Donor oocyte recipients do not benefit from preimplantation genetic testing for aneuploidy to improve pregnancy outcomes

Abstract

STUDY QUESTION

Do donor oocyte recipients benefit from preimplantation genetic testing for aneuploidy (PGT-A)?

SUMMARY ANSWER

PGT-A did not improve the likelihood of live birth for recipients of vitrified donor oocytes, but it did avoid embryo transfer in cycles with no euploid embryos.

WHAT IS KNOWN ALREADY

Relative to slow freeze, oocyte vitrification has led to increased live birth from cryopreserved oocytes and has led to widespread use of this technology in donor egg IVF programs. However, oocyte cryopreservation has the potential to disrupt the meiotic spindle leading to abnormal segregation of chromosome during meiosis II and ultimately increased aneuploidy in resultant embryos. Therefore, PGT-A might have benefits in vitrified donor egg cycles. In contrast, embryos derived from young donor oocytes are expected to be predominantly euploid, and trophectoderm biopsy may have a negative effect relative to transfer without biopsy.

STUDY DESIGN, SIZE, DURATION

This is a paired cohort study analyzing donor oocyte-recipient cycles with or without PGT-A performed from 2012 to 2018 at 47 US IVF centers.

PARTICIPANTS/MATERIALS, SETTING, METHODS

Vitrified donor oocyte cycles were analyzed for live birth as the main outcome measure. Outcomes from donors whose oocytes were used by at least two separate recipient couples, one couple using PGT-A (study group) and one using embryos without PGT-A (control group), were compared. Generalized estimating equation models controlled for confounders and nested for individual donors contributing to both PGT-A and non-PGT-A cohorts, enabling a single donor to serve as her own control.

MAIN RESULTS AND THE ROLE OF CHANCE

In total, 1291 initiated recipient cycles from 223 donors were analyzed, including 262 cycles with and 1029 without PGT-A. The median aneuploidy rate per recipient was 25%. Forty-three percent of PGT-A cycles had only euploid embryos, whereas only 12.7% of cycles had no euploid embryos. On average 1.09 embryos were transferred in the PGT-A group compared to 1.38 in the group without PGT-A (P < 0.01). Live birth occurred in 53.8% of cycles with PGT-A versus 55.8% without PGT-A (P = 0.44). Similar findings persisted in cumulative live birth from per recipient cycle.

LIMITATIONS, REASONS FOR CAUTION

Pooled clinical data from 47 IVF clinics introduced PGT-A heterogeneity as genetic testing were performed using different embryology laboratories, PGT-A companies and testing platforms.

WIDER IMPLICATIONS OF THE FINDINGS

PGT-A testing in donor oocyte-recipient cycles does not improve the chance for live birth nor decrease the risk for miscarriage in the first transfer cycle but does increase cost and time for the patient. Further studies are required to test if our findings can be applied to the young infertility patient population using autologous oocytes.

STUDY FUNDING/COMPETING INTEREST(S)

No external funding was used for this study. There are no conflicts of interest to declare.

TRIAL REGISTRATION NUMBER

N/A.

Introduction

The proportion of ART cycles utilizing oocyte from donors has been steadily growing and in 2016 represented over 22 000 IVF cycles and almost 9% of all IVF cycles in the USA (https://www.cdc.gov/art/pdf/2016-national-summary-slides/art_2016_graphs_and_charts.pdf#page= 44). Donor oocyte-recipient cycles are considered one of the best-prognosis groups and generally produce some of the highest pregnancy and live birth rates of any in-vitro fertilization procedure (Barad et al., 2017). This is especially attractive to women who struggle with diminished ovarian reserve, premature ovarian insufficiency and age-related oocyte quality decline. Due to continued technological advances in oocyte cryopreservation, vitrified donor oocytes have been increasingly used in donor oocyte-recipient cycles. However, some have reported that oocyte cryopreservation can disrupt the meiotic spindle, which could theoretically lead to abnormal chromosome segregation and ultimately increased aneuploidy in resultant embryos (Bromfield et al., 2009; Coticchio et al., 2009). Therefore, preimplantation genetic testing for aneuploidy (PGT-A) may have benefit in vitrified donor oocytes cycles. On the other hand, aneuploidy is associated with maternal age and therefore embryos from young donors are expected to be predominantly euploid (Hassold and Hunt, 2001). Trophectoderm biopsy is also not without risks and the procedure could potentially be harmful to the viability of the developing embryo and impact implantation and progression to delivery (Rosenwaks et al., 2018). It has been shown that the rate of aneuploidy in embryos generated from donated oocytes varies widely across treatment centers (Franasiak et al., 2014). Given the uncertainty about self-correction of the embryo and the difference in chromosome abnormality rates based on the testing technology used, false-positive PGT-A results as well as the accuracy and unclear interpretation of mosaic diagnosis, there is concern that one may be discarding embryos that would have resulted in healthy babies.

