Comment on “Trisomy 21 noninvasive prenatal testing for twin pregnancies”

Abstract

We read with great interest the article by Wald et al. entitled “Prenatal screening for Down syndrome in twin pregnancies: Estimates of screening performances based on 61 affected and 7302 unaffected twin pregnancies.” Using real twin pregnancies data, they were able to establish that β‐hCG and PAPP‐A markers were in different ranges from what could be theoretically expected from singleton pregnancies data extrapolation. They provide simulated results of Down syndrome screening with test performances being better than for singletons in monochorionic cases but clearly lower for dichorionic pregnancies. However, Wald et al. approach still has an important false positive rate leading to numerous invasive procedures while the risk of miscarriage from invasive testing is known to be higher in twin pregnancies compared to singletons. Noninvasive prenatal testing (NIPT) in singleton pregnancies has a detection rate of 99.2% for trisomy 21 with a false positive rate of 0.09% but test performances are still under debate for twin pregnancies. Lyon University Hospital Genetics department and Eurofins Biomnis laboratory have used semiconductor sequencing and short‐read sequencing by synthesis technologies respectively for low pass whole genome sequencing. Between October 2014 and September 2017, 1566 tests were performed for twin pregnancies as a first‐tier test (8% of total activity). Mean maternal age was 33.4 years, significantly different than for singletons (35.8 years) (p < 0.001). Mean age of pregnancy was also significantly lower than for singletons (16.1 vs 16.8 weeks of gestation (p < 0.05). This is likely due to the position of NIPT in the validated French screening strategy after first trimester screening results and not as a first‐tier test for singletons. Of the fourteen positive tests for trisomy 21 (0.89%), eleven have been confirmed through karyotyping. Two cases were not confirmed including a dichorionic pregnancy with one fetal demise during testing. One patient refused invasive testing and was lost for follow‐up. The frozen plasma sample of the single false negative test was retested on the other platform blindly and was positive for trisomy 21. No test failure was observed. Overall, positive (91.7%) and negative (99.9%) predictive values of the test for twin pregnancies were not statistically different from what we observed for singletons (98.7% and 99.99% respectively) (n = 18,298) and are above every simulation made by Wald et al. using combined screening.