central databases were searched until 28 June 2020 for RCTs in which omega 3 supplementation was used versus placebo to reduce PTB risk. Data from 37 trials were analyzed. We found an 11% reduction in PTB risk (RR(risk ratios), 0.89; 95% CI (confidence intervals), 0.82 to 0.97) in trials using omega 3 supplements versus placebo. Regarding early PTB (ePTB), there was a 27% reduction in the risk of ePTB (RR, 0.73; 95% CI, 0.58 to 0.92). However, after sensitivity analyses, there were no significant differences in PTB and ePTB risk (PTB RR, 0.92; 95% CI, 0.83 to 1.01, ePTB RR, 0.82; 95% CI, 0.61 to 1.09). We conclude that omega 3 supplementation during pregnancy does not reduce the risk of PTB and ePTB. More studies are required to determine the effect of omega 3 supplementations during pregnancy and the risk of detrimental fetal outcomes.

Routine screening for SARS CoV-2 in unselected pregnant women at delivery.

BACKGROUND: South America has become the epicenter of coronavirus pandemic. It seems that asymptomatic population may contribute importantly to the spread of the disease. Transmission from asymptomatic pregnant patients' needs to be characterized in larger population cohorts and symptom assessment needs to be standardized. OBJECTIVE: To assess the prevalence of SARS CoV-2 infection in an unselected obstetrical population and to describe their presentation and clinical evolution. METHODS: A cross-sectional study was designed. Medical records of pregnant women admitted at the Obstetrics & Gynecology department of Clínica Dávila for labor & delivery, between April 27th and June 7th, 2020 were reviewed. All patients were screened with RT-PCR for SARS CoV-2 at admission. After delivery, positive cases were inquired by the researchers for clinical symptoms presented before admission and clinical evolution. All neonates born from mothers with confirmed SARS CoV-2 were isolated and tested for SARS CoV-2 infection. RESULTS: A total of 586 patients were tested for SARS CoV-2 during the study period. Outcomes were obtained from 583 patients which were included in the study. Thirty-seven pregnant women had a positive test for SARS CoV-2 at admission. Cumulative prevalence of confirmed SARS CoV-2 infection was 6.35% (37/583) [CI 95%: 4.63-8.65]. From confirmed cases, 43.2% (16/37) were asymptomatic. From symptomatic patients 85.7% (18/21) had mild symptoms and evolved without complications and 14.3% (3/21) presented severe symptoms requiring admission to intensive care unit. Only 5.4% (2/37) of the neonates born to mothers with a positive test at admission had a positive RT-PCR for SARS CoV-2. CONCLUSION: In our study nearly half of pregnant patients with SARS CoV-2 were asymptomatic at the time of delivery. Universal screening, in endemic areas, is necessary for adequate patient isolation, prompt neonatal testing and targeted follow-up.

Increased Circulating Levels of Tissue-Type Plasminogen Activator Are Associated with the Risk of Spontaneous Abortion During the First Trimester of Pregnancy.

Spontaneous abortion is a common complication in early pregnancy, with an incidence of around 20%. Ultrasound scan and measurement of human chorionic gonadotropin are used to identify patients at risk of spontaneous abortion; however, there is a clinical need to find new biomarkers to prospectively identify patients before the onset of clinical symptoms. Here, we aim to investigate potential biomarkers of spontaneous abortion taken in the first clinical appointment of pregnancy. A case-control study was conducted based on a prospectively collected cohort in which cases and controls were retrospectively stratified based on pregnancy outcome: normal healthy pregnancies (controls = 33) and pregnancies that ended in spontaneous abortion (cases = 10). We evaluated extracellular vesicles isolated by precipitation with ExoQuick™ and protein concentrations of tissue plasminogen activator, leptin, and adiponectin measured by ELISA. The extracellular vesicles showed the typical morphology and membrane proteins: CD63, Alix, and Flotilin-1. The size distributions of the isolated extracellular vesicles were 112 ± 27 and 118 ± 28 nm in diameter for controls and spontaneous abortion, respectively, and the total amount of extracellular vesicles did not show any difference between controls and the spontaneous abortion group. The tissue plasminogen activator showed a significant difference (p = 0.0004) between both groups, although neither adiponectin nor leptin revealed significant changes, indicating that women who had spontaneous abortions have significantly higher levels of tissue plasminogen activator than women who had normal pregnancies.

