Evaluating non-invasive Preimplantation Genetic Testing for Aneuploidy (niPGT-A) from spent culture medium: prolonged embryo culture to day 6 increases informativity without negatively impacting quality and live birth rates

Abstract

Purpose

Embryos release cell-free DNA into spent culture medium (SCM), mostly from day 5 to day 6, making extended culture to day 6 (something done in IVF only for slow embryos) a preferred option in non-invasive preimplantation genetic testing for aneuploidy (niPGT-A). The impact of day 6 culture on embryo quality remains unclear. This study investigates whether prolonged culture to day 6 in niPGT-A improves informativity without compromising embryo quality and live birth rates (LBR).

Methods

In this retrospective cohort study, we investigate the impact of day 6 culture on embryo quality and LBR. After preliminary concordance assessment on 18 embryos, a total of 336 SCM from day six blastocysts were collected. According to the prioritisation criteria applied, 103 embryos were transferred in 91 frozen embryo transfer cycles (January 2021–February 2024). Informativity rates in day 5/6, levels of expansion/quality, female age impact and effects of prolonged culture to day 6 on embryo quality were assessed.

Results

Higher informativity rate was observed in day 6 versus day 5 blastocysts (p < 0.001). Informativity was unaffected by different age groups (p = 0.22), blastocyst expansion (p = 0.51), blastocyst quality (p = 0.67), or incubator setting (p = 0.74). Day 6 prolonged culture did not compromise embryo viability and quality and better quality embryos had higher probability of euploidy (p = 0.01). We report a LBR of 33.3%, higher than the reported 10% by HFEA for a similar age group (40–42 years).

Conclusions

Day 6 extended culture in niPGT-A improves informativity and does not compromise embryo viability and quality. LBR significantly increased compared to international benchmarks of non-tested embryos.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10815-025-03564-9.

Introduction

With the continuing debate on the relative benefits of the use of embryo biopsy (cell removal) to obtain DNA samples for preimplantation genetic testing (PGT), efforts are continuing to try and develop less-invasive means of sampling DNA. The ongoing concerns that the biopsy procedure might compromise development (especially if performed sub-optimally) is potentially counter-indicative to the raison d'être of preimplantation genetic testing for aneuploidy (PGT-A), whose purpose is to improve pregnancy and live birth rates per embryo transfer. Most IVF clinics worldwide now make use of blastocyst biopsy as the standard of care in PGT-A cycles. This allows for more cells (and hence more DNA) in total to be analysed than in the older cleavage stage biopsy technology. Paradoxically, fewer cells proportionally (i.e. ~ 5/150 for blastocyst compared to ~ 1/8 for cleavage stage) are analysed, meaning that blastocyst biopsy is potentially much less harmful than cleavage stage. Evidence from two studies [1, 2] indicates that there is no negative impact of blastocyst biopsy on ongoing pregnancy rates; however, both were conducted at the same experienced centre, and we cannot be certain that it is entirely without risk in all clinics who perform it. Indeed, it is reasonably well established that IVF outcomes vary widely between centres, and this inconsistency, most likely, extends to embryo biopsy procedures generally, even among individuals within the same centre. Little is standardised globally, leading to significant differences in pregnancy rates for similar maternal ages across clinics. It seems reasonable to suggest therefore that sub-optimal biopsy techniques have the potential to harm embryo development or lead to the discarding of normal embryos [3], even if this is not the case in many clinics. A small decline in cumulative live birth rates (CLBR) with PGT-A has been reported in young patients, implying possible adverse effects of biopsy (or possibly misdiagnoses) [4], at least in some centres. A randomised controlled trial (RCT) performed over five years ago indicated a 28% reduction in ongoing pregnancy rates (OPR) simply by performing a biopsy, indicating that embryo development can be compromised or lead to discarding of normal embryos if blastocyst biopsy is not performed correctly [5]. It has also been suggested that blastocyst biopsy might increase the risk of complications such as pre-eclampsia, with a previous study showing a statistically significant threefold increase in the odds of pre-eclampsia, although these studies are rare in the literature [6–8].

Regardless of the evidence either supporting or debunking the notion of biopsy damage, it is generally agreed that high skill levels are needed for embryo biopsy. Widespread biopsy training thus remains a challenge amid increasing worldwide IVF demand. Given these complexities, non-invasive methods to identify aneuploid embryos would be advantageous. Promising approaches include machine vision with artificial intelligence (AI), metabolomics and minimally invasive PGT-A through the sampling of blastocoele fluid (BF) or non-invasive PGT-A (niPGT-A), with the analysis of the spent culture media (SCM).

