Effect of the endometrial thickness on the live birth rate: insights from 959 single euploid frozen embryo transfers without a cutoff for thickness

OBJECTIVE

To investigate whether endometrial thickness (ET) independently affects the live birth rate (LBR) after embryo transfer.

DESIGN

Retrospective study.

SETTING

Private assisted reproductive technology center.

PATIENT(S)

A total of 959 single euploid frozen embryo transfers.

INTERVENTION(S)

Vitrified euploid blastocyst transfer.

MAIN OUTCOME MEASURE(S)

Live birth rate per embryo transfer.

RESULT(S)

The conditional density plots did not demonstrate either a linear relationship between the ET and LBR or a threshold below which the LBR decreased perceivably. Receiver operating characteristic curve analyses did not suggest a predictive value of the ET for the LBR. The area under the curve values were 0.55, 0.54, and 0.54 in the overall, programmed, and natural cycle transfers, respectively. Logistic regression analyses with age, embryo quality, day of trophectoderm biopsy, body mass index, and ET did not suggest an independent effect of the ET on the LBR.

CONCLUSION(S)

We did not identify a threshold of the ET that either precluded live birth or under which the LBR decreases perceivably. Common practice of cancelling embryo transfers when the ET is <7 mm may not be justified. Prospective studies, in which the management of the transfer cycle would not be altered by ET, would provide higher-quality evidence on the subject.

P-589 Natural endometrial preparation for single euploid frozen embryo transfer increases the likelihood of live birth in obese patients

Study question

Does a natural endometrial preparation for single euploid frozen embryo transfer increase the odds of live birth (LB) in obese patients?

Summary answer

Compared to an artificial approach, a natural endometrial preparation for single euploid frozen embryo transfer improves the odds of LB in obese patients.

What is known already

Several protocols to prepare the endometrium for frozen embryo transfer have been described with comparable clinical outcomes, and no individual protocol has been demonstrated to be superior. Likewise, no specific patient group has been defined that can substantially benefit from a natural endometrial preparation. Despite obesity being linked to adverse clinical outcomes, observational studies and clinical trials have traditionally included participants with normal weight when comparing protocols, resulting in obese patients being underrepresented in clinical research. Therefore, studies focusing on endometrial preparation protocols for frozen embryo transfer that include obese patients are required.

Study design, size, duration

This study involved the analysis of a retrospective dataset including 975 single euploid frozen embryo transfer cycles performed at two tertiary referral centers between March 2017 and November 2019. The primary outcome was LB after natural and artificial endometrial preparation. Participants were stratified according to the World Health Organization Body mass index (BMI) classification into three groups (G): G1: Normal-weight ≤24.9 (n = 390), G2: Overweight 25 to 29.9 (n = 332), and G3: Obese ≥30 (n = 253).

Participants/materials, setting, methods

Cycles included infertility patients attempting a single euploid frozen blastocyst transfer under natural (n = 324) or artificial (n = 651) endometrial preparation. PGT-A was performed after trophectoderm biopsy using Next-Generation Sequencing. Spontaneous LH rise confirmed ovulation for natural endometrial preparation, and embryo transfer occurred five days after initial progesterone elevation. Hormones were measured from blood samples. For artificial preparation, embryo transfer was performed after oral estradiol and 120-hour vaginal progesterone exposure. All participants received luteal phase support.

Main results and the role of chance

Within each BMI group, no statistically significant differences in age were observed in patients between a natural and artificial endometrial preparation: G1 (33.5±5.1 vs. 32.7±5.2 years; p = 0.13), G2 (34.5±4.9 vs. 33.9±5.8 years; p = 0.39), and G3 (35.6±4.9 vs. 34.3±5.6 years; p = 0.1). Similarly, no statistically significant differences were observed regarding implantation rates: G1 (60% vs 57.1%; p = 0.57), G2 (54.8% vs 55.7%; p = 0.88), and G3 (64.3% vs 51.4%; p = 0.07). LB rates remained constant across all BMI groups when a natural preparation was conducted: G1 [56.66% (n = 85/150)], G2 [49.03% (n = 51/104)], and G3 [58.57% (n = 41/70)]. We observed a decrease in LB rates in G3 when an artificial endometrial preparation was performed: G1 [47.5% (n = 114/240)], G2 [44.73% (n = 102/228)], and G3 [34.42% (n = 63/183)]. The positive impact of a natural preparation on LB rates was more significant in G3, showing a 24% difference of effect between the natural and artificial endometrial preparations. Univariate logistic regression analysis showed a statistically significant difference in the primary outcome (LB) when a natural endometrial preparation was conducted in G3 (OR 2.69, 95% CI 1.53-4.74; p = 0.001), with no differences found in G1 (OR 1.45, 95% CI 0.96-2.18; p = 0.08) and G2 (OR 1.19, 95% CI 0.75-1.89; p = 0.73).

