Non-invasive Prenatal Testing in Pregnancies Following Assisted Reproduction

Abstract

In the decade since non-invasive prenatal testing (NIPT) was first implemented as a prenatal screening tool, it has gained recognition for its sensitivity and specificity in the detection of common aneuploidies. This review mainly focuses on the emerging role of NIPT in pregnancies following assisted reproductive technology (ART) in the light of current evidence and recommendations. It also deals with the challenges, shortcomings and interpretational difficulties related to NIPT in ART pregnancies, with particular emphasis on twin and vanishing twin pregnancies, which are widely regarded as the Achilles’ heel of most pre-natal screening platforms. Future directions for exploration towards improving the performance and extending the scope of NIPT are also addressed.

1. INTRODUCTION

Over the past four decades, the utilization of assisted reproductive technology (ART) as a treatment option for infertility has progressively increased across the world, with more than 8 million babies being born using this technology [1]. Notably, in some European countries, the percentage of births following ART reaches up to 6% of the total recorded births, and the upward trend is likely to continue as it has become safer and more successful than ever before [2].

The rising trend of ART has unleashed a unique set of concerns and challenges relevant to prenatal testing. Women resorting to ART are likely to be older, and those who conceive frequently have multiple gestations. They are also less likely to take up invasive prenatal testing as compared to women who conceive naturally in view of the risk of procedure-related miscarriage [3, 4]. Furthermore, due to altered levels of biochemical markers in pregnancies following ART, as compared to natural pregnancies [5-7], the false-positive rates of biochemical screening are higher, and interpretation in a prenatal setting becomes ambiguous. The introduction of non-invasive prenatal testing (NIPT) has offered a viable alternative for the non-invasive detection of fetal aneuploidies. NIPT, however, has its own limitations in the setting of ART, which clinicians need to be aware of.

The current review describes NIPT in the context of ART pregnancies, its advantages and limitations, with a critical appraisal of the evidence derived from the available studies. Future directions towards delineating the role of NIPT in pregnancies following assisted reproduction are also touched upon.

2. PRENATAL SCREENING - BACKGROUND

Chromosomal abnormalities have emerged as one of the leading causes of birth defects worldwide and are estimated to occur at a frequency of 1 in 150 live births [8]. The frequency of chromosomal abnormalities in ART pregnancies is similar to pregnancies occurring through normal conception [9-11], with aneuploidies being the most frequent aberration. Most aneuploid fetuses are non-viable, and these pregnancies often adversely result in early miscarriage, while certain aberrations, such as trisomies 21, 18, and 13, and those involving the sex chromosomes carry the potential to survive to the neonatal period or even to adult life. The health and welfare concerns related to affected individuals have paved a path towards the establishment of prenatal screening and testing procedures as a widely accepted clinical resource to identify chromosomal defects in the fetus. Screening allows couples to pre-emptively explore options about the pregnancy itself, as well as post-natal care in case of continuation of pregnancy.

2.1. The Role of Biochemical Screening Tests in Normal Conception versus ART Pregnancies

The combined first-trimester screening is the most frequently used non-invasive approach to gauge the risk of carrying a fetus with Down syndrome in normal conception pregnancies. This screening method has a detection rate of 82-87% and a screen-positive rate of 5% [12]. Second-trimester screening has a relatively inferior performance. Another approach (integrated or sequential screening) amalgamates one or more parameters of first-trimester screening with a selective set of second-trimester markers. This provides a detection rate of 88-96% with a screen-positive rate of 5% [12].

On the flipside, the feasibility of combined first-trimester screening in ART pregnancies remains unclear. Some studies recommend first-trimester screening as a reasonable approach to detect trisomy 21, since no significant differences have been noted in terms of accuracy between ART and normal conception pregnancies [5, 6]. Others have observed a slightly higher false-positive rate (7% vs. 5%), resulting in an overestimation of trisomy 21 in ART pregnancies when compared to normal conceptions [7, 13]. When the performance characteristics of second-trimester screening were evaluated, a much higher false-positive rate was obtained in ART pregnancies (26% vs. 5%) as compared to normal conceptions, with an exception of in-vitro fertilization (IVF) pregnancies conceived through oocyte donation [14, 15]. These discrepancies could be attributed to underlying infertility and the procedures involved in assisted reproductive technology [14]. Since none of these screening modalities offer an absolute positive predictive value (PPV), a positive screen, irrespective of the modality, necessitates confirmation by invasive testing before a definite diagnosis can be reached. The use of screening approaches with lower PPV means that many pregnant women will be subjected to invasive diagnostic testing, with attendant procedural risks, cost and anxiety.

