Preimplantation genetic testing for aneuploidy (PGT-A)—a single-center experience

Abstract

Purpose

The aim of this study was to determine the prevalence and nature of human embryonic aneuploidy based on the preimplantation genetic testing for aneuploidy (PGT-A), the distribution of aneuploidy across the individual chromosomes, and their relationship to maternal age.

Methods

This is a retrospective cohort study conducted at a single center. The study includes subjects who opted for PGT-A in their in vitro fertilization (IVF) cycle from 2016 to 2020. PGT-A was performed on 1501 embryos from 488 patients in 535 cycles. PGT-A was performed using NGS-based technique on Ion Torrent PGM (Life Technologies). Analysis was performed to determine the (i) frequency of the aneuploidy, (ii) the chromosome most commonly affected, (iii) relationship between maternal age and the rate of aneuploidy, and (iv) incidence of segmental aneuploidy.

Results

The overall frequency of aneuploidy was observed to be 46.8%. The incidence of aneuploidy rate was ~ 28% at maternal age < 30 years which steadily increased to ~ 67% in women above 40 years. High frequency of aneuploidy was observed in chromosomes 16, 22, 21, and 15. Segmental abnormalities, involving loss or gain of chromosomal fragments, were observed at a frequency of 5.3%, and highest incidence of segmental gain was observed on the q-arm of chromosome 9.

Conclusion

The study provides important information regarding the frequency of the aneuploidy in IVF cohort and the most frequent chromosomal abnormality. The study further emphasizes the relationship between maternal age and aneuploidy. This study has important implications which help clinicians and genetic counselors in providing information in patient counseling.

Introduction

Embryonic aneuploidy is one of the major limiting factors in achieving a successful pregnancy during in vitro fertilization (IVF) procedures [1]. Aneuploidy is naturally occurring, and is observed at a frequency of more than 50% in blastocyst stage embryos [2]. The observed high rate of chromosomal aneuploidies in preimplantation human embryos arises due to abnormal meiotic divisions, especially in women with advanced maternal age (AMA) while segmental aneuploidies and mosaicism can also arise as a result of aberrant mitotic divisions [3].

Assisted reproductive technologies (ARTs) have emerged as useful therapeutic options for infertile couples in order to achieve successful pregnancy. The success of ART depends on multiple factors, such as the quality and genetic status of the embryo, endometrial receptivity, and embryo transfer techniques [4, 5]. Although multiple factors contribute to the success of ART, evidence supports the notion that aneuploidy is one of the most common limiting factors. In human embryos, 65% of aneuploidies are lethal, leading to implantation failure or miscarriage. Although a majority of chromosomally abnormal embryos culminate in spontaneous miscarriages, few of them survive to term and is the leading genetic cause of developmental disabilities and intellectual disability [6]. This has led to the development of preimplantation genetic screening (PGS), a process of detecting the chromosomal status of embryos and thus transferring euploid embryos, which can increase pregnancy and live birth rates and reduce miscarriages in couples undergoing IVF [7].

Different strategies have been used for selecting an embryo with the highest implantation potential. Historically, the microscopic morphological appearance of the embryo was the only criteria available for selection of “healthy” embryos. Unfortunately, there is no correlation between appearance of the embryo and its genetic constitution [8]. Hence, perfect-appearing embryos may carry chromosomal abnormalities. PGS is a way to detect chromosomal abnormalities in the embryo. The original technique used fluorescence in situ hybridization (FISH) which was limited to just a few chromosomes and was performed on early-stage embryos. PGS technologies have since improved and now include assessment of all the chromosomes, using advanced techniques such as array comparative genomic hybridization (aCGH) or next-generation sequencing (NGS) on trophectoderm stage embryos [9]. Refinement in blastocyst culture techniques as well as advancements in molecular diagnostic technology has resulted in the revision of PGS terminology to preimplantation genetic testing for aneuploidy (PGT-A) [10]. NGS has become the method of choice in reproductive medicine for PGT-A. Our center performs IVF/PGT-A in a selected group of couples using the NGS-based approach. This study aims a retrospective analysis of PGT-A results from a single center to determine the (i) rate of aneuploidy in patients undergoing IVF (ii) relationship between maternal age and the aneuploidy rate and (iii) analyze the type and chromosomal distribution of segmental aneuploidy.

