By Deborah Borfitz
January 5, 2022 | A global collaborative of researchers has used a novel platform based on cell-free RNA (cfRNA) to monitor patterns of gene expression as pregnancy progresses—including, surprisingly, genes that appear to originate with the developing fetus. They recently demonstrated the platform’s potential as a pregnancy screening tool, using preeclampsia as a test case, but it might logically also be brought to bear in the future treatment and prevention of any number of pregnancy-related complications.
The investigation resulted in a framework to assess healthy pregnancy progression and, by extension, maternal and fetal complications that were differently expressed, according to Thomas McElrath, M.D., Ph.D., in the division of maternal-fetal medicine at Brigham and Women's Hospital. “Once you have a pattern of progression for normal, you can start to look at what abnormal looks like.”
For preeclampsia specifically, the method produced a seven-fold improvement in correctly predicting the disease compared to current methods. Preeclampsia is typically diagnosed based on a combination of clinical tests (i.e., blood pressure and urine or blood test) and ultrasound.
Ultrasound and cfDNA-based screening are currently mainstays of prenatal testing for birth defects and chromosomal abnormalities, says McElrath. “This represents a novel use of cell-free RNA in the pregnancy arena for women’s health.”
Clinicians currently don’t have any great means of identifying women at risk of preeclampsia, he says. The only solid clue is if they had preeclampsia previously, which is of no help with the one-third of first-pregnancy cases.
The best doctors can do is look at clinical factors such as hypertension and diabetes to assess risk and proceed accordingly. “[RNA profiling] allows us to be much more specific … based on the gene set being expressed by an individual woman.”
Maternal-Fetal Crosstalk
In a research article newly published in Nature (DOI: 10.1038/s41586-021-04249-w), McElrath and his colleagues describe the use of RNA profiles as harbingers of future health and disease in pregnancy. The enabler was the RNA platform of biotechnology company Mirvie, which was applied to plasma samples from 1,840 pregnant women across eight diverse patient cohorts.
The RNA platform correctly identified 73% of pregnant women who ultimately had a delivery complicated by preeclampsia and did so several months before clinical symptoms appeared, says McElrath. It performed nearly as well at predicting preterm birth.
Intriguingly, information about the fetal gastrointestinal tract and heart were detectable in maternal blood although the underlying mechanisms aren’t well understood, he adds. “It’s almost as if the placenta is purposefully shunting some of this information [probably via exosomes] … across the placenta onto the maternal side.”
This contradicts what many doctors were taught in medical school; that is, “what happens on the baby side of the placenta stays on the baby’s side of the placenta and what happens on the mother’s side stays on the mother’s side,” says McElrath. “Clearly, there appears to be some crosstalk that we are just beginning to discern.”
The cfRNA approach also sheds light on the molecular functioning of the placenta itself as well as maternal tissues composing the cervix and uterus. As pointed out in the paper, gene sets involved with collagen would make potential reference to how the cervix is doing, which might be important information when determining a woman’s risk of preterm birth.
As a clinician, McElrath continues, he is aware of the acute need for better diagnostic and screening tools in the pregnancy realm. The research community has given scant attention to the beginning of life—arguably one of the most important areas of human health—relative to diseases such as cancer and heart disease that strike many years later.
Robust Indicator
RNA sequencing was used in the study to monitor pregnancy health week to week, creating a “roadmap” of the normal gene set expression, McElrath says. Researchers analyzed 2,539 plasma samples to learn which maternal, fetal, and placental genes were activated at sequential points in pregnancy.
McElrath’s lab contributed samples from a biobank it runs in Boston, as did the larger group of collaborators across the three continents. The samples were all sent directly to Mirvie for analysis, and a subset of the group (including McElrath) did the interpretation and bioinformatics work.
To make an early prediction of preeclampsia, researchers first identified signatures that separated cases and controls. Seven genes were consistently identified: CLDN7, PAPPA2, SNORD14A, PLEKHH1, MAGEA10, TLE6, and FABP1.
The transcriptome expression dataset used is one of the largest ever studied, he notes. Samples came from cohorts in North America, Europe, and Africa, creating a diverse representation of the human pregnancy experience.
When standard clinical variables such as race, age, and body mass index were introduced, the gene variables were unaffected, he points out, indicating the cfRNA profiles are a robust indicator of preeclampsia by themselves and generalizable across human populations. “This suggests we’re finally getting down to what is probably a deeper level of the underlying mechanisms of the disease.”
It also makes the case for moving away from the controversial use of race variables in clinical risk scores, which he describes as a “black box on the human experience and probably not a box that is evenly distributed among all of us.” Diseases might instead be understood, free of bias, at the individual molecular level.
Clinical Needs
In terms of estimating pregnancy progression, the novel cfRNA technique performed on par with a second trimester ultrasound and better than a third trimester ultrasound, McElrath says. The takeaway here is that the platform might be used to assess gestational age for women who begin prenatal care later in pregnancy.
While 10% or less of pregnant women in the Boston area begin prenatal care in the late second or third trimester, he adds, in more underserved locations around the globe the proportion can be significantly higher. Only when prenatal care happens in a timely fashion would the RNA platform be useful for the early detection of conditions like preeclampsia that have a long latency period prior to their clinical presentation.
The platform might also allow clinical trials of pregnancy conditions to be more efficiently conducted by homing in on women who are at heightened risk, says McElrath. The need for such studies is dire but the cost of doing them is often prohibitively expensive. Preeclampsia, for example, occurs in only about 7% of all pregnancies so proving a therapeutic medication has efficacy would require enrolling many thousands of participants to get a representative sample of the disorder in its different forms.
In terms of therapeutic development, the seven genes associated with preeclampsia can now be further explored to better understand their functioning and look for any existing therapeutics that modify their activity—or other genes or proteins in the same pathway that may be susceptible to modification, he says.
“None of us working with preeclampsia think it is a single disease,” adds McElrath. “It is probably a syndrome, multiple related conditions that share a common clinical presentation.” Teasing that apart might identify potential therapeutics best suited to certain versions of the disease so that patients might be enrolled in therapeutic trials that are more likely to succeed.
Consider the situation with breast cancer, he cites as an example. Three decades ago, it was commonly thought to come in several basic forms. But today, women diagnosed with breast cancer routinely get their tumor genotyped and treated with a targeted therapeutic. “Not everyone gets the same treatment anymore, and that is hopefully the way we go with conditions in pregnancy [not only preeclampsia].”
The cfRNA-based testing approach remains under active development by Mirvie, where McElrath serves on the scientific advisory board. McElrath also holds an equity interest in the company.
While the platform is unlikely to supplant ultrasound, it might well augment the standard diagnostic arsenal for prenatal care, he says. It might also find utility in multiple areas of concern in clinical obstetrics, including gestational diabetes and fetal growth restriction, based on “deviations from normal” gene set expression.