Clinical Application of Noninvasive Prenatal Testing for Sex Chromosome Aneuploidies in Central China

Background The relatively high incidence and the clinical symptoms of sex chromosome aneuploidies (SCAs) make prenatal screening of SCAs an attractive option for pregnant women. However, limited studies have assessed the clinical performance of noninvasive prenatal testing (NIPT) for screening SCAs. This study was performed to evaluate the clinical performance of NIPT for SCAs in singleton pregnancies in central China. Methods Noninvasive prenatal testing was performed using next-generation sequencing. Standard Z-score analysis was used to identify fetal SCAs. NIPT-positive results were confirmed by invasive prenatal diagnosis (IPD). Results A total of 42,164 pregnant women with singleton pregnancies were recruited in this study. They were divided into the following five groups with different clinical indications: with ultrasound soft index abnormalities (9.23%, 3,892/42,164); with advanced maternal age (22.14%, 9,336/42,164); with high risk for maternal serum screening (MSS) (18.35%, 7,738/42,164); with an intermediate risk for MSS (26.6%, 11,215/42,164); and with low risk (23.68%, 9,983/42,164). In all, 223 women had a high risk for SCAs by NIPT with a positive rate of 0.53%. There was no significant difference associated with the five groups in the positive rate. Of all of the positive results, 89 were 45,X (39.91%), 38 were 47,XXX (17.04%), 31 were 46,XY,del(X) (13.90%), 50 were 47,XXY (22.42%,), and 15 were 47,XYY (6.73%). Finally, 147 participants (65.92%) chose to undergo IPD, and 47 cases were confirmed. The combined positive predictive value (PPV) of NIPT for SCA was 31.97% (47/147). PPV was high for 47,XYY (100%, 11/11), moderate for 47,XXX (42.86%, 9/21) and 47,XXY (45.45%, 15/33), but low for 45,X (16.13%, 10/62) and 46,XY,del(X) (10%, 2/20). The termination rates of Turner syndrome and 47,XXY syndrome were higher than 47,XXX and 47,XYY syndromes. Conclusion In this relatively large cohort, we evaluated the value of NIPT for SCAs. Our data showed that with informed consent and subsequent professional genetical consulting, NIPT can be a useful method to screen SCAs.

INTRODUCTION risk, and limitations of NIPT. All the participants provided written informed consent and chose to undergo NIPT after understanding the purpose of the test and study. Then, post-test clinical genetic counseling was offered to the women with the suspected risk of SCAs by qualified clinical geneticists. Further IPD was carried out for high-risk pregnant women after obtaining informed consent. Finally, we assessed the clinical value of NIPT for screening SCAs from multiple aspects.

Cohort
The retrospective study was conducted from January 2016 to June 2019 in pregnant women undergoing NIPT at the First Affiliated Hospital of Zhengzhou University, Zhengzhou, Henan Province, Central China. Women were categorized into five groups based on different clinical indications. Ultrasound soft index abnormalities group (including echogenic intracardiac focus, choroid plexus cysts [uni-or bilateral cysts], pyelectasis, thickened nuchal translucency or nuchal fold, hyperechogenic bowel, absent or shortened nasal bone length, single umbilical artery); Advanced maternal age group (aged ≥ 35 years); The high-risk group for MSS (a value of ≥ 1/270 means high risk for trisomy 21

Noninvasive Prenatal Testing
Approximately 10 ml of the peripheral blood was collected in a cell-free DNA TM BCT tube (Streck, Omaha, NE, USA), according to protocol of the manufacturer. The collected blood samples were centrifuged for 10 min at 1,600 g and 4 • C. Then, the supernatant was centrifuged for 10 min at 16,000 g and 4 • C to remove any residual blood cells. Cell-free DNA (cfDNA) was extracted from the plasma using a cfDNA extraction isolation kit (Berry Genomics Corporation, Beijing, China) according to instructions of the manufacturer. The sequencing library was then constructed and purified using an NIPT library prep kit (Berry Genomics Corporation, Beijing, China). A NextSeq CN500 MPS kit (Berry Genomics Corporation, Beijing, China) was used to sequence the sample using the NextSeq CN500 sequencer (Berry Genomics Corporation, Beijing, China). At least 2 million 36-bp clean reads were generated for each sample. Sequencing results were mapped and processed using the data analysis system (Berry Genomics Corporation, Beijing, China). Chromosomes were classified as abnormal if an absolute Z score was greater than 3.