Materials and methods

Study subjects

This is a retrospective study performed on 1501 embryos from couples who underwent all chromosomes IVF-PGT-A at Gunasheela Surgical and Maternity hospital from January 2016 to December 2020. A total of 488 patients exhibited one or more of the following indications: advanced maternal age (AMA) (≥ 35 years), recurrent implantation failure (RIF) (≥ 2 IVF failure), recurrent miscarriages (RM) (≥ 2 pregnancy losses), or male factor (MF) and chose to undergo PGT-A. Cases where either of the couple showed an abnormal karyotype or were undergoing IVF/ICSI with donor egg or sperms were excluded from the study. The study was approved by the institutional ethical committee, and written informed consent was obtained from all the participants before undergoing IVF or PGT-A procedure.

Ovarian stimulation and embryo biopsy

All the participants underwent controlled ovarian stimulation, using either the long agonist, antagonist, or flare protocol depending on the clinical scenario of the patient. The follicular response was monitored at regular intervals by transvaginal pelvic ultrasound scans, and ovulation was induced with hCG (250mcg) when at least two follicles had reached a diameter of 17 mm. Ultrasound-guided oocyte retrieval was performed 34–36 h after hCG injection. The aspirated follicular fluid was screened under the microscope to identify and to assess the cumulus oocyte complex (COC). The COC was denuded mechanically using a 150um Denupet. Simultaneously, the semen sample obtained from the husband was processed using density gradient centrifugation or wash method in combination with the swim-up technique depending on the semen parameter. The intracytoplasmic sperm injection (ICSI), involving the injection of matured oocyte with single sperm, was carried out using standard protocol. The embryos were incubated in a trigas incubator and cultured until day 5 in One-Step media (Vitromed, USA). Expanded blastocysts equivalent to Gardner blastocele expansion score of 3 to 6 were biopsied for PGT-A. The trophectoderm biopsy was performed using laser-assisted hatching. Six to seven trophectoderm cells were carefully removed from the portion of the embryo opposite the inner cell mass and transferred into thin walled 0.2-mL PCR tubes containing 0.5ul 1X phosphate buffer solution.

PGT-A using NGS

PGT-A for all 23 chromosome pairs was performed on the Ion Personal Genome Machine® (PGM™) System (Thermo-Fisher Scientific, MA, USA). The biopsied samples were whole genome amplified (WGA) and barcoded using the ION single Seq kit (Thermo Fisher Scientific, USA) according to manufacturer’s protocol. Barcoded WGA products were pooled, purified, quantified, and processed by following the Thermo Fisher Scientific Ion Reproseq PGS library preparation kit protocol. The sequencing run was performed using the Ion PGM Sequencing 500 Kit v2, and samples were loaded on to Ion 316 Chip Kit v2. The data generated was analyzed using the Ion Reporter Software version 5.0 for read filtering, base calling, barcode filtering, and alignment to the human genome hg19 reference. Aneuploidy detection was performed on the Ion Reporter software using default high-sensitive ReproSeq low-pass whole-genome aneuploidy workflow v5.2 that can detect whole-chromosome copy number variation (CNV) events and sub-chromosomal CNV events of ≥ 8 Mb. The decimal-level copy number gain or loss calls are enabled in the mosaic detection workflow, which filters out the false positive calls. The workflow ReproSeq Mosaic PGS w1.1 v 5.10 was used for mosaic detection. The embryos that had proportions of aneuploid cells ranging from 20 to 80% were reported as mosaic aneuploid and those over 80% as truly aneuploid, based on the guidelines issued by Preimplantation Genetic Diagnosis International Society (PGDIS) [11]. The analysis can be visualized using the Integrated Genome Viewer (IGV) light version 5.0, and scoring of aneuploidy was based on visualization of the IGV profile indicating losses and gains of the whole chromosome coupled with confidence and precision metrics. Embryos were further evaluated and scored based on the Median Average Pairwise Differences (MAPD) value (< 0.3), number of reads obtained (above 100 K), and the coverage value (low pass coverage-0.01X). An embryo was considered abnormal or aneuploid if the ploidy deviated from the reference baseline, i.e., 2 on autosomes and X in females and 1 on sex chromosomes in males.

