Non-invasive prenatal testing for Down's better than standard screening

2 April 2015

Newly published research reports that the performance of non-invasive prenatal testing (NIPT) for abnormal fetal chromosome number (aneuploidy) by analysis of cell-free DNA (cfDNA) from maternal blood samples is better than that of current gold-standard screening approaches  in a routine, low-risk pregnant population.

Thus far, most large-scale studies of NIPT for aneuploidy using fetal cfDNA from maternal blood have taken place in high-risk antenatal populations – women whose risk has been assessed as high on the basis of maternal age, previous medical history or routine screening results. The technique has demonstrated high specificity and sensitivity in these studies.

However, the real test is how well it performs in unselected antenatal populations, the hope being that more accurate screening by this method could reduce the number of false positives, suspected cases that on invasive testing are found to be normal. Invasive forms of fetal chromosomal analysis carry an associated risk of miscarriage and reducing the need for them would be of real value.

A previous US-based study of 1,900 women found that cfDNA performed better than standard screening approaches. Now a much larger study has found similarly encouraging outcomes.

Writing in the NEJM, researchers report from the US Noninvasive Examination of Trisomy (NEXT) study, comparing the performance of NIPT with normal first-trimester screening (biochemical serum analysis plus measurement of nuchal translucency via fetal ultrasound). The NIPT used was Ariosa Diagnostics’ Harmony test, and performance for identifying trisomies 21, 18 and 13 were considered.

Results from almost 16,000 women were examined. Overall, the cfDNA testing correctly identified 38/38 cases of trisomy 21 compared with just 30/38 for standard screening. False positive rates were 0.06% in the cfDNA group and 5.4% in the standard-screening group. The researchers therefore conclude that cfDNA testing for trisomy 21 had higher sensitivity, a lower false positive rate, and higher positive predictive value than standard screening.

For trisomy 18, cfDNA testing identified 9/10 cases compared with 8/10 cases for standard screening, but the number of false positives were one and 49, respectively. Similarly for trisomy 13, cfDNA testing identified 2/2 with one false positive, and standard screening 1/2 but with 28 false positive results.

Overall, the positive predictive value (PPV) of testing - that is, the probability that a positive test result is true - was 80.9% for cfDNA testing compared with 3.4% for standard screening, a difference of around 24-fold. 

The study had limitations – around 3% of cfDNA tests failed to generate results due to inadequate levels of fetal DNA, a phenomenon associated with maternal obesity and also with fetal aneuploidy; it also used only first trimester serum screening as a comparator, whereas combined first and second trimester serum screening performs better. However, it seems clear that cfDNA nevertheless offers significant benefits for testing for common forms of aneuploidy, and the capacity to substantially reduce levels of un necessary invasive testing

An editorial accompanying the publication by UK expert Professor Lyn Chitty notes that ‘ test uptake, economic aspects, and clinical utility will depend on local cultural and societal factors, including attitudes with respect to disability, laws around termination of pregnancy, and the existing health care structure’ as well as current screening programmes. Indeed, the relative ease and low cost of cfDNA testing compared with fetal ultrasound and nuchal translucency measurement might allow some less well-resourced health systems to introduce it and make substantial improvements in antenatal screening.

The authors themselves caution that maternal serum and nuchal translucency screening ‘can identify risk for a broad array of abnormalities that are not detectable on cfDNA testing’, including rarer forms of aneuploidy for which the performance of cfDNA is considerably less well characterised, as well as other chromosomal and physical abnormalities.

More from us

Genomics and policy news