Limitations, reasons for caution

This study could be subject to bias due to its retrospective nature. Sperm quality was not considered for the present study.

Wider implications of the findings

Obese women appear to be the patient group that can obtain benefit from a natural endometrial preparation for frozen embryo transfer by increasing the odds of LB. Future prospective studies should confirm this hypothesis. A natural approach might be unfeasible in a subgroup of obese patients due to anovulatory cycles.

Trial registration number

not applicable

P-358 Effect of endometrial thickness on biochemical pregnancy rate: an analysis of 1534 frozen euploid embryo transfers

Study question

Can a thick endometrial lining measured prior to embryo transfer be considered a protective factor against Biochemical Pregnancy (BP)?

Summary answer

The risk of BP is independent of Endometrial Thickness (EMT), but rather dependent of the type of endometrial preparation and parity.

What is known already

Higher EMT prior to embryo transfer is associated with better clinical outcomes in general, like higher implantation and livebirth, and lower miscarriage rates. But up to our knowledge, no studies evaluated the effect of EMT on BP per say.

Study design, size, duration

This is a two-center retrospective observational study including a total of 1534 euploid Frozen Embryo Transfer (FET) cycles between March 2017 and March 2020 at ART Fertility Clinics Muscat, Oman and Abu Dhabi, UAE. BP is defined as blood beta-hCG &gt;15 mIU/ml on day 12 post FET, that is progressively decreasing, with no evidence of gestational sac on ultrasound.

Participants/materials, setting, methods

The study group consisted of 112 cases of BP, while the control group consisted of the remaining 1422 FET’s that led to different clinical outcomes. EMT was measured by transvaginal ultrasound on the day of progesterone rise (±1 day); that rise was either spontaneous in Natural Cycles (NC), or iatrogenic in Hormone Replacement Therapy (HRT) cycles. Euploidy status of the embryos was assessed by NGS analysis of trophectoderm biopsies. Bivariate and multivariate analyses were conducted.

Main results and the role of chance

There was no difference in mean EMT between the study and the control groups (7.55 vs. 7.68 mm, p = 0.154).

Looking at the association of different variables with the rate of BP, there was no effect of age, BMI, AMH, number of embryos transferred, degree of blastocyst expansion, inner cell mass or trophectoderm grade, day of biopsy, nor presence of blood or mucus on the transfer catheter. However, patients on HRT cycles had significantly higher rates of BP compared to NC (8.42% vs. 4.99%, p = 0.015). Also, those with a previous livebirth had higher rates of BP compared to nulliparous women (8.7% vs. 5.39%, p = 0.014).

The distribution of BP showed that 54.5% occurred with EMT &lt;7.5 mm, 34.8% with EMT 7.5-9 mm, and 10.7% with EMT &gt;9 mm. These represents respectively 8.16%, 6.68%, and 5.94% of the total sample. This decreasing trend of BP with increasing EMT didn’t reach statistical significance (p = 0.429). Univariate analysis comparing the risk of BP in FET’s done with lower and higher EMT to those performed at 7.5-9 mm yielded similar conclusion: OR = 1.24 [0.82-1.88] for &lt;7.5 mm, and OR = 0.88 [0.45-1.72] for &gt;9 mm.

Controlling for different confounders, HRT cycles and multiparity remained as independent risk factors for BP.

Limitations, reasons for caution

Inter-observer variability in EMT measurement and the transfer technique, the retrospective nature of the study, and the lack of data on the mode of delivery of parous women could all have interfered with the conclusion.

Wider implications of the findings

The reduced adverse clinical outcomes with NC shed light on the role of the corpus luteum in the early phases of implantation, and some potential secreted mediators other than progesterone. Besides, the effect of previous deliveries on the endometrium and its receptivity needs further investigation.

Trial registration number

not applicable

Segmental duplications and monosomies are linked to in vitro developmental arrest

PURPOSE

To verify which genetic abnormalities prevent embryos to blastulate in a stage-specific time.

METHODS

A single center retrospective study was performed between April 2016 and January 2017. Patients requiring Preimplantation Genetic Testing for Aneuploidies (PGT-A) by Next Generation Sequencing (NGS) were included. All embryos were cultured in a time-lapse imaging system and single blastomere biopsy was performed on day 3 of development. Segmental duplications and deletions as well as whole chromosome monosomies and trisomies were registered. Embryo arrest was defined if the embryo failed to blastulate 118 h post-injection. A logistic regression model was applied using the time to blastulate as the response variable and the different mutations as explanatory variables. A p value < 0.05 was considered significant.