2.2. Invasive Diagnostic Testing

Invasive pre-natal testing by sampling chorionic villus, amniotic fluid, or cord blood explores genomic modalities like karyotyping and chromosomal microarray, or by targeted methods including Fluorescent In-Situ Hybridization (FISH) or quantitative fluorescent polymerase chain reaction (QF-PCR) to determine chromosomal aberration. The diagnostic accuracy of these tests is higher than traditional screening, but it carries a procedure-related risk of a miscarriage of approximately 1% [16-21].

3. NON-INVASIVE SCREENING: A HISTORICAL PERSPECTIVE ON NIPT

In 1947, Mandel and Metais [22] discovered the presence of extracellular DNA material in the plasma of both healthy as well as diseased individuals. In 1997, Lo et al. [23] isolated cell-free fetal DNA (cffDNA) from maternal plasma, although its biological basis and role were unclear. In 2001, Jahr et al. [24] demonstrated that cffDNA derives from apoptotic cells, while consequently, other groups confirmed that the circulating DNA originates from apoptotic trophoblastic cells of the embryo [25, 26].

In 2008, Fan et al. [27] published a landmark paper that described the successful application of massively parallel sequencing (MPS) for high-throughput DNA sequencing of cell-free DNA isolated from maternal plasma and its utilization in the detection of trisomies 21, 13 and 18. Further studies validated NIPT as a reliable tool and confirmed its efficacy and applicability, with 2011 being the year that marked the establishment of NIPT into routine clinical practice as a prenatal screening tool for the detection of mainly trisomy 21. Thereafter, it progressively incorporated the detection of other aneuploidies, such as 18, 13 and sex chromosomes [28].

3.1. Current Role and Recommendations for NIPT

The joint recommendations by the American College of Obstetricians and Gynecologists (ACOG) and the Society for Maternal-Fetal Medicine (SFMF), released in 2012, endorsed NIPT as one of the prenatal screening options for the detection of trisomies 21, 13 and 18 among high-risk obstetric populations carrying singleton pregnancies [29]. These regulatory bodies and several others (Table 1) have since then made amendments based on various levels of evidence, which has further expanded the scope of NIPT, keeping in mind the welfare of the mother and the fetus [30-39].

Table 1.

Recommendations by the various regulatory bodies on the usage of non-invasive prenatal testing in the obstetric population.

Year	Regulatory Bodies	Recommended Beneficiaries
2019	DEGUM, OGUM, SGUM & FMF Germany	All pregnant women
2018	HGSA & RANZCOG	All pregnant women
2017	SOGC & CCMG	All pregnant women
2017	PGS & PHGS	Pregnant women with screening risk ranging from 1:100 to 1:1000
2017	ISUOG	Pregnant women with high or intermediate risk.
2016	ACMG	All pregnant women.
2016	ACOG & SMFM	All pregnant women.
2016	ESHG & ASHG	All pregnant women.
2016	SFOG	Pregnant women with a prior risk of 1:1000.
2015	ISPD	All pregnant women

Note: DEGUM: German Society of Ultrasound in Medicine and Biology, OGUM: Austrian Society for Ultrasound in Medicine, SGUM: Swiss Society of Ultrasound in Medicine, FMF:Fetal Medicine Foundation, HGSA: Human Genetics Society of Australasia, RANZCOG: Royal Australian and New Zealand College of Obstetricians and Gynaecologists, SOGC: Society of Obstetricians and Gynaecologists of Canada, CCMG: Canadian College of Medical Geneticists, PGS: Polish Gynecological Society, PHGS: Polish Human Genetics Society, ISUOG: International Society of Ultrasound in Obstetrics and Gynecology, ACMG: American College of Medical Genetics and Genomics, ACOG: American College of Obstetricians and Gynecologists, SMFM: Society for Maternal-Fetal Medicine, ESHG : European society of Human Genetics, ASHG: American Society of Human genetics, SFOG: Swedish Society of Obstetrics and Gynecology, ISPD: International Society of Prenatal Diagnosis.

