Even as prenatal screening for Down syndrome has improved markedly, false-positive rates still hover around 2–5%, posing a dilemma both for patients and their physicians as they weigh the significant risks associated with invasive diagnostic tests. Discussed in this issue of Strategies, a new test based on fetal cell-free DNA in maternal plasma now promises a dramatically lower false-positive rate..
Prenatal screening for Down syndrome since the early 1990s has combined maternal age and sonographic measurement of nuchal translucency. This strategy produces a detection rate of approximately 65%, with a 5% false-positive rate. In addition, reports have shown that maternal serum markers can significantly improve screening with nuchal translucency and age. First trimester markers include pregnancy-associated plasma protein A (PAPP-A), and free beta or total human chorionic gonadotropin (hCG). For the second trimester, alpha-fetoprotein (AFP), chorionic gonadotropin (hCG), unconjugated estriol (uE3), and, increasingly, inhibin A are used. Putting the first and second trimester markers together can increase the detection rate of Down syndrome screening to as high as 90% with fewer false-positives—as low as 2%.
However, even with the added benefit of maternal serum markers, the next step in clinical practice—diagnostic testing—remains risky: about one in 200 amniocentesis or chorionic villus sampling procedures will result in fetal loss, a slight but significant statistic for patients to consider. In addition, even with the very best screening protocols, only one out of 16 women having an invasive diagnostic test will be a true positive. Now, with the advent of more cost-effective next-generation DNA sequencing, researchers may have dramatically shifted the future of screening for Down syndrome. Recently, a clinical validation study using fetal cell-free DNA from maternal plasma found that a next-generation DNA sequencing method can push the false-positive screening rate to as low as 0.2% while detecting almost 99% of affected pregnancies (Genet Med 2011;13:913-20).
Although circulating fetal cell-free DNA in maternal blood was first described in 1997, until recently, the high cost of DNA sequencing prohibited practical application of the discovery to test for Down syndrome. The new study delivers on the promise of that auspicious discovery, and also demonstrates the potential of such testing to shift the current paradigm of Down syndrome screening, said Glenn Palomaki, PhD, lead author of the study. “This test provides for the first time an additional option for women who are identified as being at high risk for Down syndrome that doesn’t involve invasive procedures,” he said. “Because it’s based on just a maternal blood sample, it stands between our routine screening tests and invasive diagnostic tests that carry a risk of fetal loss.” Palomaki is associate director of the Division of Medical Screening and Special Testing at Women & Infants Hospital in Providence, R.I.
Palomaki and his co-principal investigator, Jacob Canick, PhD, conducted the study with data from a cohort of 4,664 high-risk pregnancies at 27 prenatal diagnostic centers worldwide. They compared fetal karyotyping of 212 Down syndrome and 1,484 matched euploid pregnancies in a blinded, nested case-control design. All of the samples were tested at the Sequenom Center for Molecular Medicine (SCMM) in San Diego, Calif. using Sequenom’s lab-developed test. A subset of the samples was also tested independently by the Orphan Disease Testing Center at the University of California at Los Angeles. SCMM and UCLA performed the test, described in an earlier study, using massively parallel shotgun sequencing (MPSS) on the Illumina HiSeq 2000 platform.
The study found that the test identified 98.6% of the Down syndrome pregnancies with a 0.2% false positive rate. A positive result increased Down syndrome risk by 490-fold, and a negative result reduced risk by 72-fold. The test was unable to provide a clinical interpretation only 0.8% of the time, usually due to a low fetal fraction of cell-free DNA in the sample. The fetal fraction must be within 4–50% for the interpretation to be considered reliable.
For labs that choose to offer this test to patients via SCMM, sample collection and transportation will be the biggest concern, said Nathalie Lepage, PhD, associate professor of pathology and laboratory medicine and head of the division of biochemical genetics at Children’s Hospital of Eastern Ontario in Ottawa. “The pre-analytical phase of this could be challenging,” said Lepage, who was not associated with the study. “For example, for screening assays we perform in the lab, we always collect serum, and they need the cell-free DNA in plasma. In addition, if it’s just the one lab performing the test, reimbursement could be a problem.” For MPSS, samples must be processed, frozen, and shipped on dry ice. The authors of the paper recommend demonstration projects to show efficacy in clinical settings.
While the study demonstrates that the test is ready for clinical practice, it will likely take a long time for MPSS testing to be available outside of the commercial specialty lab, Palomaki said. “As an analogy, if you were to have said 15 years ago that tandem mass spectrometry was going to be routinely done by a small box in newborn screening laboratories, people would have said you were crazy. But that’s how far the technology has evolved,” he said. “So, will hundreds of labs someday have a platform in house that performs MPPS testing? Some say that will be true in a few years. Right now it isn’t. But I think it’s possible.”
The most difficult challenge facing a lab performing such an assay is the bioinformatics, Palomaki emphasized. “Even if you can generate these incredible amounts of data on an MPSS platform, finding out what that all means is extremely difficult, and requires a lot of computing power,” he said. In MPSS, the instrument sequences millions of tiny fragments of DNA at a time. For each patient in the study, the instrument produced, on average, 25 million sequences, after which the computer had to map each base pair to a model human genome, referred to as alignment. “After that point, the mathematics becomes a little more straightforward,” Palomaki said. “You know you have a certain number of fragments of each chromosome, and you’re looking for the proportion that map to chromosome 21 to be just ever so slightly larger if the fetus has Down syndrome than if the fetus is normal.” The difference in the number of 21st chromosomes that translates to a positive result is tiny—about 5%.
Lepage agreed that the bioinformatics component could be the biggest obstacle to implementing such a test in other labs. “Labs may be very comfortable with their instruments, but many do not have access to bioinformatics professionals that can handle this data,” she said. “This will be a major change for the lab: instead of relying on markers, levels, and comparison with medians, we would need to rely on other specialists that we may not currently have close relationships with.”
Should MPSS testing for Down syndrome become routine, as an element of prenatal screening, it will still be part of a larger program, Lepage emphasized. “When we perform the biochemical testing now, we are part of a program that offers counseling for the screen positive patients, and the same principles will need to apply, even if we move away from biochemistry to cytogenetics,” she said. “We will still need to have this program approach, where every time that you have a positive results, the patient will be able to access appropriate counseling.”