For nearly 50 years, karyotyping has been the backbone of prenatal diagnostic testing for congenital anomalies, but rapid advances in genomic technologies are outpacing both this tried-and-true method and others poised to take its place. Pivotal research showing the superiority of microarrays in comparison to karyotyping is hastening the obsolescence of the latter; however, at the same time, microarrays also appear to be only a milestone along the path to even more sophisticated non-invasive testing approaches. The growing pains of adopting these new technologies require diligence by medical teams and likely will spell changes for lab practice.

"There's never a history of going backwards. This technology is here, it's going to move forward and replace what we're presently doing, but we're going to have to be very careful," advised Ronald Wapner, MD, professor, director of reproductive genetics, and vice chair of research in the department of obstetrics and gynecology at Columbia University in New York City. "This is a giant revolution and it's gotten very complicated for practicing physicians and patients. So as we transition the technology from the lab into clinical practice we really have to work as a team to use it appropriately."

Critical Comparisons

Wapner was the principal investigator of a recent landmark clinical trial that compared microarray analysis against karyotyping in more than 4,000 women referred for prenatal diagnosis based on maternal age, abnormal Down syndrome screening results, anomalies identified on ultrasound, or other indications (N Engl J Med 2012;367:2175–84). The participants, seen at 29 centers across the U.S., underwent chorionic villus sampling or amniocentesis. The study protocol involved halving each specimen collected from these procedures, with a single cytogenetics lab performing all karyotyping. Four different labs completed the microarray analyses using either Agilent or Affymetrix platforms. The arrays, designed to maximize detection of well-characterized microdeletions and duplications, also had probes for regions throughout the genome to identify other chromosomal imbalances.

This long-awaited National Institute of Child Health and Human Development (NICHD)-funded study demonstrated that single nucleotide polymorphism (SNP)-based and comparative genomic hybridization microarrays found all the aneuploidies and unbalanced rearrangements identified by karyotyping. Additionally, microarrays detected clinically relevant deletions or duplications in 6% of samples referred due to anomalies on ultrasound and in 1.7% referred because of advanced maternal age or positive screening results, all of which had normal karyotype results.

Taiwanese researchers also recently reported similar findings, in a smaller population (BJOG 2012;119:614–25). After testing of more than 3,100 cases, the authors found that microarray analysis had an additional 8.2% diagnostic yield in comparison to karyotyping.

A First-Line Test?

These findings present a solid case for using microarrays as the first-line prenatal diagnostic test, according to Aaron Caughey, MD, PhD, chair of the department of obstetrics and gynecology and associate dean for women's health research and policy at the Oregon Health and Science University in Portland, who was not involved in either study. "This is a big step forward, monumental even, because prenatal diagnostics have been focused for so long on Down syndrome detected by karyotyping, which is a very crude way to look at someone's genome," he explained. "However, there are so many genetic syndromes related to microdeletions or duplications that we've not been diagnosing prenatally for decades, and now with microarrays there's a way to do so by looking at much smaller parts of DNA."

The researchers also found that microarray results generally were available sooner than those of karyotypes, owing to most microarray specimens not needing to be cultured, which is necessary for karyotyping. Initially they performed microarray analysis on paired cultured and uncultured specimens to see whether the different procedures led to different diagnostic yields, 259 in the case of chorionic villus sampling and 275 with amniotic fluid. Finding comparable results, the investigators subsequently used culture with microarray only when they encountered difficulties with uncultured samples. They plan to publish a report about their lab methods.

Furthering the appeal of microarrays, the methods cost about the same, $1,500 to $2,000 for microarrays and $1,000 to $1,500 for karyotyping, according to one estimate.

The Problem With Balance

Microarrays fell short of karyotyping only in detecting triploidies and balanced translocations, which the investigators anticipated going into the study, since these variations cause no gain or loss in genetic material and microarray technology is designed to detect copy number changes.

"We knew there might be cases where a balanced translocation would be identified by karyotyping which microarray would not detect. But our study was trying to identify chromosome imbalances that could cause human disease, and if there's no imbalance, most likely that wouldn't cause any issues," explained co-author Christa Lese Martin, PhD, FACMG, director and senior investigator of Geisinger Health System's Autism and Developmental Medicine Institute in Lewisburg, Penn.

She added that balanced translocations usually don't cause abnormal clinical outcomes. "It's something we look for in couples with infertility, but that's a different indication than what this study was being done for. In fact, most cases of balanced translocations in our study were inherited from a normal parent, giving more evidence that they were not related to phenotype."

At the time the study took place, Martin was an associate professor of human genetics at Emory University School of Medicine in Atlanta, and her lab was one of the four that performed the study's microarray analysis.

The Troubling Issue of VOUS

While microarrays came out looking pretty good in the studies, that does not mean they are a perfect diagnostic vehicle. One issue Wapner's study brought to light is variants of unknown significance (VOUS). Because they are at significantly higher resolution than karyotypes—50–200 Kb versus 3–10 Mb—microarrays detect submicroscopic gains, losses, and rearrangements, compared with karyotypes' broad brush look at chromosome size and structure.