Embryonic cell-free DNA was first identified in BF in 2013 by Palini et al. [9] and Stigliani et al. [10], and later in 2016 in SCM by Xu et al. [11] and Shamonki et al. [12]. DNA concentrations range from 3 to 36 ng/µL in BF and SCM, compared to 25–46 ng/µL for TE biopsies. Moreover, SCM from frozen-thawed blastocysts is thought to contain more DNA (15 ng/µL) than SCM collected from fresh blastocysts (22 ng/µL) potentially due to cell lysis during thawing [13]. The origin of cell-free DNA is widely investigated: Handayani et al. in a recent review in 2023 suggested that apoptotic pathways eliminate aneuploid cells in mosaic embryos, facilitating DNA release into SCM and blastocoel [14]. Griffin et al. suggested that DNA from aneuploid cells present in earlier cleavage stages preferentially sequester away from the developing foetus (inner cell mass—ICM) and into both BF and degenerate cells on the periphery of the embryo [15]. Some authors have found aneuploid embryos exhibit more DNA in SCM, resulting in higher karyotype concordance rates with blastocoele fluid and TE than with ICM [16]. However, other authors have not found any increase in cfDNA concentration in aneuploid blastocysts [17].

Extensive research has explored the use of BF sampling—known as blastocentesis—as a minimally invasive approach to DNA analysis. However, even under optimal conditions, blastocentesis has demonstrated concordance rates with trophectoderm biopsies of only up to 93.6% [18]. Moreover, further discrepancies exist in chromosomal comparisons between BF and whole embryos, with amplified DNA yields limited by the process of apoptosis. The general feeling at the moment is that, while BF could serve as a clinical biomarker for embryo competence, it is still invasive and it is subject to mitochondrial DNA contamination concerns [19] and is thus not likely to be a reliable and effective diagnostic tool for PGT-A.

The use of SCM as a template for PGT-A has also received great attention in recent years. Here, the main concern is maternal DNA contamination from corona cells and polar bodies and whether DNA in the SCM is truly indicative of the ploidy status of the whole embryo (or ICM). Chen et al. found that up to 50% of samples contaminated with corona cell DNA, and had a poor karyotype concordance (47% contaminated vs. 74% uncontaminated) [20, 21]. To mitigate against contamination, rigorous corona cell denudation and extra embryo washes is thus advised. It has also been reported that day 6 embryos yield more SCM DNA than day 5, and increased concordance in karyotypic terms than sampling at earlier stages [22–24]. Nakhuda et al. found a positive predictive value for whole chromosome aneuploidies in niPGT-A, indicating the potential value of niPGT-A [25].

Overall, reports of concordance between non-invasive PGT-A and TE biopsy are variable throughout the literature, probably due to sampling inconsistencies: Aside from two significant multi-centre reviews by Rubio et al. [26] and Navarro-Sánchez et al. [27], most studies have been smaller and/or conducted at single centres. As expected, the most favourable results have reported in smaller, tightly controlled studies with a small number of embryos, achieving up to 100% concordance [20, 28]. Some investigations combined SCM analysis with BF, while others involved sampling SCM after thawing. Although both BF sampling and embryo thawing can increase the amount of embryonic DNA present in the media, there are concerns that this increases the invasiveness of the procedure. Additionally, studies have been variable in terms of when the SCM was sampled, day 5 and day 6 post-insemination being the most common [26, 27].

The pros of niPGT-A from SCM lie in the fact that it is non-invasive and less skills are required by the embryologist compared to blastocyst biopsy [26, 27]. The drawbacks include the accuracy (which is variable in the different studies) and the need for prolonged culture.

There are several studies that have assessed the impact of the days in culture in informativity and concordance of the results [26, 27, 29]. In all these studies, there is an apparent improvement in day 6 vs. day 5 results. However, the existing literature provides limited information on how the prolonged culture to day 6 impacts embryos and how embryo quality impacts informativity (defined as the proportion of embryos that yield a readable PGT-A result). That is, are good quality embryos (or indeed poor-quality ones) relatively more likely to lead to an accurate diagnosis?

With the above in mind, the purpose of this study was to test the hypothesis that an additional day of culture to day 6, specifically for embryos that reach the blastocyst stage on day 5 does not negatively impact the embryo viability and quality, but nonetheless improves informativity (i.e. the proportion of embryos that yield a readable results by NGS). Furthermore, we describe the association between embryo expansion/quality with informativity and euploidy and our live birth findings to date. Our study is particularly important as it is one of the few to include live birth reporting, an essential outcome measure that has often been overlooked in previous studies. By addressing these gaps, we hope to contribute valuable insights into how niPGT-A can be implemented and optimised for improved clinical outcomes, potentially leading to better live birth rates and more reliable results.

Materials and methods

Study design

A retrospective cohort study of patients undergoing IVF with niPGT-A were analysed between January 2021 and February 2024 at the Orchid Reproductive and Andrology Services Laboratory, Dubai. The study was approved by Dubai Scientific Research Ethics Committee (Ref: DSREC-01/2025_17).

In the first set of analyses and for the test validation, we compared concordance rates of a total of 5 × day 5 blastocysts and 13 × day 6 blastocysts. This step was important to assess concordance rates between SCM and niPGT-A before clinical application. Here, the embryos were cultured using the niPGT-A protocol, and on day 6, the donated blastocyst was placed into a PCR tube. The SCM NGS result was compared for aneuploidy/euploidy against the whole embryo. The outcome measure here noted was concordance, i.e. when the results from the SCM and the whole embryo matched one another. “Non-concordant” result were returned if both results were informative (SCM and whole embryo) but different form one another in terms of ploidy status. Upon successful validation, the protocol could be implemented in clinical cases without concerns.

