Cytogenetic and Microarray Analysis Follow‐Up of PGT‐A Mosaic and Sex Discrepant Embryos During Pregnancy: Absence of Confirmation and Follow‐Up Recommendations

ABSTRACT

Objectives

The objective of this study was to examine the concordance of results between mosaic pre‐implantation genetic testing for aneuploidy (PGT‐A) and diagnostic testing in resulting pregnancies conceived by IVF with PGT‐A.

Methods

This study retrospectively provides analysis of 328 patients ascertained due to a mosaic PGT‐A finding and 25 with a discordant reported sex, studied prenatally, utilizing standard karyotype and/or SNP microarray analysis.

Results

None of the 328 mosaic PGT‐A findings were confirmed by the diagnostic prenatal studies with karyotype and/or microarray analysis. Additionally, of the 298 prenatal microarray studies performed, five demonstrated pathogenic abnormalities (1.7%) that were unrelated to PGT‐A referral. Thirty‐four of these cases were also studied to determine if UPD occurred; all were biparental.

Conclusions

This work emphasizes the importance of PGT‐A follow‐up and demonstrates that the vast majority will have normal diagnostic results and only a small number will be confirmed using prenatal diagnostic analysis. This work also suggests the most efficient way to proceed with diagnostic follow‐up studies. Additionally, it shows the importance of fetal sex confirmation in all normal PGT‐A studies.

key Points

• What is already known about this topic?

  • ◦PGT‐A is used to test embryos for chromosomal abnormalities in IVF cycles resulting in abnormal, normal, and mosaic embryos.
  • ◦Mosaic embryos are sometimes transferred but there is limited follow‐up published.

• What does this study add?

  • ◦This is one of the largest cohorts of patients following PGT‐A demonstrating an embryo with a mosaic result, studied prenatally with karyotype and/or microarray analysis.
  • ◦Transferring a mosaic embryo almost always results in a cytogenetically normal fetus at the time of prenatal diagnosis.

1. Introduction

Preimplantation genetic testing for aneuploidy (PGT‐A) has become an integral component of assisted reproductive technology (ART) and is used to identify chromosomally normal embryos prior to transfer, particularly to address aneuploidy related to maternal age. PGT‐A provides valuable early recognition of possible chromosome abnormalities before embryo transfer occurs. This approach increases the likelihood of successful implantation and delivery of a healthy baby achieved by IVF [1, 2]. PGT‐A involves analyzing a biopsy from the trophectoderm of a blastocyst‐stage embryo, which is then screened for chromosomal abnormalities. It detects full chromosome aneuploidy as well as segmental gains and losses, with the resolution varying from 2 to 20 Mb across different laboratories. The typical reporting threshold is approximately 10 Mb, but laboratories have various criteria for reporting mosaicism on PGT‐A [3, 4]. This reporting could involve a 20%–80% threshold or a 30%–70% threshold, while a 50% classification has been suggested by some [5]. Identification of euploid embryos has become an important aspect of successful in vitro fertilization (IVF) cycles [1, 2].

An embryonic or fetal chromosome abnormality could result in failure to implant, pregnancy loss, complications during gestation, or the birth of an infant with a complex structural and/or neurodevelopmental anomaly. The outcomes often depend on the specific chromosome involved; however, the vast majority of these will result in a pregnancy loss. In ongoing pregnancies, Trisomy 13, 18 and 21 are the most common abnormalities that are screened for utilizing cfDNA technology, although the vast majority of gains and losses can be detected by diagnostic tests, such as karyotype or microarray analysis following amniocentesis or chorionic villus sampling [6]. Patients who do not have euploid embryos for transfer after PGT‐A may be offered the option to transfer embryos with abnormal PGT‐A results, following thorough counseling and informed consent. PGT‐A results may be utilized to prioritize the abnormal embryos, with the knowledge that successful implantation and chance of ongoing pregnancy are lower [7].

The concordance of PGT‐A results with subsequent prenatal and postnatal diagnostic testing remains an area of interest and concern and has tremendous implications when following IVF pregnancies, particularly following mosaic embryo transfer. Surrey (2021) observed that Viotti's large retrospective study on 1000 transferred mosaic embryos, and other prior studies are lacking follow up regarding pregnancy outcome and results of prenatal or postnatal genetic diagnostic testing [3, 8]. Several studies have reported discordance between PGT‐A and follow‐up diagnostic testing results [2, 3, 9]. Factors contributing to the discordance may include technical issues, such as biopsy sampling errors, and biological factors, such as embryonic self‐correction or confined placental mosaicism (CPM).

In this manuscript, we report one of the largest cohorts of prenatal diagnostic follow up of PGT‐A embryos with mosaic results based on the experience of one large diagnostic laboratory. The objective of this study was to examine the concordance of PGT‐A results with subsequent diagnostic testing in a large cohort of pregnancies with mosaic PGT‐A results. By focusing on diagnostic outcomes, we aim to provide practical data to guide clinical management and expectations for these high‐risk pregnancies. Additionally, we explore the potential utility of uniparental disomy (UPD) testing and evaluated incidental abnormalities unrelated to the original PGT‐A indication. The results of this study contribute valuable information for both patients and clinicians about possible benefits, limitations, and recommended follow‐up testing for these pregnancies.

2. Methods

2.1. Ascertainment of Patients and Classification of PGT‐A Findings

We report on the analysis of 328 prenatal patients (297 singleton, 11 twin pairs where both twins were studied and 9 twins where only one was studied) ascertained due to a mosaic PGT‐A finding and subsequently studied utilizing standard karyotyping (at the 450 band level or greater) and/or SNP microarray analysis, collected from 2013–2025.

In addition, there were 25 other patients (all singletons) referred because of a discrepancy between the fetal sex identified by PGT‐A studies and that denoted either by ultrasound (US) or cfDNA testing.

2.2. Chromosome and Microarray Methodology

All karyotyping was done using standard G‐banded analysis. The chromosomes were analyzed at the 450‐band level or greater, allowing the detection of alterations that were 10 Mb or greater (although often smaller changes could be detected).

The majority of the microarray studies were performed utilizing the Thermo Fisher Scientific CytoScan HD array [CytoScan HD are Registered Trademarks of Thermo Fisher Scientific]. This array contains approximately 2.695 million markers across the entire human genome. There are approximately 743,000 SNPs and 1,953,000 non‐polymorphic probes (NPCNs).

On average, there is approximately 0.88 kb between markers. DNA was extracted utilizing standard methods, and 250 ng of total genomic DNA was digested with NspI, ligated to adaptors and amplified using Titanium Taq with a GeneAmp PCR System 9700. PCR products were purified using AMPure beads and quantified using NanoDrop 8000.

From February 2024—February 2025, SNP microarray analysis was performed using the Cytoscan HD Accel platform, which uses 2,029,441 nonpolymorphic copy number probes and 743,130 SNP probes. The Accel array has an average intragenic spacing of 0.818 kb and an average intergenic spacing of 1.51 kb. Total genomic DNA was extracted from the sample type provided, digested with Xce1, and then ligated to Xce1 adaptors.

For both systems, purified DNA was fragmented, and biotin labeled and hybridized to the Affymetrix GeneChip. Data were analyzed using Chromosome Analysis Suite (ChAS 4.0—Cytoscan HD or ChAS 4.4—Cytoscan Accel). The analysis is based on the GRCh37/hg19 assembly.

2.3. Detection of CNVs

Copy number variation was identified by examination of the log2 ratio, smooth signal and both B‐allele and allele difference patterns. Examination of these parameters identified regions of interest at 25 kb or greater for both gains and losses. Based on the patterns, mosaicism, involving at least 10% of the sample for any abnormality > 10 Mb, could also be identified.

2.4. Detection of Homozygosity

Homozygosity was identified in these cases by examination of both B‐allele and allele difference patterns. Homozygosity of more than 8 Mb observed in at least two chromosomes suggests common descent. Patient results demonstrating a long contiguous region of homozygosity (ROH) in a single chromosome of more than 20 Mb interstitially or more than 10 Mb telomerically (15 and 8 Mb, respectively, for imprinted chromosomes) were recommended for follow‐up uniparental disomy (UPD) testing.