With the genome far from completely annotated, microarray-detected VOUS were expected at the outset, and indeed, initially 3.4% of cases had VOUS that karyotyping also did not pick up. With genomic variant data being added daily to the literature, the authors toward the end of the study reanalyzed the original VOUS against updated databases and were able to whittle them down so that only 1.4% of cases had such findings. As genomic discoveries continue, VOUS should be even less of an issue going forward, experts agreed.

In the meantime, VOUS pose challenges for clinicians and patients alike. Clinicians not only need to make sense of any VOUS themselves but also to educate patients about the likelihood of each method to detect VOUS. "The main concern is the potential anxiety that VOUS causes or the potential of patients making decisions about whether to continue a pregnancy based on diagnostic test results," explained David T. Miller, MD, PhD, a clinical geneticist at Boston Children's Hospital and assistant professor of medicine at Harvard Medical School. "If you're delivering ambiguous results it makes it very difficult for both the physicians and the patients themselves to make those management decisions." (See Box, below).

Microarrays Versus Karyotyping in Prenatal Diagnostics
What Should Expectant Parents Know?

Microarrays have become the preferred lab method to help diagnose in infants and children developmental delays, intellectual disability, and autism spectrum disorders, and now research has demonstrated their utility in the prenatal realm. This technology, which has significantly higher resolution than the current diagnostic standard, karyotyping, raises the specter of expectant parents being informed about their fetus carrying genetic variants of unknown significance (VOUS). As microarrays make their way into routine clinical practice, clinicians, laboratorians, and geneticists will need to collaborate closely about how to educate themselves and parents about this new wealth of information, according to experts.

"There needs to be a very active discussion of how much information one should get in a prenatal test because with microarrays you'll begin to be able to identify adult-onset disorders, things that will have a much milder phenotype," explained Ronald Wapner, MD, professor, director of reproductive genetics, and vice chair of research in the department of obstetrics and gynecology at Columbia University in New York City, and the principal investigator of a major study comparing microarrays against karyotyping in prenatal diagnosis of congenital anomalies.

Although the microarray-versus-karyotype debate brings the VOUS issue into sharper focus, it is not a new concern in prenatal diagnostics, said David T. Miller, MD, PhD, a clinical geneticist at Boston Children's Hospital and assistant professor of medicine at Harvard Medical School. "There's always going to be a little bit of that, the balance between maximum diagnostic yield and minimum VOUS. That's even true with karyotypes, and it doesn't mean that people don't use the test. It just means that with microarrays as with karyotyping, they'll have to use it with their eyes wide open."

Even if the VOUS issue did not arise solely from microarrays, the wealth of information available from this method calls out for an improved across-the-board understanding of genetics, contended Aaron Caughey, MD, PhD, chair of the department of obstetrics and gynecology and associate dean for women's health research and policy at the Oregon Health and Science University in Portland. "With microarrays and other ongoing genomic research, new microdeletions are going to be discovered every day. Physicians will be faced with telling patients, ‘Your baby is missing a piece of DNA in the region where these seven genes are, but we don't know what it means to have that missing DNA.' There's going to be a lot of that going on and it's going to be very hard to deal with," he suggested. "Physicians barely understand this, so helping patients understand it is going to be very challenging. It calls out for genetic literacy for all."

Wapner suggested that while microarray technology brings out more VOUS than karyotyping in the prenatal setting, it doesn't fundamentally change patients' perception of or ability to process uncertainty. "The fear of giving people these unknown findings is it does create some anxiety. But it's pretty amazing how smart people are and that they understand about not knowing things," he observed. "So there's absolutely no reason that we shouldn't make the switch from karyotyping to microarrays, but we have to do it taking into consideration not just the technical ability but all the other social, psychological, and practical implications."

A Maturing Technology

While Wapner and his colleagues made a big splash with microarrays in prenatal diagnostics, their findings in some ways merely reflect the maturation of this technology. For several years, microarrays have been the preferred method of postnatal diagnosis of genetic abnormalities involved in developmental delays, intellectual disability, and autism spectrum disorders. In fact, the American College of Medical Genetics and Genomics (ACMG) in 2010 published a practice guideline endorsing microarray as the first-line test in the work up of anomalies not clearly tied to well-described syndromes, developmental delay and intellectual disabilities, and autism spectrum disorders.

Despite this positive view of microarrays, the maternal-fetal medicine community needed more robust evidence to use the method routinely in prenatal diagnostics, according to Martin. "It became very clear from using microarrays in the postnatal setting that they could detect not only rearrangements or imbalances that were detectible by karyotype but also sub-microscopic rearrangements that weren't readily noticeable or even detectible by karyotyping. But a person presenting with some phenotype is very different than the prenatal setting where there may be an abnormal serum screen but not abnormal ultrasound findings," she explained. "Prenatal microarrays not only have a different indication, but also different specimen types than postnatal. With these differences, the field was ready to see whether microarrays would be superior to karyotyping in the prenatal setting."

Whether this new evidence will lead the American College of Obstetricians and Gynecologists to change its prenatal testing recommendations remains to be seen. In 2009 the group endorsed microarray analysis as a second-line diagnostic when ultrasound results are abnormal but karyotype is normal. Yet leading centers like Wapner's already are offering microarray testing to all patients scheduled for invasive procedures, whether or not an abnormality has been detected on ultrasound.

