Clinical performance of Non-invasive prenatal testing for sex chromosomal aneuploidies in Northeast China

Background: Along with the discovery of cell-free DNA (cfDNA) and the invention of next-generation sequencing (NGS), non-invasive prenatal testing (NIPT) had appeared and been applied for detecting common aneuploidies such as trisomy 21, 18, and 13, with low false-negative and false-positive rates. Recently, it had also been used for sex chromosome aneuploidies (SCAs). To assess the clinical utility of NIPT for SCAs in Northeast China, we collected NIPT data from BGI 500 sequencing platform in the Center for Reproductive Medicine, Center for Prenatal Diagnosis of the First Hospital of Jilin University, and calculate the positive predictive value (PPV) and false positive rate (FPR). Results: A cohort of 14936 samples were analyzed by NIPT, and revealed 70 cases with SCAs high-risk, among them, 40 women agreed to undergo amniocentesis, but as many as 30 ones refused further diagnose. Based on veried fetal karyotype, 30.0% (12/40) were conrmed to be a true positive. Unluckily, the PPV for monosomy X performed 0%. Besides, positive 47,XXX were 46.67% (7/15), 40.00% (2/5) were positive for 47, XYY, and 42.86% (3/7) were positive for 47, XXY. Conclusions: In conclusion, our present results conrmed that NIPT sequenced by BGI 500 demonstrated lowest PPVs for 45,X, but the more accurate prediction for other SCAs, it is still a potential method for SCAs screening. Henceforth, we should focus on how to improve the test utility and provide better services for pregnant women in need.


Background
With high sensitivity and speci city, non-invasive prenatal testing (NIPT) is an unparalleled screening method for fetal aneuploidies [1]. As we know, it has now been widely used in the clinical screening for trisomy diseases, such as Patau, Edwards, and Down syndrome [2,3], and the analysis of cell-free DNA (cfDNA) in maternal blood showed high accuracy in trisomy 21, 18, and 13 [4]. However, false-positive, false-negative, and failure-reportable cases can take place, and which may have some biological reasons.
Besides these goal aneuploidies, the aneuploidies occurred in the X and Y chromosomes should also be taken seriously. In 2012, NIPT of sex chromosome aneuploidy (SCAs) was introduced in the United States. At rst, there was little information on the effects of these trials; however, its development was rapid [5,6]. The in uence of SCAs containing Turner (45, X), Klinefelter (47, XXY), Jacob syndromes (47,XYY) and Triple X syndrome (47,XXX) on new births is about 0.3% [7,8,9]. Turner, Klinefelter and some cases of 47,XXX could lead to delayed or abnormal sexual growth, and infertility, especially Turner and Triple X syndrome might show mental retardation [10]. To grasp the correct diagnostic information of neonatal SCAs is very important for early treatment. For many cases of SCAs, initial screening results in a better prognosis [11].
Nowadays, most laboratories provided the results of the SCAs as part of the report. However, recent researches had shown that the NIPT is less accurate in SCAs detection than autosomal aneuploidy [12,13]. But, a few of false-positive cases and low positive predict value still existed.
Our institution set up BGISEQ-500 sequencing platform in 2018, which was powered by Combinatorial Probe-Anchor Synthesis (CPAS) and modi ed DNA Nanoballs (DNBs) methods. In previous studies, BGISEQ-500 is a recently launched high-through put sequencing protocol; NIPT based on CPAS seemed to be a suitable method for routine clinical screening for SCAs [14]. Therefore, in our study, we try to investigate the accuracy of NIPT in SCAs screening among pregnancies admitted to our lab from Oct 2018 to May 2020, to evaluate our clinical performance of CPAS in detecting SCAs. Simultaneously, l0ml uncultured amino uid cells were used to detect by Chromosome microarray analyses based on Affymetrix CytoScan 750K array chips (Affymetrix, Santa Clara, CA). Copy number variations (CNVs) and heterozygosis loss (LOH) in the human genome were revealed by ≥50 probe markers and ≥200Kb resolution, the probe located on 22 pairs of autosomes and sex chromosomes. The results of CNVs and LOH were analyzed by comparing with the public databases or published works of literature.

Statistics
Online software (http://vassarstats.net/clin1.html) was used to calculate the FPR and PPV, the amniocentesis results regarded as the gold standard, based on the standard normal distribution, the 95% con dence interval (CI) was calculated by the cloper-Pearson method.

Results
NIPT was used to evaluate our fetal DNA in detecting SCAs performance. As a result, the analysis of 14936 pregnancies revealed 70 cases with SCAs high-risk results; the total frequency for the SCAs-positive was 0.50%  (Table 1).
At the same time, the comparison of the whole results between NIPT and amniocentesis for fetal SCAs was showed in Table 2. We had known that NIPT screening for Monosomy X in our study showed the highest FPR of
Several reasons might explain the poor PPV of monosomy X: at rst, as we know, there were 1098 genes located on the X chromosome, but Y chromosome contained only 78 genes, among them, there were also homologous 58 genes on both X chromosomes and Y chromosomes, and 29 genes were at the ends of sex chromosomes.
Secondly, the sequencing read length was only 35 bases, which might easily lead to the misplacing of X and Y chromosome. The non-random inactivation of X chromosomes in placental tissues may be the third reason, in XX female trophoblast cells, X chromosomes from father tended to inactivate more easily [20,21,22].
But, a small number of tests could give false-positive results or a complete failure for several reasons, including maternal chromosomal mosaicism [25], con ned placental mosaics [26], maternal malignancy [2,27,28,29], cotwin demise [30], and so on. In the same way, these causes could also lead to false positive results for SCAs.
Generally speaking, our results of all types SCAs performed on the lower side than the above mentioned, because of fewer subjects (57.14%, 40/70) would to accept invasive surgery, who afraid of miscarriage or intrauterine infection, its small size limited our study, in the future, we'll keep watching and updating the results.
Otherwise, it is worth mentioning that among our revealed false-positive SCAs cases, 2 cases were found CNVs of the X chromosome, and there were 3 cases found with autosomal CNVs detected by CMA, including microdeletion and microduplication. It is worth considering whether other fetal CNVs will cause different NIPT risk values to increase, and it is necessary to carry out CMA and karyotype analysis simultaneously, if possible.  Table 2 The result comparison among NIPT, conventional fetal karyotype by G-banding and CMA.