NEW YORK (GenomeWeb) – Noninvasive prenatal testing (NIPT) for trisomies and other chromosomal aneuplodies using cell-free DNA from the mother's blood has been spreading quickly in maternal-fetal medicine in recent years and is poised to move from high-risk to average-risk pregnancy populations.
Three new studies, published online in the New England Journal of Medicine today, shed light on how these tests may be best integrated into clinical care for all pregnant women.
The risk of false-positive results from NIPT, which may lead to irreversible decisions of termination if women do not seek diagnostic confirmation, is still substantial, according to a study from researchers at Baylor College of Medicine, the Chinese University of Hong Kong, and Hong Kong University. They caution that despite their high sensitivity and specificity, the new tests should only be used for screening and not to replace confirmatory diagnostic karyotyping, a position that is supported by recommendations from the American College of Obstetricians and Gynecologists and the Society for Maternal-Fetal Medicine.
Another study, led by a team at the University of Washington, delved into one of several reasons why NIPT may yield false-positive results — the presence of copy number variants in the mother's genome.
Despite these shortcomings, NIPT trumps standard first-trimester screening for trisomy 21detection in all types of pregnancies, according to a large clinical study involving test results from almost 16,000 women that was conducted by investigators at the University of San Francisco, Columbia University, Ariosa Diagnostics, which was acquired by Roche this year, and others. Results from that study were already presented a year ago at the American College of Medical Genetics and Genomics annual meeting but were formally published today.
At this year's ACMG meeting, held in Salt Lake City last week, Ariosa CEO Ken Song provided a cost-effectiveness analysis, making the case that replacing current first-trimester screening for trisomies with Ariosa's test in the general pregnancy population could provide cost savings if the price of the test was under $450.
To study the performance of various noninvasive prenatal tests, the Baylor and Hong Kong researchers focused on 307 women from three centers who had received positive NIPT results for trisomy 21, 18, 13, monosomy X, XXX, XXY, or XYY and were referred for confirmatory invasive diagnostic testing.
Results included those from Sequenom's MaterniT21 and SafeT21 tests, BGI's NIFTY, Illumina's Verifi, Natera's Panorama, and Ariosa's Harmony. A single result came from LabCorp's Integrated Genetics InformaSeq test.
For 13 of the cases, karyotyping of cells from amniocentesis or chorionic villus sampling revealed mosaicism. Of the remaining 294 NIPT-positive cases with nonmosaic karyotype, 56, or 19 percent, turned out to be false positives.
For trisomy 21, the most commonly detected condition, the false positive rate was 9 percent, followed by 17 percent for XXX, XXY, or XYY; 23 percent for trisomy 18; 46 percent for trisomy 13; and 62 percent for monosomy X.
While there is a chance that some of the diagnostic karyotyping results were incorrect, none of the centers were informed of any discrepancies after the babies were born.
The Chinese University of Hong Kong researchers did note, however, that at least eight pregnancies with false-positive NIPT results — most for trisomy 21 — resulted in the birth of healthy babies.
There are several reasons why NIPT — which currently all assay circulating cell-free DNA in the mother's blood, but differ in how they analyze the material — may deliver false-positive results, including somatic maternal mosaicism, placental mosaicism, vanishing twins, undetected cancer in the mother, and copy number variants in the mother.
The Washington University researchers explored the last reason in greater detail. For their study, they analyzed blood samples from four women who had received positive results from Illumina's Verifi test, three for trisomy 18 and one for trisomy 13, but delivered healthy babies.
In two of the patients who received positive NIPT results for trisomy 18, they identified duplicated regions on chromosome 18. These duplications likely confounded the analysis, which is based on an excess of sequencing reads for chromosome 18 over reads for other chromosomes.
Most of the cell-free DNA in the blood derives from the mother, and a maternal duplication increases the number of reads from that chromosome, boosting the chance for a false-positive result, whereas a chromosomal deletion decreases the number of reads, increasing the probability of a false-negative result.
The size of the effect depends on the fetal fraction of cell-free DNA, the size of the duplication or deletion, and whether or not it is inherited by the fetus. For one of the patients, who had a 1.15-megbase duplication on chromosome 18 that was also present in the fetus, the researchers calculated that a false-positive result was 16,000-fold more likely, so "in the absence of fetal aneuploidy, the test was nearly equivalent to flipping a coin."
Based on the frequency of copy number variations in the general populations, the researchers estimated that they "may contribute substantially" to the risk of obtaining a false positive result from NIPT.
But the risk can be reduced, they wrote, by taking maternal copy number variants into account. For example, once such a variant has been detected, reads from that region could be at least partially discarded, or the effective size of the chromosome could be increased or decreased as needed.
Also, once a copy number variant has been detected in a woman, it could be taken into account for all of her future pregnancies.