Therefore, the question of whether PGT-A in donor oocyte-recipient cycles improves IVF outcomes has been debated in the literature. Some studies have concluded that PGT-A may be harmful to IVF outcomes and is in fact associated with significantly lower pregnancy and live birth rates compared to controls without PGT-A testing (Barad et al., 2017). Others have argued that genetic testing might lead to improved IVF outcomes with embryos derived from donor oocytes and thus could be beneficial in this group of patients (Coates et al., 2017). There are also data showing that PGT-A has no effect on IVF outcomes and therefore may not be necessary in most cases (Deng and Wang, 2015). Most of these studies included small numbers of patients and had challenges with study heterogeneity including use of both slow freezing and vitrification protocols, combining embryos derived from fresh and frozen oocytes, and evaluation with Day 3 and 5 embryo biopsies for PGT-A testing.

Given this controversy and the limitations of previous studies, our objective was to evaluate the utility of PGT-A in donor oocyte recipients in a large patient population with blastocysts derived from cryopreserved oocytes only and in whom vitrified eggs from the same donor were used in multiple egg thaw and transfer cycles, both with and without PGT-A, so that the oocyte donor could serve as her own control in the analysis.

Materials and methods

Ethical statement

Institutional review board approval was obtained for review of donor oocyte-recipient cycle data within the Donor Egg Bank USA database. All oocyte donors were anonymous.

Oocyte donor and recipient selection

Data from oocyte donor-recipient cycles were obtained from Donor Egg Bank USA. Oocyte donors underwent ovarian stimulation and oocyte retrieval at 47 centers with standardized methodology for donor recruitment and screening. Qualified donors were between the ages of 21 and 33, non-smoking, had baseline ovarian reserve testing within normal limits which included an Anti-Müllerian Hormone >2 and an antral follicle count >16. Donors successfully completed a three-level multi-generational family medical history, expanded carrier testing of >107 genotypes, infectious disease testing, drug screening and psychological evaluation. Applicants found to be carriers of cystic fibrosis or X-linked diseases were not permitted to donate.

A protocol of controlled ovarian hyperstimulation with gonadotropins, coupled with GnRH antagonist suppression and GnRH agonist (leuprolide acetate) trigger for final oocyte maturation was utilized to reduce the risk of ovarian hyperstimulation in this young and presumably high responder patient population. Oocyte vitrification commenced 38 h post-trigger following a modified Kuwayama protocol described elsewhere (ASRM, 2004; Kuwayama et al., 2005). All mature (meiosis two) oocytes were vitrified for future patient use. Vitrified oocytes from a single donor were split into lots of various sizes based on total number of available oocytes and donor performance history. Oocytes were split into egg lots using an automated algorithm that determined the number of oocytes per egg lot by the number of mature oocytes frozen. The average lot size at the egg bank is 6.6 oocytes. Straws were then pulled randomly based on the number of oocytes required for the recipient. The splitting of ooyctes into multiple egg lots affords the unique ability to analyze the performance of one donor’s eggs amongst multiple recipients.

Recipients selected donated oocytes through a web-based catalog and the oocytes were shipped directly between retrieval laboratory and recipients. If the oocyte donor was a genetic carrier, the sperm source received reciprocal screening to determine match suitability or waived genetic screening post physician and/or genetic counselor counseling. Sperm from the recipient’s partner or donor sperm was used for insemination utilizing ICSI in all cases.

