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.

A Randomized Trial of Point-of-Care Early Infant HIV Diagnosis in Zambia

BACKGROUND

Point-of-care (POC) early infant diagnosis (EID) provides same-day results and the potential for immediate initiation of antiretroviral therapy (ART).

METHODS

We conducted a pragmatic trial at six public clinics in Zambia. HIV-exposed infants were individually randomized to either: (a) POC EID - on-site testing with the Alere q HIV-1/2 Detect or (b) enhanced standard of care (SOC) EID - off-site testing at a public laboratory. HIV-infected infants were referred for ART and followed for 12 months. Our primary outcome was defined as alive, in care, and virally suppressed at 12 months.

RESULTS

Between March 2016 and November 2018, we randomized 4,000 HIV-exposed infants to POC (n=1,989) or SOC (n=2,011). All but two infants in the POC group received same-day results, while the median time to result in the SOC group was 27 (IQR: 22-30) days. Eighty-one (2%, 95% CI: 1.6-2.5%) infants were diagnosed with HIV. Although ART initiation was high, there were 15 (19%) deaths, 15 (19%) follow-up losses, and 31 (38%) virologic failures. By 12 months, only 20 of 81 (25%, 95% CI: 15-34%) HIV-infected infants were alive, in care, and virally suppressed: 13 (30%, 95% CI: 16-43%) infants in the POC group vs. 7 (19%, 95% CI: 6-32%) in the SOC group (RR: 1.56, 95% CI: 0.7-3.50).

CONCLUSIONS

POC EID eliminated diagnostic delays and accelerated ART initiation but did not translate into definitive improvement in 12-month outcomes. In settings where centralized EID is well functioning, POC EID is unlikely to improve pediatric HIV outcomes.

Weekly 17 alpha-hydroxyprogesterone caproate to prevent preterm birth among women living with HIV: a randomised, double-blind, placebo-controlled trial

BACKGROUND

Women with HIV face an increased risk of preterm birth. 17 alpha-hydroxyprogesterone caproate (17P) has been shown in some trials to reduce early delivery among women with a history of spontaneous preterm birth. We investigated whether 17P would reduce this risk among women with HIV.

METHODS

We did a randomised, double-blind, placebo-controlled trial in pregnant women with HIV at the University Teaching Hospital and Kamwala District Health Centre in Lusaka, Zambia. Eligible patients were women aged 18 years or older with confirmed HIV-1 infection, viable intrauterine singleton pregnancy at less than 24 weeks of gestation, and were receiving or intending to commence antiretroviral therapy during pregnancy. Exclusion criteria were major uterine or fetal anomaly; planned or in situ cervical cerclage; evidence of threatened miscarriage, preterm labour, or ruptured membranes at screening; medical contraindication to 17P; previous participation in the trial; or history of spontaneous preterm birth. Eligible participants provided written informed consent and were randomly assigned (1:1) to receive 250 mg intramuscular 17P or placebo once per week, starting between 16 and 24 weeks of gestation until delivery, stillbirth, or reaching term (37 weeks). Participants and study staff were masked to assignment, except for pharmacy staff who did random assignment and prepared injections but did not interact with participants. The primary outcome was a composite of delivery before 37 weeks or stillbirth at any gestational age. Patients attended weekly visits for study drug injections and antenatal care. We estimated the absolute and relative difference in risk of the primary outcome and safety events between treatment groups by intention to treat. This trial is registered with ClinicalTrials.gov, NCT03297216, and is complete.

FINDINGS

Between Feb 7, 2018 and Jan 13, 2020, we assessed 1042 women for inclusion into the study. 242 women were excluded after additional assessments, and 800 eligible patients were enrolled and randomly assigned to receive intramuscular 17P (n=399) or placebo (n=401). Baseline characteristics were similar between groups. Adherence to study drug injections was 98% in both groups, no patients were lost to follow-up, and the final post-partum visit was on Aug 6, 2020. 36 (9%) of 399 participants assigned to 17P had preterm birth or stillbirth, compared with 36 (9%) of 401 patients assigned to placebo (risk difference 0·1, 95% CI -3·9 to 4·0; relative risk 1·0, 95% CI 0·6 to 1·6; p=0·98). Intervention-related adverse events were reported by 140 (18%) of 800 participants and occurred in similar proportions in both randomisation groups. No serious adverse events were reported.