After validation and protocol implementation in clinical cases, a total of 29 day 5 blastocysts from 11 patients were assessed for informativity compared to a total of 336 day 6 blastocysts from 118 patients in order to ask whether there was a greater informativity at day 6 compared to day 5. Informativity was defined as the proportion of embryos that gave a “readable” result. Thus, a result was “non-informative” if either there was no DNA amplification, if there was sub-optimal DNA or if the results were noisy upon analysis. Subsequent analyses were then performed on the day 6 embryos alone. These patients had a mean maternal age of 40.9 ± 3.5. Of the 336 day 6 blastocysts, a total of 96 were cultured in a standard benchtop incubator and the remaining 240 in a time lapse incubator.

Various factors were assessed to investigate their impact on informativity: age group (≤ 38, 39–40, 41–42 and above 42 years old), embryo expansion and quality and different incubator settings: CBI (conventional benchtop incubator) and time-lapse incubators (TLI). Finally, we assessed how the embryo quality changes from day 5 to day 6, in order to investigate whether prolonged culture to day 6 negatively impacts embryo quality and viability. The association of embryo expansion/quality to euploidy was also examined. Live birth outcomes were analysed, too.

Ovarian stimulation protocol

Ovarian stimulation for each ART cycle was initiated on the 2nd or 3rd day of menstruation, in line with established global antagonist protocol standards. Prior to stimulation, an ultrasound was performed to assess follicular count and ensure there were no ovarian cysts. A complete hormonal profile, including FSH, LH, AMH, estradiol (E2), progesterone and BHCG, was also carried out. The gonadotrophin type (recombinant or hMG) and dosage were chosen based on the hormonal findings, the levels of AMH, the antral follicle count, as well as the response to any previous stimulation cycles. Antagonist (cetrorelix or ganirelix) was introduced on day 5 of stimulation. Ovarian stimulation continued for 10–12 days, with regular monitoring of growth of follicular size and hormone levels (E2 and LH). Oocyte retrieval was scheduled 36 h after HCG or GnRH administration. The time of trigger was determined on the day that the largest follicles reached 18 mm in size, with E2 levels correlating to the follicular size and number. To prevent ovarian hyperstimulation, patients with E2 levels above 10,000 pmol/L were given an agonist trigger only (triptorelin 0.2 or 0.3 mg), avoiding BHCG trigger.

Embryo culture and vitrification

To minimise the risk of external contamination, the protocol required the use of gloves and masks while handling and conducting procedures within a laminar flow hood. Blastocysts were generated using standardised intracytoplasmic sperm injection. Out of the 336 resulting embryos, 96 were cultured in CBI and 240 in TLI. Embryos were cultured individually in drops of SAGE one-step culture media containing human serum albumin (HSA) (CooperSurgical, Denmark), covered with mineral oil, and incubated at 37 °C with 6% CO₂ and 5% O₂. The embryos cultured in CBI were cultured individually in 25-μL drops in a 60-mm round culture dish, while the embryos cultured in TLI were cultured individually in 20-μL drops in an ic8 embryoslide from the time of ICSI till the morning of day 4. Embryos were washed thoroughly in the morning of day 4 in 6 drops of 20 μL, with a new capillary used for each embryo and individually transferred to a new culture dish in individual drops, with the same volume of drops as described earlier. Embryos were graded on day 5 and subsequently on day 6 using the Gardner’s criteria for blastocyst scoring [30]: stage 3 indicates a fully formed blastocyst with a well-defined cavity, stage 4 represents an expanded blastocyst with thinning zona pellucida, stage 5 signifies a hatching blastocyst and stage 6 denotes a fully hatched blastocyst. For the morphological quality we used Gardner’s A, B or C for the inner cell mass (ICM) and trophectoderm (TE) individually, where “A” denotes the highest quality with tightly packed cells for the ICM and a cohesive epithelium for the TE, “B” indicates slightly looser cells or less cohesive epithelium and “C” reflects the poorest quality with few cells or irregular structures [30]. Embryos were classified as good quality when ICM and TE were A or B quality, and as poor quality when either ICM or TE or both were grade C.

All embryos were collapsed using a laser before vitrification; however, collapsing was done in a separate drop and not in the one where the embryo was cultured and that was sent for testing. Blastocyst vitrification was performed on day 6 once the blastocysts reached expansion grade 3 or higher, and presented with any combination of inner cell mass (ICM) or trophectoderm (TE) grades A or B, according to Gardner’s scoring system [30]. Blastocysts were vitrified using the Cryotop method vitrification system (Kitazato).

Collection of SCM, sequencing and analysis

On day 6, 20 μL of SCM were collected per blastocyst from the embryos cultured in CBI (from the 25-μL drop) and 17 μL from the embryos cultured in TLI (from the 20-μL drop). The larger culture volume and aspirated volume in the CBI system were necessary to avoid aspirating oil, as the drops are standalone and more exposed, whereas the enclosed well structure of the TLI culture dish allowed for smaller volumes. The same protocol was also followed for the concordance analysis validation on day 5 blastocysts. A new tip was used per each drop, and a new capillary was used to remove each embryo from the drop. A negative control drop for each embryo was also included, ensuring no direct contact with an embryo, in order to confirm that there was no contamination. The spent media were frozen at − 20 °C for further analysis. The analysis was performed using whole-genome amplification (WGA), and DNA barcoding was performed using the Ion ReproSeq PGS Kit (Thermo Fisher Scientific, MA), with a modified protocol for spent media because of the specific characteristics of embryonic cfDNA. Libraries were quantified, diluted and loaded onto the Ion Chef (Thermo Fisher Scientific) for automated template preparation and chip loading. Sequencing was done using S5 XL sequencer (Thermo Fisher Scientific).

