Caroline Helwick

February 09, 2015

SAN DIEGO — Cell-free (cf)DNA analysis appears to be less effective as a primary test to detect fetal chromosomal abnormalities than sequential screening, new research has shown.

"Overall, when considering all chromosome abnormalities and including cases with no test result, sequential screening has better test performance than cfDNA," said Mary Norton, MD, from the University of California, San Francisco.

"Disorders other than trisomy 21 are important considerations in prenatal testing," she explained here at the Society for Maternal-Fetal Medicine (SMFM) 2015 Annual Pregnancy Meeting. "Consideration of all outcomes is important."

Dr. Norton and her team evaluated the detection of all fetal chromosomal abnormalities in a general prenatal cohort. They compared cfDNA with the current standard of sequential screening, which involves serum analysis and the assessment of nuchal translucency with ultrasound during the first and second trimesters.

The cfNDA test was introduced into clinical practice in 2011, and there was rapid uptake in high-risk populations. It is currently used mainly in women considered to be at risk for aneuploidy or other malformations. Recent data demonstrate, however, that the test's sensitivity and specificity are also high in low-risk patients.

Although the test "has potential as primary screening for all pregnant women," according to Dr. Norton, it does have some limitations.

It tests only for trisomies 13, 18, 21, and sex chromosomal aneuploidies, but not for some of the alterations associated with intellectual and other disabilities, she explained. In contrast, sequential screening determines risk in the presence of a broad array of fetal abnormalities.

The sequencing of fetal DNA requires that a minimum amount of fetal DNA be present in maternal blood. Low fetal DNA fraction, therefore, can lead to a failed test. The implications of a "no-result" test can be significant, increasing the odds of aneuploidy by 2.5 to 6.2, according to previous studies.

But because "there have been no prospective cohort studies that include all chromosomal abnormalities and failed tests, the implications are unknown for cfDNA as a primary screen," Dr. Norton pointed out. This study fills that gap by taking into account these limitations and characteristics.

The researchers modeled the expected performance of cfDNA as a primary screening test for aneuploidy, considering all chromosomal abnormalities, in a large population-based cohort. Cases with no result were considered to be high risk and referred for follow-up. The findings were compared with those from sequential screening.

The study involved all participants in the California Prenatal Screening Program from April 2009 to December 2012 with singleton pregnancies and available sequential screening results.

Cytogenetic abnormalities are collected by state-mandated reporting. Chromosome abnormalities considered to be detectable with cfDNA were trisomies 13, 18, 21, and sex chromosomal aneuploidies. Abnormalities considered to be undetectable with cfDNA were rare trisomies, unbalanced structural rearrangements, deletions and duplications, triploidy, marker chromosomes, and mosaicism. A no-result test was considered to be an indication of high risk for trisomy 13, trisomy 18, and triploidy.

In the cohort of 452,901 women, 74% were younger than 35 years of age. Abnormal karyotypes, detected in 2575 women, were less likely to be identified with cfDNA than with sequential screening (75.4% vs 81.6%).

Table. Detection of Abnormal Karyotypes

Abnormality n cfDNA Test Sequential Screen
Trisomy 21 1276 1263 1185
Trisomy 18 336 326 313
Trisomy 13 143 132 115
Sex chromosomal aneuploidies 256 220 185
Other 560 0 302

If no-result cases are excluded, the cfDNA detection rate is lower, at 71.4%; the undetectable rate, which includes false-negative results, is 24.5%; and the no-result rate is 4.1%. In this scenario, the difference in detection rates between cfDNA and sequential screening is highly significant (P < 0.001).

The rate of "other" abnormalities detected with sequential screening is 53.8%, which ranges from 24.2% for Robertsonian translocations to 91.0% for triploidy.

False-positive rates were 4.5% for sequential screening, 5.1% for cfDNA if no-result cases were considered high risk and referred for follow-up, and 1.0% for cfDNA if no-results cases received no follow-up.

Dr. Norton acknowledged that detection was modeled using current cfDNA screening techniques, that not all patients completed sequential screening, that outcomes of no-result cases were not typically reported, and that not all cfDNA laboratories reported the same abnormalities or used the same approach.

She concluded by emphasizing that all abnormalities are important considerations in prenatal testing, and that no-result cfDNA cases warrant counseling and follow-up.

Experts Disagree on the Use of cfDNA

"Dr. Norton did a beautiful job describing the sensitivity and the detection rates for sequential screening and cfDNA, and showing that sequential screening has higher detection rates. It's also less expensive," said Vincenzo Berghella, MD, from Thomas Jefferson University in Philadelphia, who is president of the SMFM and who moderated of the plenary session.

"At this point, we don't see an advantage to the universal use of cfDNA. ACOG recommends it be used only in high-risk patients, not the vast majority of women; this is what we do in our practice," Dr. Berghella said.

This view is not universal. For the common aneuploidies, the cfDNA test is more accurate, said Laxmi Baxi, MD, from the New York University Langone Medical Center and Columbia University in New York City.

"For prenatal testing, cfDNA has 99.0% accuracy for detecting Down syndrome and 98.9% accuracy for detecting trisomy 13," Dr. Baxi told Medscape Medical News. "And patients do not want a risk score, they want a yes or no result."

Dr. Baxi questioned the importance of identifying all abnormalities, particularly those that will lead to fetal demise. "What's the point in that? I don't want to take chances with invasive testing [such as follow-up choriovillous sampling or amniocentesis] in that circumstance," she said.

How many patients are undergoing choriovillous sampling or amniocentesis with sequential screening? "I have almost stopped doing these, because with cfDNA, I don't have to," Dr. Baxi explained.

"I believe that cfDNA should be patient's choice and should be universally available, regardless of age. Why shouldn't it be?" she asked.

Dr. Norton reports receiving research support from Ariosa Diagnostics and Natera. Dr. Berghella and Dr. Baxi have disclosed no relevant financial relationships.

Society for Maternal-Fetal Medicine (SMFM) 2015 Annual Pregnancy Meeting: Abstract 2. Presented February 5, 2015.


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.
Post as: