Predicting gestational age using neonatal metabolic markers

Early Newborn Metabolic Patterning and Sudden Infant Death Syndrome

Importance

Sudden infant death syndrome (SIDS) is a major cause of infant death in the US. Previous research suggests that inborn errors of metabolism may contribute to SIDS, yet the relationship between SIDS and biomarkers of metabolism remains unclear.

Objective

To evaluate and model the association between routinely measured newborn metabolic markers and SIDS in combination with established risk factors for SIDS.

Design, Setting, and Participants

This was a case-control study nested within a retrospective cohort using data from the California Office of Statewide Health Planning and Development and the California Department of Public Health. The study population included infants born in California between 2005 and 2011 with full metabolic data collected as part of routine newborn screening (NBS). SIDS cases were matched to controls at a ratio of 1:4 by gestational age and birth weight z score. Matched data were split into training (2/3) and testing (1/3) subsets. Data were analyzed from January 2005 to December 2011.

Exposures

Metabolites measured by NBS and established risk factors for SIDS.

Main Outcomes and Measures

The primary outcome was SIDS. Logistic regression was used to evaluate the association between metabolic markers combined with known risk factors and SIDS.

Results

Of 2 276 578 eligible infants, 354 SIDS (0.016%) cases (mean [SD] gestational age, 38.3 [2.3] weeks; 220 male [62.1%]) and 1416 controls (mean [SD] gestational age, 38.3 [2.3] weeks; 723 male [51.1%]) were identified. In multivariable analysis, 14 NBS metabolites were significantly associated with SIDS in a univariate analysis: 17-hydroxyprogesterone, alanine, methionine, proline, tyrosine, valine, free carnitine, acetyl-L-carnitine, malonyl carnitine, glutarylcarnitine, lauroyl-L-carnitine, dodecenoylcarnitine, 3-hydroxytetradecanoylcarnitine, and linoleoylcarnitine. The area under the receiver operating characteristic curve for a 14-marker SIDS model, which included 8 metabolites, was 0.75 (95% CI, 0.72-0.79) in the training set and was 0.70 (95% CI, 0.65-0.76) in the test set. Of 32 infants in the test set with model-predicted probability greater than 0.5, a total of 20 (62.5%) had SIDS. These infants had 14.4 times the odds (95% CI, 6.0-34.5) of having SIDS compared with those with a model-predicted probability less than 0.1.

Conclusions and Relevance

Results from this case-control study showed an association between aberrant metabolic analytes at birth and SIDS. These findings suggest that we may be able to identify infants at increased risk for SIDS soon after birth, which could inform further mechanistic research and clinical efforts focused on monitoring and prevention.

Post-natal gestational age assessment using targeted metabolites of neonatal heel prick and umbilical cord blood: A GARBH-Ini cohort study from North India

Background

Accurate assessment of gestational age (GA) and identification of preterm birth (PTB) at delivery is essential to guide appropriate post-natal clinical care. Undoubtedly, dating ultrasound sonography (USG) is the gold standard to ascertain GA, but is not accessible to the majority of pregnant women in low- and middle-income countries (LMICs), particularly in rural areas and small secondary care hospitals. Conventional methods of post-natal GA assessment are not reliable at delivery and are further compounded by a lack of trained personnel to conduct them. We aimed to develop a population-specific GA model using integrated clinical and biochemical variables measured at delivery.

Methods

We acquired metabolic profiles on paired neonatal heel prick (nHP) and umbilical cord blood (uCB) dried blood spot (DBS) samples (n = 1278). The master data set consists of 31 predictors from nHP and 24 from uCB after feature selection. These selected predictors including biochemical analytes, birth weight, and placental weight were considered for the development of population-specific GA estimation and birth outcome classification models using eXtreme Gradient Boosting (XGBoost) algorithm.

Results

The nHP and uCB full model revealed root mean square error (RMSE) of 1.14 (95% confidence interval (CI) = 0.82-1.18) and of 1.26 (95% CI = 0.88-1.32) to estimate the GA as compared to actual GA, respectively. In addition, these models correctly estimated 87.9 to 92.5% of the infants within ±2 weeks of the actual GA. The classification models also performed as the best fit to discriminate the PTB from term birth (TB) infants with an area under curve (AUC) of 0.89 (95% CI = 0.84-0.94) for nHP and an AUC of 0.89 (95% CI = 0.85-0.95) for uCB.

Conclusion

The biochemical analytes along with clinical variables in the nHP and uCB data sets provide higher accuracy in predicting GA. These models also performed as the best fit to identify PTB infants at delivery.

Overview of the Global and US Burden of Preterm Birth

Preterm birth (PTB) is the leading cause of morbidity and mortality in children globally, yet its prevalence has been difficult to accurately estimate due to unreliable methods of gestational age dating, heterogeneity in counting, and insufficient data. The estimated global PTB rate in 2020 was 9.9% (95% confidence interval: 9.1, 11.2), which reflects no significant change from 2010, and 81% of prematurity-related deaths occurred in Africa and Asia. PTB prevalence in the United States in 2021 was 10.5%, yet with concerning racial disparities. Few effective solutions for prematurity prevention have been identified, highlighting the importance of further research.

Estimating Gestational Age and Prediction of Preterm Birth Using Metabolomics Biomarkers

Preterm birth (PTB) is a leading cause of morbidity and mortality in children aged under 5 years globally, especially in low-resource settings. It remains a challenge in many low-income and middle-income countries to accurately measure the true burden of PTB due to limited availability of accurate measures of gestational age (GA), first trimester ultrasound dating being the gold standard. Metabolomics biomarkers are a promising area of research that could provide tools for both early identification of high-risk pregnancies and for the estimation of GA and preterm status of newborns postnatally.

Accuracy of prenatal and postnatal biomarkers for estimating gestational age: a systematic review and meta-analysis

Background

Knowledge of gestational age (GA) is key in clinical management of individual obstetric patients, and critical to be able to calculate rates of preterm birth and small for GA at a population level. Currently, the gold standard for pregnancy dating is measurement of the fetal crown rump length at 11-14 weeks of gestation. However, this is not possible for women first presenting in later pregnancy, or in settings where routine ultrasound is not available. A reliable, cheap and easy to measure GA-dependent biomarker would provide an important breakthrough in estimating the age of pregnancy. Therefore, the aim of this study was to determine the accuracy of prenatal and postnatal biomarkers for estimating gestational age (GA).

Methods

Systematic review prospectively registered with PROSPERO (CRD42020167727) and reported in accordance with the PRISMA-DTA. Medline, Embase, CINAHL, LILACS, and other databases were searched from inception until September 2023 for cohort or cross-sectional studies that reported on the accuracy of prenatal and postnatal biomarkers for estimating GA. In addition, we searched Google Scholar and screened proceedings of relevant conferences and reference lists of identified studies and relevant reviews. There were no language or date restrictions. Pooled coefficients of correlation and root mean square error (RMSE, average deviation in weeks between the GA estimated by the biomarker and that estimated by the gold standard method) were calculated. The risk of bias in each included study was also assessed.

Findings

Thirty-nine studies fulfilled the inclusion criteria: 20 studies (2,050 women) assessed prenatal biomarkers (placental hormones, metabolomic profiles, proteomics, cell-free RNA transcripts, and exon-level gene expression), and 19 (1,738,652 newborns) assessed postnatal biomarkers (metabolomic profiles, DNA methylation profiles, and fetal haematological components). Among the prenatal biomarkers assessed, human chorionic gonadotrophin measured in maternal serum between 4 and 9 weeks of gestation showed the highest correlation with the reference standard GA, with a pooled coefficient of correlation of 0.88. Among the postnatal biomarkers assessed, metabolomic profiling from newborn blood spots provided the most accurate estimate of GA, with a pooled RMSE of 1.03 weeks across all GAs. It performed best for term infants with a slightly reduced accuracy for preterm or small for GA infants. The pooled RMSEs for metabolomic profiling and DNA methylation profile from cord blood samples were 1.57 and 1.60 weeks, respectively.

Interpretation

We identified no antenatal biomarkers that accurately predict GA over a wide window of pregnancy. Postnatally, metabolomic profiling from newborn blood spot provides an accurate estimate of GA, however, as this is known only after birth it is not useful to guide antenatal care. Further prenatal studies are needed to identify biomarkers that can be used in isolation, as part of a biomarker panel, or in combination with other clinical methods to narrow prediction intervals of GA estimation.

