No safety differences between cfDNA screening and diagnostic testing?

One of the major selling points for prenatal cell-free DNA (cfDNA) screening has been that it will reduce the number of miscarriages resulting from invasive diagnostic tests–amniocentesis and CVS. However, a randomized controlled trial in France, just published (paywall) in the Journal of the American Medical Association, has found no evidence of this promised benefit.

The SAFE21 Study Group randomized “high risk” pregnant women in 57 French clinics to either go straight to invasive diagnosis or, alternately, to have cfDNA screening first. Those in the cfDNA arm of the trial received invasive testing only if the cfDNA screen returned positive results. There were just over 1000 women assigned to each arm of the trial. In the cfDNA arm, nearly all the women had cfDNA screening, and 84 women (8.3%) had follow-up invasive diagnosis. Yet the miscarriage rate for the cfDNA arm was identical to that for the invasive testing arm, in which 751 women (76.5%) had invasive diagnosis. In each arm 8 women (0.8%) experienced miscarriage. Furthermore, in the invasive testing arm, researchers discovered 11 chromosomal anomalies (1.5%) that would not have been found using cfDNA screening.

Research has shown that the risk of test-related miscarriage from amnio and CVS has often been overestimated based on early studies when the procedures were still somewhat new. The most recent ACOG guidance on invasive diagnostic testing (Practice Bulletin 162) notes that procedure-related losses likely occur in only about 1 of 900 cases (0.11%). Both amnio and CVS require skill, and safety rates are much higher when the procedures are performed by an experienced provider. Some researchers have even suggested that the risk is practically zero when conducted by well-qualified providers.

It’s important to note, too, that miscarriage rates are higher among pregnancies affected by a chromosomal condition such as trisomy 21 (Down syndrome). Some miscarriages that have been blamed on invasive tests may actually have been due to the fetus’s underlying condition.

So what are we to make of the lack of evidence in this study that cfDNA is safer than amnio or CVS? There are (at least) 3 takeaways here, from our point of view:

1. Invasive testing really is very safe, and it gives more and better information than cfDNA. In the interest of fully informed consent, doctors have to inform their patients that there’s a tiny but measurable risk of test-related loss, but under the care of an experienced provider, this risk approaches zero. While cfDNA labs capitalize on the scariness of the amnio needle, in practice there is little to worry about.
2. Nevertheless, while over 2000 women sounds like a lot, miscarriage after invasive testing is very rare, so the authors note that this study “may have been underpowered to detect clinically important differences in miscarriage rates.”
3. But there’s one last, important takeaway here, and that’s about women’s preferences. Of the 1034 women in the cfDNA arm, all but 6 followed through with cfDNA as they were assigned. However, in the invasive testing arm, 272 of the 1017 women (26.7%) chose not to have invasive testing. In other words, even after agreeing to research in which they knew they might be asked to have invasive testing, over a quarter of these women decided not to follow through. This speaks volumes about how many women feel about invasive testing, and the undeniable appeal to them of a noninvasive option, even if it delivers less information.

While the original JAMA article is available only to subscribers, you can read the abstract at https://jamanetwork.com/journals/jama/article-abstract/2697008, or listen to the editor’s audio summary at https://jamanetwork.com/learning/audio-player/16709883.