Blastocyst warming and transfer

All transfers were synchronised to a day 5 patient endometrium. Embryos were transferred 2–3 h post-warming. A total of 91 frozen embryo transfers (FET) of 103 embryos were performed, out of which 79 were single embryo transfers (SET) and 12 were double embryo transfers (DET). Among 79 SET, 65 were transfers of euploid embryos, 8 of non-informative embryos and 6 of aneuploid embryos. In the DET, no mixed transfer of euploid embryos with inconclusive or aneuploid took place.

Embryo selection for patients with more than one euploid embryos was based on embryo morphology with prioritisation given to better quality embryos. For patients without euploid embryos, non-informative embryos were prioritised over the aneuploid ones. For patients with aneuploid embryos, genetic counselling was provided to the patients prior to the transfer. The decision to proceed with transfer was made after consultation with both the genetic counsellor and patient’s physician and appropriate consent forms were signed by the patients to acknowledge the risks involved.

Pregnancy and live birth outcomes

Pregnancy was confirmed through a blood test 10 days after FET. Clinical pregnancy was confirmed via ultrasound scan at 6 weeks to detect foetal cardiac activity. Follow up on live births was conducted through phone calls made by either the nurses or patient’s physician. The pregnancy rate was calculated as the percentage of patients with positive bHCG over the total number of transfers, the ongoing pregnancy rate as the percentage of patients with a heartbeat after ultrasound scan at 6 weeks over the total number of transfers and the live birth rate (LBR) as the percentage of patients with a live birth over the total number of transfers.

Statistical analysis

Categorical variables were expressed in percentages and compared using Chi-square. Nominal variables were analysed using the Cochran-Armitage test to detect trends and Wilcoxon rank-sum test was used to compare continuous variables. Multivariate analysis was used to analyse the variables related to liver birth rate and implantation rate. Patient age, incubator, blast expansion and embryo quality were used as covariates. The adjusted odds ratio and the corresponding confidence interval for each coefficient were computed. Statistical software R (version 4.2) was used to perform the statistical analysis.

Ethical considerations

This retrospective study was approved by the Dubai Scientific Research Ethics Committee (Ref: DSREC-01/2025_17). All patients undergoing IVF who signed informed consent for niPGT-A were included in the study, regardless of age. While the study did not involve any intervention beyond routine clinical care, we recognise the importance of transparency when implementing new genetic technologies. Patients were fully informed by their physician and embryology team about the limitations and implications of niPGT-A, and genetic counselling is always available prior to embryo transfer to support informed decision-making. Genetic counselling was mandatory before transferring any embryos with an aneuploid result.

Results

Concordance analysis for test validation on day 5 vs. day 6—nonclinical cases

In a total of five informative donated blastocysts analysed at day 5, only one was concordant (20%) between SCM and whole embryo. By contrast in a total of 13 informative donated blastocysts analysed at day 6, 12 were fully concordant (92.3%) and one not concordant. There was a significant increase in terms of concordance on day 6 (p = 0.002). No maternal contamination was detected by analysing the control samples. Mosaicism could not be assessed, as the current niPGT-A methodologies are not able to reliably detect mosaicism.

Informativity—clinical cases

In embryos studied on day 5, the overall informativity rate was 37.9% (11 out of 29 blastocysts), which was significantly lower than the informativity rate on day 6, which was 85.7% (288 out of 336 blastocysts); p < 0.0001 (Supplemental Fig. 1). Due to the low informativity and concordance of day 5 blastocysts, we decided to exclude the day 5 embryos from further analysis and continue the analysis with the 336 day 6 blastocysts only to avoid bias.

Age specific comparisons showed no significant differences in informativity on day 6 blastocysts across four different age groups (p = 0.22). That is, the informativity rate was 95.6% for ≤ age 38; 85.7% for age 39–40; 85.5% for age 41–42 and 89.9% for > 42 years old (Supplemental Fig. 2).

Informativity rates on day 6 were not influenced by blastocyst quality nor expansion. Specifically, the informativity rate was not affected by the level of blastocyst expansion: 90.91% for expansion stage 3 (full blastocyst), 91.89% for expansion stage 4 (expanded), 88.56% for expansion stage 5 (hatching) and 80% for expansion stage 6 (hatched) (p = 0.51) (Supplemental Fig. 3).

Likewise, no significant differences were observed based on ICM quality: 97.6% for quality A, 89.2% for quality B and 96.3% for quality C (p = 0.48), or TE quality: 88.4% for quality A, 89.8% for quality B and 88.9% for quality C (p = 0.99). When combining ICM and TE quality, the informativity rate for embryos of good quality (A and B) was 88.5%, compared to 90.9% for embryos of poor quality C (p = 0.67).

