3 points about the Illumina Down syndrome prenatal testing study

verinata months ahead of the curveA study released in February 2014 did lead to non-invasive prenatal screening (NIPS) being offered to all expectant women. Here’s three points about the new study:

  1. The study compared the positive predictive values for traditional screening, i.e. nuchal translucency combined, quad, and integrated screening, with NIPS in a general population of expectant mothers. The study showed the positive predictive value for NIPS was far better than traditional screening. If the test returned suggestive of Down syndrome, a NIPS result meant you were more likely to have a child with Down syndrome than if the test was traditional screening. However, the positive predictive value of NIPS was still only 45.5%, meaning a positive NIPS result still meant you were more likely to not have a child with Down syndrome. For this reason the authors noted:

The positive predictive values of cfDNA testing (45.5% for trisomy 21 and 40.0% for trisomy 18) underscore the need for follow-up diagnostic testing to confirm true positive results before decisions are made about irrevocable clinical intervention and to resolve discordant results.

I’ll let you translate what “irrevocable clinical intervention” means.

2. The study concludes that:

Our findings, however, suggest that cfDNA testing merits serious consideration as a primary screening method for fetal autosomal aneuploidy.

    But, as covered in the first bullet, the study’s findings are simply that:

In a general obstetrical population, prenatal testing with the use of cfDNA had significantly lower false positive rates and higher positive predictive values for detection of trisomies 21 and 18 than standard screening.

The key distinction being the study was of the general population, not specifically the low-risk population. Professional guidelines already allow for cfDNA to be offered to mothers considered high risk. So, the only remaining question is whether cfDNA is a reliable screen for the low risk population. As shown in studies of the low risk population, cfDNA does not have the same stated accuracy as it does for the high risk population. The new study, however, only reports numbers for the general population, not the low-risk population. Of the tested population, there were five pregnancies actually pregnant with a child with Down syndrome, but there were six more falsely identified as positive for Down syndrome. If those 6 false positives were all in the low-risk population, then the study’s finding could instead support the existing recommendations of using cfDNA as a secondary screen, not a primary screen.

3.  Finally, the study discloses that all of the tests were done by Illumina, who offers the verifi test, and its authors receive fees from Illumina. It should not be surprising then that the conclusion is to suggest that tests like Illumina’s verifi be offered to more women, not just those who are high risk.

UPDATE: This post has been updated (3/2/16). Since Illumina’s study was published, the ISPD updated its recommended screening protocol to include cfDNA screening as a primary screen and ACOG has recognized that women may choose cfDNA screening, even though they are not considered high risk. ACOG still recognizes that conventional screening is most appropriate for low-risk pregnancies.


  1. Did I read recently that Giana Bianchi is now with Illumina? They sound like the Illuminanti.

  2. varoutsi says:

    we are in May 2014 now and women are doing this test every day. what do the latest results show in terms of number of false negative ?

    • The labs have not reported the results of their commercial experience, only from structured studies. See my post here and upcoming posts on posters presented at ACMG which suggest lower accuracy rates as more people have the testing.


  1. […] covered in this post, the New England Journal of Medicine published a study by researchers affiliated with the NIPS lab […]

  2. […] Do you see a theme in these findings? The prevalence of misunderstanding based on misconceptions–often prompted by the marketing of NIPS labs–about the accuracy of NIPS; the need for counseling on these limitations; the lack of resources needed to counsel patients; and the absolute need to confirm through diagnostic testing “before decisions are made about irrevocable clinical intervention.“ […]

  3. […] accuracy in the general population whose lead author was also the lead researcher for verifi stated […]

Leave a Reply

%d bloggers like this: