Maternal Duplication Mimics Fetal Trisomy in DNA Screen

Ricki Lewis, PhD

April 15, 2015

Noninvasive prenatal testing (NIPT), which relies on cell-free fetal DNA in the mother's blood, is increasingly common. Yet two studies published online Apri 1 in the New England Journal of Medicine highlight the risk for false-positives, particularly in low-risk women.

In one study, by Matthew Snyder, MS, a PhD candidate in the laboratory of Jay Shendure, MD, PhD, from the University of Washington, Seattle, and colleagues, the authors uncover the source of false-positives in two pregnancies. The authors note that the overall low rate of aneuploidy in a low-risk population limits the positive predictive value of the tests. Because of that limited positive predictive value, another research team, led by Sau W. Cheung, PhD, from the Baylor College of Medicine, Houston, Texas, argue that the procedure should be clearly described as screening to emphasize the need for confirmatory testing of any putative anomalies found through cfDNA testing.

NIPT has been used since 2011 to identify pregnancies at high risk for fetal aneuploidy, and several companies offer the tests. As reported previously by Medscape Medical News, the medical community has more recently accepted the value of the test for use in lower-risk population.

Maternal Duplications Increase False-Positives

To better understand what drives false-positives, Snyder and colleagues examined four pregnancies in which the cfDNA test suggested a trisomy but amniocentesis, newborn exam, or both were normal. Three of the four patients had fetuses with trisomy 18 according to the cfDNA test, and one had a fetus that the test indicated had trisomy 13.

The researchers found that maternal duplications were responsible for the false-positive in two of the three cases of putative trisomy 18 cases. The maternal duplications were 1.15 Mb (patient 1), and the other was just 487 kb (patient 3). The authors did not identify the reason for the remaining two false-positives.

In the second part of the study, the investigators tested how the size of a duplication affects the likelihood of false-positive NIPT results. On the basis of those data and a reference panel of 19,584 people, mostly European, they estimate that a 1.15-Mb duplication, which the fetus inherited, increased the probability of a false-positive by a factor of 15,650 and makes "the test...nearly equivalent to flipping a coin," the authors write. The smaller duplication, which was not inherited by the fetus, increased the probability of a false-positive by a factor of 128 to 262, depending on the proportion of fetal DNA obtained in the testing sample.

Snyder and colleagues note several study limitations, including reliance on a European population as a reference standard and evaluating only four cases.

Confirmatory Testing Needed

In the study by Dr Cheung and colleagues, the researchers investigated the false-positive rate among women from the general population, based on samples referred to a clinical laboratory for cytogenetic analysis after amniocentesis to confirm a positive cfDNA result.

After ruling out cases of mosaicism, Dr Cheung and colleagues found that 238 (81%) of 294 cases were true positives, in which the fetus carried a trisomy. However, they also found that 16 (9%) of 177 initial reports of trisomy 21 were false, as were 12 (23%) of 52 cases of trisomy 18 and 12 (46%) of 26 reports of trisomy 13. In addition, 13 (62%) of 21 results of monosomy X and 3 (17%) of 18 reports of XXX, XXY, or XYY were falsely positive on initial cfDNA testing.

Their outcomes contrasted reports of high detection rates previously reported for NIPT (99.0% for trisomy 21, 96.8% for trisomy 18, and 92.1% for trisomy 13), according to the authors.

They note that they have heard of anecdotal cases in which a woman terminated a pregnancy because of a NIPT test later found to be a false-positive. Given their new results and other data, Dr Cheung and colleagues emphasize the need for confirmatory testing. "As recommended by the American College of Obstetricians and Gynecologists and the Society for Maternal-Fetal Medicine, positive findings on noninvasive prenatal screening must be followed by invasive prenatal diagnostic testing before any irreversible decisions are made," they conclude.

False-Positives in Context

The problem of false-positives is particularly relevant in a low-risk population, Snyder explains. "For a test for a disease with reasonably high prevalence, a small number of false-positives or [false-]negatives is a drop in the bucket. But for a situation like prenatal screening in a low-risk population, with a small number of true positives, positive predictive value is really impacted by a small number of positive tests," Snyder told Medscape Medical News.

In 2014, Diana Bianchi, MD, executive director, Mother Infant Research Institute, Tufts University, Boston, Massachusetts, and colleagues reported a 45.5% positive predictive value for detecting trisomy 21 from cfDNA in an average-risk population ( N Engl J Med. 2014:370:799-808).

"That means that a positive test result is roughly equivalent to flipping a coin," Snyder said.

However, cfDNA screening is extremely accurate in detecting when trisomy 21 is not present, and thus spares women needing the more invasive procedures, Dr Bianchi told Medscape Medical News. That is its primary value.

Despite the new findings, the future for cfDNA testing is bright, Snyder and Dr Bianchi said. Negative results spare women from more invasive testing, and the approach has better specificity and sensitivity than maternal serum screening, especially for trisomy 21. Cell-free DNA screening is also safer, albeit less informative, than amniocentesis and chorionic villus sampling. In addition, the price of DNA sequencing continues to fall, so the price of NIPT might do so as well.

Snyder sums up the state of affairs thus: "We are going to move toward a world where cfDNA-based tests are commonplace, whether it will be population screening for any pregnancy or will include a calculation to look for elevated risk factors first using ultrasound or serum screening."

Dr Bianchi calls the Snyder paper "really interesting and not surprising," and cautions that patients not be too concerned. "In a clinical setting, not investigational, 0.2% of all cases that have aneuploidy detected are false-positive. So patients should realize that false-positives command a lot of attention, but are a relatively rare situation."

One of Snyder's coauthors reports receiving fees for being on an advisory board for Pacific Biosystems, DNA Nexus, and SynapDX Corp. Another coauthor reported being a scientific adviser to Ariosa Diagnostic and has a research agreement with Illumina. Snyder and the other authors have disclosed no relevant financial relationships. Dr Cheung and colleagues have disclosed no relevant financial relationships. Dr Bianchi reports serving as a director, officer, partner, employee, advisor, consultant, or trustee for Illumina Inc and receiving a research grant from Illumina.

N Engl J Med. Published online April 1, 2015. Snyder full text, Cheung full text


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.