Blood tests to predict preterm birth risk raise excitement -- and questions
Posted June 6, 2018 6:03 p.m. EDT
(CNN) — Researchers are looking into the possibility of using blood tests to predict a woman's risk of premature birth, which could change the future of prenatal care.
A pregnant woman's risk of delivering early can depend on factors such as a previous preterm birth, certain pregnancy or medical complications, preeclampsia, stress or diabetes. A woman can also have a premature birth even if she has no known risk factors.
"It's really intriguing to be able to identify when a woman is at higher risk than the general population for a serious complication like preterm birth. That would allow us to focus on them in a way we can't do right now," said Dr. Larry Rand, an associate professor and the Marc and Lynne Benioff Endowed Chair in Maternal Fetal Medicine at the University of California, San Francisco. "On the flip side, in terms of predicting (medically) indicated preterm birth, what you're really predicting is the likelihood that mom or baby will develop a complication like preeclampsia or growth restriction that may require ending the pregnancy early in order to keep mom and baby safe," said Rand, who is also principal investigator of the California Preterm Birth Initiative.
Two separate studies -- one published last month in the Journal of Perinatology and another published Thursday in the journal Science -- are among the latest efforts to develop such a test. Rand is an author of the Journal of Perinatology study.
The tests described in those studies are still in very early stages of development, and much more research is needed to determine and validate their accuracy, safety and efficacy -- but they represent where preterm birth testing seems to be heading.
In the United States, preterm birth affected about one of every 10 infants born in 2016, according to the Centers for Disease Control and Prevention.
Around the world every year, an estimated 15 million babies are born too early -- before 37 completed weeks of gestation -- and this number is rising, according to the World Health Organization. Preterm birth complications are the leading cause of death among children under 5, responsible for about 1 million deaths worldwide in 2015.
In 2010, the 10 countries with the highest total numbers of preterm births were India, China, Nigeria, Pakistan, Indonesia, the United States, Bangladesh, Philippines, Democratic Republic of Congo and Brazil, according to a study published in the Lancet in 2012.
However, some of those data might not be entirely accurate since some women in low-income countries lack the resources to determine the gestational age of a pregnancy.
"Especially in low- and middle-income countries with really high preterm birth rates and huge neonatal mortality rates, one of the biggest problems is, there is no ultrasound, and women are not empowered to keep track of their last menstrual cycle. It's often taboo to keep track of or talk about your period," Rand said.
"That is one of the biggest problems we have worldwide, and lots of folks have been working on trying to find another way -- ideally an inexpensive test -- to figure out what the gestational age is for any given pregnancy."
What testing for preterm birth risk looks like
The blood test described in the Journal of Perinatology aimed to predict the risk of preterm birth among pregnant women with and without preeclampsia, a condition characterized by high blood pressure.
The test, developed and studied in 400 women during their second trimester, screened for 25 biomarkers or substances in the blood that were signs of inflammation and immune system activation, as well as certain protein levels, indicative of a possible preterm birth risk.
Those 25 biomarkers were selected from a panel of 63 that had been shown to be related to preterm birth or preeclampsia.
The researchers found that screening for those biomarkers, along with taking into account a pregnant woman's age and poverty status, could identify the risk of preterm birth with or without preeclampsia in most of the women in the study during their second trimester.
"Our test was able to predict 80.3% of women who went on to have any preterm birth, at 15 to 20 weeks gestation," said Laura Jelliffe-Pawlowski, an associate professor and director of Precision Health and Discovery at the University of California, San Francisco's Preterm Birth Initiative, who was first author of the study.
She added that because the test identified women who are at high risk for preterm birth with preeclampsia -- identifying about 95% of women who went on to have preterm birth with preeclampsia before 32 weeks -- it offers an opportunity for intervention, such as for starting those moms-to-be on a low-dose aspirin to help lower their risk of preeclampsia.
"Our test also suggests that women with elevated risks based on the test might benefit from additional testing," she said. "I think our test could be rapidly translated into clinical settings if it is found to validate in a larger set of pregnancies. The next step for us is to identify the right partners for conducting a large, rapid, clinical validation study."
In other words, more research is needed.
As for cost, Jelliffe-Pawlowski thinks the test could probably be run for about $50 to $100.
"A low-cost, highly accurate test has the most potential for helping blunt the current epidemic of preterm birth in the United States and around the world, an epidemic that disproportionally affects women of color, especially black women, who are as much as twice as likely to deliver preterm than their white peers," she said.
The other blood test in development, described in a small pilot study in the journal Science, aimed to predict not only preterm delivery but the gestational age of a pregnancy. Currently, the gold standard for confirming gestational age is to use ultrasound.
For that study, researchers assessed RNA in the bloodstream of pregnant women to determine the health and stage of their pregnancies. RNA, or ribonucleic acid, is a single-stranded molecule similar to DNA.
"For things like preterm birth, which we believe to be that something went wrong at some point in pregnancy, DNA would look constant over time, and it wouldn't give you an idea of how things are changing," said Mira Moufarrej, a graduate student of bioengineering at Stanford University who was a co-first author of the new pilot study. "Whereas RNA gives you an idea of what cells in the body are doing at different points in time, and sometimes the RNA found in the blood is indicative of what's going wrong."
In 31 healthy pregnant women from Denmark, the researchers measured nine cell-free RNA molecules to predict gestational age with comparable accuracy to ultrasound. Those measurements were based on a machine-learning model that the researchers had previously built to understand what RNA molecules look like in a healthy pregnancy.
In the study, the newly developed test accurately estimated gestational age within 14 days of the actual gestational age at delivery for 45% of women in the study. When ultrasound was used in the study, the ultrasound predicted gestational age in 48% of women.
The 45% accuracy for the blood test was based on an average of samples taken during the second and third trimester. The 48% accuracy for ultrasound was based on ultrasound done during the first trimester of pregnancy.
Often, using ultrasound to estimate gestational age can be costly for some women -- about $200 to $300, depending on insurance and where it's done -- so a blood test could be a more affordable option.
Then, in 38 women at risk of delivering preterm, the researchers identified seven other cell-free RNA molecules that could accurately distinguish between those women who ended up delivering preterm and those who did not. The test measured up to two months in advance of when any woman delivered preterm, using maternal blood samples drawn late in the second trimester.
Among those 38 women, 25 had full-term and 13 had preterm deliveries. The women who delivered preterm did so between 23 and 33 weeks, Moufarrej said.
"Essentially, the findings are that you can measure gestational age and predict risk of preterm delivery in this pilot study," Moufarrej said, adding that the study still needs to be replicated in a larger, more diverse sample of women, including asymptomatic women.
"The next steps would be to get an ethnically diverse cohort of women prospectively collected and then see if these findings validate there, both findings about gestational age and the ones about preterm delivery," she said. "Maybe, hopefully in the next five years, this might be available to the general public, if things validate."
Stephen Quake, professor of bioengineering at Stanford University and lead author of the study, said it would take around two to three years to do such a trial.
"It's important to emphasize that this is a proof of principle, and it's not a clinical test yet," he said, adding that if it's made available in the future, "it's hard to tell right now" how much the blood test could cost, but it could be offered for around $100.
The state of preterm birth testing
Among the major differences between those two blood test studies are sample size and methodology, but their goals were the same: to predict preterm birth.
"As a neonatologist, we see many infants who are born preterm with little warning or known risk factors during the mother's pregnancy," said Dr. Noelle Younge, assistant professor of pediatrics at Duke University School of Medicine, who was not involved in the studies.
"The ability to noninvasively monitor fetal development and predict the risk of preterm birth through circulating markers could have a major impact on pregnancy monitoring and management," she said. "These studies show that a lot of progress is being made. Further work is needed not only to understand how these tests perform in large studies but also to determine how they should be used clinically."
After all, preterm birth is often unexpected and significantly increases an infant's risk of long-term health and developmental problems, so new approaches to identify women at risk are needed, Younge said.
"Currently, diagnostic tools to predict the risk of preterm birth are limited," she said. "There are blood tests in different stages of development, but none have been widely adopted in clinical practice. Continued work is needed to understand the utility and applications of these tests in prenatal care."
There have been many efforts to predict preterm births. For example, the company Sera Prognostics developed a blood test to predict premature birth risk, the PreTRM Test, which became commercially available in the United States in 2016.
The test costs $945, as listed on its website, and because it is new, it is not covered by insurance.
Last year, a study published in the medical journal JAMA found that routine screening methods to predict spontaneous preterm birth before 37 weeks -- by measuring cervical length, assessing a certain protein called fetal fibronectin or both -- had low predictive accuracy for spontaneous preterm birth in a sample of 9,410 pregnant women in the United States.
The women had never given birth before and carried their pregnancies to 20 weeks or more between 2010 and 2014.
Drs. Steven Bloom and Kenneth Leveno, both of the University of Texas Southwestern Medical Center in Dallas, wrote an editorial that accompanied that study in JAMA.
"The health and economic consequences of preterm birth are enormous, and it is understandable why physicians feel tremendous pressure to implement new practices to prevent preterm birth. However, that pressure does not eliminate the need to introduce new practices only when based on fully vetted evidence," they wrote.
"Physician consensus committees should not legitimize new interventions until it is certain that those interventions will have clinical utility."