Statistical analysis

The descriptive data analysis was performed to determine the absolute counts and its associated percentage of biopsy samples that were euploid and the number that were aneuploid relative to maternal age of the woman. The data based on age variable was grouped into five categories < 30 years, 31 to 34, 35 to 37, 38 to 40, and > 40 years. The euploid and aneuploid rates across age groups were compared using contingency table analyses with the chi-square test. In all cases, the levels of significance were set at 0.05 to be statically significant.

Results

This study included a total of 1501 blastocyst stage embryos from 488 patients in 535 cycles. Ages of the female patients ranged from 22 to 47 years, and the mean maternal age was observed to be 33.23 ± 4.39. Of the total embryos tested, 50.3% (755/1501) were found to be euploid, the overall aneuploidy (whole-chromosome and segmental) rate was 36.7% (551/1501), 10.1% (152/1501) were found to have mosaicism, and no results were obtained for 2.9% (43/1501) due to amplification failure or insufficient data for analysis. Among the aneuploid embryos, the percentage of embryos with whole-chromosome aneuploidy was 31.4% (472/1502), the percentage of embryos carrying segmental alterations alone was 4.3% (64/1501), and 1% (15/1501) had segmental and whole-chromosome aneuploidy (Fig. 1).

Fig. 1.

NGS-based PGT-A results

Distribution of aneuploidy detected per chromosome

The frequency of aneuploidy (i.e., trisomy and monosomy together) and distributions of abnormalities across chromosomes are shown in Fig. 2a. The frequency of a particular chromosome exhibiting monosomy and trisomy are shown in Fig. 2b. The frequency of each chromosome aneuploidy varied greatly. The most frequently observed aneuploidy was seen for chromosomes 16, 22, and 21. Trisomies were more frequent than monosomies: trisomy 16 was the most common (N = 24; 18.5% of all trisomies) followed by trisomy 21 (N = 13; 10% of all trisomies) and trisomy 22 (N = 12; 9.2%).

Fig. 2.

The distribution (%) of chromosomal errors detected by PGT-A. a The most-affected chromosomal errors were observed in chromosomes 16, 22, and 21. b The percentage (%) of monosomy and trisomy chromosome errors

Aneuploidy with advancement of maternal age

In order to provide more information on the effect of maternal aging on aneuploidy, the overall chromosome errors observed were analyzed by age group. The maternal age at the time of trophectoderm biopsy was grouped into the age groups of < 30, 31–34, 35–38, 38–40, and > 40. As expected, the prevalence of whole-chromosome aneuploidy rose steadily with age (Fig. 3a). The whole-chromosome aneuploidy rate was observed to be 27.8% in women below the age of 30 which steadily increased to 67.5% in age group of women above 40 years. However, segmental abnormalities did not show a maternal age-dependent increase (Fig. 3a).

Fig. 3.

Nature of aneuploidy across different age groups of women. a Incidence of whole-chromosome and segmental errors. b The complexity of the aneuploidies

The impact of age on the complexity of the aneuploid errors is presented in Fig. 3b. Age not only increased aneuploidy rate but also impacted the complexity of aneuploidy. The aneuploidy errors found in women above 40 years of age were more likely to involve either two chromosomes or three or more chromosomes than in the aneuploid embryos in women under 30 years of age. However, data showed relatively constant rates of mosaicism ranging from 11.4 to 8.3%, suggesting that there was no obvious relationship between advancing maternal age and the incidence of mosaicism.

Incidence and type of segmental aneuploidies

Segmental aneuploidy was observed in 5.3% of the analyzed blastocysts (79 of 1501), which included embryos with one or more segmental chromosome aneuploidies and some with segmental as well as whole-chromosome aneuploidy. Pure segmental aneuploidy (segmental aneuploidies in the absence of whole-chromosome aneuploidies) was observed in sixty-four embryos (4.3%), and fifteen embryos tested (1%) had segmental aneuploidies along with whole-chromosome aneuploidy (Fig. 1). In the pure segmental aneuploidy group, single segmental aneuploidy was observed in 59 embryos (3.9%), and the remaining 5 embryos (0.33%) showed segmental aneuploidies in two different chromosomes.

