NEW YORK – Noninvasive prenatal diagnosis based on the genetic analysis of circulating fetal or placental cells appears feasible, though clinical adoption will likely take several more years, according to a new study by researchers at Baylor College of Medicine and Columbia University.
While cell-free DNA-based NIPT has taken off in recent years, a number of research groups have been focusing their efforts on fetal cell-based methods, which would enable the analysis of pure fetal DNA instead of a mixture of maternal and fetal genetic material and possibly yield more accurate results with higher resolution. However, despite decades of research, it has proven difficult to isolate sufficient numbers of high-quality fetal-derived cells from blood or other noninvasive samples.
"Single-cell analysis has the potential to allow more sophisticated and complete analyses and has the potential to eventually replace diagnostic procedures that cell-free fetal DNA likely will never realistically achieve," said Mark Evans, a professor at Mount Sinai School of Medicine. Evans was involved in a National Institutes of Health-funded study called NIFTY in the 1990s that aimed to isolate fetal cells from blood for analysis by fluorescent in situ hybridization.
In a proof-of-concept study involving almost 100 pregnant women, published two weeks ago in the American Journal of Human Genetics, researchers led by Arthur Beaudet at Baylor have now demonstrated that they could find trophoblasts — placenta-derived cells — in almost all blood samples tested. In addition, cells from more than half of the samples had genetic material suitable for single-cell sequencing and analysis for aneuploidy and copy number changes.
"These cells are definitely there, and when we find them, the analysis using sequencing technology works very well," said Ignatia Van den Veyver, an author of the study and a professor in the Departments of Obstetrics and Gynecology as well as Molecular and Human Genetics at Baylor.
For their project, the researchers collected two sets of blood samples from a total of 95 pregnant women who were recruited during prenatal visits at Baylor and Columbia — one set of 42 samples and another of 53 samples. Most came were from singleton pregnancies, and gestational age ranged from 8 weeks to 29 weeks.
To enrich trophoblasts, the scientists used a previously published positive selection protocol that they had since optimized. It involved antibodies against three trophoblast cell-surface antigens — HLA-G, TROP-2, and EpCAM — and magnetic bead selection. This was followed by immunostaining with anti-cytokeratin and CD45 antibodies. The samples were then spread on a slide and scanned using RareCyte's CyteFinder. After manually identifying trophoblast candidates under the microscope, based on their cytokeratin patterns and the absence of CD45, the researchers picked individual cells using the RareCyte CytePicker.
On average, they identified five to seven trophoblasts per blood sample, with only two samples where no such cells could be found. Subsequent sequencing and genotyping showed that 94 to 96 percent of the cells scored as trophoblasts under the microscope were indeed of fetal origin.
In addition, a little over half the samples had at least two high-quality trophoblasts where the sequencing results could be scored for both aneuploidy and small copy number variants.
Interestingly, Van den Veyver said, there seem to be more trophoblasts in the blood until about week 15 of pregnancy, after which their number appears to decline.
Also, almost half the trophoblasts were present in clusters of several cells. "It's one parameter we can look at to say that these cells are indeed trophoblasts," she said. "It helps us identify the cells."
Next, DNA from single trophoblasts or clusters of cells underwent whole-genome amplification, followed by Illumina sequencing and analysis.
For a total of 45 samples, the researchers had diagnostic results from amniocentesis or chorionic villus sampling (CVS) available. For 34 of these, they saw concordance with the single circulating trophoblast (SCT) sequencing results, and for eight cases, the diagnostic results were normal but SCT testing failed.
In the remaining three cases, the diagnostic results were normal but SCT had an abnormal finding, which the researchers interpreted as examples of confined placental mosaicism. "This is not unexpected and I don't think this is a false positive," Van den Veyver said. "These cells are probably there but they are confined to the placenta."
Although the study provides a good first start, the method is still far from clinical use, according to Evans. "It needs more work to prove it can work reliably and can be done in a cost-efficient manner," he said. "If they can do three specimens a week, it's useless. First you have to prove the science, then you have to prove the engineering — that you can make it work — then you have to prove the economics."
"They will eventually get there, I have no doubt about that, but the question is when, and I don't think it will be anytime soon," Evans said.
To gain more experience with SCT, the team is currently conducting a second study, which will involve 300 women carrying fetuses with congenital anomalies. All of them will have undergone invasive diagnostic testing, and half will have a known finding from this test, while the other half won't. The study, planned for about three years, will allow the researchers to compare the results from SCT with the diagnostic test results.
After that, a trophoblast-based NIPT might gradually be introduced into clinical care, initially for women at high risk for fetal abnormalities, as a possible alternative to invasive diagnostic testing. "The paper and the work to date have pretty much substantiated that this is a very reliable test," said Ron Wapner, another study author and director of reproductive genetics at Columbia University, although there may be cases where not enough fetal cells with high-quality DNA are available.
According to the AJHG paper, Baylor's Beaudet is also a founder of a 2018 startup called Luna Genetics, which states on its website that it is "developing new technologies to capture fetal cells in maternal blood early in pregnancy and use advanced DNA analytic techniques to search for serious genetic abnormalities." Beaudet declined to provide further information about Luna Genetics at this time, including whether it plans to develop the method described in the paper.
In terms of technical improvements, the number of cells recovered intact could be further increased, Van den Veyver said. "These cells are rare but they're certainly there," she said. In addition, the quality of the sequencing results could be improved by eliminating cells from the analysis that are undergoing apoptosis or DNA replication.
Scaling the test through automation will be another challenge, but both Van den Veyver and Wapner said they're confident this will be possible.
"The ultimate goal would be, any pregnant woman who wants as much genetic information as she can obtain about her pregnancy would be able to have this test," Wapner said, adding that there needs to be discussion about what type of information is appropriate to provide and what exactly patients want to know.
Van den Veyver added that women who don't have access to invasive diagnostic testing at a major medical center could receive the test to obtain diagnostic genetic results for their fetus.
According to Evans, a main advantage of a noninvasive fetal cell-based test would be the ability to detect many abnormalities that current NIPT misses, but invasive testing does not. "There are tens of thousands of serious abnormalities that are currently being missed because people are drinking the Kool-Aid that NIPT can do everything that can be done with a diagnostic procedure and proper analysis," he said. "The utilization of CVS and amnio[centesis] has plummeted by more than 50 percent because patients and many physicians don't realize that what NIPT provides is less than what was available 10 years ago [by invasive testing], and is considerably less than what will be available over the next few years."
It is unclear, though, whether cell-based NIPT would eventually replace cell-free NIPT. "It's difficult to predict because we have to be aware that in parallel, all cell-free DNA-based testing continues to be researched and people are looking at how to improve accuracy of that for smaller and smaller changes," Van den Veyver said. "But one thing that cell-free DNA-based NIPT can never get us, or it would be very challenging, is pure fetal genomes. That's a big advantage of this test."
In the meantime, other groups are also pursuing cell-based NIPT, focusing on trophoblasts isolated from blood like the Baylor-Columbia team, fetal cells isolated from blood, or trophoblasts from cervical swabs.
For example, Danish company Arcedi Biotech published a proof-of-concept study two years ago, showing that it could detect aneuploidy and subchromosomal aberrations in trophoblasts isolated from blood samples of five pregnant women. The Baylor team previously collaborated with Arcedi but is no longer working directly with the company, Van den Veyver said.
Also, Illumina Ventures-backed Cradle Genomics recently raised $17.1 million in a Series A financing round to develop cell-based NIPT, likely by isolating trophoblasts from cervical samples.