INTERPRETATION

Although 17P seems to be safe and acceptable to participants, available data do not support the use of the drug to prevent preterm birth among women whose risk derives solely from HIV infection. The low risk of preterm birth in both randomisation groups warrants further investigation.

FUNDING

US National Institutes of Health and the Bill and Melinda Gates Foundation.

Association of mid-trimester maternal angiogenic biomarkers with small-for-gestational-age infants in an urban Zambian cohort: a nested case-control study

Objective

To investigate whether angiogenic biomarker concentrations differ between women who deliver small‐for‐gestational‐age (SGA) infants (<10th centile birth weight for gestational age) compared with controls, because identifying SGA risk early could improve outcomes.

Methods

This case‐control study compared serum concentrations of angiogenic biomarkers before 24 weeks of pregnancy from 62 women who delivered SGA infants (cases) and 62 control women from an urban Zambian cohort. Odds of delivering an SGA infant were calculated using conditional logistic regression.

Results

Placental growth factor (PlGF), soluble fms‐like tyrosine kinase (sFLT‐1) and soluble endoglin (sEng) in controls were 37.74 pg/mL (interquartile range [IQR] 23.12–63.15), 2525.18 pg/mL (IQR 1502.21–4265.54) and 2408.18 pg/mL (IQR 1854.87–3017.94), respectively. SGA cases had higher PlGF (40.50 pg/mL, IQR 22.81–67.94) and sFLT‐1 (2613.06 pg/mL, IQR 1720.58–3722.50), and lower sEng (2038.06 pg/mL, IQR 1445.25–3372.26). Participants with sEng concentration below and concomitant sFLT‐1 concentration above their respective thresholds (n = 40) had five‐fold higher odds of SGA (adjusted odds ratio 4.77, 95% confidence interval 1.61–14.1; P = 0.005).

Conclusion

Biomarker concentrations were similar between cases and controls. Participants with concomitant low sEng and high sFLT‐1 had the highest odds of SGA, suggesting that a combination of biomarkers may better for predicting SGA than single biomarkers.

Prediction of SGA risk using the interaction of multiple angiogenic biomarkers may perform better than using a single angiogenic biomarker concentration before 24 weeks gestation.

Circulating angiogenic factors and HIV among pregnant women in Zambia: a nested case-control study

BACKGROUND

Maternal HIV increases the risk of adverse birth outcomes including preterm birth, fetal growth restriction, and stillbirth, but the biological mechanism(s) underlying this increased risk are not well understood. We hypothesized that maternal HIV may lead to adverse birth outcomes through an imbalance in angiogenic factors involved in the vascular endothelial growth factor (VEGF) signaling pathway.

METHODS

In a case-control study nested within an ongoing cohort in Zambia, our primary outcomes were serum concentrations of VEGF-A, soluble endoglin (sEng), placental growth factor (PlGF), and soluble fms-like tyrosine kinase-1 (sFLT-1). These were measured in 57 women with HIV (cases) and 57 women without HIV (controls) before 16 gestational weeks. We used the Wilcoxon rank-sum and linear regression controlling for maternal body mass index (BMI) and parity to assess the difference in biomarker concentrations between cases and controls. We also used logistic regression to test for associations between biomarker concentration and adverse pregnancy outcomes (preeclampsia, preterm birth, small for gestational age, stillbirth, and a composite of preterm birth or stillbirth).

RESULTS

Compared to controls, women with HIV had significantly lower median concentrations of PlGF (7.6 vs 10.2 pg/mL, p = 0.02) and sFLT-1 (1647.9 vs 2055.6 pg/mL, p = 0.04), but these findings were not confirmed in adjusted analysis. PlGF concentration was lower among women who delivered preterm compared to those who delivered at term (6.7 vs 9.6 pg/mL, p = 0.03) and among those who experienced the composite adverse birth outcome (6.2 vs 9.8 pg/mL, p = 0.02). Median sFLT-1 concentration was lower among participants with the composite outcome (1621.0 vs 1945.9 pg/mL, p = 0.04), but the association was not significant in adjusted analysis. sEng was not associated with either adverse birth outcomes or HIV. VEGF-A was undetectable by Luminex in all specimens.

