Non-invasive prenatal test findings in 41,819 pregnant women: results from a clinical laboratory in southern China

This paper evaluated the clinical utility of massively parallel sequencing-based non-invasive prenatal testing (NIPT) for detecting trisomy 21 (T21), T18, T13, sex chromosome aneuploidies (SCA), and rare chromosome aneuploidies (RCA) among the data collected by a clinical laboratory in southern China. In a 3-year period between January 2017 and December 2019, over 40,000 pregnant women underwent NIPT clinical screening test for fetal T21, T18, T13, SCA, and RCA in our laboratory. NIPT samples were processed using the NextSeq CN500 platform. The positive results were confirmed by karyotyping, and chromosomal microarray analysis (CMA) or copy number variants (CNV) sequencing. Details of the pregnancy outcomes were collected via telephone interview. NIPT results were available for 41,819 cases; 691 positive cases were reported. The overall sensitivity for detection of T21, T18, T13, SCA, and RCA was 99.21, 100.00, 100.00, 98.55, and 100.00%, and the specificity was 99.95, 99.94, 99.98, 99.69, and 99.92%, respectively. The positive predictive values (PPVs) for detection of T21, T18, T13, SCA, and RCA were 85.62, 45.24, 40.00, 34.17, and 13.51%, respectively, and those for detection of 45,X, 47,XXY, 47,XXX, 47,XYY, and 46,XY(delX) 20.00, 59.18, 28.95, 61.54, and 25.00%, respectively. Regarding pregnancy outcomes, 92.38% of the pregnancies with confirmed aneuploidies were terminated, and 91.20% of those identified as having a false-positive result were carried to term. Among 252 unconfirmed cases, 24.60% of the pregnancies were terminated and 38.10% carried to term, while 37.30% declined interview. NIPT is widely used to screen fetal aneuploidies based on its high sensitivity and specificity. However, in this study, the PPVs of NIPT in terms of detecting T18, T13, XO, XXX and RCA were < 50%. In addition, more than one-third of NIPT-positive women did not accept invasive prenatal diagnosis. Confirmatory diagnosis is strongly recommended for women with positive NIPT outcomes before any further decision is made.


Introduction
Cell-free DNA (cfDNA) from the peripheral blood of pregnant women has been widely used to screen for fetal chromosome aneuploidies, including Down syndrome (trisomy 21, T21), Edwards syndrome (trisomy 18, T18), Patau syndrome (trisomy 13, T13), and sex chromosome aneuploidies (SCA). Lo et al., in 1997, was the first to describe fetal cfDNA in the plasma of pregnant women [1], and a prenatal testing method based on cfDNA was introduced in 2008 [2]. Compared with traditional serum screening, cfDNA testing has a higher sensitivity, lower false-positive rate, and higher positive predictive value (PPV) [3,4]. Both the American College of Obstetricians and Gynecologists (ACOG) and the American College of Medical Genetics and Genomics (ACMG) have endorsed non-invasive prenatal testing (NIPT) as a routine screening option [5,6]. The use of NIPT for detecting fetal aneuploidies has rapidly transformed the global prenatal screening landscape, with the test performed in millions of pregnant women worldwide [7][8][9][10]. In China, NIPT has additional value because of the implementation of a two-child policy, such that a sharp rise in high-risk pregnancies is expected [11]. The aim of the present study was to evaluate the performance and accuracy of NIPT, not only for common aneuploidies T21, T18 and T13, but also for SCA and rare chromosome aneuploidies (RCA), to determine its clinical utility. The knowledge generated from our study could assist clinicians in pregnancy counseling, provide reassurance to pregnant women, and guide further decision-making.

Ethics statement and sample collection
This retrospective study included pregnant women who underwent NIPT at Nanfang Hospital of Southern Medical University (Guangzhou, Guangdong Province, China) between January 2017 and December 2019. All participants received clinical counseling prior to NIPT and were informed by clinicians about the content, principle, advantages, and limitations of the test. The study was approved by the Institutional Ethics Committee of Nanfang Hospital (approval no. NFEC-2017-035). Written informed consent was obtained from all subjects or their legal guardians.