The latest joint recommendation by the ACOG and SFMF was released in 2020 [12]. These guidelines are currently in force and state that both screening and diagnostic options for prenatal genetic testing should be discussed with and offered to all pregnant women early in pregnancy irrespective of age or risk. The document affirms NIPT as the most sensitive and specific prenatal screening test for the detection of common aneuploidies and the only laboratory test to identify sex chromosome aneuploidies while reiterating that this approach is not equivalent to diagnostic testing. Among its advantages, the guidelines specify the possibility of performing the screening procedure at any gestational age, from as early as nine weeks. In line of the importance of such procedures, these recommendations emphasize that pre- and post-test genetic counseling are necessary pre-requisites for NIPT.

3.2. The Science Behind NIPT

All NIPT platforms, irrespective of the technology applied, obtain their results from the interrogating of fetal DNA from the circulating free DNA in maternal plasma. The circulating free DNA represents a mixture of maternal DNA mostly derived from the mother’s hematopoietic cells and cffDNA from the placenta released by trophoblastic cells during processes, such as apoptosis, necrosis and micro particle secretion [26, 40, 41].

cffDNA are shorter in length with a dominant peak size of 143 base-pairs (bp) as compared to 166 bp identified as a mean size among maternal fragments [42, 43]. It has been demonstrated that fragmentation of cell-free DNA does not occur in random but is sequence-dependent and the motif followed differs between maternal and fetal DNA, producing different signatures at the ends of fragments [44, 45]. These differences are utilized for physical or in silico enrichment of fetal fragments while testing for copy number variations or estimating fetal fractions (FF) [46-49].

The proportion of the free circulating DNA of fetal origin is referred to as the fetal fraction. The FF averages 15-20% between 10-20 weeks of pregnancy, although with a wide range of as low as 4% to over 30% [50]. This proportion increases with gestational age and is also influenced by various maternal and pregnancy-related factors. High maternal body mass index (BMI) decreases FF by a dilutional effect or by increasing the maternal component of circulating free DNA. Pregnancies conceived through ART and twin pregnancies demonstrate lower percentages of FFs [46-49]. Certain aneuploidies, such as trisomies of 18 and 13, are associated with lower FF, while trisomy 21 as well as the rise in certain serum markers, such as β-hCG and PAPP-A, have been associated with increased FF [51-54]. A low FF may affect the efficacy of the techniques in the detection of abnormalities due to the limited genetic material per analysis.

Physiologically, FF is rapidly eliminated from maternal blood following delivery, becoming undetectable within a day. Therefore, it does not interfere with NIPT testing in subsequent pregnancies [55].

3.3. Platforms Used for NIPT and their Underlying Principles

Most current NIPT platforms use Massive Parallel Sequencing (MPS) technology. MPS enables the performance of sequencing across the genome that generates sufficient reads to allow the detection of aneuploidies or copy number variations contributed by the relatively small FF of free circulating DNA in maternal plasma.

The basic principle underlying the performance of NIPT using MPS is that in cases of aneuploidy or copy number variation, the proportion of DNA fragments that align to the chromosome or region of interest in comparison to DNA fragments aligning to other regions of the genome will be altered. In the case of trisomy 21, for example, the proportion of fragments that map to chromosome 21 will be higher than expected in normal diploid individuals [27, 56]. The extent of deviation and the cut-off for assigning positivity are expressed in terms of a Z-score [57]. A high Z-score represents a greater extent of deviation from the normal. The reliability of a positive NIPT result is affected by the Z-score obtained. A Z-score of 6 or more appears to be more indicative of true positivity while being independent of a priori risk factors, while a Z-score of 5 or less must be evaluated in the context of a priori risk [57].