CONCLUSIONS

We present preliminary findings that HIV is associated with a shift in the VEGF signaling pathway in early pregnancy, although adjusted analyses were inconclusive. We confirm an association between angiogenic biomarkers and adverse birth outcomes in our population. Larger studies are needed to further elucidate the role of HIV on placental angiogenesis and adverse birth outcomes.

Maternal HIV, antiretroviral timing, and spontaneous preterm birth in an urban Zambian cohort: the role of local and systemic inflammation

OBJECTIVE

To assess plasma and vaginal inflammation in three antenatal groups (HIV-uninfected women, HIV-infected women entering care on preconceptional ART, and HIV-infected women not on preconceptional ART) and whether these measures are associated with spontaneous preterm birth (sPTB).

DESIGN

Case--control study nested within a pregnancy cohort in Lusaka, Zambia.

METHODS

We analyzed 11 pro-inflammatory and two anti-inflammatory markers in 207 women with paired plasma and vaginal specimens collected between 16 and 20 gestational weeks. Among 51 HIV-infected women, we repeated the assays in 24-34-week samples. We used confirmatory factor analysis to create inflammation scores and compared them among the three groups.

RESULTS

At baseline, HIV-infected women not on ART had higher vaginal pro-inflammatory scores than HIV-uninfected women [mean 0.37 (95% CI -0.06 to 0.80) vs. -0.02 (-0.32 to 0.27), P = 0.02]. In repeat testing, women not on preconceptional ART had an increase in vaginal inflammation between the baseline and 24-34-week visits compared with those continuing preconceptional ART [mean 0.62 (95% CI -0.80 to 4.20) vs. -0.07 (-2.78 to 2.11), P = 0.04]. In multivariate analyses, baseline vaginal inflammation predicted sPTB (aOR 1.5; 95% CI 1.0-2.3; P = 0.02). Plasma inflammation did not differ by HIV or ART exposure and was not associated with sPTB.

CONCLUSION

Women not receiving ART at entry into pregnancy care had more vaginal inflammation than women entering on treatment. They also experienced an increase in vaginal inflammation between the two sampling timepoints, possibly as a consequence of ART initiation. Vaginal (but not systemic) inflammation was associated with sPTB and offers a potential mechanistic insight into this important adverse birth outcome.

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.

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.

Adverse birth outcomes and their clinical phenotypes in an urban Zambian cohort

Background: Few cohort studies of pregnancy in sub-Saharan Africa use rigorous gestational age dating and clinical phenotyping. As a result, incidence and risk factors of adverse birth outcomes are inadequately characterized. Methods: The Zambian Preterm Birth Prevention Study (ZAPPS) is a prospective observational cohort established to investigate adverse birth outcomes at a referral hospital in urban Lusaka. This report describes ZAPPS phase I, enrolled August 2015 to September 2017. Women were followed through pregnancy and 42 days postpartum. At delivery, study staff assessed neonatal vital status, birthweight, and sex, and assigned a delivery phenotype. Primary outcomes were: (1) preterm birth (PTB; delivery <37 weeks), (2) small-for-gestational-age (SGA; <10 ^th percentile weight-for-age at birth), and (3) stillbirth (SB; delivery of an infant without signs of life). Results: ZAPPS phase I enrolled 1450 women with median age 27 years (IQR 23-32). Most participants (68%) were multiparous, of whom 41% reported a prior PTB and 14% reported a prior stillbirth. Twins were present in 3% of pregnancies, 3% of women had short cervix (<25mm), 24% of women were HIV seropositive, and 5% were syphilis seropositive. Of 1216 (84%) retained at delivery, 15% were preterm, 18% small-for-gestational-age, and 4% stillborn. PTB risk was higher with prior PTB (aRR 1.88; 95%CI 1.32-2.68), short cervix (aRR 2.62; 95%CI 1.68-4.09), twins (aRR 5.22; 95%CI 3.67-7.43), and antenatal hypertension (aRR 2.04; 95%CI 1.43-2.91). SGA risk was higher with twins (aRR 2.75; 95%CI 1.81-4.18) and antenatal hypertension (aRR 1.62; 95%CI 1.16-2.26). SB risk was higher with short cervix (aRR 6.42; 95%CI 2.56-16.1). Conclusio ns: This study confirms high rates of PTB, SGA, and SB among pregnant women in Lusaka, Zambia. Accurate gestational age dating and careful ascertainment of delivery data are critical to understanding the scope of adverse birth outcomes in low-resource settings.