Sample collection and processing
Peripheral blood (10 mL) was collected from each of the pregnant women and stored in tubes from Streck (Irving, TX, USA) or Lakebio (Hefei, China). The blood samples were centrifuged to isolate the plasma, and cfDNA was extracted from the plasma. A Qubit fluorometer (Thermo Fisher Scientific, Inc., Waltham, MA, USA) was used to measure the cfDNA concentration, and a cfDNA sequencing library was constructed and purified. The StepOnePlus ™ real-time PCR system (Thermo Fisher Scientific, Inc.) was used to quantify the libraries. All libraries were pooled and sequenced using the NextSeq CN500 massive parallel sequencing kit and NextSeq CN500 (Illumina China, Shanghai, China). All steps were performed according to the manufacturer's instructions (Berry Genomics Corp., Beijing, China). The sequencing data were mapped and processed using the Bambino test data analysis system 'RUPA' (Berry Genomics Corp.).

Evaluation of the performance of NIPT
Abnormal chromosomal results obtained by NIPT were confirmed by karyotyping and chromosomal microarray analysis (CMA) or copy number variants (CNV) sequencing after conducting an invasive prenatal diagnostic procedure. All women with a NIPT-positive result were followed-up by telephone interview to document the pregnancy outcomes, miscarriages, terminations, and deliveries were recorded. Cytogenetic or clinical follow-up results were used to calculate the sensitivity and specificity of NIPT. The true positive (TP) was defined as those high-risk NIPT results that were confirmed by invasive prenatal or postnatal diagnostic genetic testing. The false positive (FP) was defined as highrisk NIPT results that were shown to be euploid by followup invasive diagnostic genetic testing. The false negative (FN) was defined as low-risk NIPT screening cases with an aneuploidy karyotype confirmed by prenatal or postnatal genetic testing. The true negative (TN) was defined as low-risk NIPT results confirmed by normal neonatal clinical physical examination results (barring SCA). To encourage reporting of FP and FN results, an insurance policy was provided for each participant as part of the test. The policy would reimburse patients for the cost of invasive tests in the case of a positive NIPT result, and would pay CNY400 000 to each FN case that had aneuploidy confirmed. Pregnancies without confirmatory diagnostic results were not used in the calculation of test parameters including sensitivity, specificity, positive perspective value (PPV), and negative perspective value (NPV).
The participants were categorized according to their aneuploidy risk. Those with any of the following factors were classified as high risk: advanced maternal age (AMA ≥ 35 years), a positive conventional serum screening test (cut-off of 1/270 for T21 or 1/350 for T18), abnormal ultrasound markers, a history of adverse pregnancy outcome. Participants with none of those factors were defined as low risk. The performance of NIPT for detecting T21/T18/T13 was compared between these two risk groups, and the two groups of AMA and young pregnant women. All data analyses were performed using Microsoft Excel 2007 (Microsoft Corp., Redmond, WA, USA) and SPSS version 20 software (IBM Corp., Armonk, NY, USA). Differences in proportions were tested for statistical significance using the chi-square test, and a P value < 0.05 was considered significant.

NIPT performance on different SCA
The results of a detailed analysis of the 316 SCA detected by NIPT are shown in Table 3

PPV according to pregnancy characteristics
Among the 41,819 samples, 21,164 (50.61%) were from women in the high-risk group and the remaining 20,655 from women in the low-risk group. The performance of NIPT in detecting T21/T18/T13 was compared between the two risk groups, AMA and young women (Fig. 3

Clinical outcomes of NIPT-positive cases
The outcomes of the 691 NIPT-positive pregnancies are shown in

