Discussion
This study demonstrated that cfDNA screening was preferred and useful test in pregnancies with NT between 95th and 99th percentile. After the pregnant women with NT between 95th and 99th percentile were informed according to the ACOG guideline [2], cfDNA screening (52.8%) as a first tier test was preferred over IT (37.1%). The SAFE 21 Study Group explored the pregnant women’s attitude toward IT and cfDNA screening following positive combined screening [11]. As a result, women with higher NT were more likely to prefer IT, but old age (≥35 years) did not affect the preference for IT [12]. In our cohort, there was no difference in the choice of first-tier test according to the NT value (P= 0.084). However, there was a difference in the test preferences according to the NT value range. While pregnancies with NT between 95th percentile and 3mm preferred cfDNA screening (66.7%), IT (51.1%) was the most preferred in pregnancies with NT between 3mm and 3.5mm. Analysis of the test preferences based on maternal age showed no difference in cfDNA screening selection between ≥35 years and <35 years (48.9% vs. 51.1%, P =0.481). Our study found that the NT value range rather than advanced maternal age(≥35 years) affected the patient’s choice of the first tier test.
As is well known, INT is associated with chromosomal abnormalities [3, 13, 14], and previous studies reported approximately 3.7-7.1% incidence of chromosomal abnormalities in fetuses with INT [3, 15]. ACOG and SMFM recommend cfDNA screening or diagnostic testing for women with NT≥3.0mm or 99th percentile [16]. However, cfDNA screening has limited value for detecting chromosomal abnormalities in a fetus with INT. Several studies reported that the proportion of T21 was decreased as the thickness of NT increased [3, 15, 17]. Chromosomal abnormalities undetected by cfDNA screening were reported in 2%- 10% of fetuses with NT≥99thpercentile [18 -22]. Thus, diagnostic testing rather than cfDNA screening should be considered as the first-tier test in pregnancies with NT≥3.5mm. However, it is questionable whether this strategy should be applicable to pregnancies with NT between 95th and 99th percentile. Nicolaides et al. demonstrated that the incidence of T21 was 66.1% of chromosomal abnormalities in fetuses with NT between 95th percentile and 3.5mm [15]. In this study, chromosomal abnormalities were 5.6% and T21 accounted for 80% of chromosomal abnormalities. Berger et al. reported that rare aneuploidies undetectable by cfDNA screening were found in 0.8% of pregnancies with NT between 3mm and 3.5mm [12]. A large cohort study showed chromosomal abnormalities undetected by cfDNA screening but detected by classic karyotyping in 0.6% of fetuses with NT between 95th and 99th percentile [21]. Additionally, a recent meta-analysis reported accuracy of cfDNA screening in detecting > 99% of fetuses with T21, 98% of T18 and 99% of T13 with a combined false positive rate of 0.13% [23]. Therefore, the low risk of cfDNA screening was acceptable to pregnant women with NT between 95th and 99th percentile.
In the previous study, the incidence of major fetal abnormalities was 2.5% in a chromosomally normal fetus with NT between 95th and 99th percentile [21]. Ghiet et al. reported that the prevalence of major cardiac defects in the fetus with NT between 95th and 99th percentile was 2.5% [24]. Atzei et al. reporteda 2% prevalence of cardiac defects in fetuses with NT between 95th and 99th percentile, and was similar to that of patients with a family history of cardiac defects and diabetes mellitus [25]. In our study, the prevalence of major fetal structural anomalies was 2.5%, and all cases showed ventricular septal defect. Therefore, pregnant women with NT between 95thand 99th percentile should be recommended to undergo detailed, high-resolution ultrasound examinations in the second trimester.
The weakness of our study is the small number of pregnancies with NT between 95th and 99th percentile. The incidence of NT between 95th and 99th percentile was about 1.3% in our study, lower than 4% in the previous study by Souka et al. [3]. We excluded the cases with unknown pregnancy outcomes, which could have led to a lower incidence of INT and selection bias. Also, cfDNA screening targeted trisomies 21,18,13, which was relatively not difficult to identify chromosomal abnormalities morphologically, because we defined newborn babies with normal physical examination as euploidy. The inclusion of them with undiagnosed chromosomal abnormalities might bias. However, the main strength of our study is the prospective design in a large multicenter cohort, and it was possible to find out which test pregnant women with NT between 95th and 99thpercentile preferred in the clinical field.