Adverse birth outcomes and their clinical phenotypes in an urban Zambian cohort

Background: Few cohort studies of pregnancy in sub-Saharan Africa use rigorous gestational age dating and clinical phenotyping. As a result, incidence and risk factors of adverse birth outcomes are inadequately characterized. Methods: The Zambian Preterm Birth Prevention Study (ZAPPS) is a prospective observational cohort established to investigate adverse birth outcomes at a referral hospital in urban Lusaka. This report describes ZAPPS phase I, enrolled August 2015 to September 2017. Women were followed through pregnancy and 42 days postpartum. At delivery, study staff assessed neonatal vital status, birthweight, and sex, and assigned a delivery phenotype. Primary outcomes were: (1) preterm birth (PTB; delivery <37 weeks), (2) small-for-gestational-age (SGA; <10 ^th percentile weight-for-age at birth), and (3) stillbirth (SB; delivery of an infant without signs of life). Results: ZAPPS phase I enrolled 1450 women with median age 27 years (IQR 23-32). Most participants (68%) were multiparous, of whom 41% reported a prior PTB and 14% reported a prior stillbirth. Twins were present in 3% of pregnancies, 3% of women had short cervix (<25mm), 24% of women were HIV seropositive, and 5% were syphilis seropositive. Of 1216 (84%) retained at delivery, 15% were preterm, 18% small-for-gestational-age, and 4% stillborn. PTB risk was higher with prior PTB (aRR 1.88; 95%CI 1.32-2.68), short cervix (aRR 2.62; 95%CI 1.68-4.09), twins (aRR 5.22; 95%CI 3.67-7.43), and antenatal hypertension (aRR 2.04; 95%CI 1.43-2.91). SGA risk was higher with twins (aRR 2.75; 95%CI 1.81-4.18) and antenatal hypertension (aRR 1.62; 95%CI 1.16-2.26). SB risk was higher with short cervix (aRR 6.42; 95%CI 2.56-16.1). Conclusio ns: This study confirms high rates of PTB, SGA, and SB among pregnant women in Lusaka, Zambia. Accurate gestational age dating and careful ascertainment of delivery data are critical to understanding the scope of adverse birth outcomes in low-resource settings.

The Zambian Preterm Birth Prevention Study (ZAPPS): Cohort characteristics at enrollment

Background: Sub-Saharan Africa bears a disproportionate burden of preterm birth and other adverse outcomes. A better understanding of the demographic, clinical, and biologic underpinnings of these adverse outcomes is urgently needed to plan interventions and inform new discovery.  Methods: The Zambian Preterm Birth Prevention Study (ZAPPS) is a prospective observational cohort established at the Women and Newborn Hospital (WNH) in Lusaka, Zambia. We recruit pregnant women from district health centers and the WNH and offer ultrasound examination to determine eligibility. Participants receive routine obstetrical care, lab testing, midtrimester cervical length measurement, and serial fetal growth monitoring. At delivery, we assess gestational age, birthweight, vital status, and sex and assign a delivery phenotype. We collect blood, urine, and vaginal swab specimens at scheduled visits and store them in an on-site biorepository. In September 2017, enrollment of the ZAPPS Phase 1-the subject of this report-was completed. Phase 2, which is limited to HIV-uninfected women, reopened in January 2018.  Results: Between August 2015 and September 2017, we screened 1784 women, of whom 1450 (81.2%) met inclusion criteria and were enrolled. The median age at enrollment was 27 years (IQR 23-32) and median gestational age was 16 weeks (IQR 13-18). Among women with a previous pregnancy (n=1042), 19% (n=194) reported a prior miscarriage.  Among parous women (n=992), 41% (n=411) reported a prior preterm birth and 14% (n=126) reported a prior stillbirth. The HIV seroprevalence was 24%. Discussion: We have established a large cohort of pregnant women and newborns at the WNH to characterize the determinants of adverse birth outcomes in Lusaka, Zambia. Our overarching goal is to elucidate biological mechanisms in an effort to identify new strategies for early detection and prevention of adverse outcomes. We hope that findings from this cohort will help guide future studies, clinical care, and policy.