MPS platforms that are used for NIPT are generally those with shorter read lengths but high outputs. Frequently used platforms include the Illumina platforms (Hi-seq, Miseq, Next seq), Sequenom laboratories, Roche, Natera Inc. and semi-conductor sequencing-based Ion Torrent PGM (Table 2) [58, 59].

Table 2.

Various platforms used in non-invasive prenatal testing.

Platforms for NIPT	Illumina	Ion Torrent PGM	Sequenom Laboratories	Roche	Natera Inc.	BioRad Laboratories
Testing methodology	MPS with shot gun sequencing	Semi-conductor sequencing	MPS with shot gun sequencing	Targeted multiplex seq- uencing (Digital analysis of selected regions)	Next generation targeted SNP based	Digital droplet PCR

NIPT platforms that use MPS can be classified into two broad kinds of platforms that either sequence the entire genome as fragments subsequently aligned against the reference genome (shotgun sequencing) or platforms that sequence only target regions or chromosomes [60-63]. Each method has its merits and demerits. On one hand, interrogating the entire genome involves the generation of enormous amounts of data that reduce throughput while increasing cost and storage requirements for the generated data, while on the other hand, targeted NIPT generates less data and enables higher throughput, which reduces costs, though it only interrogates the specified regions and will miss aberrations in other regions of the genome [62, 64].

Techniques that rely entirely on a number of sequencing reads without distinguishing fetal and maternal genotypes are unreliable in situations where the proportions may deviate from the expected, such as multiple gestations or vanishing twins. Another targeted approach uses single nucleotide polymorphism (SNP) data to differentiate the maternal and one or more fetal genotypes. This technique can quantify the FF as well as examine allelic patterns to detect aneuploidy in a single processing [65-67]. The use of genotype data allows the technique to be used for twin pregnancies, to determine zygosity, and detect vanishing twins. However, SNP-based platforms are not useful in the settings of surrogacy and oocyte donation, as well as when there is a high degree of parental consanguinity.

3.4. Factors Affecting the Accuracy of NIPT

Several factors influence the accuracy of NIPT testing. The performance characteristics of NIPT, as described by various studies, are depicted in Table 3 [68-71].

Table 3.

The performance characteristics of noninvasive prenatal test as described by various studies.

Studies	Population	Foetal Fraction (Median)	Test Failure at 1st Sampling (Median)	Test Failure after 2nd Sampling (Median)	Overall PPV (Median)	PPV Trisomy 21 (Median)	PPV Trisomy 13 (Median)	PPV Trisomy 18 (Median)	PPV SCA (Median)	False Negative (Median)
Talbot et al. [68]	ART	*5.6% (4.9% for fresh embryo; 6.3% for frozen embryo)	-	-	-	-	-	-	-	-
Talbot et al. [68]	Spontaneous conception with high risk on combined screening	*7.2%	-	-	-	-	-	-	-	-
Zou et al. [69]	ART with VTS	-	7.6%	1.4%	8%	11.1%	0%	0%	-	Nil reported
Tan et al. [70]	ART with twins	-	3.2%	0.9%	100%	100%	-	-	-	Nil reported
Lee et al. [71]	Singleton IVF	10.3%	5.2%	2.4%	28.6%	100%	0%	50%	0%	-
Lee et al. [71]	Spontaneous singleton	11.9%	2.2%	0.7%	73.4%	100%	25%	76.9%	31.8%	-

Abbreviations: ART: assisted reproductive technique; PPV: positive predictive value, VTS: vanishing twin syndrome,

3.4.1. Z- Score

The Z-score, as a screening output, quantifies the extent of deviation of the measured value from the mean observed values. The Z-score directly correlates with test accuracy, with higher values demonstrating higher result reliability [57, 72].