Discussion
This study was based on the clinical data from 41,819 pregnant women who underwent NIPT at our center, including detailed pregnancy outcome data for all cases with positive NIPT results. Detection sensitivity and specificity were high for the common aneuploidies T21/T18/T13 and RCA, but lower for detection of SCA, as shown in other studies [7][8][9][10][12][13][14][15]. The PPVs for detection of T18, T13, SCA, and RCA were low in general population, as we reviewed [16]. Differences in PPV between studies occur because study populations differ in size, demographics, clinical characteristics, and different NIPT platforms with variable sequence read depths and algorithms used by different providers. Reports from outside China showed higher PPVs for T21, ranging from 96% to 99.2% [7][8][9][10]. We think that it was caused by the participant women having a higher risk for aneuploidies, such as the mean age of 35.3 years in the report from Italy [9], 34.7 and 34 years from the reports of USA [8,10], while the mean age of 31.7 years in this study. However, the report from Netherlands had the same mean age 31.7 years, but with a different average gestation weeks in first and second trimester, 11.9 weeks in Netherlands and 16.7 weeks in the study [7]. Considering the sequencing platforms are the same NextSeq 500 from both the studies, the difference in PPVs of T21/T18 may be caused by study populations, demographics and each algorithms. As Demko et al. reviewed that the studies conducted in China tend to have several differences from the other studies, namely, Chinese women tend to be of lower maternal weight [17]. By the way, all current reports of NIPT from China were performed by the method of massively parallel shotgun sequencing (MPSS) [17]. We also noticed that the incidences of T21/T18/T13 were 3.01‰, 0.45‰ and 0.14‰ in the study, while the incidences were 3.18‰ (233/73239), 0.72‰ (53/73 239) and 0.37‰ (27/73 239) in the report from Netherlands. The higher incidences of T21/T18/T13 in Netherlands may cause the difference in PPVs. The algorithms from each provider may contribute to the divide too. However, the PPVs from this study showed consistent with most studies from China [16]. More comparison and analysis in these studies need be done between China, Europe, Australia and the USA, not only on its proportion of women [18]. For SCA, XO accounted for the majority of the positive NIPT results, but the PPV for detection of aneuploidy was the lowest. The PPVs of XYY, XXY and XXX varied greatly, which may due to the small numbers of sex aneuploidies in each study as reviewed before [16,19]. In summary, the PPV in XYY was highest, followed by XXY, XXX and XO [19], which similar with the study. Circulating cfDNA consists of both maternal and fetal DNA, and the proportion of maternal DNA is more than 90%. Hence, abnormal maternal chromosomes affect NIPT results. There is a small percentage of healthy fertile women with maternal SCA mosaicism or occasionally a full SCA, such as XO or XXX, but nonetheless they may  [20]. Maternal X chromosome CNVs have been found in association with discordant fetal SCA detected by NIPT [21]. Moreover, the gradual and preferential loss of chromosomes occurs in some AMA women, such that their blood karyotype might change from XX to XO/XX mosaic [22]. However, the maternal chromosomal complement was not investigated in this study, because it was not part of the informed consent procedure; this is a limitation of the study. We have to mention that since it is not possible to perform the gold standard method for all live births (karyotype analysis), some false negative of SCA newborn may be missed, that affects the calculation of incidence but not PPV. For RCA, the study showed that trisomy 7 was the most frequently detected, followed by trisomy 16 and trisomy 8, which is consistent with other studies, while PPVs ranged around 5.5% to 28.6% in the general population of China [13,16,23]. We also compared NIPT performance for detecting T21/T18/T13 in women with high-and low-risk pregnancies, AMA and young women; significant higher of PPVs were presented in high-risk and AMA groups. The lower PPV in the low-risk group was attributable to the lower prevalence of chromosomal abnormalities in the population, emphasizing that NIPT is only a screening test such that confirmation via invasive diagnosis testing is essential. As described by AGOG, PPV oscillates between 38 and 80% and 91 and 99% for T21, between 11 and 41% and 66 and 92% for T18, and between 5 and 13% and 45 and 71% for T13, respectively, at 20 and 40 years old [24]. The pregnancy outcomes of all women were evaluated in this study. Among the 691 cases with positive NIPT results, 252 (36.47%) were not confirmed by invasive prenatal diagnosis, similar to the other study [25]. In this study, we found that the terminate pregnancies were higher in T21/T18/T13 NIPT-positive cases without confirmed results. Pregnancy termination without confirmation by invasive diagnostic testing can be prevented by adequate pre-and post-NIPT counseling, but apparently there were many women did not get detail and knowledge counsel by the study. In our followup phone interviews, some women stated that their clinician at a local hospital did not inform them to take invasive test for confirming their RCA high-risk results; some women mentioned that they could not afford confirmatory test or have limited access to a diagnostic center (often remotely located), so they chose to ignore the results or terminated the pregnancy without invasive prenatal diagnosis confirmation. By the study we found that women in economically developed areas are likely to undergo confirmatory tests, while women in rural areas are more unlikely to undergo confirmatory test. However, clinicians should inform women that invasive prenatal diagnostic tests are important, and it is essential to confirm a high-risk NIPT result, to prevent the disaster of wrongful pregnancy termination. National recommendations or clear guideline and continuing educational support are essential for ensuring that providers and clinicians are appropriately trained in this rapidly evolving field.

Conclusion
In conclusion, our study supports the integration of NIPT into current clinical practice, but the accuracy of detecting T13, XO and RCA is still relatively poor, especially in low risk and young women. NIPT is the most sensitive and specific screening test for the common fetal aneuploidies. Nevertheless, it has the potential for false-positive and falsenegative results. NIPT is not equivalent to diagnostic test. We urge that clinicians inform women with positive NIPT results to undergo further invasive prenatal diagnosis before making any decision.