The incidence and distribution of specific segmental aneuploidies (i.e., gain and loss of q and p arms) across chromosomes were further characterized. The distribution of segmental aneuploidy across all the chromosomes is depicted in Fig. 4. The distribution of segmental errors was uneven but most frequently observed on larger chromosomes. The highest incidence of segmental gain was observed on the q-arm of chromosome 9. Segmental loss was not observed in chromosomes 10, 21, and 22, while neither segmental loss nor gain was observed in chromosomes 14 and 20.

Fig. 4.

Distribution of segmental aneuploidy per chromosome

Segmental aneuploidy losses of chromosome fragments were more common and accounted for 60.4%, while 39.6% were segmental gains. Segmental aneuploidy was more frequently located on the q-arm than on the p-arm of the chromosome (62.6% vs 37.4%, respectively).

The segmental imbalances were also assessed in terms of the site at which chromosome breakage had occurred. The distribution of chromosomal breakpoints across all the chromosomes is shown in Fig. 5. The chromosomal breakpoints in our study were compared to the known fragile sites in the human genome and published evidence. A high incidence of segmental breaks was observed on chromosome 9, and 25% of the breaks observed were in defined fragile sites (Fig. 5).

Fig. 5.

Ideogram that shows the segmental chromosomal breakage sites in embryos analyzed. Foot note: The green rectangular boxes indicate the breakpoints corresponding to known fragile sites in the human genome

Discussion

This study reports and analyzes the different types of aneuploidies observed and the effects of maternal age on aneuploidy rates in preimplantation embryos analyzed with 24-chromosome PGT-A.

As aneuploidy is a common and a limiting factor in achieving successful pregnancies in [1] infertile populations, including couples with RM [12], RIF [13], and AMA [14], aneuploidy screening has become common practice to improve outcomes. PGT-A also helps in utilizing elective single-embryo transfer in infertile patients, thus reducing the risk of multiple pregnancies [15]. Current PGT-A techniques employ comprehensive chromosome screening (CCS) in blastocysts and can detect segmental aneuploidy and embryonic mosaicism [16]. aCGH has been widely used for CCS and has shown to improve the clinical outcomes in IVF patients [17]. However, the inferior ability of aCGH array to detect segmental aneuploidies and mosaicism when compared to NGS has demonstrated the NGS platform as being more efficient and superior for PGT-A compared to other methods [18]. Friedenthal et al. has also shown that NGS-based PGT-A significantly improves pregnancy outcomes when compared to using aCGH [19]. Further, the study by García-Pascual CM et al. has shown the robustness of diagnosing segmental (≥ 10 Mb) and mosaic aneuploidies using NGS techniques [20]. Hence, NGS-based testing has been widely used and is the choice for aneuploidy screening by a number of PGT-A laboratories. We have used NGS method in our center for aneuploidy screening of embryos.

The conventional cytogenetic studies of human oocytes have reported remarkably high rates of aneuploidy of 50% or more [21]. Later, the genome-based methods of aneuploidy detection—aCGH and SNP array analysis—showed aneuploidy rates of 30–60%, which is consistent with the estimates from conventional cytogenetic studies [22, 23]. The average percentage of embryos affected by aneuploidy ranged from 40 to 57% in NGS studies [24, 25]. In the present study, 46.8% of the embryos tested were found to be aneuploid (which includes whole-chromosome, segmental, and mosaic) (Fig. 1) which was within the range observed in earlier studies. However, direct comparison is difficult because of the differences in the patient population across the studies.

With the advancement of technology, it is now possible for refinement of aneuploidy categories into whole-chromosome aneuploidy, segmental aneuploidy, and mosaic abnormalities. Embryonic mosaicism in the PGT-A context is defined as the co-presence of euploid and aneuploid cells within the embryo. The prevalence of mosaicism in prenatal specimens has been estimated as 1 to 2% [26, 27]. However, the frequency of embryonic mosaicism reported in literature varies greatly, ranging from 4 to 22% [28, 29]. This wide range of variation can be attributed to biological and technical variations [30]. Embryos with proportions of aneuploid cells ranging from 20 to 80% were used by most laboratories to define mosaic embryos, including us. García-Pascual CM et al. carried out a study to determine percentage of mosaic aneuploidies in embryos using different thresholds. Their study showed that using a threshold of 20–80% to define mosaicism detected mosaicism in up to 17% of embryos, whereas the 30–70% threshold range decreased the mosaicism rate to 5% [20]. Our study showed a frequency of 10.1%, and we used a threshold of 20–80% used to define mosaicism.