Like the Baylor and Hong Kong study, the authors concluded that despite the good performance of NIPT compared to conventional first trimester screening for trisomies, "follow-up diagnostic testing remains essential."
"The effects of false-positive screening results go beyond the clinical risks and financial costs of diagnostic testing and include potentially substantial psychological stress for patients," they wrote.
While caution may be called for in how NIPT is implemented, evidence is mounting that it is superior to standard first trimester screening for trisomy 21 in pregnancies both high-risk and low-risk.
In their study of almost 16,000 women, the UCSF, Columbia University, and Ariosa researchers found that Ariosa's Harmony test had better sensitivity, a lower false-positive rate, and a higher positive predictive value for trisomy 21 than standard screening with nuchal translucency and biochemical analysis.
The researchers enrolled 18,955 patients from 35 centers in six countries. More than 3,000 samples were excluded for a variety of reasons, among them 488 where NIPT yielded no result because of low fetal fraction, assay variation, or assay failure.
The mean age of the remaining 15,851 women was 31 years, and they included both low-risk and high-risk pregnancies. Of these, 38 carried fetuses with trisomy 21, which were all discovered by the Harmony test, for a sensitivity of 100 percent, whereas standard screening only picked up 30 of the cases, translating to a sensitivity of 79 percent.
However, the analysis did not account for the 488 cases where NIPT had yielded no results. As it turned out, those cases included 13 undetected aneuploidies, among them three cases of trisomy 21. Including those cases in the analysis would have lowered the detection rate of the test, the authors acknowledged, or, if they were all counted as positives, it would have increased the number of false positives.
Further study is needed to determine the best approach for no-calls, the authors wrote, such as repeat testing, adjusting the initial test for maternal weight, adding screening by another approach, and recommending invasive diagnostic testing for women with no results from NIPT.
NIPT also identified nine false positive trisomy 21 cases, for an overall false-positive rate of 0.06 percent. Standard screening, on the other hand, identified 854 false-positive cases, for a false-positive rate of 5.3 percent.
The positive predictive value of the Harmony test was 80.9 percent, comparing favorably to just 3.4 percent for standard screening. For women under the age of 35 only, the positive predictive value of NIPT decreased to 76 percent, and for women considered low-risk based on standard screening, it was 50 percent, still outperforming current first trimester screening.
The researchers also analyzed the test for trisomy 18 and 13, though the study was not sufficiently powered for this. Of 10 true cases of trisomy 18, Harmony detected nine and it delivered one false-positive result, for a positive predictive value of 90 percent, compared to eight detected cases and 49 false-positive results from standard screening, for a positive predictive value of 14 percent.
Harmony also detected the two cases of trisomy 13 and delivered one false-positive result, whereas standard screening identified one of the two cases and delivered 28 false-positive results.
Overall, the researchers concluded that NIPT represents "a highly accurate screening option" for trisomy 21 and other common aneuploidies but cannot identify the risk for "a broad array of abnormalities" that standard screening picks up. For example, trisomy 21 cases only accounted for a little more than 50 percent of all aneuploidies present in their test population.
While the results support the use of NIPT in women regardless of age or risk, "further cost-utility studies are warranted," the authors wrote.
In a presentation at the ACMG annual meeting in Salt Lake City last week, Ariosa CEO Ken Song took a first stab at such an analysis.
Assuming a 70 percent screening adherence rate in a population of 4 million pregnant women — the approximate number of annual births in the US — he calculated that standard first trimester screening, also called first trimester combined screening or FTS, would detect about 7,800 cases of trisomies 21, 18, or 13, whereas NIPT would detect almost 9,000.
Because of its higher false-positive rate, FTS would also lead to about 1,400 mostly unnecessary invasive procedures, whereas NIPT would result in fewer than 100 such procedures.
Song calculated that NIPT could help save overall costs per trisomy case identified, compared to FTS, if it was priced at $453 or less. This calculation assumes prices of $171 for FTS, $1,300 for invasive diagnostic testing, $120 for an office visit, and various prices for pregnancy termination and for the birth of a child affected with one of the trisomies, as well as an average termination rate of 75 percent for trisomy 21 and 90 percent for trisomies 13 and 18 in the US.
The price for NIPT could be higher while maintaining the same cost savings, however, if the screening rate in the population increased to more than 70 percent, he added.
In an editorial on the Ariosa study, Lyn Chitty from the UCL Institute of Child Health and the Great Ormond Street Hospital for Children NHS Foundation Trust in London wrote that the uptake of NIPT will eventually "depend on local cultural and societal factors, including attitudes with respect to disability, laws around termination of pregnancy, and the existing health care structure," such as existing screening programs.
Also, the use of NIPT to screen for other chromosomal abnormalities, such as sex chromosome aneuploidies, microdeletions, and duplications will require further validation, Chitty wrote.