Study design

This was a retrospective paired cohort study of vitrified donor oocyte-recipient cycles. Choosing cryopreserved oocytes derived from a large egg bank allowed for a unique study design in which all oocytes from one donor were split between two or more recipient couples. Outcomes between couples using PGT-A (study group) and couples transferring embryos without PGT-A (control group) were compared. As oocytes from the same donor were used by multiple recipient couples with and without PGT-A, this analysis enabled a single donor to serve as her own control (Fig. 1). In other words, compared embryos (study group and control group) came from the same cohort of oocytes and hence, there was no difference in the donor’s demographic factors, genetics, gonadotropin protocol, retrieval or cryopreservation technique. Only oocyte cohorts that had at least one PGT-A recipient and at least one non-PGT-A recipient were included. No slow-freezing cycles were included, and all oocytes were vitrified prior to banking. Preimplantation genetic testing for monogenic disorders (PGT-M) cycles were excluded and no embryos diagnosed as mosaic were transferred. Recipients who utilized donor sperm for insemination were included.

Figure 1.

A schematic representation of the study design. Embryos from vitrified oocytes were derived from the same donor for both the study and control group. Only donor cycles in which oocytes were contributed to both the preimplantation genetic testing for aneuploidy (PGT-A) and no PGT-A cohort were included. Donors could contribute multiple cycles into both the PGT-A and non-PGT-A cohorts. Numbers given for donors and recipient and 1:1 splitting of donor oocytes are examples, as multiple combinations exist in the dataset.

Preimplantation genetic testing for aneuploidies

PGT-A was performed at the blastocyst stage of embryonic development by array comparative genomic hybridization (array CGH), DNA microarray or next generation sequencing (NGS), the latter accounting for 80% of PGT-A testing. The decision to perform PGT-A was made at the discretion of the physician and the recipient and the factors leading to this decision were unknown and not recorded in the donor oocyte bank datasets.

Embryo transfer

Embryos that underwent PGT-A were generally cryopreserved and underwent a fresh embryo transfer in only 2.6% of cycles. Conversely, non-PGT-A cycles received a fresh embryo transfer in 96.6% of cycles. Luteal support and frozen embryo transfer protocols were chosen by the clinics performing the IVF cycle.

Statistical analysis

Generalized estimating equation (GEE) models accounted for multiple transfer cycles within the same recipient and were used to paired cycles both with and without PGT-A from the same donor. PGT-A cycles with no euploid embryos available for transfer were included in the primary analysis, in order to evaluate the intention-to-treat effect of PGT-A testing. A subanalysis was performed excluding cycles with no euploid embryos for transfer to evaluate the effect of known euploid embryo transfer.

Cycles that have not yet undergone embryo transfer were excluded from the dataset. In univariate GEE models, only the number of embryos transferred was positively associated with live birth and was included in multivariate modeling assessing the effect of PGT-A on live birth. While the primary model was adjusted for the number of embryos transferred, a subanalysis was performed to evaluate the cohorts based on single embryo transfer cycles only. Aneuploid rates varied both by genetics laboratory and embryology laboratory site and these variables were included in GEE subanalyses models, to determine if the comparison changed when accounting for laboratory level data. Data were also analyzed per year, to determine if change over time was a confounder. The primary outcome was live birth which was defined as delivery >24 weeks of gestation. Secondary outcomes were clinical pregnancy loss (defined as intrauterine pregnancy demise confirmed by ultrasound or histology) and biochemical loss (defined as positive serum beta hCG without clinical sonographic detection of gestational sac, yolk sac or fetal pole). All outcomes are reported per first transfer and total transfers.

Results

A total of 1291 donor oocytes-recipient cycles, 262 cycles with PGT-A testing and 1029 control cycles without PGT-A testing from 223 donors were included in the analysis. Couples underwent IVF between 2012 and 2018. The dataset was derived from a large US donor egg bank that serves as a resource for frozen oocytes for numerous fertility centers across the USA. Vitrified donor eggs were sent to 47 different facilities. The mean age of the recipients was 42.4 in the non-PGT-A group and 42.8 in the PGT-A group (P = 0.40). The mean age of the donors was 25.6 (range 21–32) and was the same in both groups given the paired design. The number of cycles included in the study ranged from a minimum of 3 in 2012 to a maximum of 515 in 2017. The majority of the cycles occurred over a 3-year period from 2015 to 2107 (1139 cycles, 88.2% of the study cycles).