No differences in informativity rates were observed between the two incubator settings on day 6, with CBI incubators showing 86.4% and TLI incubators 84.3% (p = 0.74).

Impact on blastocyst quality in extended culture to day 6

The change in blastocyst expansion and quality from day 5 to day 6 is represented in Figs. 1, 2, and 3 as stacked bar plots. Each vertical axis represents the embryos categorised into three or four groups, according to their appearance on day 5, e.g. “no blastocyst, full, expanded, hatching, or hatched blastocyst (stages 3–6 above, Gardner scale)” (for Fig. 1) or grade A, B or C (Gardner scale) for Figs. 2 and 3. In each of the figures, the horizontal axis represents the subsequent progression of the same embryos on the following day (day 6). For instance, they were expanded or hatching/hatched blastocyst (for Fig. 1) or A–C on the Gardner scale (Figs. 2 and 3). Figure 2 pertains to the ICM and Fig. 3 the TE. We thus measured whether blastocysts continued to develop after day 5 and if their quality increased or decreased with an extra day in culture.

Fig. 1.

Embryo expansion status from day 5 to day 6. The vertical axis shows the day 5 expansion stage; the horizontal axis shows progression by day 6

Fig. 2.

Changes in inner cell mass (ICM) quality from days 5 to 6. The vertical axis shows the day 5 ICM quality; the horizontal axis shows ICM quality by day 6

Fig. 3.

Changes in trophectoderm (TE) quality from day 5 to day 6. The vertical axis shows the day 5 TE quality; the horizontal axis shows TE quality by day 6

As shown in Fig. 1, of 114 embryos that had not reached blastocyst by day 5, all had done so by day 6, with 8 (7%) full blastocysts, 42 (37%) expanded and 64 (56%) hatching by day 6; none were fully hatched. Of the 50 that were full blastocysts on day 5, a total of only three (6%) remained at the same stage, 22 (44%) were expanded, 24 (48%) hatching and one (2%) had fully hatched. Of the 111 embryos that were expanded, nine (8%), 94 (85%) and eight (7%) were still expanded, hatching and fully hatched respectively. Of the 61 embryos that were hatching on day 5, only one (2%) appeared not to be hatching any more, with the remainder 54 (89%) either still hatching or completely hatched six (10%). These results are expressed in Fig. 1.

Scoring according to the Gardner scoring system is compared on day 5 (vertical axis) with the same embryos on day 6 in Figs. 2 (for the ICM) and 3 (for the TE). Considering the ICM, of the 154 embryos that were grade A on day 5, a total of 105 (68.2%) remained so, 45 (29.2%) were grade B and four (2.6%) degraded to grade C. Of the 57 that were grade B on day 5, 15 (23.2%) improved to grade A, 35 (61.4%) remained at grade B, with only seven (12.3%) degrading to grade C. In total therefore, of the 211 embryos that were grades A or B on day 5, 200 (94.8%) remained so, with only 11 (5.2%) degrading to grade C. Of the 11 that were originally grade C on day 5, only three (27.3%) remained so, with the remaining eight (72.7%) improving to grades A (1 (9.1%)) or B (7–63.6%). In a total of 222 embryos therefore, 166 (74.8%) maintained or improved their quality, with only 56 (23.2%) having a poorer quality on day 6. Of these 56 embryos, 45 (80.3% or 20.3% of the total) were nonetheless still good quality (grade B) on day 6. These results are expressed in Fig. 2.

Now considering the TE, of the 122 embryos that were grade A on day 5, a total of 97 (79.5%) remained so on day 6, 22 (18.0%) were grade B and three (2.5%) degraded to grade C. Of the 75 that were grade B on day 5, a total of 19 (25.3%) improved to grade A by day 6, 49 (65.3%) remained at grade B, with only seven (9.3%) degrading to grade C. In total therefore, of the 197 embryos that were grades A or B on day 5, 187 (94.9%) remained so, with only 10 (5.1%) degrading to grade C. Of the 25 that were originally grade C on day 5, 11 (44%) remained so, with the remaining 14 (56%) improving to grades A (2 (8%)), or B (12–48%). In a total of 222 embryos therefore, 190 (85.6%) maintained or improved their quality, with only 32 (14.4%) having a poorer quality on day 6. Of these 32 embryos, 45 (68.8% or 9.9% of the total) were nonetheless still good quality (grade B) on day 6. These results are expressed in Fig. 3.

Overall, therefore, the evidence suggests that extended culture to day 6 does not detriment development of the embryo. If anything, in morphological terms, there is a slight improvement afforded by this extended culture.

Chromosomal analysis for aneuploidy

Overall, the euploidy rate (percentage of diagnoses with no chromosomal abnormality) was 40%, with no maternal contamination detected in any of the cases after analysis of the control samples. No significant differences were seen between different incubator settings (p = 0.63). A significant trend to higher aneuploidy with increased female age was observed (p = 0.03 or 0.02 depending on the analysis; model vs. age groups).