2.5. Confirmation of UPD

Confirmation of UPD was performed by one of three methods, depending on which chromosomes were involved. For non‐imprinted chromosomes, Mendelian Inheritance Errors (MIEs) in parent‐offspring trios were calculated to determine whether the genotype in the proband follows Mendelian rules of inheritance [10]. In those cases where all other chromosomes followed the normal inheritance pattern, but the chromosome of interest did not, UPD was detected. Similarly, for most of the imprinted chromosomes, Chromosomal Variable Number Tandem Repeats (VNTRs) located along the chromosome of interest were amplified by polymerase chain reaction (PCR) and sized by polyacrylamide gel electrophoresis. For a specific chromosome, whole chromosome UPD is established if chromosomal inheritance of the VNTRs is uniparental and whole chromosome UPD is excluded if chromosomal inheritance of the VNTRs is biparental.

2.6. Methylation

In some cases, molecular analysis of the SNRPN gene was used to assess UPD for chromosome 15, which is performed by methylation‐specific PCR and gel electrophoresis. Methylation‐specific PCR will determine the presence of both maternal and paternal alleles consistent with a normal imprinting pattern in the Prader‐Willi‐Angelman critical region on chromosome 15q11‐13. This assay nearly detects all cases of AS and PWS arising from UPD, microdeletions, and imprinting defects.

3. Results

This study primarily looks at the confirmation of findings involving reported mosaic findings detected in the PGT‐A analysis. Secondarily, it examines the frequency of pathogenic abnormalities detected by both karyotype and microarray analyses in prenatal patients referred with mosaic results detected by PGT‐A, where the diagnostic results are unrelated to the reason of referral. The study also has broad specific implications regarding the best approaches for follow‐up of pregnancies referred with the finding of mosaic results by PGT‐A.

3.1. Classification of PGT‐A Findings and Follow‐Up Testing

As previously indicated, 328 prenatal patients with reported PGT‐A mosaic results were studied. The vast majority of these specimens were submitted as amniotic fluid (312/328) and submitted after 15 weeks of gestation. Only 4.9% (16 specimens) were submitted as CVS samples at an earlier gestation.

Overall, these cases comprised 130 that included only aneuploid mosaic results, 9 with both an aneuploid and segmental (e.g., deletion and duplication) chromosome mosaic result, 7 with a deletion and duplication, 95 with a deletion, 52 with a duplication, 11 with complex findings (3 or more mosaic results including at least two different types), and 24 without an indication (categorized as unspecified) with respect to mosaic results reported by PGT‐A (Table 1; Figure 1).

TABLE 1.

Type of PGT‐A mosaic result and subsequent follow‐up.

Type of abnormality	Total	Array only	Chroms only	Both
Aneuploid	130	7	18	105
Aneuploid/segmental	9	1	1	7
Deletion/duplication	7	1		6
Deletion	95	14	1	80
Duplication	52	5	2	45
Complex	11	2		9
Unspecified	24	4	8	12
Total	328	34	30	264

Note: This Table lists the different types and distribution of mosaic results detected by PGT‐A and the follow‐up that was done to confirm the abnormality.

FIGURE 1.

This Figure shows the distribution of different types of mosaic results detected by the PGT‐A analysis that were referred for cytogenetic and/or array analysis.

While information was received on most patients delineating the mosaic results detected by PGT‐A, specific reports were received on only 210 patients. These reports revealed that the PGT‐A analysis was performed at least at 24 different laboratories. While the specific technologies may have varied from 2013–2025, all laboratories utilized NGS based approaches.

Most cases were assessed by both karyotype and microarray analysis (264 studies out of a total of 328). There were 30 cases which were studied by karyotype only and 34 cases were studied by microarray only. While only 9.1% of cases in the overall cohort were analyzed via karyotyping alone, within the aneuploid PGT‐A referrals and unspecified PGT‐A referrals, karyotyping only was requested by the referral physician in 13.8% (18/130) of the mosaic aneuploid PGT‐A referrals and 33.3% (8/24) of unspecified PGT‐A referrals (Table 1).

3.2. Concordance With PGT‐A Findings

Analysis of the 328 cases revealed that none of the mosaic results seen by PGT‐A analysis were confirmed by prenatal diagnostic studies. While this included 24 patients in whom the precise mosaic result was not delineated, all of these were normal as detected by karyotype/microarray analysis.

3.3. Examination of Uniparental Disomy

Thirty‐four of these patients (involving a total of 39 chromosomes) underwent follow‐up studies to rule in/out uniparental disomy (UPD). Thirty‐two of the 39 UPD follow‐up studies involved imprinted chromosomes (11 involving chromosome 7; two involving chromosome 11; 10 involving chromosome 14; five involving chromosome 15; and four involving chromosome 20). The seven cases referred to rule in/out UPD of a non‐imprinted chromosomes looked at chromosomes 1,8,16,18 and 22.