Euploidy frequency distribution per patient

Results were derived from pooled clinical data of 47 different recipient clinics and 262 cycles using PGT-A for 665 recipients. On average, there were 2.9 embryos per patient and per PGT-A cycle to biopsy. The median euploidy rate was 75% per recipient. As depicted in Fig. 2, the euploidy frequency distribution per patient ranged from 0% to 100%. In 43% of patients, all embryos were euploid and 12.7% of patients had no euploid embryos available. The vast majority of patients (87.3%) had at least one euploid embryo available for transfer. Mosaic embryos were diagnosed in 3% of embryos and there were no mosaic embryo transfers. Every patient with an embryo diagnosed as mosaic also had a euploid embryo for transfer. Euploidy rates per genetics diagnosis laboratory varied from 40.9% to 79.2% (P < 0.001). Euploidy rates per embryology laboratory varied from 0% to 100% (P < 0.001). Embryology labs with euploidy rates of 0% (n = 6) and 100% (n = 7) had only had a single patient biopsied and between 1 and 4 blastocysts total for evaluation, explaining their extreme euploidy rates. Thirty-two percent of the embryos biopsied came from a single embryology laboratory with euploidy rate of 78.6%, similar to the overall study euploidy rate of 75%. Euploidy rate did not vary by age as a linear variable (P = 0.89), given the restriction of age donors from 21 to 32 years of age with 83% of donor cycles clustered between 24 and 30 years of age.

Figure 2.

Euploidy frequency distribution per patient among 47 clinical centers.

Number of transferred embryos and multiple gestation

The average number of transferred embryos was significantly higher in the non-PGT-A group with 1.38 embryos compared to 1.09 embryos transferred in the PGT-A group (P < 0.01). Multiple gestation occurred less frequently per cycle start in the PGT cohort compared to the without PGT-A cohort (3.0% versus 9.8%, P < 0.001). Multiple gestation also occurred less frequently in the cohort with PGT compared to without PGT-A cohort when excluding the 12.7% of cycles with no euploid embryos for transfer (3.4% versus 9.8%, P = 0.002).

First embryo transfer outcomes

The first planned embryo transfer cycle and all subsequent embryo transfers were analyzed. Patients who had no embryos to transfer due to negative PGT-A results of all embryos were included in the analysis (denominator is planned rather than actual embryo transfer). In the first transfer cycle and controlling for the number of embryos transferred, live birth was 53.8% in the PGT-A group versus 55.8% in the no PGT-A group (P = 0.44; odds ratio (OR) 0.79; 95% CI, 0.45–1.42) (Fig. 3). When the genetic diagnosis laboratory and embryology laboratory were included in the model, there was no difference in live birth between the PGT-A and control cohorts (P = 0.34; OR 0.85; 95% CI 0.50–1.48). In the single center performing the 32% of all the PGT-A cycles in the study, live birth was 69.0% in the PGT-A cohort and 63.6% in the no PGT-A arm (P = 0.82, OR 1.27; 95% CI 0.46–3.47).

Figure 3.

Live birth (%) in the first embryo transfer cycle within the same recipient. (Blue; 53.8% in the PGT-A group versus 55.8% in the no PGT-A control group, P = 0.44; OR 0.79; 95% CI, 0.45–1.42) and all embryo transfers (red; 48.4% in the PGT-A group versus 47.2% in the no PGT-A control group, P = 0.7; odds ratio (OR) 0.95; 95% CI, 0.73–1.23.)

Similarly, for the first transfer, there was no difference in terms of clinical pregnancy loss and chemical pregnancy loss per transfer between the PGT-A group and the control group, and both were overall very low. Clinical pregnancy loss occurred in 6.5% of first transfers with PGT-A and in 8% of the transfers without PGT-A testing (P = 0.44; OR 0.86; 95% CI, 0.47–1.55). Comparable results were found for biochemical loss with no difference between the groups, 6.4% biochemical pregnancies in the PGT-A group and 7.9% in the no PGT-A, respectively (P = 0.45; OR 0.86; 95% CI, 0.49–1.55).

There was also no difference in total pregnancy loss, defined as biochemical and clinical pregnancy losses, between the PGT-A study group with 13% total losses and the non-PGT-A control group with 15.9% total losses (P = 0.29; OR 0.85; 95% CI, 0.54–1.32) (Fig. 4). Total pregnancy loss, miscarriage and clinical pregnancy comparisons were unchanged when further adjusting for genetic testing and embryology laboratory sites. Live birth by year varied by <8% from 2013 to 2018 between the study and control cohorts and year was not associated with the probability of live birth in either cohort (P = 0.58 in PGT-A arm and P = 0.76 in control arm).