The euploidy rate was 60.0% (6/10) for full blastocysts, 38.2% (26/28) for expanded blastocysts, 37.3% (78/209) for hatching blastocysts and 75.0% (9/12) for fully hatched blastocysts (Supplemental Fig. 4). The blastocyst expansion grade on day 6 was significantly correlated to euploidy (p = 0.04), with more embryos assessed as euploid at stages 3 (full blastocyst) and 6 (fully hatched). These results are represented in Supplemental Fig. 4. Because of the small numbers in embryos that were grades 3 and 6, we grouped together grades 3 and 4, as well as grades 5 and 6. Among the full/expanded blastocysts, the euploidy rates were 41.0% (32/78), and among the hatching/hatched blastocysts, the euploidy rates were 39.4% (87/221). No statistical difference was observed between the two groups and the euploidy rates on day 6 (p = 0.79).

For day 6 blastocysts, the euploidy rates were 43.5% (104/239) for good quality blastocysts and 25.0% (15/60) for poor quality ones. Thus, there were more chromosomally normal good quality blastocysts than poor quality ones (p = 0.01). Moreover, among good quality blastocysts there were statistically more aneuploid vs. euploid embryos (56.5% vs. 43.5%; p = 0.045), and there were more aneuploid than euploid embryos among the poor-quality embryos (75.0% vs. 25.0%; p = 0.0001). A relatively higher euploidy rate was associated with good embryo quality.

Live birth rate analysis

The overall pregnancy, ongoing pregnancy and live birth outcomes are summarised in Table 1. A total of 103 day 6 niPGT-A analysed blastocysts were transferred in 79 single embryo transfers (SET) and 12 double embryo transfers (DETs) in 68 patients. The overall pregnancy rates were 62.3% (43/69), 80% (8/10) and 50% (6/12) for euploid, non-informative and aneuploid embryos respectively (p = 1.0). Regarding the types of chromosomal abnormalities in the transferred aneuploid embryos, two involved segmental alterations, one exhibited a complex abnormality affecting four full chromosomes, and the remaining cases involved abnormalities in one or two whole chromosomes. Also, the ongoing pregnancy rates were 36.23% (25/69), 60% (6/10) and 8.3% (1/12) for transfers of euploid, non-informative and aneuploid embryos respectively (p = 0.9). We report a LBR of 33.3% (23/69) for transfers of euploid embryos, 50% (5/10) for transfers of non-informative embryos and 8.3% (1/12) for transfers of aneuploid embryos (p = 0.79). The aneuploid embryo that led to a healthy live birth was the complex abnormal with four full chromosomes affected. The 33.3% LBR of euploid embryos compares to an age-matched, not tested cohort of 10% reported by the HFEA in 2022 (p < 0.001) (HFEA data for 40–42 year olds, as reported above, the mean age of 40.9) [31]. Also, it exceeds the 29% LBR observed in our control group, which includes patients aged 38 and older who underwent transfers of untested embryos in 2023 (p = 0.58). Overall, the LBR is higher when euploid embryos are transferred compared to aneuploid, with only one live birth reported after transfer of an embryo diagnosed as complex abnormal.

Table 1.

Pregnancy, ongoing pregnancy and live birth outcomes

	SET (79 transfers)			DET (12 transfers)			Overall (91 transfers)
niPGT-A result	Pregnancy (%)	Ongoing pregnancy (%)	Live birth (%)	Pregnancy (%)	Ongoing pregnancy (%)	Live birth (%)	Pregnancy (%)	Ongoing pregnancy (%)	Live birth (%)
Euploid (mean age 40 ± 3.9)	39/65 (60%)	22/65 (33.8%)	20/65 (30.8%)	4/4 (100%)	3/4 (75%)	3/4 (75%)	43/69 (62.3%)	25/69 (36.23%)	23/69 (33.33%)
Non-informative (mean age 41 ± 3.2)	6/8 (80%)	5/8 (62.5%)	4/8 (50%)	2/2 (50%)	1/2 (50%)	1/2 (50%)	8/10 (80%)	6/10 (60%)	5/10 (50%)
Aneuploid (mean age 42 ± 2.3)	3/6 (50%)	0/6 (0%)	0/6 (0%)	3/6 (50%)	1/6 (16.7%)	1/6 (16.7%)	6/12 (50%)	1/12 (8.3%)	1/12 (8.3%)

Discussion

This study highlights the utility of niPGT-A using SCM, as it investigates the potential risks to embryo viability associated with prolonged day 6 embryo culture and sequential washes. Specifically, we report that prolonged day 6 culture is more informative than day 5 in terms of the proportion of embryos that yielded a readable result (informativity), with high levels concordance on day 6 (92.3%) between SCM and the whole embryo and our results agree with previous publications [23, 24]. Prolonged day 6 culture revealed no age-specific differences in terms of informativity, nor differences in terms of blastocyst expansion and quality. The informativity rates are thus comparable with those previously reported by our centre based on previous validations and publications [32]. We observed no significant difference in informativity between embryos cultured in a benchtop incubator and those in a time-lapse incubator. Although Sakkas et al. (2024) did not directly compare different incubator settings, their multicentre study—spanning diverse laboratory conditions and culture systems—similarly reported no negative impact on clinical outcomes following the implementation of a standardised niPGT-A protocol [29]. Also, we did not find that there was any detrimental effect to prolonging culture. That is, the degree of expansion and the morphological quality of the embryos was at least as good, if not better, following prolonged culture to day 6. This finding is significant in the world of niPGT-A as prolonged culture to this stage is not usually a facet of a regular IVF cycle. That is, most embryo transfers are typically on day 5 and transfer of day 6 embryos is restricted to slow embryos that have not reached blastocyst stage on day 5. Prolonged culture to day 6, with subsequent vitrification and transfer of day-6 thawed blastocysts are requirements for optimal niPGT-A cases, so that informative, meaningful results can be obtained.