Funding

The research was funded by the Bill and Melinda Gates Foundation (INV-000368). ATP is supported by the Oxford Partnership Comprehensive Biomedical Research Centre with funding from the NIHR Biomedical Research Centre funding scheme. The views expressed are those of the authors and not necessarily those of the UK National Health Service, the NIHR, the Department of Health, or the Department of Biotechnology. The funders of this study had no role in study design, data collection, analysis or interpretation of the data, in writing the paper or the decision to submit for publication.

Associations between Smoking and Smoking Cessation during Pregnancy and Newborn Metabolite Concentrations: Findings from PRAMS and INSPIRE Birth Cohorts

Newborn metabolite perturbations may identify potential biomarkers or mechanisms underlying adverse, smoking-related childhood health outcomes. We assessed associations between third-trimester smoking and newborn metabolite concentrations using the Tennessee Pregnancy Risk Assessment Monitoring System (PRAMS, 2009-2019) as the discovery cohort and INSPIRE (2012-2014) as the replication cohort. Children were linked to newborn screening metabolic data (33 metabolites). Third-trimester smoking was ascertained from birth certificates (PRAMS) and questionnaires (INSPIRE). Among 8600 and 1918 mother-child dyads in PRAMS and INSPIRE cohorts, 14% and 13% of women reported third-trimester smoking, respectively. Third-trimester smoking was associated with higher median concentrations of free carnitine (C0), glycine (GLY), and leucine (LEU) at birth (PRAMS: C0: adjusted fold change 1.11 [95% confidence interval (CI) 1.08, 1.14], GLY: 1.03 [95% CI 1.01, 1.04], LEU: 1.04 [95% CI 1.03, 1.06]; INSPIRE: C0: 1.08 [95% CI 1.02, 1.14], GLY: 1.05 [95% CI 1.01, 1.09], LEU: 1.05 [95% CI 1.01, 1.09]). Smoking cessation (vs. continued smoking) during pregnancy was associated with lower median metabolite concentrations, approaching levels observed in infants of non-smoking women. Findings suggest potential pathways underlying fetal metabolic programming due to in utero smoke exposure and a potential reversible relationship of cessation.

A Cohort Study of Gastric Fluid and Urine Metabolomics for the Prediction of Survival in Severe Prematurity

Predicting survival in very preterm infants is critical in clinical medicine and parent counseling. In this prospective cohort study involving 96 very preterm infants, we evaluated whether the metabolomic analysis of gastric fluid and urine samples obtained shortly after birth could predict survival in the first 3 and 15 days of life (DOL), as well as overall survival up to hospital discharge. Gas chromatography-mass spectrometry (GC-MS) profiling was used. Uni- and multivariate statistical analyses were conducted to evaluate significant metabolites and their prognostic value. Differences in several metabolites were identified between survivors and non-survivors at the time points of the study. Binary logistic regression showed that certain metabolites in gastric fluid, including arabitol, and succinic, erythronic and threonic acids, were associated with 15 DOL and overall survival. Gastric glyceric acid was also associated with 15 DOL survival. Urine glyceric acid could predict survival in the first 3 DOL and overall survival. In conclusion, non-surviving preterm infants exhibited a different metabolic profile compared with survivors, demonstrating significant discrimination with the use of GC-MS-based gastric fluid and urine analyses. The results of this study support the usefulness of metabolomics in developing survival biomarkers in very preterm infants.

Association of Maternal Metabolites and Metabolite Networks with Newborn Outcomes in a Multi-Ancestry Cohort

The in utero environment is important for newborn size at birth, which is associated with childhood adiposity. We examined associations between maternal metabolite levels and newborn birthweight, sum of skinfolds (SSF), and cord C-peptide in a multinational and multi-ancestry cohort of 2337 mother-newborn dyads. Targeted and untargeted metabolomic assays were performed on fasting and 1 h maternal serum samples collected during an oral glucose tolerance test performed at 24-32 week gestation in women participating in the Hyperglycemia and Adverse Pregnancy Outcome (HAPO) Study. Anthropometric measurements were obtained on newborns at birth. Following adjustment for maternal BMI and glucose, per-metabolite analyses demonstrated significant associations between maternal metabolite levels and birthweight, SSF, and cord C-peptide. In the fasting state, triglycerides were positively associated and several long-chain acylcarnitines were inversely associated with birthweight and SSF. At 1 h, additional metabolites including branched-chain amino acids, proline, and alanine were positively associated with newborn outcomes. Network analyses demonstrated distinct clusters of inter-connected metabolites significantly associated with newborn phenotypes. In conclusion, numerous maternal metabolites during pregnancy are significantly associated with newborn birthweight, SSF, and cord C-peptide independent of maternal BMI and glucose, suggesting that metabolites in addition to glucose contribute to newborn size at birth and adiposity.

Development and external validation of machine learning algorithms for postnatal gestational age estimation using clinical data and metabolomic markers

BACKGROUND

Accurate estimates of gestational age (GA) at birth are important for preterm birth surveillance but can be challenging to obtain in low income countries. Our objective was to develop machine learning models to accurately estimate GA shortly after birth using clinical and metabolomic data.

METHODS

We derived three GA estimation models using ELASTIC NET multivariable linear regression using metabolomic markers from heel-prick blood samples and clinical data from a retrospective cohort of newborns from Ontario, Canada. We conducted internal model validation in an independent cohort of Ontario newborns, and external validation in heel prick and cord blood sample data collected from newborns from prospective birth cohorts in Lusaka, Zambia and Matlab, Bangladesh. Model performance was measured by comparing model-derived estimates of GA to reference estimates from early pregnancy ultrasound.

RESULTS

Samples were collected from 311 newborns from Zambia and 1176 from Bangladesh. The best-performing model accurately estimated GA within about 6 days of ultrasound estimates in both cohorts when applied to heel prick data (MAE 0.79 weeks (95% CI 0.69, 0.90) for Zambia; 0.81 weeks (0.75, 0.86) for Bangladesh), and within about 7 days when applied to cord blood data (1.02 weeks (0.90, 1.15) for Zambia; 0.95 weeks (0.90, 0.99) for Bangladesh).

CONCLUSIONS

Algorithms developed in Canada provided accurate estimates of GA when applied to external cohorts from Zambia and Bangladesh. Model performance was superior in heel prick data as compared to cord blood data.

Metabonomics profile analysis in inflammation-induced preterm birth and the potential role of metabolites in regulating premature cervical ripening

BACKGROUND

Preterm birth (PTB) is the primary cause of infant morbidity and mortality. Moreover, previous studies have established that PTB is related to premature cervical ripening. However, the underlying mechanism remains to be elucidated. This study sought to identify differentially expressed metabolites and investigate their potential biological functions in PTB.

METHODS

Pregnant C57BL/6 J mice were treated with either LPS or normal saline and cervical alterations before labor were detected by staining. Metabolic profiles in the plasma of PTB and control mice were examined through non-targeted metabonomics analyses, quantitative polymerase chain reaction and immunofluorescence staining were performed on human cervical smooth cells.

RESULTS

The study demonstrated that the mRNA and protein levels of α-SMA, SM-22, and calponin in cervical smooth muscle cells of PTB mice were lower while OR was higher at both mRNA and protein levels compared to the CTL group. A total of 181 differentially expressed metabolites were analyzed, among them, 96 were upregulated, while 85 were downregulated in the PTB group. Differentially expressed metabolites may play a role in STAT3, RhoA, mTOR, TGF-β, and NK-κB signaling pathways. Furthermore, when treated with taurine, the levels of α-SMA and SM-22 in human cervical smooth muscle cells were elevated, whereas that of connexin-43 was decreased.

CONCLUSION

Our study highlighted the changes of metabolites in the peripheral blood changed prior to PTB and revealed that these differentially expressed metabolites might participate in the development of premature cervical ripening. Taurine was identified as an important metabolite may modulate human cervical smooth muscle cells. Our study provided new insights into the mechanism underlying premature cervical ripening in PTB.

Machine learning prediction of gestational age from metabolic screening markers resistant to ambient temperature transportation: Facilitating use of this technology in low resource settings of South Asia and East Africa

Background

Knowledge of gestational age is critical for guiding preterm neonatal care. In the last decade, metabolic gestational dating approaches emerged in response to a global health need; because in most of the developing world, accurate antenatal gestational age estimates are not feasible. These methods initially developed in North America have now been externally validated in two studies in developing countries, however, require shipment of samples at sub-zero temperature.