Diagnostic Performance of First Trimester Screening of Preeclampsia Based on Uterine Artery Pulsatility Index and Maternal Risk Factors in Routine Clinical Use.

Preeclampsia is a pregnancy-specific disorder defined by new onset of hypertension and proteinuria after 20 weeks of gestation. The early detection of patients at risk of developing preeclampsia is crucial, however, predictive models are still controversial. We aim to evaluate the diagnostic performance of a predictive algorithm in the first trimester of pregnancy, in order to identify patients that will subsequently develop preeclampsia, and to study the effect of aspirin on reducing the rate of this complication in patients classified as high risk by this algorithm. A retrospective cohort including 1132 patients attending prenatal care at Clínica Dávila in Santiago, Chile, was conceived. The risk of developing preeclampsia (early and late onset) was calculated using algorithms previously described by Plasencia et al. Patients classified as high risk, in the first trimester of pregnancy, by these algorithms, were candidates to receive 100 mg/daily aspirin as prophylaxis at the discretion of the attending physician. The overall incidence of preeclampsia in this cohort was 3.5% (40/1132), and the model for early onset preeclampsia prediction detected 33% of patients with early onset preeclampsia. Among the 105 patients considered at high risk of developing preeclampsia, 56 received aspirin and 49 patients did not. Among those who received aspirin, 12% (7/56) developed preeclampsia, which is equal to the rate of preeclampsia (12% (6/49)) of those who did not receive this medication. Therefore, the diagnostic performance of an algorithm combining uterine artery Doppler and maternal factors in the first trimester predicted only one third of patients that developed preeclampsia. Among those considered at high risk for developing the disease using this algorithm, aspirin did not change the incidence of preeclampsia, however, this could be due either to the small study sample size or the type of the study, a retrospective, non-interventional cohort study.

Vesículas extracelulares como predictores tempranos de diabetes gestacional / Extracellular vesicle concentration in maternal plasma as an early marker of gestational diabetes

Background During pregnancy, there is an increase in the amount of extracellular vesicles, especially placental exosomes, in maternal plasma. Aim To isolate and characterize extracellular vesicles from blood during the three trimesters of pregnancy and to evaluate their capacity to identify patients at risk of developing gestational diabetes. Material and Methods A case-control study was conducted in a cohort of 50 pregnant women with plasma samples taken in each trimester. Six women who developed gestational diabetes were paired with three healthy controls per case (a total of 19). Clinical characteristics were recorded at first prenatal appointment, and blood samples were obtained during the first, second and third trimesters. Extracellular vesicles were isolated from plasma by the commercial kit, ExoQuick™. Nanoparticle tracking analysis, was used to characterize the obtained extracellular vesicles. Results The total concentration of extracellular particles isolated from maternal plasma increased along with gestational age. The size of the extracellular vesicles obtained in the first trimester of pregnancy was very similar between groups (144 ± 37 nm for controls and 143 ± 34 nm for patients with gestational diabetes mellitus). Moreover, the concentration of extracellular vesicles collected in the first trimester, was significantly higher in patients who developed gestational diabetes mellitus later in pregnancy compared to normoglycemic pregnant women (7.94 x 10 8 and 5.15 x 10 8 , p = 0.03). Conclusions Our results provide an insight into the potential capacity of first trimester plasma extracellular vesicles as early biomarkers for the prediction of gestational diabetes mellitus.

Oral extracellular vesicles in early pregnancy can identify patients at risk of developing gestational diabetes mellitus.