3.4.2. Fetal Fraction

The performance of NIPT is heavily influenced by FF levels. A low fetal fraction can reduce the extent of deviation produced by aneuploidy in the fetus. This, in turn, will reduce the Z-score and can lead to false-negative or indeterminate results [28]. Therefore, estimation of the FF is a crucial step in the performance of NIPT. An FF of less than 3-4% is often considered too low to provide accurate results resulting in a “no-call” or test failures [50]. Conditions associated with a low FF, such as high maternal BMI, are therefore more likely to result in failed tests [73]. Low FF has also been associated with aneuploidy. Aneuploidy rates in cases with low FF range from 2.7% to 23.3% [74-76]. In tests that use genotyping data, a low FF could result in false-positive as well as false-negative results [50].

3.4.3. Chromosomal Abnormality of Interest

NIPT is primarily used to detect trisomies of chromosomes 21, 18, and 13. While it has also been applied for the detection of other autosomal trisomies and sex chromosomal aneuploidies, the accuracy of NIPT in the detection of these conditions is lower. NIPT is also available in the market as a prenatal screening tool to detect common microdeletions. However, due to low PPV, the ACOG so far has not recommended this test as a suitable prenatal screening option for the detection of common microdeletions [12]. Single-gene cell-free DNA testing has also been studied, but data on its accuracy and PPV and NPV are still insufficient for clinical validation [77].

Based on recent meta-analyses, trisomy 21 has a false positive rate of 0.05% and a detection rate of 99.5%, trisomy 18 has a false positive rate of 0.04% and a detection rate of 97.7%, and trisomy 13 has a false positive rate of 0.06% and a detection rate of 96.1%. By contrast, monosomy X has only a 90.3% detection rate with a 0.23% false-positive rate. For trisomies of the sex chromosomes, the detection rate is 93.0% and the false-positive rate is 0.14% [78].

3.4.4. Fetoplacental Mosaicism

The use of NIPT for detecting chromosomal abnormalities of the fetus by testing cffDNA lies on the presumption that the FF of the cffDNA is representative of the fetal genome. This is not always true since mosaicism may be confined to either the placenta or the fetus itself. In either of these situations, as the FF originates from the placenta, NIPT could prove to be lacking accuracy [79]. It must be reiterated that NIPT is not a diagnostic test, and no irreversible management decisions are to be based on a positive NIPT without confirming its findings using an invasive diagnostic test. However, fetoplacental mosaicism is a concern even in invasive diagnostic testing, more so with chorionic villus samples. In situations where there is a discrepancy between the NIPT results and the results of invasive testing, or where mosaicism is identified on invasive testing, a detailed and focused ultrasound examination, further cord blood sampling, examination of the newborn or peripheral blood sampling of the child may help to verify fetal involvement.

3.4.5. Interference from Other Sources of Cell-free DNA

The presence of multiple gestations, vanishing twins, maternal mosaicism or chimerism, maternal transplantation or recent transfusion, and even the presence of tumors may affect the accuracy of the produced results. An SNP-based NIPT test may be more accurate in some of these situations [71, 79].

Maternal malignancies, both benign and malignant, release DNA fragments in the maternal circulation. These DNA fragments can produce discordant NIPT results. In a study by Bianchi et al., 10 cases of occult maternal cancers were identified due to discordant NIPT results from a cohort of 125,426 asymptomatic pregnant women, who underwent NIPT for the detection of fetal aneuploidy [80].

4. CHALLENGES FOR NIPT IN ART PREGNANCIES

4.1. Low Fetal Fraction

Several studies have linked ART with lower FF [68, 71, 74, 81, 82]. Talbot et al. [68] compared fetal fractions from 54 pregnancies following fresh (n = 23) and frozen embryo transfers (n = 26) against a control group (n = 238) with natural conceptions but at a high risk for aneuploidies following combined first-trimester screening. The mean FF for ART pregnancies was 5.6%, which was significantly lower than natural conceptions, with a mean fetal fraction of 7.2%. A potential limitation of this study could be the utilization of an MPSS platform with a bio-informatics algorithm based on fragment size (seq FF) to compute FF, which appears less accurate when compared to genotyping approaches, especially at low FF levels [83]. However, the difference in FF was statistically highly significant (p<0.0001) even on multivariate analysis adjusted for BMI and PAPP-A levels, indicating that ART was almost certainly an independent predictor of low FF. In a secondary analysis, the authors reported significantly lower FF following fresh embryo transfers compared to frozen embryo transfers following modified natural cycles (mean FF 4.9% versus 5.3%; p=0.02).