Methods

A subset of 330 pairs of heel prick dried blood spot samples were shipped on dry ice and in ambient temperature from field sites in Tanzania, Bangladesh and Pakistan to laboratory in Iowa (USA). We evaluated impact on recovery of analytes of shipment temperature, developed and evaluated models for predicting gestational age using a limited set of metabolic screening analytes after excluding 17 analytes that were impacted by shipment conditions of a total of 44 analytes.

Results

With the machine learning model using all the analytes, samples shipped in dry ice yielded a Root Mean Square Error (RMSE) of 1.19 weeks compared to 1.58 weeks for samples shipped in ambient temperature. Out of the 44 screening analytes, recovery of 17 analytes was significantly different between the two shipment methods and these were excluded from further machine learning model development. The final model, restricted to stable analytes provided a RMSE of 1.24 (95% confidence interval (CI) = 1.10-1.37) weeks for samples shipped on dry ice and RMSE of 1.28 (95% CI = 1.15-1.39) for samples shipped at ambient temperature. Analysis for discriminating preterm births (gestational age <37 weeks), yielded an area under curve (AUC) of 0.76 (95% CI = 0.71-0.81) for samples shipped on dry ice and AUC of 0.73 (95% CI = 0.67-0.78) for samples shipped in ambient temperature.

Conclusions

In this study, we demonstrate that machine learning algorithms developed using a sub-set of newborn screening analytes which are not sensitive to shipment at ambient temperature, can accurately provide estimates of gestational age comparable to those from published regression models from North America using all analytes. If validated in larger samples especially with more newborns <34 weeks, this technology could substantially facilitate implementation in LMICs.

Targeted newborn metabolomics: prediction of gestational age from cord blood

OBJECTIVE

Our study sought to determine whether metabolites from a retrospective collection of banked cord blood specimens could accurately estimate gestational age and to validate these findings in cord blood samples from Busia, Uganda.

STUDY DESIGN

Forty-seven metabolites were measured by tandem mass spectrometry or enzymatic assays from 942 banked cord blood samples. Multiple linear regression was performed, and the best model was used to predict gestational age, in weeks, for 150 newborns from Busia, Uganda.

RESULTS

The model including metabolites and birthweight, predicted the gestational ages within 2 weeks for 76.7% of the Ugandan cohort. Importantly, this model estimated the prevalence of preterm birth <34 weeks closer to the actual prevalence (4.67% and 4.00%, respectively) than a model with only birthweight which overestimates the prevalence by 283%.

CONCLUSION

Models that include cord blood metabolites and birth weight appear to offer improvement in gestational age estimation over birth weight alone.

Gestational Age, Newborn Metabolic Markers and Academic Achievement

BACKGROUND

Gestational age is associated with greater school achievement and variation in newborn metabolic markers. Whether metabolic markers are related to gestational age differences in achievement is unknown. This study examines whether newborn metabolic markers are associated with gestational age differences in performance on standardized school tests.

METHODS

This retrospective cohort study linked birth certificates of children born in Iowa between 2002 and 2010 to newborn screening records and school tests between 2009 and 2018. The analytical sample includes up to 229,679 children and 973,247 child-grade observations. Regression models estimate the associations between gestational age and 37 newborn metabolic markers with national percentile ranking (NPR) scores on math, reading comprehension, and science tests.

RESULTS

An additional gestational week is associated with 0.6 (95% CI: 0.6, 0.7), 0.5 (95% CI: 0.4, 0.5), and 0.4 (95% CI: 0.4, 0.5) higher NPRs on math, reading, and science, respectively. Compared to full term children (37-44 weeks), preterm children (32-36 weeks) have 2.2 (95% CI: -2.6, -1.8), 1.5 (95% CI: -1.9, -1.1), and 1.0 (95% CI: -1.4, -0.7) lower NPRs on math, reading comprehension, and science. Very preterm children (20-31 weeks) have 8.3 (95% CI: -9.4, -7.2), 5.2 (95% CI: -6.2, -4.0), and 4.7 (95% CI: -5.6, -3.8) lower NPRs than full term children on math, reading, and science. Metabolic markers are associated with 27%, 36%, and 45% of gestational age differences in math, reading, and science scores, respectively, and over half of the difference in test scores between preterm or very preterm and full term children.

CONCLUSIONS

Newborn metabolic markers are strongly related to gestational age differences in school test scores, suggesting that early metabolic differences are important markers of long-term child development.

Evaluation of heparinized syringes for measuring newborn metabolites in neonates with a central arterial line

Newborn metabolic screening is emerging as a novel method for predicting neonatal morbidity and mortality in neonates born very preterm (<32 weeks gestation). The purpose of our study was to determine if blood collected by an electrolyte-balanced dry lithium heparin syringe, as is routine for blood gas measurements, affects targeted metabolite and biomarker levels. Two blood samples (one collected with a heparinized syringe and the other with a non-heparinized syringe) were obtained at the same time from 20 infants with a central arterial line and tested for 49 metabolites and biomarkers using standard procedures for newborn screening. Overall, the median metabolite levels did not significantly differ by syringe type. However, there was wide variability, particularly for amino acids and immunoreactive trypsinogen, for individual paired samples and therefore, consideration should be given to sample collection when using these metabolites in prediction models of neonatal morbidity and mortality.

Immunoreactive Trypsinogen and Free Carnitine Changes on Newborn Screening after Birth in Patients Who Develop Type 1 Diabetes

Are free carnitine concentrations on newborn screening (NBS) 48–72 h after birth lower in patients who develop type 1 diabetes than in controls? A retrospective case-control study of patients with type 1 diabetes was conducted. NBS results of patients from a Sydney hospital were compared against matched controls from the same hospital (1:5). Multiple imputation was performed for estimating missing data (gestational age) using gender and birthweight. Conditional logistic regression was used to control for confounding and to generate parameter estimates (α = 0.05). The Hommel approach was used for post-hoc analyses. Results are reported as medians and interquartile ranges. A total of 159 patients were eligible (80 females). Antibodies were detectable in 86. Median age at diagnosis was 8 years. Free carnitine concentrations were lower in patients than controls (25.50 µmol/L;18.98–33.61 vs. 27.26; 21.22–34.86 respectively) (p = 0.018). Immunoreactive trypsinogen was higher in this group (20.24 µg/L;16.15–29–52 vs. 18.71; 13.96–26.92) (p = 0.045), which did not persist in the post-hoc analysis. Carnitine levels are lower and immunoreactive trypsinogen might be higher, within 2–3 days of birth and years before development of type 1 diabetes as compared to controls, although the differences were well within reference ranges and provide insight into the pathogenesis into neonatal onset of type 1 diabetes development rather than use as a diagnostic tool. Given trypsinogen’s use for evaluation of new-onset type 1 diabetes, larger studies are warranted.

Epidemiology of cytomegalovirus infection in pregnancy in Israel: Real-world data from a large healthcare organization

Congenital cytomegalovirus infection (cCMVi) is the leading cause of nonhereditary sensorineural hearing loss among newborns. Women newly acquiring cytomegalovirus infection (CMVi) during pregnancy have the highest risk of vertical transmission. This study aimed to describe the epidemiology of CMVi in pregnancy in a large healthcare database. A retrospective cohort study was performed using the Maccabi Healthcare Services database (Israel). Women aged 18-44 years old on July 1, 2013 with no record of pregnancy in the prior 6 months were followed through December 31, 2017 for first pregnancy occurrence. Pregnancy outcomes (live birth, spontaneous/therapeutic abortions, stillbirth, and uncertain outcomes) were captured. CMV test results were obtained to assess serostatus at the start of pregnancy (SoP) and primary CMV infection (CMVi) during pregnancy. Associations of demographic and reproductive factors with pCMVi were investigated (multivariable logistic regression). The study included 84 699 pregnant women (median age = 31 years; interquartile range = 28-35). Live birth, fetal loss, and uncertain pregnancy outcomes accounted for 76.8%, 18.2%, and 5.0%, respectively. The seroprevalence of CMV at the start of pregnancy in this cohort was 63.4% (95% confidence interval [CI]: 63.1-63.7). Among seronegative women with available test results (n = 10 657), CMVi incidence was 14.5 per 1000 (95% CI = 12.2-16.7). In multivariate logistic regression models adjusting for maternal age, CMVi was significantly associated with having one or more prior live births (odds ratio [OR]: 3.8 [95% CI: 2.6-5.4]) and having a child less than 6 years of age (OR: 4.3 [95%CI: 3.0-6.1]). One in three pregnant women in Israel is at risk for primary CMVi. This study demonstrates that real-world electronic healthcare data can be leveraged to support clinical management and development of interventions for congenital CMV by identifying women at high risk for CMVi during pregnancy.