All UPD studies of the mosaic PGT‐A indication cases revealed biparental inheritance. Additionally, microarray analysis of the 39 chromosomes requested from the mosaic PGT‐A cases and studied for UPD (which were biparental) did not reveal any large regions of homozygosity.

3.4. Detection of Unexpected/Incidental Microarray Findings

Clinically significant pathogenetic copy number variants unrelated to the PGT‐A findings were detected by microarray analysis in 5 of the 298 (1.7%) cases tested (Table 2). These included a distal 16p11.2 microdeletion [11], a 1q21.1 microduplication [12] and a de novo deletion of 5p that included the gene SDHA, which is associated with Hereditary Paraganglioma‐Pheochromocytoma Syndromes with low penetrance [13]. In addition, there were two mosaic deletions seen in the microarray analysis that were not detected by standard karyotyping. One involved a 26.48‐Mb terminal deletion of Xq26.1‐ > q28 detected in 32% of the cells and the other a 6.85‐Mb interstitial deletion of 13q21.32‐ > q22.1 in 40% of the cells. In the latter case, there was also a concomitant 50% mosaicism for 13q allele homozygosity extending from q21.32 to the telomere of the long arm. This was possibly due to post‐zygotic loss of the 13q segment in early development and subsequent repair.

TABLE 2.

Microarray abnormalities detected.

PGT‐A abnormality	PGT‐A chrom	Microarray abnormality	Array nomenclature	Parental studies
		Pathogenic deletion or duplication
Deletion	6Q	Pathogenic deletion	16p11.2 (28, 689, 085‐29, 088, 624) × 1	Not studied
Aneuploid/deletion	10, 2Q	Pathogenic deletion	5p15.33 (113, 576‐1, 232, 981) × 1	De novo
Aneuploid	15	Pathogenic duplication	1q21.1q21.2 (146, 018, 964‐147, 926, 347) × 3	Not studied
Aneuploid	10	Mosaic pathogenic deletion (32%)	Xq26.1q28 (128, 751, 938‐155, 233, 731) × 1[0.32]	Not studied
Deletion	14Q	Mosaic pathogenic deletion (40%) Segmental UPD	13q21.32q22.1 (68, 406, 974‐75, 258, 150) × 1 [0.6] 13q21.32q34 (68, 406, 974‐115, 169, 878) × 2 mos hmz	Not studied

Note: This Table demonstrates the different types of microarray copy number variations that have been detected in this study following up the abnormal embryos. None were related to the reason for referral.

3.5. Detection of Unexpected/Incidental Chromosome Findings

Karyotyping, as with microarray studies, in addition to being able to confirm the PGT‐A findings can reveal other changes in the genome, possibly unrelated to the referral indications. Results from the 294 PGT‐A patients with the follow‐up karyotype revealed that none of the patients had a gain or loss of chromosomal material. However, a balanced translocation was identified in six patients (2.0%) (Table 3). One of these patients also had a balanced inversion. In three of these patients, the translocation was familial (all paternally inherited), whereas in two patients, the translocation was de novo and in one case, there was no follow‐up study.

TABLE 3.

Chromosome abnormalities detected.

PGT‐A abnormality	PGT‐A chrom	Chromosome abnormality	Parental studies
Unknown	Unknown	46, XY, t (1; 7) (q12; q31.3)	Paternal
Aneuploid	21	46, XY, t (3; 8) (p24.2; p21.3),inv (12) (p13.3q15)	Paternal
Aneuploid	8	46, XY, t (7; 18) (p11.2; p11.2)	Not studied
Deletion	9Q	46, XX, t (1; 3) (p32; 21)	De novo
Deletion	2P, 5P	46, XX, t (1; 7) (q22; q35)PAT	Paternal
Aneuploid	17	46, XY, t (1; 16) (p32; p13.3) dn	De novo

Note: This Table demonstrate the different types of chromosome abnormalities that have been detected in this study following‐up the abnormal embryos. None were related to the reason for referral.

3.6. Evaluation of Sex Discrepancies

During this study, there were 25 patients referred because of a discrepancy between the sex designated by the PGT‐A analysis and that specified from a subsequent cfDNA study or US evaluation. The vast majority of these samples were amniotic fluid (23/25 specimens) received after 15 weeks of gestation. Only 2 specimens were chorionic villus samples.

