NEW YORK – New research suggests circulating, cell-free DNA (cfDNA) found in blood samples collected during pregnancy can be used to predict preeclampsia risk, particularly the risk of severe preeclampsia linked to preterm birth.
"Our study demonstrates the possibility to leverage existing clinical prenatal cfDNA screening tests given during early pregnancy to predict the risk of developing preeclampsia (PE) many months in advance," Gavin Ha, a researcher affiliated with the Fred Hutchinson Cancer Center, the Brotman Baty Institute for Precision Medicine, and the University of Washington, said in an email.
As they reported in Nature Medicine on Wednesday, Ha, University of Washington researcher Raj Shree, and their colleagues used low-coverage whole-genome sequencing on 1,854 blood plasma samples collected at a median of 12.1 weeks, searching for maternal and fetal cfDNA sequences that portend development of the placental dysfunction condition.
"Given that the placenta is inaccessible during pregnancy," the study's authors reasoned, "noninvasive approaches are essential to identify molecular features indicative of preeclampsia risk, improve the understanding of underlying biology, and develop strategies to mitigate the risks of preeclampsia."
The team analyzed the newly generated prenatal cfDNA sequences with a computational framework known as "preeclampsia early assessment of risk from liquid biopsy" (PEARL) to predict preeclampsia risk based on tissue-specific sequence analyses and on the proportion of cfDNA sequences stemming from fetal DNA based on nucleosome profiles, nucleosome-depleted regions, and sequence coverage.
"We developed new computational methods that enabled us to learn several important insights," Ha said. "We accurately predicted the fetal fraction agnostic of fetal sex, and we observed the enrichment of endothelial cfDNA shed into circulation in those who eventually developed [preeclampsia]. Both these features were important for guiding the development of a classifier to accurately predict [preeclampsia] risk."
In 831 prenatal cfDNA samples collected at a median of 12.2 weeks of gestation, for example, the investigators were able to predict preeclampsia risk with 85 percent accuracy with a sensitivity of 81 percent and 80 percent specificity.
In an external validation cohort composed of 141 samples collected at a median of 12.1 weeks gestation, meanwhile, they were able to predict preeclampsia cases requiring preterm delivery with 84 percent accuracy, 79 percent sensitivity, and 80 percent specificity. The accuracy increased to 92 percent for predicting early preeclampsia and dipped to 79 percent when it came to predicting the risk of late preeclampsia with preterm birth.
Based on the findings so far, Ha suggested that there "are opportunities to include preeclampsia risk prediction directly into routine prenatal screening tests during early pregnancy."
"Such a test can be useful for informing interventions, including increased monitoring and treatment decisions," he explained. "Furthermore, a test for accurately stratifying [preeclampsia] risk may be useful in studies or trials testing new investigational therapeutic strategies."
Members of the team are now designing larger studies aimed at refining their cfDNA-based preeclampsia risk prediction strategy. While the current approach appeared particularly adept at finding severe and early forms of preeclampsia, for example, Ha highlighted the importance of flagging the risk of preeclampsia development later in pregnancy.
"We are working to improve our methods for predicting the risk of developing [preeclampsia] even later in pregnancy," he said. "Eventually, we hope to implement this approach into the clinical prenatal cfDNA screening platform."