Machine learning guided postnatal gestational age assessment using new-born screening metabolomic data in South Asia and sub-Saharan Africa

BACKGROUND

Babies born early and/or small for gestational age in Low and Middle-income countries (LMICs) contribute substantially to global neonatal and infant mortality. Tracking this metric is critical at a population level for informed policy, advocacy, resources allocation and program evaluation and at an individual level for targeted care. Early prenatal ultrasound examination is not available in these settings, gestational age (GA) is estimated using new-born assessment, last menstrual period (LMP) recalls and birth weight, which are unreliable. Algorithms in developed settings, using metabolic screen data, provided GA estimates within 1-2 weeks of ultrasonography-based GA. We sought to leverage machine learning algorithms to improve accuracy and applicability of this approach to LMICs settings.

METHODS

This study uses data from AMANHI-ACT, a prospective pregnancy cohorts in Asia and Africa where early pregnancy ultrasonography estimated GA and birth weight are available and metabolite screening data in a subset of 1318 new-borns were also available. We utilized this opportunity to develop machine learning (ML) algorithms. Random Forest Regressor was used where data was randomly split into model-building and model-testing dataset. Mean absolute error (MAE) and root mean square error (RMSE) were used to evaluate performance. Bootstrap procedures were used to estimate confidence intervals (CI) for RMSE and MAE. For pre-term birth identification ROC analysis with bootstrap and exact estimation of CI for area under curve (AUC) were performed.

RESULTS

Overall model estimated GA had MAE of 5.2 days (95% CI 4.6-6.8), which was similar to performance in SGA, MAE 5.3 days (95% CI 4.6-6.2). GA was correctly estimated to within 1 week for 85.21% (95% CI 72.31-94.65). For preterm birth classification, AUC in ROC analysis was 98.1% (95% CI 96.0-99.0; p < 0.001). This model performed better than Iowa regression, AUC Difference 14.4% (95% CI 5-23.7; p = 0.002).

CONCLUSIONS

Machine learning algorithms and models applied to metabolomic gestational age dating offer a ladder of opportunity for providing accurate population-level gestational age estimates in LMICs settings. These findings also point to an opportunity for investigation of region-specific models, more focused feasible analyte models, and broad untargeted metabolome investigation.

Using AMANHI-ACT cohorts for external validation of Iowa new-born metabolic profiles based models for postnatal gestational age estimation

BACKGROUND

Globally, 15 million infants are born preterm and another 23.2 million infants are born small for gestational age (SGA). Determining burden of preterm and SGA births, is essential for effective planning, modification of health policies and targeting interventions for reducing these outcomes for which accurate estimation of gestational age (GA) is crucial. Early pregnancy ultrasound measurements, last menstrual period and post-natal neonatal examinations have proven to be not feasible or inaccurate. Proposed algorithms for GA estimation in western populations, based on routine new-born screening, though promising, lack validation in developing country settings. We evaluated the hypothesis that models developed in USA, also predicted GA in cohorts of South Asia (575) and Sub-Saharan Africa (736) with same precision.

METHODS

Dried heel prick blood spots collected 24-72 hours after birth from 1311 new-borns, were analysed for standard metabolic screen. Regression algorithm based, GA estimates were computed from metabolic data and compared to first trimester ultrasound validated, GA estimates (gold standard).

RESULTS

Overall Algorithm (metabolites + birthweight) estimated GA to within an average deviation of 1.5 weeks. The estimated GA was within the gold standard estimate by 1 and 2 weeks for 70.5% and 90.1% new-borns respectively. Inclusion of birthweight in the metabolites model improved discriminatory ability of this method, and showed promise in identifying preterm births. Receiver operating characteristic (ROC) curve analysis estimated an area under curve of 0.86 (conservative bootstrap 95% confidence interval (CI) = 0.83 to 0.89); P < 0.001) and Youden Index of 0.58 (95% CI = 0.51 to 0.64) with a corresponding sensitivity of 80.7% and specificity of 77.6%.

CONCLUSION

Metabolic gestational age dating offers a novel means for accurate population-level gestational age estimates in LMIC settings and help preterm birth surveillance initiatives. Further research should focus on use of machine learning and newer analytic methods broader than conventional metabolic screen analytes, enabling incorporation of region-specific analytes and cord blood metabolic profiles models predicting gestational age accurately.

External validation of machine learning models including newborn metabolomic markers for postnatal gestational age estimation in East and South-East Asian infants

Background: Postnatal gestational age (GA) algorithms derived from newborn metabolic profiles have emerged as a novel method of acquiring population-level preterm birth estimates in low resource settings. To date, model development and validation have been carried out in North American settings. Validation outside of these settings is warranted.  

Methods: This was a retrospective database study using data from newborn screening programs in Canada, the Philippines and China. ELASTICNET machine learning models were developed to estimate GA in a cohort of infants from Canada using sex, birth weight and metabolomic markers from newborn heel prick blood samples. Final models were internally validated in an independent sample of Canadian infants, and externally validated in infant cohorts from the Philippines and China. 

Results: Cohorts included 39,666 infants from Canada, 82,909 from the Philippines and 4,448 from China.  For the full model including sex, birth weight and metabolomic markers, GA estimates were within ±5 days of ultrasound values in the Canadian internal validation (mean absolute error (MAE) 0.71, 95% CI: 0.71, 0.72), and within ±6 days of ultrasound GA in both the Filipino (0.90 (0.90, 0.91)) and Chinese cohorts (0.89 (0.86, 0.92)). Despite the decreased accuracy in external settings, our models incorporating metabolomic markers performed better than the baseline model, which relied on sex and birth weight alone. In preterm and growth-restricted infants, the accuracy of metabolomic models was markedly higher than the baseline model.

Conclusions: Accuracy of metabolic GA algorithms was attenuated when applied in external settings.  Models including metabolomic markers demonstrated higher accuracy than models using sex and birth weight alone. As innovators look to take this work to scale, further investigation of modeling and data normalization techniques will be needed to improve robustness and generalizability of metabolomic GA estimates in low resource settings, where this could have the most clinical utility

Gestational age-dependent development of the neonatal metabolome

BACKGROUND

Prematurity is a severe pathophysiological condition, however, little is known about the gestational age-dependent development of the neonatal metabolome.

METHODS

Using an untargeted liquid chromatography-tandem mass spectrometry metabolomics protocol, we measured over 9000 metabolites in 298 neonatal residual heel prick dried blood spots retrieved from the Danish Neonatal Screening Biobank. By combining multiple state-of-the-art metabolome mining tools, we retrieved chemical structural information at a broad level for over 5000 (60%) metabolites and assessed their relation to gestational age.

RESULTS

A total of 1459 (~16%) metabolites were significantly correlated with gestational age (false discovery rate-adjusted P < 0.05), whereas 83 metabolites explained on average 48% of the variance in gestational age. Using a custom algorithm based on hypergeometric testing, we identified compound classes (617 metabolites) overrepresented with metabolites correlating with gestational age (P < 0.05). Metabolites significantly related to gestational age included bile acids, carnitines, polyamines, amino acid-derived compounds, nucleotides, phosphatidylcholines and dipeptides, as well as treatment-related metabolites, such as antibiotics and caffeine.

CONCLUSIONS

Our findings elucidate the gestational age-dependent development of the neonatal blood metabolome and suggest that the application of metabolomics tools has great potential to reveal novel biochemical underpinnings of disease and improve our understanding of complex pathophysiological mechanisms underlying prematurity-associated disorders.

IMPACT

A large variation in the neonatal dried blood spot metabolome from residual heel pricks stored at the Danish Neonatal Screening Biobank can be explained by gestational age. While previous studies have assessed the relation of selected metabolic markers to gestational age, this study assesses metabolome-wide changes related to prematurity. Using a combination of recently developed metabolome mining tools, we assess the relation of over 9000 metabolic features to gestational age. The ability to assess metabolome-wide changes related to prematurity in neonates could pave the way to finding novel biochemical underpinnings of health complications related to preterm birth.