Ten embryos were classified as female by PGT‐A studies and male by cfDNA or US studies. In all cases, the microarray analysis confirmed the male classification from cfDNA/US. However, in one case of a sex discrepancy referral, the microarray analysis demonstrated the presence of a pathogenic microdeletion of 1q21.1 [14]. Thirteen embryos were classified as male by PGT‐A studies and female by cfDNA or US studies. In all cases, the microarray and/or karyotyping confirmed the female assignment from cfDNA/US. Lastly, two fetuses were assigned male by PGT‐A studies and cfDNA screening indicated a 45,X genotype, but a male by US. In one case, the microarray analysis revealed a normal female result. However, in the other case, initial karyotyping on a CVS demonstrated a 45,X[37]/46,X,+mar [3] karyotype. Subsequent microarray studies on amniotic fluid demonstrated both a mosaic terminal gain of Yp and non‐mosaic terminal Yq loss consistent with an isodicentric Y chromosome. The isodicentric Y was detected in ∼80% of the cells and in the other 20% of cells there was a single X chromosome.

4. Discussion

In the United States, some form of infertility is experienced by approximately 11% of women and 9% of men of reproductive age [15, 16]. Couples with infertility issues, same‐sex couples and individuals who are single but wish to have a child may pursue assisted reproductive technology (ART) to build their family. It is estimated that approximately 86,146 infants were born in 2021 using ART. Preimplantation genetic testing (PGT‐A) is commonly used in pregnancies achieved through IVF to screen embryos for chromosome aneuploidy and to prioritize embryos for transfer [17].

The optimal goal of PGT‐A following IVF is to implant a cytogenetically normal and viable embryo. However, studies have repeatedly indicated that a cytogenetically normal embryo is not always available for transfer. Instead, in some instances, only an embryo with a mosaic or nonmosaic abnormality detected by PGT‐A studies is available for transfer. Depending on the criteria used to define mosaicism detected by PGT‐A (which does vary between PGT‐A labs), it is estimated that the overall incidence of mosaicism in trophectoderm is between 4% and 32% [18]. While the transfer of an aneuploid embryo is generally not expected to result in a pregnancy, if a mosaic embryo transfer does result in pregnancy which progresses to the point where prenatal diagnosis could be performed, this already strongly suggests that whatever abnormality was present in the original PGT‐A test was not sufficient to cause embryo implantation failure. The purpose of this manuscript is not to assess PGT laboratory performance but to provide clinically relevant outcome data for providers and genetic counselors managing pregnancies following reported mosaic or sex discrepant PGT‐A results, regardless of the underlying cause of the initial finding. The overall goal is to also provide some guidance to providers and diagnostic laboratories regarding the best approaches for follow‐up testing in the prenatal period.

4.1. Confirmation of PGT‐A CNV Findings

In this study, there were a total of 328 transferred embryos with documented mosaic results that were followed up with diagnostic prenatal tests, either microarray or karyotype. None of these analyses confirmed the mosaic finding from the PGT‐A studies. In addition, 34 of these (involving 39 chromosomes) were studied to determine if UPD was present. All of these demonstrated biparental inheritance.

This is not the first diagnostic study to follow‐up mosaic PGT‐A findings, but it is the largest prenatal follow‐up by a single diagnostic laboratory. The results of this study largely match the results in the published literature to date. Many of the previous studies looked for the presence of a chromosome abnormality at birth. Almost 1500 babies born from mosaic embryos were examined at birth, in five studies, and none described a live‐born child with a chromosome abnormality [3]. The initial report of a mosaic result was in 2020 when Kahraman reported on the transfer of a mosaic (35%) monosomy 2 [19]. Amniocentesis did not confirm monosomy 2 but identified mosaicism for trisomy 2 in a fetus with no pathologic findings on ultrasound. At birth, a phenotypically normal baby was born; however, the karyotype revealed 2% mosaic monosomy (2/100 cells). Another patient, reported by Schlade‐Bartusiak in 2022, transferred an embryo with mosaic trisomy 15 and a mosaic deletion of 20q11.23‐qter [20]. Postnatal analysis, prompted by feeding difficulties in the newborn period, revealed a supernumerary chromosome involving the proximal arm of chromosome 15 and UPD for chromosome 15. The authors surmised that this likely resulted from incomplete trisomy rescue of chromosome 15.