Invasive Prenatal Diagnosis
Women with positive NIPT results were offered ultrasoundguided amniocentesis. The karyotype analysis of fetal amniotic fluid exfoliated cells was performed with amniocyte culture using Giemsa-banding techniques. For each sample, more than 30 metaphase cells were analyzed at a resolution of 320 G-bands at least. The accuracy of NIPT for SCA was assessed based on the results of fetal chromosomal karyotyping. CNV-seq was performed according to standard procedures. In short, DNA extracted from fetal amniotic fluid or uncultured peripheral blood samples was fragmented. Then, sequencing libraries constructed were sequenced on the Next-Seq CN500 platform. Approximately 5 million raw-sequencing reads with 36 bp in length were generated. The results were analyzed using the previously described algorithms to detect the microdeletions and duplications (24).

Follow-Up
The positive cases were followed up within 3 months after we sent the reports, which was defined as a prenatal follow-up. A second follow-up was conducted to know the pregnancy outcomes of those cases within 12 months after delivery, which was defined as a postpartum follow-up.

Statistical Analysis
Statistical Product and Service Solutions (SPSS) software (version 21, IBM Corporation, Armonk, NY, USA) was used for all data analyses. The chi-squared test was used to test the differences of proportions for statistical significance. P < 0.05 was considered to indicate statistical significance. Positive predictive value (PPV) was calculated as the number of cases for which the results of NIPT and IPD were concordant.

Characteristics of Pregnancies for NIPT
Excluding the samples that could not be detected, 42,164 singleton pregnant women were recruited in this study (Figure 1). The singleton pregnancies included 40,092 (95.09%)

The Invasive Prenatal Diagnosis Results of Positive NIPT Cases
After the pretest clinical counseling, all 223 (100%) participants acknowledged the risks associated with SCAs in their fetuses. These positive cases were followed up within 3 months after we sent the reports. The prenatal follow-up results showed that of the 223 participants, 41 (18.39%) rejected further prenatal diagnosis; 31 (13.90%) were lost to follow-up; 4 (1.79%) chose to medically terminate their pregnancies due to other reasons including gestation-induced hypertension and structural abnormalities detected during the late trimester (Figure 1)  polymorphism or other chromosomal abnormalities (Figure 1).  Table 3). The false-negative rate for SCAs in our study was unavailable, as most of the affected children do not present obvious symptoms in their childhood and the majority of them had no neonatal karyotype information.

Detection of Sex Chromosome Mosaicism
In the 147 high-risk cases who underwent IPD, nine mosaicism cases were confirmed ( Table 3). For all the mosaicism cases, the mean ± SD maternal age was 29.6 ± 7.3 years, the mean gestational age at testing was 17.2 ± 1.7 weeks. The proportion of mosaic anomalies was between 14.00% and 83.33% (Table 4). Cases 1, 4, 6, 7, 8, and 9 chose to terminate the pregnancy. Cases 2 and 5 chose to continue the pregnancy. These two children are now developing normally at the time we followed up. Case 3 was lost to the second postpartum follow-up ( Table 4).  (Figure 2), the false-positive NIPT results were because of the background of maternal chromosome mosaicism. For case 15, the fetus had a deletion of 7.56 Mb at chromosome 18, which was de novo deletion as both of the parents were normal. The fetus was more likely to develop into 18p partial deletion syndrome in the future (25). For cases 16 and 17, they all had copy number variation with no pathogenicity. For case 16, the microdeletion of chromosome 16 was inherited from his mother with no abnormal clinical manifestations. The follow-up results showed that cases 10, 11, 12, 13, 14, 16, and 17 chose to continue the pregnancy and are now developing normally at the time we followed up. Case 15 chose to terminate the pregnancy ( Table 5).