Newborn metabolic vulnerability profile identifies preterm infants at risk for mortality and morbidity

BACKGROUND

Identifying preterm infants at risk for mortality or major morbidity traditionally relies on gestational age, birth weight, and other clinical characteristics that offer underwhelming utility. We sought to determine whether a newborn metabolic vulnerability profile at birth can be used to evaluate risk for neonatal mortality and major morbidity in preterm infants.

METHODS

This was a population-based retrospective cohort study of preterm infants born between 2005 and 2011 in California. We created a newborn metabolic vulnerability profile wherein maternal/infant characteristics along with routine newborn screening metabolites were evaluated for their association with neonatal mortality or major morbidity.

RESULTS

Nine thousand six hundred and thirty-nine (9.2%) preterm infants experienced mortality or at least one complication. Six characteristics and 19 metabolites were included in the final metabolic vulnerability model. The model demonstrated exceptional performance for the composite outcome of mortality or any major morbidity (AUC 0.923 (95% CI: 0.917-0.929). Performance was maintained across mortality and morbidity subgroups (AUCs 0.893-0.979).

CONCLUSIONS

Metabolites measured as part of routine newborn screening can be used to create a metabolic vulnerability profile. These findings lay the foundation for targeted clinical monitoring and further investigation of biological pathways that may increase the risk of neonatal death or major complications in infants born preterm.

IMPACT

We built a newborn metabolic vulnerability profile that could identify preterm infants at risk for major morbidity and mortality. Identifying high-risk infants by this method is novel to the field and outperforms models currently in use that rely primarily on infant characteristics. Utilizing the newborn metabolic vulnerability profile for precision clinical monitoring and targeted investigation of etiologic pathways could lead to reductions in the incidence and severity of major morbidities associated with preterm birth.

Designing a bed-side system for predicting length of stay in a neonatal intensive care unit

Increased length of stay (LOS) in intensive care units is directly associated with the financial burden, anxiety, and increased mortality risks. In the current study, we have incorporated the association of day-to-day nutrition and medication data of the patient during its stay in hospital with its predicted LOS. To demonstrate the same, we developed a model to predict the LOS using risk factors (a) perinatal and antenatal details, (b) deviation of nutrition and medication dosage from guidelines, and (c) clinical diagnoses encountered during NICU stay. Data of 836 patient records (12 months) from two NICU sites were used and validated on 211 patient records (4 months). A bedside user interface integrated with EMR has been designed to display the model performance results on the validation dataset. The study shows that each gestation age group of patients has unique and independent risk factors associated with the LOS. The gestation is a significant risk factor for neonates < 34 weeks, nutrition deviation for < 32 weeks, and clinical diagnosis (sepsis) for ≥ 32 weeks. Patients on medications had considerable extra LOS for ≥ 32 weeks’ gestation. The presented LOS model is tailored for each patient, and deviations from the recommended nutrition and medication guidelines were significantly associated with the predicted LOS.

Investigating the Metabolic Model in Preterm Neonates by Tandem Mass Spectrometry: A Cohort Study

The changes of metabolite profiles in preterm birth have been demonstrated using newborn screening data. However, little is known about the holistic metabolic model in preterm neonates. The aim was to investigate the holistic metabolic model in preterm neonates. All metabolite values were obtained from a cohort data of routine newborn screening. A total of 261 758 newborns were recruited and randomly divided into a training subset and a testing subset. Using the training subset, 949 variates were considered to establish a logistic regression model for identifying preterm birth (<37 weeks) from term birth (≥37 weeks). Sventy-two variates (age at collection, TSH, 17α-OHP, proline, tyrosine, C16:1-OH, C18:2, and 65 ratios) entered into the final metabolic model for identifying preterm birth from term birth. Among the variates entering into the final model of PTB [Leucine+Isoleucine+Proline-OH)/Valine (OR=38.36], (C3DC+C4-OH)/C12 (OR=15.58), Valine/C5 (OR=6.32), [Leucine+isoleucine+Proline-OH)/Ornithine (OR=2.509)], and Proline/C18:1 (OR=2.465) have the top five OR values, and [Leucine+Isoleucine+Proline-OH)/C5 (OR=0.05)], [Leucine+Isoleucine+Proline-OH)/Phenylalanine (OR=0.214)], proline/valine (OR=0.230), C16/C18 (OR=0.259), and Alanine/free carnitine (OR=0.279) have the five lowest OR values. The final metabolic model had a capacity of identifying preterm infants with >80% accuracy in both the training and testing subsets. When identifying neonates ≤32 weeks from those >32 weeks, it had a robust performance with nearly 95% accuracy in both subsets. In summary, we have established an excellent metabolic model in preterm neonates. These findings could provide new insights for more efficient nutrient supplements and etiology of preterm birth.

Cost-effectiveness of a gestational age metabolic algorithm for preterm and small-for-gestational-age classification

Background

Preterm birth complications are the leading cause of death among children under 5 years of age, and this imposes a heavy burden on healthcare and social systems, particularly in low- and middle-income countries where reliable estimates of gestational age may be difficult to obtain. Metabolic analyte data can aid in accurately estimating gestational age. However, important costs are associated with this approach, which are related to the collection and analysis of newborn samples, and its cost-effectiveness has yet to be determined.

Objective

This study aimed to evaluate the cost-effectiveness of an internationally validated gestational age estimation algorithm based on neonatal blood spot metabolite data in combination with clinical and demographic variables (birthweight, sex, and multiple birth status) compared with a basic algorithm that uses only clinical and demographic variables in classifying infants as preterm or term (using a 37-week dichotomous preterm or term classification) and determining gestational age.

Study Design

The cost per correctly classified preterm infant and per correctly classified small-for-gestational-age infant for the metabolic algorithm vs the basic algorithm were estimated with data from an implementation study in Bangladesh.

Results

Over 1 year, the metabolic algorithm correctly classified an average of 8.7 (95% confidence interval, 1.3–14.7) additional preterm infants and 145.3 (95% confidence interval, 128.0–164.7) additional small-for-gestational-age infants per 1323 infants screened compared with the basic algorithm using only clinical and demographic variables. The incremental annual cost of adopting the metabolic algorithm was $100,031 (95% confidence interval, $86,354–$115,725). If setup costs were included, the cost was $120,496 (95% confidence interval, $106,322–$136,656). Compared with the basic algorithm, the incremental cost per preterm infant correctly classified by the metabolic algorithm is $11,542 ($13,903 with setup), and the incremental cost per small-for-gestational-age infant is $688 ($829 with setup).

Conclusion

This research quantifies the cost per detection of preterm or small-for-gestational-age infant in the implementation of a newborn screening program to aid in improved classification of preterm and, in particular, small-for-gestational-age infants in low- and middle-income countries.

External validation of ELASTIC NET regression models including newborn metabolomic markers for postnatal gestational age estimation in East and South-East Asian infants

Background: Postnatal gestational age (GA) algorithms derived from newborn metabolic profiles have emerged as a novel method of acquiring population-level preterm birth estimates in low resource settings. To date, model development and validation have been carried out in North American settings. Validation outside of these settings is warranted.   Methods: This was a retrospective database study using data from newborn screening programs in Canada, the Philippines and China. ELASTICNET machine learning models were developed to estimate GA in a cohort of infants from Canada using sex, birth weight and metabolomic markers from newborn heel prick blood samples. Final models were internally validated in an independent group of infants, and externally validated in cohorts of infants from the Philippines and China.  Results: Cohorts included 39,666 infants from Canada, 82,909 from the Philippines and 4,448 from China.  For the full model including sex, birth weight and metabolomic markers, GA estimates were within 5 days of ultrasound values in the Canadian internal validation (mean absolute error (MAE) 0.71, 95% CI: 0.71, 0.72), and within 6 days of ultrasound GA in both the Filipino (0.90 (0.90, 0.91)) and Chinese cohorts (0.89 (0.86, 0.92)). Despite the decreased accuracy in external settings, our models incorporating metabolomic markers performed better than the baseline model, which relied on sex and birth weight alone. In preterm and growth-restricted infants, the accuracy of metabolomic models was markedly higher than the baseline model. Conclusions: Accuracy of metabolic GA algorithms was attenuated when applied in external settings.  Models including metabolomic markers demonstrated higher accuracy than models using sex and birth weight alone. As innovators look to take this work to scale, further investigation of modeling and data normalization techniques will be needed to improve robustness and generalizability of metabolomic GA estimates in low resource settings, where this could have the most clinical utility.

