The other method, the ultrasound, can also be problematic because it gives less reliable information as a pregnancy progresses and doesn't predict spontaneous preterm birth (not to mention the equipment and trained technicians needed makes this option really expensive). Recently released provisional data for 2017 from the National Center for Health Statistics show that the preterm birth rate in the US has reached 9.93 percent, up from 9.86 in 2016, the third consecutive annual increase after steady declines over the previous seven years.
So now, new research funded by the USA infant research group March of Dimes has published its findings on an experimental blood test that might be able to detect when a pregnant woman is likely to deliver her baby prematurely.
The team of 18 researchers from multiple universities was led by Stephen Quake of Stanford University, a pioneer in genomic diagnostics and in developing new approaches to biological measurement.
The tests measure the activity of maternal, placental and foetal genes by assessing maternal blood levels. "To date, no test on the market can reliably predict which pregnant moms will go on to preterm labor", comments Stacey D. Stewart, president of March of Dimes.
Premature birth is the leading cause of newborn death in the United States.
The research is still at a nascent stage and was based on data obtained from a small number of women.
Encouragingly, the cfRNA blood test could predict the gestational age (where the birth occurred within 14 days of the estimated due date) with 45% accuracy, which is in the same range as estimates of gestational age and due date based on ultrasound, at 48%.
He said the findings affirmed the existence of a "transcriptomic clock of pregnancy" that could serve as a new way to access the gestational age of a foetus. The New York Times reported that the research, which is still in preliminary stages, detects changes in RNA in a pregnant woman's blood and can estimate due dates within 2 weeks in almost half the cases. Further development and validation of an initial model resulted in the identification of a panel nine placenta-specific RNAs (CGA, CAPN6, CGB, ALPP, CSHL1, PLAC4, PSG7, PAPPA, and LGALS14) in maternal blood that could predict gestational age.
In a related study of another 38 women, all at elevated risk of delivering preterm, the researchers identified seven nucleic acids that accurately classified women who delivered preterm up to two months in advance of labour. "They can be applied across the globe as a complement to or substitute for ultrasound, which can be expensive and inaccurate during the second and third trimester...." The gestational age blood test did not do a great job of predicting which women would deliver prematurely, suggesting those particular genes "may not account for the various outlier physiological events that may lead to preterm birth", the study said.
The new tests however need to be validated in larger cohorts of pregnant women before they can be made available for widespread use, the researchers said. Existing tests at best can provide a premature delivery to women at high risk, who have already given birth prematurely, but again their accuracy is only around 20%.