The most comprehensive study to date has been a multi‐center study of PGT‐A mosaic embryos (20–80% mosaicism) reported by Viotti et al. (2023) [21]. In general, this multi‐center study demonstrated that for mosaic embryos, which were transferred, there were (1) a lower implantation rate of the embryos and (2) a higher miscarriage rate, and (3) those embryos that went to term resulted in healthy neonates. Of the 488 babies that were born, only one had overt phenotypic abnormalities. There was prenatal testing for 250 of these 488 embryos. They reported that there were 9 nonmosaic abnormalities that were either benign findings or variants of uncertain significance (VUS); 2 mosaic findings were unrelated to the PGT‐A, and 3 mosaic findings were consistent with the PGT‐A findings. Taken together, these studies have demonstrated that when mosaic PGT‐A embryos are transferred, the preponderant outcome is a normal birth, which has led to several proposed models regarding mosaic embryos. These include the (1) embryonic mortality model, where embryos with high mosaicism do not survive implantation; (2) the clonal depletion model, where there is either apoptosis or reduced propagation of aneuploid cells; and (3) the trisomic/monosomic rescue model, where aneuploid cells give rise to diploid cells via mitotic chromosome gain or loss [22]. These models could account for the low number of mosaic results confirmed through diagnostic prenatal testing. However, it has also been suggested that there may be false positive mosaicism detection in PGT‐A due to several possible factors. These include bias in the whole genome amplification, number of laser pulses, number of biopsied cells, mechanical damage to cells, and introduction of cellular fragments into PCR tubes [22]. This remains an area of ongoing study.

4.2. Detection of Unexpected/Incidental Microarray Findings

In this study, a total of 298 of these patients underwent prenatal diagnostic microarray analysis. There were five (1.7%) pathogenic abnormalities; 2 which involved neurodevelopmental disorder microdeletion/duplication syndromes, one cancer susceptibility with low penetrance and two mosaic deletions that were not detected by standard karyotyping. Again, these frequencies are consistent with both our internal data and literature studies involving unselected patients referred without any ultrasound abnormalities [23]. This suggests that the background risk of a CNV unrelated to the PGT‐A result is likely similar to that of the general obstetric population undergoing prenatal diagnosis. It also highlights that both patients and ordering providers need to be aware of the possibility that these unexpected results may be detected.

Also in this study, a total of 294 of these patients had prenatal diagnostic karyotyping. There were six patients (2.0%) identified with a balanced translocation. In three of these patients, the translocation was familial, in two the translocation was de novo, and in one there was no follow‐up study. The two de novo translocations were greater than expected, but neither had any detectable associated copy number changes. The three familial cases are not unexpected, as some of these patients may have had IVF due to a familial translocation.

4.3. Evaluation of Sex Discrepancies

One byproduct of this study was the referral of 25 patients because of a sex discrepancy between the sex designated by the PGT‐A analysis and that specified from a subsequent cfDNA study or US evaluation. In most cases, the microarray study was consistent with the cfDNA and/or US findings, and not the PGT‐A findings. However, in one case, after initial karyotype on a CVS demonstrated a 45,X[37]/46,X,+mar [3] karyotype, the subsequent microarray studies on amniotic fluid demonstrated an isodicentric Y in ∼80% of the cells and a single X chromosome in the other 20% of cells.

These findings are not believed to indicate a problem with embryo transfer as they are only a small portion of the IVF transfer cases (estimated to be < 1%) but to highlight the importance of follow‐up in cases of sex discrepancy by a diagnostic test [24]. As shown, these will not only confirm the sex of the fetus but can also illustrate the presence of a sex chromosome abnormality resulting in a sex discrepancy.