Newborn Screening Samples for Diabetes Research: An Underused Resource

Inborn errors of metabolism and diabetes share common derangements in analytes of metabolic networks that are tested for in newborn screening, usually performed 48-72 h after birth. There is limited research examining the metabolic imprint of diabetes on newborn screening results. This paper aims to demonstrate the links between diabetes, biochemical genetics and newborn screening in investigating disease pathophysiology in diabetes, provide possible reasons for the lack of research in diabetes in newborn screening and offer recommendations on potential research areas. We performed a systematic search of the available literature from 1 April 1998 to 31 December 2018 involving newborn screening and diabetes using OVID, MEDLINE, Cochrane and the PROSPERO register, utilizing a modified extraction tool adapted from Cochrane. Eight studies were included after screening 1312 records. Five studies reanalyzed dried blood spots (DBS) on filter paper cards, and three studies utilized pre-existing results. The results of these studies and how they relate to cord blood studies, the use of cord blood versus newborn screening dried blood spots as a sample and considerations on newborn screening and diabetes research is further discussed. The timing of sampling of newborn screening allows insight into neonatal physiology in a catabolic state with minimal maternal and placental influence. This, combined with the wide coverage of newborn screening worldwide, may aid in our understanding of the origins of diabetes.

Metabolic gestational age assessment in low resource settings: a validation protocol

Preterm birth is the leading global cause of neonatal morbidity and mortality. Reliable gestational age estimates are useful for quantifying population burdens of preterm birth and informing allocation of resources to address the problem. However, evaluating gestational age in low-resource settings can be challenging, particularly in places where access to ultrasound is limited. Our group has developed an algorithm using newborn screening analyte values derived from dried blood spots from newborns born in Ontario, Canada for estimating gestational age within one to two weeks. The primary objective of this study is to validate a program that derives gestational age estimates from dried blood spot samples (heel-prick or cord blood) collected from health and demographic surveillance sites and population representative health facilities in low-resource settings in Zambia, Kenya, Bangladesh and Zimbabwe. We will also pilot the use of an algorithm to identify birth percentiles based on gestational age estimates and weight to identify small for gestational age infants. Once collected from local sites, samples will be tested by the Newborn Screening Ontario laboratory at the Children’s Hospital of Eastern Ontario (CHEO) in Ottawa, Canada. Analyte values will be obtained through laboratory analysis for estimation of gestational age as well as screening for other diseases routinely conducted at Ontario’s newborn screening program. For select conditions, abnormal screening results will be reported back to the sites in real time to facilitate counseling and future clinical management. We will determine the accuracy of our existing algorithm for estimation of gestational age in these newborn samples. Results from this research hold the potential to create a feasible method to assess gestational age at birth in low- and middle-income countries where reliable estimation may be otherwise unavailable.

Trends of change in the individual contribution of risk factors for small for gestational age over more than 2 decades

PURPOSE

Over the past years, the prevalence of various risk factors for small for gestational age (SGA) neonates has changed. Little is known if there was also a change in the specific contribution of these risk factors to the prevalence of SGA. We aim to identify trends in the specific contribution of various risk factors for SGA by observing their odds ratios (ORs) throughout different time periods.

METHODS

A nested case-control study was conducted. The ORs for selected known risk factors for SGA occurring in three consecutive 8-year intervals between 1988 and 2014 (T1 - 1988-1996; T2 - 1997-2005; T3 - 2006-2014) were compared. Data were retrieved from the medical centre's computerized perinatal database. Multivariable logistic regression models were constructed and ORs were compared to identify the specific contribution of independent risk factors for SGA along the study period.

RESULTS

During the study period, 285,992 pregnancies met the study's inclusion criteria, of which 15,013 (5.25%) were SGA. Between 1988 and 2014, the incidence of SGA increased from 2.6% in 1988 to 2.9% in 2014. Using logistic regression models, nulliparity, maternal age, gestational age, hypertensive disorders of pregnancy, oligohydramnios and pre-gestational diabetes mellitus were found to be independently associated with SGA. While the adjusted ORs (aOR) of hypertensive disorders of pregnancy and pre-gestational diabetes mellitus had increased, aORs for nulliparity, maternal age and gestational age had remained stable over time. Oligohydramnios had demonstrated a mixed trend of change over the time.

CONCLUSION

In our study, the specific contribution of factors associated with SGA had changed over time. Having a better understating of the changes in the specific contribution of different risk factors for SGA may enable obstetricians to provide consultations.

[Application of metabolomics in neonatal clinical practice]

Metabolomics is an emerging and popular subject in the post-genome era, and a large number of studies have been noted on the application of metabolomics in health evaluation, growth and development evaluation, disease diagnosis, and therapeutic efficacy evaluation. As a special period of life, the neonatal period is characterized by rapid cell renewing, consumption of a lot of energy and materials, and changes in metabolic pathways, all of which affect the level of metabolites. However, there is still no reference standard for metabolic level and profile in neonates. This article reviews the current status of metabolic research on neonatal growth and development and common diseases and related clinical application of metabolomics, so as to provide new ideas for nutrition guidance and evaluation, selection of therapeutic regimens, and new drug research in neonates.

[Application of metabolomics in neonatal clinical practice]

Metabolomics is an emerging and popular subject in the post-genome era, and a large number of studies have been noted on the application of metabolomics in health evaluation, growth and development evaluation, disease diagnosis, and therapeutic efficacy evaluation. As a special period of life, the neonatal period is characterized by rapid cell renewing, consumption of a lot of energy and materials, and changes in metabolic pathways, all of which affect the level of metabolites. However, there is still no reference standard for metabolic level and profile in neonates. This article reviews the current status of metabolic research on neonatal growth and development and common diseases and related clinical application of metabolomics, so as to provide new ideas for nutrition guidance and evaluation, selection of therapeutic regimens, and new drug research in neonates.

Postnatal gestational age estimation via newborn screening analysis: application and potential

Introduction: Preterm birth is a major global health concern, contributing to 35% of all neonatal deaths in 2016. Given the importance of accurately ascertaining estimates of preterm birth and in light of current limitations in postnatal gestational age (GA) estimation, novel methods of estimating GA postnatally in the absence of prenatal ultrasound are needed. Previous work has demonstrated the potential for metabolomics to estimate GA by analyzing data captured through routine newborn screening. Areas covered: Circulating analytes found in newborn blood samples vary by GA. Leveraging newborn screening and demographic data, our group developed an algorithm capable of estimating GA postnatally to within approximately 1 week of ultrasound-validated GA. Since then, we have built on the model by including additional analytes and validating the model's performance through internal and external validation studies, and through implementation of the model internationally. Expert opinion: Currently, using metabolomics to estimate GA postnatally holds considerable promise but is limited by issues of cost-effectiveness and resource access in low-income settings. Future work will focus on enhancing the precision of this approach while prioritizing point-of-care testing that is both accessible and acceptable to individuals in low-resource settings.

Divergent Patterns of Thyrotropin and Other Thyroidal Parameters in Relationship with the Sex of Healthy Neonates and Infants Less than Two Years Old: A Longitudinal Study

Background: A longitudinal study was conducted in full-term healthy infants who were born between 2015 and 2017 in Athens, Greece, to elucidate the evolution of thyrotropin (TSH) and other thyroidal parameters according to sex, from their day of birth until two years old. Other thyroidal parameters that were taken into account include antithyroid peroxidase antibody (TPO-Ab) and antithyroglobulin antibody (TG-Ab), total triiodothyronine (T3), and free triiodothyronine (fT3), along with total thyroxine (T4) and free thyroxine (fT4). Methods: Blood samples were taken at 5-day intervals from the day of birth until the 31st day of life, and then every 5th month until 2 years of age. All thyroid parameters were measured by electrochemiluminescence immunoassays. The study took place at the Iaso General, Maternity and Gynecological Clinic in Athens, Greece. Results: The sample consisted of 2916 full-term healthy neonates/infants: 1507 (51.7%) boys and 1409 (48.3%) girls. There were no significant differences in TSH levels between boys and girls in all time periods from birth up to 2 years except between 11 and 15 months of age (p = 0.038). Mean TSH levels for boys exhibited much more fluctuation and variability than for girls. In boys we found a significant association between TSH levels and fT4 (p < 0.001), while we found a significant association between TSH levels and T3 in girls (p = 0.045). Furthermore, we found that mean TPO-Ab and TG-Ab levels for boys exhibited larger variability than those for girls. Conclusions: In this study, we were able to plot the development of TSH and other thyroidal parameters by sex from birth up to two years of age. In terms of clinical practice, our findings suggest the need for a re-evaluation of the reference ranges of the studied parameters according to sex, especially in the first months of life and until the first year. Furthermore, our results suggest new optimal ranges for thyroid hormone replacement for that specific period.