Figure 4.

Total pregnancy loss (%) defined as biochemical and clinical pregnancy losses after the first transfer cycle within the same recipient. (Blue; 13% in the PGT-A group versus 15.9% in the no PGT-A control group P = 0.29; OR 0.85; 95% CI, 0.54–1.32) and all embryo transfers (red; 13.4% in the PGT-A group versus 17.1% in the no PGT-A control group, P = 0.16; OR 0.85; 95% CI, 0.54–1.32.)

All embryo transfer outcomes

The analysis of all embryo transfers within the same recipient again found that the PGT-A group and the no PGT-A group performed similarly. There were 262 total embryo transfers available for analysis in the PGT-A group and 1029 total embryo transfers in the non-PGT-A control group. As expected after multiple transfers, live birth or ongoing pregnancy was comparable in both groups. Live birth per embryo transfer event was 48.4% in the PGT-A group and 47.2% in the non-PGT-A group (P = 0.70; OR 0.95; 95% CI, 0.73–1.23) and was unchanged when adjusting for genetics and embryology lab sites. As seen in the first embryo transfer, there was also no difference in terms of clinical pregnancy loss and biochemical loss between the PGT-A (P = 0.23; OR 0.73; 95% CI, 0.41–1.31) and the non-PGT-A group (P = 0.47; OR 0.85; 95% CI, 0.5–1.46). Clinical pregnancy loss occurred in 6.1% of all transfers with PGT-A and in 8.4% of all transfers without PGT-A testing. Comparable results were found for biochemical losses with 7.3% in the PGT-A group and 8.7% loss in the no PGT-A group, respectively. To determine if donors who provided a large number of oocyte recipient cycles might bias the results, an analysis of the 10 donors who contributed the most recipient cycles were analyzed. The top donors contributed oocytes which were split into 25 PGT-A and 181 no PGT-A cycles (Supplementary Table SI). Live birth did not differ between the two cohorts within any single donor and overall live birth was 48.4% in the non-PGT-A cohort and 50.0% in the PGT-A cohort in the 10 most contributing donors (P = 0.99, OR1.20 95% CI 0.53–2.83). When the results were restricted to only single embryo transfers, results remained unchanged with live birth occurring in 48.1% in the non-PGT-A cohort and 53.3% in the PGT-A cohort (P = 0.28, OR 1.25 95% CI 0.89–1.75). All of the PGT-A cycles were frozen embryo transfers and 239 of the non-PGT-A cycles were frozen embryo transfers. In the non-PGT-A group, live birth was similar between the fresh and frozen groups (48.3% fresh versus 43.3% frozen, P = 0.18). When limiting the comparison to only cycles with frozen embryo transfer, live birth was 48.4% in the PGT-A cohort versus 43.4% in the no PGT-A cohort (P = 0.19). When limiting the comparison to only PGT-A cycles performed with NGS for ploidy diagnosis, live birth remained similar between the two groups at 47.2% in the non-PGT-A group versus 51.2% in the PGT-A group (P = 0.42).

Taking biochemical and clinical pregnancy losses together and analyzing total losses for all embryo transfers, the PGT-A group had fewer losses with 13.4% compared to the non-PGT-A group with 17.1% losses; however, this difference did not reach statistical significance (P = 0.16; OR 0.85; 95% CI, 0.54–1.32). Total pregnancy loss, miscarriage and clinical pregnancy comparisons were unchanged when further adjusting for genetic testing and embryology laboratory sites.

Excluding cycles with all aneuploid embryos

One of the arguments in favor of PGT-A is that it may reduce time to pregnancy and live birth as it allows women who have no euploid embryos to start a new IVF cycle promptly without going through multiple unsuccessful embryo transfer cycles until the untested cohort of embryos is exhausted. The live birth rate in the first embryo transfer when excluding women with non-PGT-A normal blastocysts to transfer was evaluated. Once again there was no difference between groups, live birth was 61.3% in the PGT-A group compared to 55.8% in the untested group (P = 0.14; OR 1.25 (95% CI 0.92–1.59). This finding correlated with the overall high median euploid rate of 75% and that 12.7% of PGT-A cycles did have all aneuploid embryos and avoided potentially futile embryo transfer cycles.