4.4. Approaches/Recommendations for Diagnostic Evaluation of PGT‐A Pregnancies

Based on the multi‐center study published by Viotti et al. (2023), in which 3/250 (1.2%) embryos with a mosaic result studied diagnostically were found to have the same abnormality detected by PGT‐A, it was strongly suggested that mosaic IVF pregnancies that are transferred should be monitored for a cytogenetic abnormality [21]. There are three major ways to follow‐up pregnancies for cytogenetic abnormalities: interphase FISH analysis, standard karyotyping, and microarray analysis. Even though interphase FISH analysis is the quickest, it is the least practical and should not be utilized. Most laboratories do not have probes validated for prenatal interphase analysis outside chromosomes 13, 18, 21, X, and Y. Even for these probes, it is difficult to classify an abnormality as mosaic. There are established guidelines for the detection of mosaicism for karyotypes prenatally [25]. However, only 130 of the 328 pregnancies studied in this cohort demonstrated only an aneuploid mosaic result by PGT‐A. Approximately 63.5% of the referrals indicated that there was a segmental abnormal mosaic result (e.g., deletion, duplication) and often no specific breakpoint was provided. Thus, in these cases microarray analysis is preferable as it allows detection of any change that might not be visible by karyotype. While karyotypes can be utilized with aneuploid referrals, this study has indicated that 1.7% of the patients had a pathogenic abnormality detectable only by microarray studies, a finding consistent with the utilization of microarray in patients without US abnormalities [26]. Given that these pregnancies are especially valued by the parents, providing the additional information revealed by a microarray would be the most efficacious approach for the diagnostic follow‐up studies. Furthermore, this emphasizes that regardless of PGT‐A results, patients should be counseled that there is still residual risk of a chromosome abnormality in an ongoing pregnancy. Professional societies recommend that patients should still be offered prenatal screening and diagnosis, even in pregnancies conceived following PGT‐A [27, 28].

The reason why the prenatal follow‐up results are often normal after detection of an embryo with a mosaic PGT‐A result has not been completely settled. These could be due to different factors, other than chromosomal constitution of the biopsied cells consistent with false positive results. Alternatively, it may be due to loss of the abnormal cell line, which might have implications for uniparental disomy. Therefore, there is some justification for follow‐up UPD testing when an imprinted chromosome has been implicated by PGT‐A testing. In this study, there were no UPD findings among the mosaic cases. Overall, the utilization of UPD testing for non‐imprinted chromosomes has less justification but can be considered if microarray testing reveals a larger region of homozygosity that could be indicative of UPD.

Lastly, our studies revealed that 1 of 25 cases of sex discrepancy between PGT‐A results and cfDNA/US findings had an underlying sex chromosome aberration. There are a number of possible explanations outside a sex chromosome aberration resulting in a sex discrepancy. These include: cumulous cell contamination giving a euploid female result on PGT‐A when in fact the embryo is euploid male, other extraneous DNA obscuring embryo results, spontaneous pregnancy and unintended embryo transfer. Although ∼99% of all PGT‐A results are correct, it is advisable to confirm fetal sex by cfDNA or US and follow‐up any discrepant findings with microarray analysis.

5. Conclusions

In this manuscript, a large cohort of mosaic PGT‐A results was followed up with diagnostic prenatal karyotype and/or microarray analysis to confirm the CNV detected in the PGT‐A studies. None of the 328 PGT‐A mosaic findings were confirmed by diagnostic prenatal studies. Additionally, prenatal diagnostic microarray analysis detected five pathogenic abnormalities (1.7%) that were only detected by microarray analysis. Further, this study demonstrates the importance of fetal sex confirmation in all normal PGT‐A studies. Finally, this work emphasizes the importance of PGT‐A follow‐up but also shows the small number that will be confirmed using prenatal diagnostic analysis. This work also presents the most efficient way to proceed with diagnostic follow‐up studies.

Funding

The authors have nothing to report.

Ethics Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Institutional Review Board at Aspire IRB under the clinical protocol SCMM‐RND‐402 (NCT04364503).

Consent

Informed consent was not required as Aspire IRB declared that this research meets the requirements for a waiver of consent under 45 CFR 46 116(f)[2018 Requirements].

Conflicts of Interest

All authors are current or former employees of or consultants for Labcorp, a commercial diagnostic laboratory. Each individual has an option to hold stock. Vanessa Nitibhon is a current employee of Vitrolife Group.

Kline, Laura A. , Nitibhon Vanessa A., Dyr Brittany, Soster Erica L., Sussman Amanda, Weinblatt Vivian J., and Schwartz Stuart. 2026. “Cytogenetic and Microarray Analysis Follow‐Up of PGT‐A Mosaic and Sex Discrepant Embryos During Pregnancy: Absence of Confirmation and Follow‐Up Recommendations,” Prenatal Diagnosis: 46. no. 8), 1151–1158. 10.1002/pd.70132.

Data Availability Statement

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.