IVF: in-vitro fertilization; SCA: sex chromosomal abnormality.

All values expressed in median

* expressed as mean

Another group, Lee et al. [71] from Australia, retrospectively studied data from 4633 spontaneously conceived and 992 IVF singleton pregnancies that were screened through NIPT using an SNP-based assay for the detection of trisomies 21, 13, and 18, and sex chromosomal aneuploidies. They found a median FF of 10.3% in the IVF group versus 11.9% in the natural conception group. This result was statistically significant on multivariate analyses adjusted for the effects of other factors, such as BMI, gestational age and ethnicity, which have been previously reported as independent predictors. On a secondary analysis of IVF modalities, the use of donor oocytes as opposed to autologous oocytes, and hormone treatment were associated with lower FF. The test failure rate was found to be higher (5.2%) in the IVF group as compared to the spontaneous conception group (2.2%). The PPV for aneuploidies other than trisomy 21 in IVF was found to be lower. However, in view of the low numbers of the other autosomal trisomies, the findings related to PPV cannot be regarded as conclusive.

The exact cause for lower FF in IVF gestations remains uncertain. Some suggestions include a lower placental mass in IVF pregnancies or increased maternal cell-free DNA due to increased inflammation and cellular damage in women conceiving with IVF [71]. Despite the low FF, most studies report that ccfDNA with NIPT generated results from the first sampling. Therefore, IVF is not a contra-indication for NIPT testing; rather, medical specialists and candidate patients should be fully aware and informed respectively of the possibility of test failure during initial screening, which would necessitate a second sampling from maternal peripheral blood.

Talbot et al. [68] in an attempt to identify a possible reason for lower FF following fresh transfer as compared to frozen-thawed transfers, underlined the presence of supraphysiological estradiol levels during the fresh treatment cycle, which may cause placental dysfunction. Fresh ART cycles involve controlled ovarian stimulation, usually with high doses of gonadotrophin, to obtain multifollicular growth, but this, in turn, is linked to supraphysiological levels of estradiol and increased risk of ovarian hyperstimulation (OHSS) [84]. Consequently, all over the world, there is an increasing trend towards ‘freeze all’ cycles in ART practice [85, 86], and a higher number of frozen embryo transfers are being performed to avoid the risk of OHSS and for endometrial optimization in subsequent cycles leading to oocyte retrieval. This might, to an extent, help to mitigate the burden of test failures and diagnostic errors due to low FF in the setting of ART.

4.2. Multiple Pregnancies Following ART

Multiple pregnancies are a common occurrence following ART due to the practice of transferring more than one embryo during the treatment. Of note, the multiple birth rate in Europe following ART treatment was 16.9%, while in the United States, the rate was 26.4% [87, 88].

Guidelines allow or recommend NIPT screening in twin pregnancies [12, 89], even though NIPT failure rates in twins range from 1.6% to 13.2%, with a median of 3.6%, which is higher than for singletons. In multiple gestation, such as in triplets, a test failure is reported to be as high as 20% [90].

While the amount of cffDNA is higher in twin pregnancies than in singleton pregnancies, the amount contributed by each fetus is comparatively lower and may vary for each twin [90]. Canick et al. [91] found that while the average FF for twin pregnancies was 35% higher than singleton pregnancies, the average FF contributed from each twin was about 66% of the average FF from a singleton pregnancy. In dizygotic pregnancies, individual FF must be computed, and a low value of individual or combined FF may lead to test failure, while in monozygotic pregnancies, only the total FF is relevant. In dizygotic twins, the FF for one or both twins may be so low as to lead to test failure.