PGT-A occasionally fails to yield a diagnostic result. Different methods of PGT-A demonstrated that the prevalence of inconclusive results ranged from 0.86 to 3.8% [31, 32]. In this study, embryo biopsy for NGS-based PGT-A yielded inconclusive results in a small minority (2.8%) of cases, which is consistent with the literature evidence [32]. Neal et al. carried out a study to evaluate the diagnostic results following re-biopsy of blastocysts with inconclusive results. Their analysis demonstrated that re-biopsy of no-result blastocysts yields a euploid result over half of the time, suggesting that re-biopsy for repeat PGT-A should be offered to patients who have blastocysts with inconclusive results following initial biopsy [32].

Reproductive aging in the female is associated with a progressive increase in embryonic aneuploidy, due to meiotic and mitotic errors. The results from early studies demonstrated that most aneuploidies arise due to errors in maternal meiosis and increasing maternal age is a powerful contributor to the occurrence of aneuploidy. Two large cohort studies have shown the effect of maternal age on the prevalence and nature of human embryonic aneuploidy and also the probability of retrieving at least one euploid embryo as functions of maternal age [2, 14]. One study analyzed 15,112 day-5 embryos which had undergone SNP-based PGS. Their result showed that in women aged 27 to 35 years, the median proportion of euploid embryos was 55%, but it decreased rapidly after age 35. Another study analyzed 15,169 embryos undergoing comprehensive chromosomal screening using trophectoderm biopsies, demonstrating that the prevalence of aneuploidy was 20–30% in women aged 26–31 and from age 31 through age 43, the aneuploidy rose steadily. The results of the current study showed a steady proportion of ~ 67.7% euploid embryos in women < 30 years, which then rapidly declines to 55.6% in women aged above 35 and then further to 29.9% in the age above 40 years (Fig. 3a). Increasing age results not only in an increase in the proportion of aneuploid embryos but also has an impact on the complexity of the aneuploid errors presented. A large systematic report by Franasiak et al. reveals that the frequency of two or more chromosome aneuploidies was predominantly higher than single chromosome aneuploidy with increasing age [2]. The present study also examined the nature of aneuploidy in different age groups which showed that 2, 3, or more chromosomal errors increased with advancement of maternal age (Fig. 3b).

To date, much of our understanding of early aneuploidy is based upon cytogenetic analysis of products of conception (POCs). Conventional cytogenetic analysis and most recently high-resolution and high-throughput NGS techniques are used for chromosomal analysis of POCs to identify the genetic cause of miscarriages [33, 34]. Analysis has shown that aneuploidies in relatively small chromosomes 15, 16, 21, and 22 were the most common, while those in the large chromosomes 1 to 6 were less common [33–35]. McCoy et al. and Franasiak et al. have shown elevated rate of errors in chromosomes 16, 22, 15, and 21, in this order [36, 37]. Our study showed aneuploidy of chromosome 16 (14.4%) to be the highest, followed by chromosome 22 (10.7%), 21 (8.0%), and 15 (6.2%) (Fig. 2). All these studies including ours indicate that aneuploidies do not affect all chromosomes equally and suggest that some chromosomes are more prone to meiotic errors.

Segmental aneuploidy is also referred to as partial aneuploidy, where a small segment of a chromosome is gained or lost. The detection size limit of most PGT-A platforms is 5–10 Mb [38]. Cytogenetic analyses have shown that segmental aneuploidies account for approximately 6% of clinical miscarriages [39] and the frequency of the segmental aneuploidy detected in preimplantation embryos is shown to be 6–8% [3, 40]. However, more recently a study by Girardi L et al. reported a very low frequency of 2.4% of segmental aneuploidies [41]. This observed low frequency is most likely due to the segmental detection limit of above 10 Mb and non-mosaic segmental aneuploidy, which were used to classify segmental aneuploidy. Our study showed a frequency of 4.3% for segmental aneuploidies and 1.0% of embryos showed segmental along with whole-chromosome aneuploidy (Fig. 1).