External validation of postnatal gestational age estimation using newborn metabolic profiles in Matlab, Bangladesh

This study sought to evaluate the performance of metabolic gestational age estimation models developed in Ontario, Canada in infants born in Bangladesh. Cord and heel prick blood spots were collected in Bangladesh and analyzed at a newborn screening facility in Ottawa, Canada. Algorithm-derived estimates of gestational age and preterm birth were compared to ultrasound-validated estimates. 1036 cord blood and 487 heel prick samples were collected from 1069 unique newborns. The majority of samples (93.2% of heel prick and 89.9% of cord blood) were collected from term infants. When applied to heel prick data, algorithms correctly estimated gestational age to within an average deviation of 1 week overall (root mean square error = 1.07 weeks). Metabolic gestational age estimation provides accurate population-level estimates of gestational age in this data set. Models were effective on data obtained from both heel prick and cord blood, the latter being a more feasible option in low-resource settings.

Newborn Metabolic Profile Associated with Hyperbilirubinemia With and Without Kernicterus

Our objective was to assess the relationship between hyperbilirubinemia with and without kernicterus and metabolic profile at newborn screening. Included were 1,693,658 infants divided into a training or testing subset in a ratio of 3:1. Forty‐two metabolites were analyzed using logistic regression (odds ratios (ORs), area under the receiver operating characteristic curve (AUC), 95% confidence intervals (CIs)). Several metabolite patterns remained consistent across gestational age groups for hyperbilirubinemia without kernicterus. Thyroid stimulating hormone (TSH) and C‐18:2 were decreased, whereas tyrosine and C‐3 were increased in infants across groupings. Increased C‐3 was also observed for kernicterus (OR: 3.17; 95% CI: 1.18–8.53). Thirty‐one metabolites were associated with hyperbilirubinemia without kernicterus in the training set. Phenylalanine (OR: 1.91; 95% CI: 1.85–1.97), ornithine (OR: 0.76; 95% 0.74–0.77), and isoleucine + leucine (OR: 0.63; 95% CI: 0.61–0.65) were the most strongly associated. This study showed that newborn metabolic function is associated with hyperbilirubinemia with and without kernicterus.

Combining newborn metabolic and DNA analysis for second-tier testing of methylmalonic acidemia

Purpose

Improved second-tier tools are needed to reduce false-positive outcomes in newborn screening (NBS) for inborn metabolic disorders on the Recommended Universal Screening Panel (RUSP).

Methods

We designed an assay for multiplex sequencing of 72 metabolic genes (RUSPseq) from newborn dried blood spots. Analytical and clinical performance was evaluated in 60 screen-positive newborns for methylmalonic acidemia (MMA) reported by the California Department of Public Health NBS program. Additionally, we trained a Random Forest machine learning classifier on NBS data to improve prediction of true and false-positive MMA cases.

Results

Of 28 MMA patients sequenced, we found two pathogenic or likely pathogenic (P/LP) variants in a MMA-related gene in 24 patients, and one pathogenic variant and a variant of unknown significance (VUS) in 1 patient. No such variant combinations were detected in MMA false positives and healthy controls. Random Forest–based analysis of the entire NBS metabolic profile correctly identified the MMA patients and reduced MMA false-positive cases by 51%. MMA screen-positive newborns were more likely of Hispanic ethnicity.

Conclusion

Our two-pronged approach reduced false positives by half and provided a reportable molecular finding for 89% of MMA patients. Challenges remain in newborn metabolic screening and DNA variant interpretation in diverse multiethnic populations.

Initial Metabolic Profiles Are Associated with 7-Day Survival among Infants Born at 22-25 Weeks of Gestation

OBJECTIVE

To evaluate the association between early metabolic profiles combined with infant characteristics and survival past 7 days of age in infants born at 22-25 weeks of gestation.

STUDY DESIGN

This nested case-control consisted of 465 singleton live births in California from 2005 to 2011 at 22-25 weeks of gestation. All infants had newborn metabolic screening data available. Data included linked birth certificate and mother and infant hospital discharge records. Mortality was derived from linked death certificates and death discharge information. Each death within 7 days was matched to 4 surviving controls by gestational age and birth weight z score category, leaving 93 cases and 372 controls. The association between explanatory variables and 7-day survival was modeled via stepwise logistic regression. Infant characteristics, 42 metabolites, and 12 metabolite ratios were considered for model inclusion. Model performance was assessed via area under the curve.

RESULTS

The final model included 1 characteristic and 11 metabolites. The model demonstrated a strong association between metabolic patterns and infant survival (area under the curve [AUC] 0.885, 95% CI 0.851-0.920). Furthermore, a model with just the selected metabolites performed better (AUC 0.879, 95% CI 0.841-0.916) than a model with multiple clinical characteristics (AUC 0.685, 95% CI 0.627-0.742).

CONCLUSIONS

Use of metabolomics significantly strengthens the association with 7-day survival in infants born extremely premature. Physicians may be able to use metabolic profiles at birth to refine mortality risks and inform postnatal counseling for infants born at <26 weeks of gestation.

Metabolic profiles derived from residual blood spot samples: A longitudinal analysis

Background: Secondary use of newborn screening dried blood spot samples include use for biomedical or epidemiological research. However, the effects of storage conditions on archival samples requires further examination. The objective of this study was to determine the utility of residual newborn samples for deriving reliable metabolic gestational age estimates. Methods: Residual newborn dried blood spot samples that had been stored for 2-, 4-, 6-, or 12-months in temperature controlled (21°C) conditions were re-analyzed for the full panel of newborn screening analytes offered by a provincial newborn screening lab in Ottawa, Canada. Data from re-analyzed samples were compared to corresponding baseline newborn screening values for absolute agreement, and Pearson and intraclass correlation. Performance of a gestational age estimation algorithm originally developed from baseline newborn screening values was then validated on data derived from stored samples. Results: A total of 307 samples were used for this study. 17-hydroxyprogesterone and newborn hemoglobin profiles measured by immunoassay and high-performance liquid chromatography, respectively, were among the most stable markers across all time points of analysis. Acylcarnitines exhibited the greatest degree of variation in stability upon repeat measurement. The largest shifts in newborn analyte profiles and the poorest performance of metabolic gestational age algorithms were observed when samples were analyzed 12-months after sample collection. Conclusions: Duration of sample storage, independent of temperature and humidity, affects newborn screening profiles and gestational age estimates derived from metabolic gestational dating algorithms. When considering use of dried blood spot samples either for clinical or research purposes, care should be taken when interpreting data stemming from secondary use.

Biofluid Metabolomics in Preterm Birth Research

This article presents an account of the research carried out so far in the use of metabolomics to find biomarkers of preterm birth (PTB) in fetal, maternal, and newborn biofluids. Metabolomic studies have employed mainly nuclear magnetic resonance spectroscopy or mass spectrometry-based methodologies to analyze, on one hand, prenatal biofluids (amniotic fluid, maternal urine/maternal blood, cervicovaginal fluid) to identify predictive biomarkers of PTB, and on the other hand, biofluids collected at or after birth (amniotic fluid, umbilical cord blood, newborn urine, and newborn blood, maternal blood, or breast milk) to assess and follow up the health status of PTB babies. Besides advancing on the biochemical knowledge of PTB metabolism mainly during the in utero period and at birth, the work carried out has also helped to identify important requirements related to experimental design and analytical protocol that need to be addressed, if translation of these biomarkers to the clinic is to be envisaged. An outlook of possible future developments for the translation of laboratory results to the clinic is presented.

Using newborn screening analytes to identify cases of neonatal sepsis

Neonatal sepsis is associated with high mortality and morbidity, yet challenges with available diagnostic approaches can lead to delays in therapy. Our study assessed whether newborn screening analytes could be utilized to identify associations with neonatal sepsis. We linked a newborn screening registry with health databases to identify cases of sepsis among infants born in Ontario from 2010-2015. Correlations between sepsis and screening analytes were examined within three gestational age groups (early preterm: <34 weeks; late preterm: 34-36 weeks; term: ≥37 weeks), using multivariable logistic regression models. We started with a model containing only clinical factors, then added groups of screening analytes. Among 793,128 infants, 4,794 were diagnosed with sepsis during the neonatal period. Clinical variables alone or in combination with hemoglobin values were not strongly predictive of neonatal sepsis among infants born at term or late preterm. However, model fit improved considerably after adding markers of thyroid and adrenal function, acyl-carnitines, and amino acids. Among infants born at early preterm gestation, neither clinical variables alone nor models incorporating screening analytes adequately predicted neonatal sepsis. The combination of clinical variables and newborn screening analytes may have utility in identifying term or late preterm infants at risk for neonatal sepsis.