Zygosity is another factor affecting FF levels. Struble et al. [92] reported a median FF level of 14% (range 8.2 to 27%) and 7.9% (4.9 to 14%) in monozygotic and dizygotic pregnancies, respectively. In a large cohort study by Hedriana et al. [90], the FF levels for monozygotic (n = 1,454) and dizygotic twins (n = 3,161) were compared with singleton pregnancies (n = 1,21,446). The authors reported that when compared to singleton pregnancies, the total FF was 35% and 26% higher in monozygotic and dizygotic pregnancies, respectively. At an FF threshold of 2.8%, Hedriana et al. [90] reported uninterpretable results in 1.7% of singleton pregnancies, 0.8% of monozygotic and 5.6% of dizygotic pregnancies.

In situations where the FF vastly differs between the fetuses in a dizygotic twin pregnancy, the combined FF may appear satisfactory, but the result will be more representative of the twin with a higher FF, sometimes to a point where the contribution of the twin with lower FF is not detected. This particularly concerns cases where one fetus is affected by either trisomy 13 or 18 and where FF is already low. NIPT platforms that only utilize a combined FF as opposed to FF calculation for each zygote are more likely to result in incorrect calls for aneuploidy risk assessment in dizygotic twin pregnancies, and therefore, should be interpreted with caution. SNP-based platforms offer some advantages in these cases, as they can assess the zygosity as well as the individual FF [93]. However, these are not useful in the settings of surrogacy or donor oocytes since cffDNA generated from surrogate mother/donor oocyte interferes with the identification of alleles, and thus zygosity [93].

There is a paucity of data on the role of NIPT for prenatal screening in twin pregnancies following ART treatment. Huang et al. [94] evaluated the performance of NIPT in twin pregnancies in 189 women, of which the majority conceived through ART. A total of 11 aneuploidies were reported on the karyotype, of which nine were trisomy 21 and two were trisomy 18. NIPT identified all nine cases of trisomy 21 and one case of trisomy 18, while failing to detect one case of trisomy 18. This sole discordant case was an unusual case of monozygotic twins (monochorionic diamniotic) with discordant fetal karyotypes, one normal and the other trisomy 18, which presumably arose as a post-zygotic event, sparing the other fetus and the placenta. This case illustrates the afore-mentioned pitfall of testing cffDNA, which represents the placental but not always the fetal genome.

Tan et al. [70], in a prospective observational study, evaluated the role of NIPT in 565 women with twin pregnancies following ART treatment. Four trisomy 21 cases were identified by NIPT which were later confirmed by karyotype. A 100% PPV was reported. NIPT failed to generate results in 0.9% cases, while no false negatives were reported. Based on these results, the authors suggested that NIPT could be a reasonable alternative to first-trimester biochemical screening for twin ART pregnancies.

4.3. Vanishing Twin Following ART

The prevalence of a vanishing twin syndrome (VTS) following ART ranges from 10-30%, and is influenced by the practice of transferring more than one embryo during a treatment cycle [95]. NIPT in cases of a vanishing twin is associated with a lower PPV as well as higher test failure when compared to singleton or twin pregnancies, and approaches that of conventional prenatal screening.

It has been suggested that the false-positive NIPT in VTS is due to the interference from cffDNA derived from necrotic trophoblasts of the aneuploid demised fetus. Studies suggest that the influence of a vanishing twin could last for up to 7-8 weeks after the fetal demise, but not beyond 12-14 weeks [69].

A recent retrospective study evaluated the role of NIPT in 579 ART pregnancies with VTS [69]. The NIPT failure rate was 7.6% on the first sampling and 1.4% following resampling. Twelve positive NIPT results were reported, of which only one was found to be true positive, giving a PPV of 8%. No false-negative results were reported. Based on their findings, the authors suggested that NIPT could potentially be used for screening in the VTS scenario because, unlike other modalities, it did not appear to miss aneuploidy. They also suggested resampling after 15 weeks of gestation for all positive results and test failures to reduce false positivity and repeated failure.