RESULTS

Of the 285 biopsied cleavage stage embryos, 103 (36.1%) were euploid, and 182 (63.9%) were aneuploid. There was a significant difference in the developmental arrest between euploid and aneuploid embryos (8.7% versus 42.9%; p = 0.0001). Segmental duplications and whole chromosome monosomies were found to have a significant effect on developmental arrest (p = 0.0163 and p = 0.0075), while trisomies and segmental deletions had no effect on developmental arrest. In case of segmental duplications, an increase of one extra segmental duplication increases the odd of arrest by 159%. For whole chromosome monosomies, the odd will only increase by 29% for every extra chromosomal monosomy. Both chromosomal abnormalities remained significant after adding age as an explanatory variable to the model (p = 0.014 and p = 0.009).

CONCLUSION

Day 3 cleavage stage embryos with segmental duplications or monosomies have a significantly decreased chance to reach the blastocyst stage.

mtDNA dynamics between cleavage-stage embryos and blastocysts

PURPOSE

The aim was to evaluate mtDNA content and its dynamics in euploid and aneuploid embryos from cleavage to blastocyst stage following consecutive biopsies. The effect of female age on mtDNA content was evaluated by comparing reproductively younger (≤ 37 years) with older (> 37 years) women.

METHODS

A retrospective single-centre descriptive study was performed between August 2016 and January 2017. Forty patients, with 112 embryos, undergoing preimplantation genetic testing for aneuploidies (PGT-A) by next-generation sequencing (NGS) were included. Embryos that reached the blastocyst stage and were not selected for fresh embryo transfer were included following consecutive biopsies of a single blastomere on day 3 and trophectoderm biopsy of day 5 blastocysts.

RESULTS

Cleavage-stage mtDNA was significantly lower in fast cleaving embryos (p = 0.016). Based on the concordance between day 3 and day 5 biopsies, a difference was identified in blastocyst mtDNA content between groups (p = 0.019); true euploid blastocysts presented a lower mtDNA content. No association was identified between cleavage-stage mtDNA content and ploidy status (OR 1.008 [0.981-1.036], p = 0.565) nor between blastocyst mtDNA content and ploidy outcome (OR 0.954 [0.898-1.014], p = 0.129). No difference was found when comparing mtDNA content and ploidy outcome between the two reproductive age groups (p = 0.505 (cleavage stage) and p = 0.774 (blastocyst)).

CONCLUSION

Mitochondrial DNA content of cleavage-stage embryos and blastocysts is unable to predict ploidy status. Subgroup analysis based on ploidy concordance between day 3 and day 5 revealed a significantly lower mtDNA content for true euploid blastocysts. Reproductive ageing does not affect mtDNA content.

Cleavage stage mitochondrial DNA is correlated with preimplantation human embryo development and ploidy status

PURPOSE

To evaluate whether the mitoscore of cleavage stage embryos might correlate with developmental kinetics and the ploidy status.

MATERIALS

This retrospective single-center study involved all cycles between April 2016 and April 2018 in which preimplantation genetic testing for aneuploidy (PGT-A) on day 3 was performed. The mitochondrial DNA (mtDNA) content and embryo ploidy were determined on 375 single blastomere biopsies by next generation sequencing (NGS). After intracytoplasmic sperm injection, a time-lapse imaging system (embryoscope) was used to follow the development. The median mtDNA content of cleavage stage embryos (49.4) was used to stratify the embryos into two groups to compare embryo development and ploidy status: low mitoscore group (≤ 49.4) and high mitoscore group (> 49.4).

RESULTS

The total number of euploid embryos was equal between both mitoscore groups (32.1% versus 33.5%; p = 0.854). However, embryos in the low mitoscore group had a significantly higher cell number on day 3 (8.13 ± 1.59 versus 7.62 ± 1.5; p = 0.0013) and showed a significantly faster development up until the 8-cell stage. Mitoscore was not different between euploid and aneuploid embryos, with the same blastomere number at the time of biopsy. Furthermore, absence of cavitation within 118 h after insemination was correlated with higher mitoscore values (60.22 ± 42.23 versus 50.97 ± 13.37; p = 0.006) and a lower chance of being euploid (17.1% versus 47.4%; p = 0.001).