AIM: To isolate and characterize oral extracellular vesicles from gingival crevicular fluid at 11-14 weeks and evaluate their capacity to identify patients at risk of developing gestational diabetes mellitus. METHODS: A case-control study was conducted, including patients who developed gestational diabetes mellitus (n = 11) and healthy pregnant controls (n = 23). Obstetric and periodontal histories were recorded at 11-14 weeks of gestation, and samples of gingival crevicular fluid obtained. Extracellular vesicles were isolated from gingival crevicular fluid by ExoQuick. Nanoparticle tracking analysis, ELISA and transmission electron microscopy were used to characterize extracellular vesicles. RESULTS: Total extracellular vesicles isolated from gingival crevicular fluid were significantly higher in patients who developed gestational diabetes mellitus later in pregnancy compared to normoglycemic pregnant women (6.3x109 vs 1.7 x1010, p value = 0.0026), and the concentration of the extracellular vesicles delivered an area under the ROC curve of 0.81. The distribution size of extracellular vesicles obtained using ExoQuick was around 148 ± 57 nm. There were no significant differences in the periodontal status between cases and controls. The exosome transmembrane protein CD63 was also detected in the extracellular vesicles of gingival crevicular fluid. CONCLUSION: We were able to isolate extracellular vesicles from gingival crevicular fluid using a method that is suitable to be applied in a clinical setting. Our results provide an insight into the potential capacity of first trimester oral extracellular vesicles as early biomarkers for the prediction of gestational diabetes mellitus in pre-symptomatic women.

Ultrasound-Guided Extraction of Intrauterine Devices With Nonvisible Threads: 254 Consecutive Cases: An Effective, Noninvasive Technique.

Intrauterine devices (IUDs) are among the most efficient and widely used contraceptive methods available. Removal is recommended after a certain time or in some cases when adverse effects are observed. A considerable number of patients have nonvisible guides or "lost threads" on speculum examinations, hindering the extraction. In this article, we report a consecutive series of 254 nonpregnant patients referred to our center after 1 or more failed attempts at IUD removal. We describe a novel ultrasound-guided approach, using a laparoscopic forceps to safely and effectively remove IUDs in nonpregnant patients.

Patients with acute cervical insufficiency without intra-amniotic infection/inflammation treated with cerclage have a good prognosis.

Background The frequency of intra-amniotic infection/inflammation (IAI/I) in patients with midtrimester cervical insufficiency is up to 50%. Our purpose was to determine the perinatal outcomes of cervical cerclage in patients with acute cervical insufficiency with bulging membranes, and to compare the admission-to-delivery interval and pregnancy outcomes according to the results of amniotic fluid (AF) analysis and cerclage placement. Methods This was a retrospective cohort study including singleton pregnancies with cervical insufficiency between 15 and 26.9 weeks in two tertiary health centers. IAI/I was defined when at least one of the following criteria was present in AF: (a) a white blood cell (WBC) count >50 cells/mm3; (b) glucose concentration <14 mg/dL; and/or (c) a Gram stain positive for bacteria. Three different groups were compared: (1) absence of IAI/I with placement of a cerclage; (2) amniocentesis not performed with placement of a cerclage; and (3) IAI/I with or without a cerclage. Results Seventy patients underwent an amniocentesis to rule out IAI/I. The prevalence of IAI/I was 19%. Forty-seven patients underwent a cerclage. Patients with a cerclage had a longer median admission-to-delivery interval (33 vs. 2 days; P < 0.001) and delivered at a higher median gestational age (27.4 vs. 22.6 weeks; P = 0.001) than those without a cerclage. The neonatal survival rate in the cerclage group was 62% vs. 23% in those without a cerclage (P = 0.01). Patients without IAI/I who underwent a cerclage had a longer median admission-to-delivery interval (43 vs. 1 day; P < 0.001), delivered at a higher median gestational age (28 vs. 22.1 weeks; P = 0.001) and had a higher neonatal survival rate (67% vs. 8%; P < 0.001) than those with IAI/I. Conclusion The pregnancy outcomes of patients with midtrimester cervical insufficiency and bulging membranes are poor as they have a high prevalence of IAI/I. Therefore, a pre-operative amniocentesis is key to identify the best candidates for the subsequent placement of a cerclage.