4.4. Pregnancy Following Oocyte Donation

According to the European registry, oocyte donations have been gaining widespread acceptance as an infertility treatment approach, with a notable increase of more than 15% between the years 2014 and 2015 [87]. In clinical practice, oocyte donation is offered to women of advanced maternal age due to a higher success rate compared to autologous ART. Due to increased anxiety and the frequent presence of medical comorbidities, such as chronic hypertension and diabetes, women undergoing oocyte donation are more likely to explore NIPT for prenatal testing as compared to women aged less than 35 years who undergo autologous ART [96].

A study by Lee et al. [71] included 63 pregnancies following oocyte donation, only to find that the median FF in pregnancies following oocyte donation was significantly lower versus pregnancies with autologous ART (8.4% to 12.5% versus 10.5% to 13.7%; p < 0.001). The authors suggested that this finding may be due to HLA compatibility issues occurring because of ‘foreign antigen’.

A large retrospective analysis presented at the 2019 American Society of Reproductive Medicine meeting included NIPT results in 1611 pregnancies following oocyte donation, compared to a large set of reference cases matched for maternal weight and gestational age. The average FF for donor oocyte pregnancies was found to be significantly lower [71].

4.5. Role of NIPT in Preimplantation Genetics

Invasive prenatal procedure is the main stay investigative tool for confirmation of preimplantation genetic results for hereditary disorders. However, a substantial fraction of pregnant women decline invasive testing due to the small potential risk of miscarriage associated with it. A few studies have addressed this concern and are now offering NIPT as a substitute for invasive testing for those women who have declined invasive procedures, although it only serves as an advanced screening tool for verification of PGT results [97].

To summarize, current literature suggests that ART pregnancies against normal conceptions have emerged as an independent risk factor for generating low FF. Furthermore, fresh embryo transfer is associated with a lower yield of FF when compared to frozen embryo transfer. The use of exogenous hormone for endometrial preparation and donor oocytes in ART also contributed to lower FF. NIPT is advisable in twin pregnancies, although the failure rate is higher as compared to singleton pregnancies. In dizygotic twin, SNP-based platforms may be useful since they help to assess zygosity as well as FF contributed by an individual fetus. NIPT in cases of a vanishing twin is associated with a lower PPV as well as a higher test failure, and hence, resampling after 15 weeks has been suggested for these situations.

There is a need for close coordination between ART practitioner, obstetrician who advises prenatal testing, the genetic laboratory performing NIPT, and the genetic counselor, for optimal outcome. All the ART treatment-related details should be clearly documented and taken into account while interpreting the NIPT results in the prenatal setting.

CONCLUSION

NIPT is currently the most sensitive and specific screening method and an alternative to traditional screening, which may be offered to all pregnant women irrespective of risk. In spite of the evident superiority over traditional screening methods, there are factors in ART pregnancies that may affect the performance of NIPT. These include lower fetal fractions, frequent twin pregnancies, and vanishing twins. NIPT has been found to have higher failure rates and higher false-positive rates in ART pregnancies with twins, vanishing twin, oocyte donation and fresh embryos (as compared to frozen). Clinicians need to be aware of these issues so that they apply the necessary caution when acting on results. Despite these limitations, the performance of NIPT is still superior to conventional screening in the setting of ART.

Guidelines and recommendations on NIPT in the ART setting are sparse, mainly because of limited data. Further study into the causality of lower FF in ART pregnancies as compared to spontaneous conceptions is indicated as well as the impact on FF following different types of ART treatment.

As more data emerge for ongoing and future studies, it is hoped that current guidelines will be updated and will address the efficacy and limitations of NIPT in ART pregnancies.

LIST OF ABBREVIATIONS

ART

Assisted Reproductive Technology

FISH

Fluorescent In-Situ Hybridization

IVF

In-vitro Fertilization

MPS

Massively Parallel Sequencing

NIPT

Non-invasive Prenatal Testing

PPV

Positive Predictive Value

QF-PCR

Quantitative Fluorescent Polymerase Chain Reaction

AUTHORS’ CONTRIBUTION

All authors contributed to the planning and writing of the review. All authors approved the final draft.

CONSENT FOR PUBLICATION

Not applicable.

FUNDING

None.

CONFLICT OF INTEREST

The authors declare no conflict of interest, financial or otherwise.