CONCLUSION

mtDNA content of cleavage stage embryos correlates with time-lapse parameters. Early blastulation is correlated with a lower mtDNA content and a higher chance of euploidy.

Clinical reassessment of human embryo ploidy status between cleavage and blastocyst stage by Next Generation Sequencing

One of the most important limitations of genetic testing in preimplantation embryos is embryonic mosaicism, especially when performed on D3 with only a single blastomere evaluated. Previous publications, using Array-Comparative Genomic Hybridization (a-CGH) to compare day 3 (D3) biopsies versus trophectoderm biopsies for the analysis of aneuploid embryos, showed similar high concordance rates per embryo diagnosis for D3 biopsies and trophectoderm biopsies. Next generation sequencing (NGS) was introduced lately as a new technique for preimplantation genetic testing for aneuploidies (PGT-A). Using this technique, this retrospective descriptive study evaluated the degree of the concordance of the diagnosis between preimplantation human cleavage stage (D3) and blastocyst stage (D5) embryos. Double biopsies on D3 and D5 were performed on 118 embryos, reaching blastocyst stage on D5 and had not been selected for transfer. As the fertilization law of the United Arab Emirates does not allow embryo freezing, also surplus euploid embryos after D 3 biopsy were included. Analysis of the NGS results from D3 and D5 embryo biopsies showed a total concordance rate per embryo diagnosis of 85.6% for euploid and aneuploid embryos. The concordance rates per embryo chromosomal pattern for embryo diagnosed as aneuploid at both biopsy stages was 82.2%. However, the status regarding the affected chromosomes was not identical on D3 and D5. Hence, the total concordance rate between D3 biopsy and D5 biopsy was limited to 67.8%. This current study clearly demonstrated that the concordance rates between D3 and D5 biopsies in aneuploid and euploid embryos are lower than previously reported.

Type of chromosome abnormality affects embryo morphology dynamics

OBJECTIVE

To study the differences in the cleavage time between types of embryo chromosomal abnormalities and elaborate algorithm to exclude aneuploid embryos according to the likelihood to be euploid.

DESIGN

Retrospective cohort study.

SETTING

University affiliated private center.

PATIENT(S)

Preimplantational genetic screening patients (n = 112) including cases of advanced maternal age, repeated implantation failure, and recurrent miscarriage. A total of 485 embryos were analyzed.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

All biopsied embryos were cultured in an incubator with time-lapse technology, cleavage timing from insemination to day 3 and all kinetic parameters that have been described in previous studies by our group.

RESULT(S)

Logistic regression analysis were used to identify morphokinetic parameters and some were strongly associated with complex aneuploid embryos; t3 (odds ratio = 0.590, 95% confidence interval 0.359-0.971) and t5-t2 (odds ratio = 0.151, 95% confidence interval 0.082-0.278).

CONCLUSION(S)

Embryo morphokinetics are affected by chromosome aneuploidy and further analysis of the chromosome content reveals higher differences when the complexity in the chromosome disorders is increased. The use of time-lapse monitoring, although not able to detect an abnormal embryo, may be potentially useful to discard those embryos with high risk of complex chromosomal abnormalities.

Sperm DNA fragmentation index does not correlate with the sperm or embryo aneuploidy rate in recurrent miscarriage or implantation failure patients

BACKGROUND

The aneuploidy rate is higher in poor-quality sperm samples, which also have higher DNA fragmentation index values. The aim of this study was to assess the relationship between sperm DNA fragmentation in samples from infertile men belonging to couples with recurrent miscarriage or implantation failure and the aneuploidy rate in spermatozoa as well as in embryos from patients.

METHODS

This prospective study evaluated DNA damage and the aneuploidy rate in fresh and processed (density gradient centrifugation) ejaculated sperm as well as the aneuploidy rate in biopsied embryos from fertility cycles. Fluorescence in situ hybridization was used for the aneuploidy analysis. Results were compared using linear regression and analysis of variance.

RESULTS

A total of 154 embryos were evaluated from 38 patients undergoing PGD cycles; 35.2% of the embryos were chromosomally normal. Analysis of the same sperm samples showed an increased DNA fragmentation after sperm preparation in 76% of the patients. There was no correlation between DNA fragmentation and the aneuploidy rate in embryos or in fresh or processed sperm samples.

CONCLUSIONS

Sperm DNA fragmentation is not related to chromosomal anomalies in embryos from patients with recurrent miscarriage or implantation failure. However, we cannot rule out the possibility that a relationship between DNA fragmentation and aneuploidy exists for other causes of infertility. Furthermore, the different methods used to evaluate DNA fragmentation may produce different results.