The Zambian Preterm Birth Prevention Study (ZAPPS): Cohort characteristics at enrollment

Background: Sub-Saharan Africa bears a disproportionate burden of preterm birth and other adverse outcomes. A better understanding of the demographic, clinical, and biologic underpinnings of these adverse outcomes is urgently needed to plan interventions and inform new discovery.  Methods: The Zambian Preterm Birth Prevention Study (ZAPPS) is a prospective observational cohort established at the Women and Newborn Hospital (WNH) in Lusaka, Zambia. We recruit pregnant women from district health centers and the WNH and offer ultrasound examination to determine eligibility. Participants receive routine obstetrical care, lab testing, midtrimester cervical length measurement, and serial fetal growth monitoring. At delivery, we assess gestational age, birthweight, vital status, and sex and assign a delivery phenotype. We collect blood, urine, and vaginal swab specimens at scheduled visits and store them in an on-site biorepository. In September 2017, enrollment of the ZAPPS Phase 1-the subject of this report-was completed. Phase 2, which is limited to HIV-uninfected women, reopened in January 2018.  Results: Between August 2015 and September 2017, we screened 1784 women, of whom 1450 (81.2%) met inclusion criteria and were enrolled. The median age at enrollment was 27 years (IQR 23-32) and median gestational age was 16 weeks (IQR 13-18). Among women with a previous pregnancy (n=1042), 19% (n=194) reported a prior miscarriage.  Among parous women (n=992), 41% (n=411) reported a prior preterm birth and 14% (n=126) reported a prior stillbirth. The HIV seroprevalence was 24%. Discussion: We have established a large cohort of pregnant women and newborns at the WNH to characterize the determinants of adverse birth outcomes in Lusaka, Zambia. Our overarching goal is to elucidate biological mechanisms in an effort to identify new strategies for early detection and prevention of adverse outcomes. We hope that findings from this cohort will help guide future studies, clinical care, and policy.

The Zambian Preterm Birth Prevention Study (ZAPPS): Cohort characteristics at enrollment

Background:Sub-Saharan Africa bears a disproportionate burden of preterm birth and other adverse outcomes. A better understanding of the demographic, clinical, and biologic underpinnings of these adverse outcomes is urgently needed to plan interventions and inform new discovery.  Methods:The Zambian Preterm Birth Prevention Study (ZAPPS) is a prospective observational cohort established at the Women and Newborn Hospital (WNH) in Lusaka, Zambia. We recruit pregnant women from district health centers and the WNH and offer ultrasound examination to determine eligibility. Participants receive routine obstetrical care, lab testing, midtrimester cervical length measurement, and serial fetal growth monitoring. At delivery, we assess gestational age, birthweight, vital status, and sex and assign a delivery phenotype. We collect blood, urine, and vaginal swab specimens at scheduled visits and store them in an on-site biorepository. In September 2017, enrollment of the ZAPPS Phase 1 – the subject of this report – was completed. Phase 2 – which is limited to HIV-uninfected women – reopened in January 2018.  Results:Between August 2015 and September 2017, we screened 1784 women, of whom 1450 (81.2%) met inclusion criteria and were enrolled. The median age at enrollment was 27 years (IQR 23–32) and thee median gestational age was 16 weeks (IQR 13–18). Among parous women (N=866; 64%), 21% (N=182) reported a prior miscarriage, 49% (N=424) reported a prior preterm birth, and 13% (N=116) reported a prior stillbirth. The HIV seroprevalence was 24%. Discussion:We have established a large cohort of pregnant women and newborns at the WHN to characterize the determinants of adverse birth outcomes in Lusaka, Zambia. Our overarching goal is to elucidate biological mechanisms in an effort to identify new strategies for early detection and prevention of adverse outcomes. We hope that findings from this cohort will help guide future studies, clinical care, and policy.