Postpartum Follow-Up Results of the Rejected Groups and the True Positive Cases
We followed the pregnancy outcomes of the rejected groups and the confirmed true positive cases within 12 months after delivery. Of the 41 cases without IPD, 24 (58.54%) had a live birth and no visible abnormality, and 4 (9.76%) chose to terminate the pregnancies. A total of 13 (31.70%) cases lost the second followup. Of the 47 cases with true-positive SCA, 14 (29.79%) had

DISCUSSION
Noninvasive prenatal testing has high sensitivity and specificity, and it is widely used worldwide for screening trisomy 21, trisomy 18, and trisomy 13. As for SCAs, the detection accuracy varies among different research groups. The positive rate of SCA in our study was 0.53%. There was no significant difference associated with the five groups in the positive rate, which means there was no clear correlation between SCAs and these indications including ultrasound abnormalities, maternal age, and MSS. Of the 223 SCA-positive cases, 65.92% chose to undergo IPD, which was consistent with previously reported data (8,22,26). The IPD rates of SCA-positive cases were lower than that of trisomy 21/18/13 (data not shown), which was because SCAs are less severe and there exist some effective medical intervention methods for SCAs. This is consistent with the study of another research group (22). Most of the people who rejected prenatal diagnosis preferred to choose to give birth to these fetuses ( Table 6).
The overall PPV for SCAs in our study was 31.97%. The PPV for 47,XYY syndrome, 47,XXY syndrome, 47,XXX syndrome, 45,X syndrome, and 46,XY,del(X) were 100.00%, 45.45%, 42.86%, 16.13%, and 10.00%, respectively, which were similar to a previous report (8). Although the performance of PPV for SCAs was different in different studies because of varied sample sizes and the sequencing platform used, NIPT performed better in predicting sex chromosome trisomy than monosomy X in most studies (8,21,26).
In our study, the maternal peripheral blood chromosomes were detected for only a few cases according to their own accord. The maternal peripheral blood chromosomes of case 10 were proven to have triple X and the maternal karyotype of case 11 was mos 46,XXX,rob(13;14)(q10;q10) [31]/45,XX,rob(13;14)(q10;q10) [19], which led to discordant results ( Table 5). Maternal SCA (full or mosaic) is a common and important cause of discordant SCA (1,27) because the majority of plasma cfDNA of pregnant women are derived primarily from cellular apoptosis of the hematopoietic system (28). Thus, determining the maternal karyotype would increase the accuracy of NIPT results for SCA (1,29). However, as the associated costs would increase, it is difficult to make maternal karyotyping routinely available during pregnancy check-up visits. On the other hand, once the maternal copy number variation is accidentally found, it would cause unnecessary anxiety to pregnant women in the future, even if the copy number variation is benign.
As circulating cffDNA has been shown to mainly originate from the cytotrophoblast cells (30), confined placental mosaicism or placental mosaicism are other significant factors for the discordant results, such as false positive and false negative results, which occur in about 2% of all pregnancies (31). It is very difficult to prenatally confirm confined placental mosaicism in a pregnant woman. However, one study implemented a novel analytical approach that could detect the presence and estimate the degree of confined placental mosaicism, with accurate prediction in 90% of the cases (32). If this method can be further optimized, the identification of potential mosaicism would greatly improve clinical guidance during pregnancy.
The poorer performance in screening monosomy X could partially be due to a highly variable amplification of chromosome X because of its lower guanosine-cytosine content (33). Although not detected in this study, low levels of mosaicism for monosomy-X can be present in apparently healthy women (34), i.e., the maternal mosaicism for 45,X relatively more frequently influences the effectiveness in predicting the fetal 45,X karyotype. Nearly half of the true positive cases of 45,X were mosaicism in this study ( Table 3). Another potential reason could be a vanishing twin affected by monosomy-X.