Discussion

The principal question addressed in this study is whether PGT-A is a beneficial addition in ART for recipients using embryos derived from vitrified donor oocytes. These data suggest that donor oocyte PGT-A does not improve the likelihood of a live birth in the first embryo transfer or cumulative transfers but PGT-A can avoid an embryo transfer in cycles with all aneuploid embryos. Aneuploidy is strongly associated with maternal age and is expected to be low in young oocyte donors. A recent large retrospective analysis of over 15 000 trophectoderm (TE) biopsies demonstrated that the aneuploidy rate is lowest between ages 26 and 37 while both younger and older age groups are at risk for higher rates and more complex aneuploidies (Forman et al., 2012; Yang et al., 2012; Franasiak et al., 2014). Consistent with the expected low aneuploidy rate for young women, we found that the majority of recipients had at least one euploid embryo after PGT-A. Only 12.7% of patients had only aneuploid embryos and no embryo transfer. The low aneuploidy rate observed in our study cohort of young donors was also reflected in the overall very low probability for pregnancy loss, in both the PGT-A and non-PGT-A cohorts. Our data correlate with a recent publication by Munne et al., who examined chromosomal abnormalities in blastocysts from 42 fertility centers and reported a mean euploidy rate per donor of 68% (ASRM, 2017; Munne et al., 2017).

The incidence of mosaicism was 3% and this low number is at the lower end of the normal range reported in the literature. The Preimplantation Genetic Diagnosis International Society reports the range of mosaicism with NGS to vary substantially from 2% to 40% between programs, with most clinics reporting mosaicism rates of 5–10% (Cram et al., 2019). A multicenter prospective study analyzing over 16 000 human blastocysts via trophectoderm biopsy with NGS, reported just a 2.6% rate of mosaicism (Katz-Jaffe, 2017). In contrast, the Single-Embryo Transfer of Euploid Embryo (STAR) trial reported a mosaicism rate of 16% using NGS, which is higher than our data and the other studies we have cited (Munné et al., 2019). It is also possible that some of the 47 clinics chose not to receive mosaic results and limit reporting from PGT-A labs to only euploid versus aneuploid. The prevalence of mosaicism in PGT varies based on PGT-A platform, analytic threshold determinations, embryology laboratory conditions, trophectoderm biopsy technique and biases in library construction (Cram et al., 2019). Given the large number of oocyte retrieval, embryology and genetic testing sites in this present study, we were unable to make any definitive analysis of mosaicism rates in blastocysts derived from donor oocyte cycles.

The risks and benefits of PGT-A in donor oocyte-recipient cycles have been debated in the literature and there is no consensus on whether PGT-A in donor oocyte-recipient cycles improves clinical outcome measures. Given that PGT-A is performed in only 4% of donor oocyte-recipient cycles, large-scale datasets evaluating PGT-A are difficult to obtain and most published studies infer recommendations based on small numbers (Haddad et al., 2015). Most of these studies also include both slow freezing and vitrification protocols, embryos derived from fresh and frozen oocytes and cleavage stage as well as blastocyst biopsies for PGT-A testing, all of which introduced study heterogeneity. Recent published Society for Assisted Reproductive Technology (SART) data evaluating the impact of PGT-A in donor oocyte-recipient cycles over a period of 9 years concluded that live birth rates were lower for PGT-A cycles compared to control cycles, with the odds of live birth reduced by 35% (Barad et al., 2017). The SART database analysis also did not differentiate between embryos derived from frozen versus fresh oocytes, slow freeze or vitrification, or cleavage versus blastocyst biopsies. Consistent with our data, the study also found no effect of PGT-A on pregnancy loss rates. Deng and Wang (2015) reached a similar conclusion and found lower live birth rates for PGT-A cycles with blastocysts derived from vitrified donor oocytes, although a major limitation was only seven patients receiving PGT-A testing. A 2015 retrospective analysis included 31 cycles with PGT-A on blastocysts derived from fresh oocytes and found no statistically significant differences compared to control cycles in terms of clinical pregnancy rate, ongoing pregnancy or live birth, and pregnancy loss rates (Haddad et al., 2015). Results from a recent larger retrospective analysis with 99 PGT-A cycles transferring blastocysts from fresh oocytes suggests a significantly higher live birth in the PGT-A group with double embryo transfer (DET; n = 41) compared to the untested group. While a similar benefit was not observed in the single embryo category (n = 58) (Coates et al., 2017). Whenever euploid embryos are available, the American Society for Reproductive Medicine (ASRM) and SART guidelines recommend single embryo transfer, regardless of the recipient’s age, in an effort to increase singleton gestation and reduce twin gestations and high-order multiples (ASRM, 2017). Single embryo transfer is also recommended for favorable patients <38 years and certainly for embryos derived from oocytes from young donors, independent of whether PGT-A tested or not. However, these data are derived from 47 clinics with independent practice patterns and patient desires. In this real-world setting, these data demonstrate that DET was higher with untested embryos, despite meeting ASRM criteria for single embryo transfer, whereas PGT-A testing was a strong promoter for single embryo transfer and reduced the risk of multiple gestation by 3-fold. Therefore, while PGT-A was associated with an increase in single embryo transfer usage in practice, single embryo transfer should be recommended in donor oocyte cycles, regardless of whether or not PGT-A is used.