A key feature of this study is that it was conducted in older women (mean age 40.9 ± 3.5), despite niPGT-A often being recommended for younger patients as a prioritisation tool. The focus on women of advanced reproductive age is a particular strength of this study, as this is a primary PGT-A referral category. We observed an association between euploidy and good embryo morphology, and our live birth rates (LBR) suggest an improvement in LBR per embryo transfer for older women compared to cases where PGT-A was not performed, based on our control group and international benchmarks [31]. This study is unique in focusing on the primary PGT-A referral category—advanced maternal age—while also addressing concordance, informativity, the impact of prolonged culture, and subsequent LBRs. Our findings fill a literature gap in LBR reporting, with a 33.3% LBR for euploid embryo transfers, significantly higher than the 10% reported by HFEA for women aged 40–42 and with a higher trend compared to the control group. While these results suggest a potential added value of niPGT-A in treatment, they should be interpreted cautiously due to small sample sizes. As mentioned, another important limitation of niPGT-A is that a proportion of embryos fail to yield informative results, likely due to insufficient cfDNA release. Although the number of such cases in our cohort is small, it is noteworthy that six out of eight non-informative embryos transfers led to healthy pregnancies. This suggests that non-informative results do not necessarily indicate chromosomal abnormality and highlights the risk of potentially excluding viable embryos in clinical practice. Additionally, we report a live birth following the transfer of a complex aneuploid embryo, underscoring the need for protocol optimisation to improve concordance rates and the importance of genetic counselling before such transfers. Among existing studies on LBR, Sun et al. recently presented the most comprehensive findings, demonstrating that cumulative LBRs for PGT-A (blastocyst biopsy) and niPGT-A (SCM) surpassed ICSI without diagnosis (27.9% vs. 44.9% and 51.0%, p < 0.003) [33]. They concluded that both invasive and non-invasive PGT-A effectively and comparably select euploid embryos in women aged 35–40. Although live birth rate (LBR) follow-up was performed for all transferred cases, systematic follow-up of miscarriages was not possible, as these cases were managed by the treating physicians and not documented in the electronic medical record. Future prospective studies should aim to incorporate prenatal follow-up and foetal tissue analysis where feasible.

In this study, the overall 85% informativity rate, with no significant difference between different age groups is reassuring and similar to previously reported publications [26]. However, this informativity rate is still a little lower than typically reported for invasive PGT-A, although we consider as non-informative not only the SCM samples with non-conclusive results, but also media with chaotic chromosome constitution, which is not the case with the regular estimations of informativity in PGT-A, where chaotic embryos are classified as aneuploid. In time, as technology develops, however, it seems likely that this will improve. In this respect, the benefits of niPGT-A will only be realised if prior preparation of the embryo is significantly less involved than the embryo biopsy itself. Perhaps the main novel finding of this paper that is relevant to the practice of niPGT-A is that concerns about the prolonged culture to day 6 should not present a problem. Indeed, it seems likely that day 6 culture will become commonplace if niPGT-A is practiced widely, and this study is one of the few that provides reassurance by analysing the same embryos on day 5 and day 6, by performing the whole study “end to end” from ICSI to niPGT-A to transfer. The absence of demonstrable differences in informativity regardless of age, culture setting, blastocyst expansion level or grade suggests that sampling and diagnostic methods are robust and repeatable. Moreover, the culture protocol used in this study supports prolonged day 6 culture (here, we used the protocol previously published by Rubio et al., 2019 [23]), and the embryos did not degenerate and die. If anything, their quality improved on average. As with many studies however, the relatively small numbers of embryos analysed here is a limitation and can be seen as a prelude to larger studies where thousands of embryos are examined with these questions in mind [26]. Further areas of bias include the objective scoring of operators—it was not practicable to “blind” the embryos for the study as the number of days in culture was obvious from the records and from the developmental level of the embryo, the potential of selection bias and lack of a concurrent control group.

In this study, we report a pregnancy rate of 62.3% for transfers of euploid embryos. A few clinical studies however have been conducted, such as Fang et al., who reported a 58% pregnancy rate [34]. Franco et al. noted higher pregnancy rates for niPGT (61%) using SCM versus blastocyst biopsy controls (49%), although differences were not significant [35]. Chen et al. [20, 21] compared 345 TE biopsy results with niPGT, developing an algorithm to categorise niPGT results by pregnancy potential. They prospectively analysed 278 cycles with niPGT, finding group A (91% euploidy, 52% ongoing pregnancy rate) performed significantly better than groups B (76% euploidy, 49% OPR) and C (35% euploidy, 27% OPR) with p < 0.003. However, the study lacked a control group, and false positive/negative rates remained a concern for diagnostics.