Fetal mosaicism is very common among SCAs, compared with the three common autosomal aneuploidies. About 15% of cases of 47,XXY, 10% of cases of 47,XXX, and up to 50% of cases of monosomy X are mosaic, which may lead to a nonuniform distribution of sex chromosomes influencing the accuracy of classifying such anomalies (35). As in our study, 19.15% (9/47) of all the true positive results were mosaicism. Furthermore, 40% (4/10) cases of 45,X; 22.22% (2/9) cases of 47,XXX; 100% (2/2) cases of 46,XY,del(X); and 6.67% (1/15) cases of 47,XXY were mosaicism. No mosaicism was found in 47,XYY. In summary, our study showed good performance in detecting the mosaicism.
The necessity of screening SCA by NIPT is debatable with respect to their relatively mild phenotypes of SCA compared with trisomy 21/18/13. Hence, it is more complicated in counseling SCAs after diagnosis. Clinicians should exercise more patience and be empathetic when explaining the possible symptoms and prognosis to soothe the possibility of the parental anxiety. On the other hand, the affected children could be benefitted from the prenatal screening and diagnosis of SCA, providing an opportunity for early intervention and comprehensive postnatal management including hormone therapy and targeted surveillance, as hormonal treatment could induce puberty in Turner syndrome and androgen supplementation is also beneficial in Klinefelter syndrome (3,36). The termination rates were different among different types of SCAs. The termination rates of Turner syndrome and 47,XXY syndrome were higher than 47,XXX syndrome and 47,XYY syndrome, which was similar to the previous reports (26,37). The difference may be caused by the clinical manifestations related to high risks of infertility and lack of secondary sexual characteristics (26). Besides, the decision to terminate or continue pregnancy may also be related to the level of prenatal genetic counseling, the cognition level of the pregnant woman, economic situation, and cultural factors (37). Therefore, it is very important to provide professional and comprehensive genetical counseling for women with positive NIPT results for SCAs (38).
Our study adds further support for NIPT as an effective screening method for the detection of SCAs in pregnant women with different kinds of clinical indications in a large cohort of singleton pregnancies in central China. However, there are a few limitations to our study. First, the sensitivity, specificity, and negative predictive value were not calculated due to the absence of neonatal karyotype information, as most of the affected individuals do not present anomalies in the neonatal period. Second, we could not obtain all of the information about maternal or placental mosaicism contributing to the false positive rate. Third, though counseling clinicians of the participants are typically contacted in the prenatal and postnatal phases to obtain diagnostic results, many are still lost, which resulted in incomplete confirmation of NIPT positive results by karyotyping and a lack of the postnatal data.
In conclusion, our data with a relatively large sample size showed that NIPT can be used to screen SCAs with similar performance as in the previous reports (8,26), providing more comprehensive and reliable reference information in the genetic consulting for SCAs. After adequately informing pregnant women of the availability of the SCA screening along with the ability to provide accurate, balanced, and up-to-date information for subsequent genetic counseling when a fetus is diagnosed with SCA, pregnant women can benefit from NIPTbased SCA screening.

DATA AVAILABILITY STATEMENT
The original contributions presented in the study are included in the article, further inquiries can be directed to the corresponding author.

ETHICS STATEMENT
This study was conducted in accordance with the Declaration of Helsinki and was approved by the Ethics Committee of The First Affiliated Hospital of Zhengzhou University (2018-016). All participants signed the written informed consent.

AUTHOR CONTRIBUTIONS
XK designed the project and edited the manuscript. GZ was involved in project management, writing, performing the experiment, and data analysis. PD, CW, LL, and XZ performed the experiment. All the authors contributed to the article and approved the submitted version.