First trimester prediction of gestational diabetes mellitus using plasma biomarkers: a case-control study.

Objectives To evaluate the first trimester maternal biomarkers for early pregnancy prediction of gestational diabetes mellitus (GDM). Methods The study was a case-control study of healthy women with singleton pregnancies at the first trimester carried out at the Obstetrics and Gynecology Unit, Clinica Davila, Santiago, Chile. After obtaining informed consent, peripheral blood samples of pregnant women under 14 weeks of gestation were collected. At 24-28 weeks of pregnancy, women were classified as GDM (n=16) or controls (n=80) based on the results of a 75-g oral glucose tolerance test (OGTT). In all women, we measured concentrations of fasting blood glucose, insulin, glycated hemoglobin, uric acid, cholesterol, high density lipoprotein (HDL), low density lipoprotein (LDL), very low density lipoprotein (VLDL), triglycerides, aspartate aminotransferase (AST), alanine aminotransferase (ALT), γ-glutamyl transpeptidase (GGT), alkaline phosphatase (AP), sex hormone-binding globulin (SHBG), adiponectin, tissue plasminogen activator (t-PA), leptin and placental growth factor (PGF). Results The GDM group displayed an increased median concentration of cholesterol (P=0.04), triglycerides (P=0.003), insulin (P=0.003), t-PA (P=0.0088) and homeostatic model assessment (HOMA) (P=0.003) and an increased mean concentration of LDL (P=0.009) when compared to the control group. The receiver operating characteristic (ROC) curve for significant variables achieved an area under the curve (AUC) of 0.870, a sensitivity of 81.4% and a specificity of 80.0%. The OGTT was positive for GDM according to the IADPSG (International Diabetes in Pregnancy Study Group) criteria. Conclusion Women who subsequently developed GDM showed higher levels of blood-borne biomarkers during the first trimester, compared to women who did not develop GDM. These data warrant validation in a larger cohort.

[Extracellular vesicle concentration in maternal plasma as an early marker of gestational diabetes]. / Vesículas extracelulares como predictores tempranos de diabetes gestacional.

Background During pregnancy, there is an increase in the amount of extracellular vesicles, especially placental exosomes, in maternal plasma. Aim To isolate and characterize extracellular vesicles from blood during the three trimesters of pregnancy and to evaluate their capacity to identify patients at risk of developing gestational diabetes. Material and Methods A case-control study was conducted in a cohort of 50 pregnant women with plasma samples taken in each trimester. Six women who developed gestational diabetes were paired with three healthy controls per case (a total of 19). Clinical characteristics were recorded at first prenatal appointment, and blood samples were obtained during the first, second and third trimesters. Extracellular vesicles were isolated from plasma by the commercial kit, ExoQuick™. Nanoparticle tracking analysis, was used to characterize the obtained extracellular vesicles. Results The total concentration of extracellular particles isolated from maternal plasma increased along with gestational age. The size of the extracellular vesicles obtained in the first trimester of pregnancy was very similar between groups (144 ± 37 nm for controls and 143 ± 34 nm for patients with gestational diabetes mellitus). Moreover, the concentration of extracellular vesicles collected in the first trimester, was significantly higher in patients who developed gestational diabetes mellitus later in pregnancy compared to normoglycemic pregnant women (7.94 x 10 8 and 5.15 x 10 8 , p = 0.03). Conclusions Our results provide an insight into the potential capacity of first trimester plasma extracellular vesicles as early biomarkers for the prediction of gestational diabetes mellitus.