Performance of a postnatal metabolic gestational age algorithm: a retrospective validation study among ethnic subgroups in Canada

OBJECTIVES

Biological modelling of routinely collected newborn screening data has emerged as a novel method for deriving postnatal gestational age estimates. Validation of published models has previously been limited to cohorts largely consisting of infants of white Caucasian ethnicity. In this study, we sought to determine the validity of a published gestational age estimation algorithm among recent immigrants to Canada, where maternal landed immigrant status was used as a surrogate measure of infant ethnicity.

DESIGN

We conducted a retrospective validation study in infants born in Ontario between April 2009 and September 2011.

SETTING

Provincial data from Ontario, Canada were obtained from the Institute for Clinical Evaluative Sciences.

PARTICIPANTS

The dataset included 230 034 infants born to non-landed immigrants and 70 098 infants born to immigrant mothers. The five most common countries of maternal origin were India (n=10 038), China (n=7468), Pakistan (n=5824), The Philippines (n=5441) and Vietnam (n=1408). Maternal country of origin was obtained from Citizenship and Immigration Canada's Landed Immigrant Database.

PRIMARY AND SECONDARY OUTCOME MEASURES

Performance of a postnatal gestational age algorithm was evaluated across non-immigrant and immigrant populations.

RESULTS

Root mean squared error (RMSE) of 1.05 weeks was observed for infants born to non-immigrant mothers, whereas RMSE ranged from 0.98 to 1.15 weeks among infants born to immigrant mothers. Area under the receiver operating characteristic curve for distinguishing term versus preterm infants (≥37 vs <37 weeks gestational age or >34 vs ≤34 weeks gestational age) was 0.958 and 0.986, respectively, in the non-immigrant subgroup and ranged from 0.927 to 0.964 and 0.966 to 0.99 in the immigrant subgroups.

CONCLUSIONS

Algorithms for postnatal determination of gestational age may be further refined by development and validation of region or ethnicity-specific models. However, our results provide reassurance that an algorithm developed from Ontario-born infant cohorts performs well across a range of ethnicities and maternal countries of origin without modification.

Postnatal gestational age estimation using newborn screening blood spots: a proposed validation protocol

Background

Knowledge of gestational age (GA) is critical for guiding neonatal care and quantifying regional burdens of preterm birth. In settings where access to ultrasound dating is limited, postnatal estimates are frequently used despite the issues of accuracy associated with postnatal approaches. Newborn metabolic profiles are known to vary by severity of preterm birth. Recent work by our group and others has highlighted the accuracy of postnatal GA estimation algorithms derived from routinely collected newborn screening profiles. This protocol outlines the validation of a GA model originally developed in a North American cohort among international newborn cohorts.

Methods

Our primary objective is to use blood spot samples collected from infants born in Zambia and Bangladesh to evaluate our algorithm’s capacity to correctly classify GA within 1, 2, 3 and 4 weeks. Secondary objectives are to 1) determine the algorithm's accuracy in small-for-gestational-age and large-for-gestational-age infants, 2) determine its ability to correctly discriminate GA of newborns across dichotomous thresholds of preterm birth (≤34 weeks, <37 weeks GA) and 3) compare the relative performance of algorithms derived from newborn screening panels including all available analytes and those restricted to analyte subsets. The study population will consist of infants born to mothers already enrolled in one of two preterm birth cohorts in Lusaka, Zambia, and Matlab, Bangladesh. Dried blood spot samples will be collected and sent for analysis in Ontario, Canada, for model validation.

Discussion

This study will determine the validity of a GA estimation algorithm across ethnically diverse infant populations and assess population specific variations in newborn metabolic profiles.

Acylcarnitine Profiles Reflect Metabolic Vulnerability for Necrotizing Enterocolitis in Newborns Born Premature

OBJECTIVE

To evaluate the association between newborn acylcarnitine profiles and the subsequent development of necrotizing enterocolitis (NEC) with the use of routinely collected newborn screening data in infants born preterm.

STUDY DESIGN

A retrospective cohort study was conducted with the use of discharge records for infants born preterm admitted to neonatal intensive care units in California from 2005 to 2009 who had linked state newborn screening results. A model-development cohort of 94 110 preterm births from 2005 to 2008 was used to develop a risk-stratification model that was then applied to a validation cohort of 22 992 births from 2009.

RESULTS

Fourteen acylcarnitine levels and acylcarnitine ratios were associated with increased risk of developing NEC. Each log unit increase in C5 and free carnitine /(C16 + 18:1) was associated with a 78% and a 76% increased risk for developing NEC, respectively (OR 1.78, 95% CI 1.53-2.02, and OR 1.76, 95% CI 1.51-2.06). Six acylcarnitine levels, along with birth weight and total parenteral nutrition, identified 89.8% of newborns with NEC in the model-development cohort (area under the curve 0.898, 95% CI 0.889-0.907) and 90.8% of the newborns with NEC in the validation cohort (area under the curve 0.908, 95% CI 0.901-0.930).

CONCLUSIONS

Abnormal fatty acid metabolism was associated with prematurity and the development of NEC. Metabolic profiling through newborn screening may serve as an objective biologic surrogate of risk for the development of disease and thus facilitate disease-prevention strategies.

Postnatal Prediction of Gestational Age Using Newborn Fetal Hemoglobin Levels

Introduction

In many parts of the developing world procurement of antenatal gestational age estimates is not possible, challenging provision of appropriate perinatal care. This study aimed to develop a model for postnatal gestational age estimation utilizing measures of the newborn hemoglobin levels and other metabolic analyte data derived from newborn blood spot samples.

Methods

We conducted a retrospective cohort analysis of 159,215 infants born January 2012–December 2014 in Ontario, Canada. Multivariable linear and logistic regression analyses were used to evaluate the precision of developed models.

Results

Models derived from a combination of hemoglobin ratios and birthweight were more precise at predicting gestational age (RMSE1·23 weeks) than models limited to birthweight (RMSE1·34). Models including birthweight, hemoglobin, TSH and 17-OHP levels were able to accurately estimate gestational age to ± 2 weeks in 95·3% of the cohort and discriminate ≤ 34 versus > 34 (c-statistic, 0·98). This model also performed well in small for gestational age infants (c-statistic, 0·998).

Discussion

The development of a point-of-care mechanism to allow widespread implementation of postnatal gestational age prediction tools that make use of hemoglobin or non-mass spectromietry-derived metabolites could serve areas where antenatal gestational age dating is not routinely available.

•

Mechanisms for postnatal gestational age estimation are required to guide care in low resource settings.

•

Newborn fetal/adult hemoglobin ratio and other non-mass spectrometry derived data can be used to provide precise estimates of gestational age.

•

Hemoglobin derived postnatal gestational age prediction models also performed comparatively well in small for gestational age infants.

Three research groups including our own have recently published on the development of postnatal gestational age prediction algorithms derived from newborn screening metabolic profiles. Expanded newborn screening practices relying on tandem mass spectrometry instrumentation are not common place in many low resource settings, thus limiting the utility of such prediction models. Newborn fetal and adult hemoglobin levels are known to vary by gestational age of birth, and may be derived by methods other than mass spectrometry. In this study we used a retrospective cohort study design to develop and validate the precision of postnatal gestational age prediction models derived from fetal and adult hemoglobin levels, and readily available perinatal characteristics obtained from the Better Outcomes Registry & Network and the Newborn Screening of Ontario program. Final models were able to accurately predict postnatal gestational age to within 2 weeks of true gestational age, with excellent precision to discriminate the gestational age of average and small for gestational age infants. We have built upon our existing postnatal gestational age prediction algorithm to demonstrate both the stand-alone and additive predictive potential of newborn hemoglobin levels to the model. Methods to predict gestational age based on newborn screening markers have the potential to provide accurate postnatal assessments of gestational age in settings where gold standard first trimester ultrasounds are limited.