Further evidence of the utility of micro-array technology was published in the same New England Journal of Medicine issue as Wapner's and his colleagues' study, in which researchers at NICHD reported that microarray testing outperformed karyotyping in providing genetic diagnosis in stillbirth (N Engl J Med 2012;367:2185–93). They found that microarrays had a higher diagnostic yield than karyotyping—87.4% versus 70.5%—and detected more genetic abnormalities, including both aneuploidy and pathogenic copy-number variants, 8.3% versus 5.8%.

The End of Karyotyping?

If these studies seem to sound the death knell for karyotyping, experts—Miller among them—cautioned not to count out this method anytime soon. "There are still going to be complex rearrangements that need to be evaluated with traditional techniques like karyotyping and fluorescence in situ hybridization. There are going to be indications where a traditional karyotype is still the best test for conditions like Turner syndrome, Down syndrome, and in couples with a history of multiple miscarriages where you suspect there's a rearrangement that's leading to an imbalance," he explained.

Experts also emphasized that microarrays are not a prenatal diagnostic panacea. In a third study published in the New England Journal of Medicine, Harvard researchers reported prenatal diagnosis of CHARGE syndrome—colomboma of the eye, heart anomaly, atresia of the choanae, retardation, and genital and ear anomalies—using whole genome sequencing (N Engl J Med 2012;367:2226–32). The authors emphasized that ultrasound, karyotyping, and microarray all would have missed this problem.

The Rise of Non-Invasive Testing

If microarrays with rare exceptions outperform karyotyping in prenatal diagnosis, does this method have the same 5-decade staying power as karyotyping? In a word, no. The experts who spoke with CLN emphasized that microarrays one day will be sidelined by non-invasive testing methods like the one pioneered by Y. M. Dennis Lo, DM, DPhil, which use massively parallel sequencing of cell-free fetal DNA in maternal plasma to detect trisomy 21 (CLN April 2011). Already, commercial tests based off this amazing technology are available, and all indications are that this form of testing has taken off like a rocket, to the chagrin of maternal-fetal medicine, laboratory, and genetics specialists. Their concern is that microarrays are being bypassed inappropriately in favor of non-invasive testing. Rapid adoption of non-invasive prenatal testing prompted the ACMG in April to issue a policy statement, pointedly emphasizing that these tests are screening, not diagnostic, in nature, and identifying at least 10 limitations of the technology (CLN May 2013).

"Non-invasive prenatal testing is seductive. Patients think they'll be able to know all they want to learn about their fetus without having an invasive procedure. But with non-invasive testing right now we're talking about learning about just five chromosomes versus with microarray getting the sub-microscopic changes and other copy number variants that we know are linked to disease," said Martin. "A lot of people are calling non-invasive testing the new karyotype, but it's really a step backward. Microarrays are the new karyotype—for now."

In a recent commentary on the new era of non-invasive prenatal testing authors Stephanie Morain, MPH, Michael Greene, MD, and Michelle Mello, JD, PhD, argued that cell-free fetal DNA testing "seems to be drifting into routine practice ahead of the evidence" (N Engl J Med 2013;doi:10.1056/NEJMp1304843). They suggested that this is partially due to test providers building up consumer demand for the tests in the absence of a requirement by the U.S. Food and Drug Administration to provide evidence of clinical utility.

Wapner agreed that non-invasive prenatal testing is beginning to outpace its current diagnostic utility. However, he suggested that by the time this technology matures, experience gained from working with microarrays will have prepared the field to make the best use of it. "In the long run I believe that sequencing will replace microarrays. But we've learned from our study and from working with microarrays all the problems, all the issues, so that as we move into sequencing we've begun to understand how introducing genomics information into prenatal testing should be done."

The Role of Labs in Prenatal Testing

The shifting sands of prenatal diagnostics have enveloped the lab community as well as clinicians. Karyotyping traditionally has been the domain of cytogeneticists operating in specialized labs, owing in part to the need to culture cells prior to staining them. Now Wapner's team has demonstrated that nearly all microarray samples can be processed successfully without culture, and non-invasive prenatal testing has bolted out of the starting block. Does this mean that prenatal diagnostics might no longer be concentrated in cytogenetics? Sarah South, PhD, medical director of cytogenetics and microarray at ARUP Laboratories and associate professor of pediatrics at the University of Utah in Salt Lake City, envisions such a possibility.

"The technologies in this area and genomics in general seem to be moving so quickly that we're all students again. We will have to work more as teams, train each other, share our expertise, and think outside the box of our respective fellowships and certifications," she said. "In the past we've been defined by our technologies. But the technologies are moving so fast now that they're taking us into different areas. This is causing us to think more about how our various colleagues understand biology and medicine and what we can teach each other. Then we can apply that knowledge as a team to better utilize the technologies that are coming up."

Disclosures: Dr. Caughey receives salary/consultant fees and stock/bonds from Ariosa and Cellscape. Dr. Miller receives salary/consultant fees from Claritas Genomics.