Segmental aneuploidies arise as a consequence of DNA double-stranded breaks (DSBs) during DNA synthesis, a phenomenon known as chromothripsis [42]. The available evidence suggests that the mechanisms implicated in segmental aneuploidy differ from those put forward to explain whole-chromosome aneuploidies. In general, segmental aneuploidy appears to be caused by disturbances during mitosis and is independent of maternal age [3, 40, 43]. Studies by other groups including ours did not show association between the incidence of segmental aneuploidies and maternal age (Fig. 3a), suggesting a distinct etiology [3, 41, 43].

Studies have also assessed segmental imbalances based on the location of gains or losses either on the p or q chromosome arms and their breakage site [3, 43]. It was apparent from these studies that losses of chromosome fragments were more frequent and they were more frequently located on the q-arm of the chromosome. The incidence of segmental gain in q-arm of chromosome 9 was the most commonly observed [3, 41]. Our study also showed a similar trend with high incidence of segmental gains in the q-arm of chromosome 9. This observation may relate to the presence of a large heterochromatic region in chromosome 9 [3]. The heterochromatin blocks are comprised of highly repetitive DNA sequences, which may hinder replication fork progression, thereby causing DNA breaks and thus resulting in high frequency of segmental aneuploidies [3, 44]. A study by Babariya et al. provided valuable evidence on the distribution of chromosomal breakpoints across all chromosomes [3]. Their study showed that the location of breakages was not entirely random but had a tendency to occur at distinct hotspots, which corresponded to known fragile sites or novel fragile sites which might be specific to gametogenesis or embryogenesis. The findings of the current study showed a similar pattern, with occurrence of chromosomal breaks most often associated with defined fragile sites and similar sites as observed in the above article.

This study may have some limitations like the associated factors including paternal age, drug exposure, smoking, and alcohol use that may have contributed to aneuploidy not being considered. In addition, certain maternal effect variants might have also contributed to rate of aneuploidy. The patients included were only those who had viable gametes and in whom fertilized embryos were able to reach blastocyst stage. Thus, higher responders with a large number of blastocysts are over represented compared to low responders who may have only a single embryo to contribute to the study population.

In conclusion, the data presented here provides important information regarding the relationship between maternal age and aneuploidy and the most frequently observed aneuploidy at the chromosomal level. Our study showed an incidence of aneuploidy rate of ~ 28% at maternal age less than 30 years which steadily increased to ~ 67% when maternal age was above 40 years which confirm that aneuploidy rates increase with increase in maternal age. Our study shows a high frequency of aneuploidy in chromosomes 16, 22, 21, and 15, which are the most common non-viable trisomies causing spontaneous miscarriages. The frequency of segmental and mosaic aneuploidies was consistent with observations from other studies which employed NGS protocols. The study further provided evidence on frequencies and location of segmental aneuploidies which were similar to those reported in previous large studies. The findings from this study have important implications which help clinicians and genetic counselors in providing information to patients regarding aneuploidy rates across age groups and most frequently observed aneuploidies.

Author contribution

Dr. Swathi Shetty, Dr. Devika Gunasheela, Dr. Rajsekhar Nayak, and Dr. Jiny Nair contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Dr. Jiny Nair, Sachin Shetty, Cynthia Irene Kasi, Nirmala Thondehalmath, Deepanjali Ganesh, Anjana R, Vidyalakshmi R Bhat, and Sajana M. The first draft of the manuscript was written by Dr. Jiny Nair, and review and editing of the manuscript was done by Dr. Swathi Shetty, Sachin Shetty and Dr. Devika Gunasheela. All authors read and approved the final manuscript.

Declarations

Ethics approval and consent to participate

Ethical approval was waived by the Ethics Committee of Gunasheela Surgical and Maternity Hospital in view of the retrospective nature of the study and as all the procedures being performed were part of the routine care. Written informed consent was obtained from all the study subjects undergoing PGT-A and IVF procedure.

Competing interests

The authors declare no competing interests.