In our centre, we have observed that patients are increasingly concerned about the invasiveness associated with conventional PGT-A, which has generated substantial interest in non-invasive alternatives like niPGT-A. Patients are more willing to undergo niPGT-A precisely because it circumvents the need for an embryo biopsy. However, one of the main challenges we face in implementing niPGT-A is the necessity for prolonged culture to day 6. The approximate 15% risk of non-concordance is not such a concern, given that the test is used as an extra biomarker to help to prioritise embryos for transfer and does not dictate which embryo should be transferred. Our centre’s extensive experience with niPGT-A, including robust validation data and detailed analysis of factors affecting informativity and euploidy rates, provides critical insights for optimising this workflow. By sharing our findings, we aim to assist other centres in successfully adopting niPGT-A. Importantly, our promising live birth rates, particularly among older patients, underscore the potential of niPGT-A to enhance clinical outcomes while alleviating concerns about procedural invasiveness.

A key consideration is which “standard” protocol should be used in order to harvest the maximum quantity of DNA from the SCM, as the studies performed so far have not seen an increase in the amount of DNA after LAH or biopsy is performed. As mentioned above, we used the protocol previously published by Rubio et al. [23] but, of course, there are others and, in the long term, procedures should be standardised. Hanson et al. conducted a prospective non-selection study wherein SCM was collected prior to both embryo vitrification and biopsy (two processes that may elevate DNA levels in the culture media) [16]. In this study, no special measures were applied to minimise the risk of maternal contamination. This study compared the Yikon niPGT test with targeted NGS (next generation sequencing) of TE biopsy samples, revealing a 15% amplification failure rate for SCM samples collected on day 5, compared to 0% for those collected on day 6, these are slightly better than the rates we report here. The concordance rate for ploidy was however disappointingly low, with only 60% of SCM correlating with TE biopsy, indicating that 13% of viable euploid embryos might be inappropriately discarded. Here, although numbers are small, we only observed one out of 13 embryos that were not concordant (SCM vs whole embryo; 7.7%), and this could theoretically be a predominantly normal embryo that nonetheless shed some aneuploid DNA into the SCM.

Some studies have gone so far as suggesting that niPGT-A by day 6 SCM may offer greater accuracy than PGT-A using TE biopsy results when correlating to the ICM [22]. Here, we do not find direct evidence of this as it was not inherent in the study design. It is notable however that, if anything, our results do seem to show a slight improvement in embryo quality (alongside the highly significant improvement in informativity) by extended culture to day 6. A comprehensive review of literature indicates that the error rates for niPGT-A and TE compared to the ICM are similar. In 2024, a series of studies emerged highlighting the efficacy of niPGT-A. Nakhuda et al. conducted a non-selection pilot study evaluating the clinical predictive value of copy number variations detected through next-generation sequencing of cfDNA in spent culture media. They reported a negative predictive value of 57.3% (43/75) for presumed euploid samples and a positive predictive value of 94.4% (17/18) for samples identified as aneuploid, based on subsequent clinical outcomes. Their study also highlighted a potential concern for maternal contamination, reflected in an uneven sex ratio. Importantly, they reported ongoing pregnancy rates of 57.3% in the euploid group and 5.6% in the aneuploid group, which are consistent with the trend observed in our findings. Ardestani et al. found that culturing frozen/thawed blastocysts until late day 5 or early day 6 provided more significant insights than earlier assessments, reporting concordance rates of 90.5% (day 5 short), 93.6% (day 5 long) and 92.3% (day 6 short) with respect to the whole blastocyst chromosome constitution [32]. Chow et al. examined various factors including SCM collection times, rinsing protocols and the use of IVF versus ICSI within the context of niPGT-A, using conventional PGT-A as a control [36]. Their results indicated higher concordance rates between day 6 samples as opposed to day 5, especially with a sequential rinsing method outperforming single-step rinsing, achieving concordance rates of 82.5–85%, without notable differences between IVF and ICSI. A consistent finding is that later embryonic development stages, from around day 6, may yield more reliable concordance rates, with further variations being due to factors such as collection timing and clinic-specific methodologies. Despite the promising results shown in this study, as well as previous studies, there is still room for improvement in concordance rates; therefore, niPGT-A could be offered as a prioritisation tool for embryo selection till then. Also, the results are not yet conclusive as to whether invasive PGT-A can be definitely replaced by non-invasive approaches.

Conclusion

Our findings support the continued practice of prolonged culture to day 6 for the purposes of niPGT-A. Additionally, live birth analysis suggests a potential added value of niPGT-A in patient treatment, with rates exceeding those reported internationally. While further advancements are needed before niPGT-A can fully replace PGT-A by trophectoderm biopsy, the results presented here represent a significant step toward that goal.

Supplementary Information

Below is the link to the electronic supplementary material.

Author contribution

MB and DK conceived and designed research. MB and EVZ conducted laboratory work. MB and MA compiled and analysed data. MB and DG wrote the manuscript. All authors read and approved the manuscript.

Data availability

Patient data available as part of patient treatment.

Declarations

Ethics approval

The study was approved by the Dubai Scientific Research Ethics Committee (Ref: DSREC-01/2025_17).

Competing interests

The author declare no competing interests.