They aimed to identify a small set of detectable signals in the first 16 weeks of pregnancy that could form the basis of a simple, inexpensive diagnostic test that is feasible for use in low-, middle-, and high-income countries. To assess the accuracy of the machine learning models, the researchers initially constructed the models with data from the discovery cohort and then validated the results by testing their performance on data from women in the validation cohort.
When you reduce preeclampsia, you are likely to reduce preterm birth. It’s a double whammy of good influences.
A prediction model using a set of nine urine metabolites was highly accurate, the researchers found. These urine markers, in samples collected before 16 weeks of pregnancy, strongly predicted who later developed preeclampsia. Test performance was measured by a statistical standard used in machine learning known as the area under the characteristic curve. An AUC of 1 for a two-outcome test indicates perfect prediction, while an AUC of 0.5 indicates no predictive value, just like the results obtained by flipping a coin. For urine markers, the AUC was 0.88 in the discovery cohort and 0.83 in the validation cohort, indicating high predictive ability.
Measurement of the same set of urinary metabolites in samples collected during pregnancy yielded similar predictive power, with an AUC of 0.89 in the discovery cohort and 0.87 in the validation cohort.
The researchers confirmed that their model had stronger predictive power than using only clinical characteristics associated with a pregnant woman’s risk of preeclampsia, such as chronic hypertension, high body mass index, and carrying twins.
A set of nine proteins measured in blood performed almost as strongly, with an AUC of 0.84.
The researchers also created a predictive model that combined participants’ clinical characteristics with urinary metabolites, which allowed them to predict early-onset preeclampsia with an AUC of 0.96. Clinical characteristics in the combined model are data already collected as part of standard medical records, such as patient age, height, body mass index, and prepregnancy hypertension.
“This data collection is routine and could serve as a first level of triage,” Agheeapour said. “We hypothesize that patients who the data show are at risk could receive a more extensive urine test.”
Discovering the biology of disease
Stanford Medicine researchers are also opening windows into the biology of preeclampsia. Another study, published in February the natureused cell-free RNA measurements to uncover biological clues about how preeclampsia occurs.
“Being able to eavesdrop on conversations during pregnancy, synchronously measuring molecules from the pregnant woman, the fetus and the placenta, is very helpful in giving us clues about which biological changes contribute to disease,” said Dr. Meera Mufaray, lead author of the study. the nature of work, who was a graduate student in bioengineering when the research was conducted. The paper’s senior author is Stephen Quake, DPhil, professor of bioengineering and applied physics.
“The most striking changes occurred before 20 weeks of gestation, while the diagnosis of preeclampsia is usually made at 30 and more weeks of pregnancy,” said Moufarey. “That was surprising.” We would expect changes in gene signals when you see clinical symptoms, and that was happening much earlier in pregnancy.
Using 404 blood samples from 199 pregnant women, Mufarej and her colleagues identified a set of 18 genes whose activity in early pregnancy predicted the development of preeclampsia.
The genes are consistent with what is known about the biology of how the disorder develops, she noted.
Scientists hypothesize that in preeclamptic pregnancies the placenta does not fully develop; its blood vessels may be too small. At first this is fine because the fetus is small and does not need much nutrition.
“But later in the pregnancy, the fetus grew, sending signals for more nutrition,” Mufaray said. “At that point, the only solution for the small blood vessels is more blood flow, so we see high blood pressure.” In severe cases, the pressure can cause the placenta to separate from the lining of the uterus prematurely, creating an emergency in which the baby must be delivered immediately.
The gene activity signals that Moufaray and her colleagues identified came from genes involved in pathways consistent with the development of preeclampsia, such as tissues related to the endothelial system, the placenta, and the brain. (The brain is relevant because full-blown eclampsia causes seizures.) The scientists plan to use the work as a basis for future research into how the condition develops.
Scientists involved in both studies will validate their predictive tests in much larger, more diverse populations of women, with the goal of creating tests for universal use.
Knowing more about how preeclampsia develops and how to predict it could have profound benefits for the world’s most vulnerable mothers, the researchers said, noting that an estimated 86% of maternal deaths worldwide occur in Asia and sub-Saharan Africa.
“This kind of testing is really needed here, where resources are very scarce,” said Marić. Unlike women in high-income countries, many women in low-income regions give birth far from hospitals, limiting their access to emergency care when they show symptoms of preeclampsia or eclampsia. “If we can identify which pregnancies are at high risk early, we can help get those women to health facilities and prevent death.”
The Patterns The study was supported by the March of Dimes Prematurity Research Center at Stanford University School of Medicine, the Stanford Maternal and Child Health Research Institute, the Christopher Hess Research Fund, the National Institutes of Health (grants 1R01HL139844, 5RM1HG00773507 and R35GM138353), Burroughs Wellcome, the Alfred E. Mann, The Bill & Melinda Gates Foundation, The Thomas C. and Joan M. Merrigan at Stanford University and Chan Zuckerberg’s Microbiome Initiative.
The the nature The study was supported by the Chan Zuckerberg Biohub, the Global Alliance to Prevent Prematurity and Stillbirth, the March of Dimes Foundation, the National Science Foundation (grant DGE 1656518), the Benchmark Stanford Graduate Fellowship, the Stanford ChEM-H Chemical Biology Interface Training Program, and the H&H Evergreen Fund.