Because PGT-A cannot change the chromosomal status of an embryo, it stands to reason that PGT-A would not be expected to improve the cumulative live birth per egg thaw. Our results confirm no significant difference in cumulative live birth between the PGT-A group and the non-PGT-A group. While there seems to be no benefit of PGT-A per in terms of overall live birth, it has been suggested that women who choose genetic testing will potentially undergo fewer transfer cycles and save the associated additional cost and emotional strain as they can avoid transfers of aneuploid embryos and subsequent failed implantation or pregnancy loss. Less time to achieve an ongoing pregnancy is a major concern for patients and has been identified as one of the gains associated with PGT-A as knowledge of the embryo’s ploidy status enables selection of a genetically normal embryo in the first transfer whereas the selection of a euploid embryo in untested embryo cohorts is due to chance and therefore could potentially require multiple cycles (Forman et al., 2013; Grifo et al., 2013). This was not the case in this current study. Live birth was not statistically significantly different between the PGT-A and non-PGT-A group and implies that PGT-A and transfer of a euploid embryo in the first transfer cycle did not result in higher odds for a child compared to couples who transferred an embryo with unknown ploidy status. However, we were not able to directly measure time to pregnancy in this dataset.

Strengths of this study included utilizing data from an egg bank, which allowed us to address some of the challenges with previous study designs. Data from 47 centers increased the generalizability of the study. We included only vitrified frozen oocytes, narrowing the study design to the most commonly employed modality for current oocyte donation cycles. All biopsies for genetic testing were performed at the blastocyst stage of embryo development. A limitation of the study was its retrospective cohort design. We attempted to reduce bias from the lack of randomization, by analyzing vitrified split donor oocyte cycles, where oocyte from a single donor was divided between couples using PGT-A testing and couples who did not. This design eliminates the risk of bias resulting from donor-specific characteristics. However, this study design cannot eliminate the potential that recipients who elect to have PGT-A are different from those who do not. The factors contributing to the decision to proceed with PGT-A testing were unknown to us. Our results are derived from pooled clinic data of 47 different facilities, each having their own aneuploidy rates in their embryology labs, each using different PGT-A companies that also all have their own aneuploidy rates and each using different PGT-A testing platforms that are known to also influence aneuploidy. However, adjusting for variability in genetic testing and embryology laboratories did not change the results. Finally, given that the dataset was derived from the donor oocyte bank itself, we lack patient specific data other than age, PGT-A usage and results, embryo transfer, the number of embryos transferred and pregnancy outcomes. While the study designed controlled for donor characteristics, it is possible that unmeasured recipient variables represent a chance for residual confounding.

In conclusion, our findings demonstrate no benefit for PGT-A in donor oocyte recipients for improving live birth or reducing miscarriages, neither in the first nor all transfer cycles.

Authors’ roles

N.D. and M.H. were responsible for the conception, design, data interpretation and manuscript writing. M.H. and S.J. performed the data analysis. A.E. and M.G. contributed to the literature search. H.H. supervised the data collection. J.D., M.T., K.D., A.D. and M. L. participated in the interpretation of the study data and in revisions to the manuscript.

Funding

No funding was received for this research.

Conflict of interest

None of the authors have any conflict of interests to declare.