A proof of concept for a deep learning system that can aid embryologists in predicting blastocyst survival after thaw

The ability to understand whether embryos survive the thaw process is crucial to transferring competent embryos that can lead to pregnancy. The objective of this study was to develop a proof of concept deep learning model capable of assisting embryologist assessment of survival of thawed blastocysts prior to embryo transfer. A deep learning model was developed using 652 labeled time-lapse videos of freeze-thaw blastocysts. The model was evaluated against and along embryologists on a test set of 99 freeze-thaw blastocysts, using images obtained at 0.5 h increments from 0 to 3 h post-thaw. The model achieved AUCs of 0.869 (95% CI 0.789, 0.934) and 0.807 (95% CI 0.717, 0.886) and the embryologists achieved average AUCs of 0.829 (95% CI 0.747, 0.896) and 0.850 (95% CI 0.773, 0.908) at 2 h and 3 h, respectively. Combining embryologist predictions with model predictions resulted in a significant increase in AUC of 0.051 (95% CI 0.021, 0.083) at 2 h, and an equivalent increase in AUC of 0.010 (95% CI -0.018, 0.037) at 3 h. This study suggests that a deep learning model can predict in vitro blastocyst survival after thaw in aneuploid embryos. After correlation with clinical outcomes of transferred embryos, this model may help embryologists ascertain which embryos may have failed to survive the thaw process and increase the likelihood of pregnancy by preventing the transfer of non-viable embryos.

Diminished ovarian reserve is associated with reduced euploid rates via preimplantation genetic testing for aneuploidy independently from age: evidence for concomitant reduction in oocyte quality with quantity

OBJECTIVE(S)

To determine whether women with diminished ovarian reserve (DOR) (quantitatively) had lower rates of euploid blastocysts, as a proxy for oocyte quality.

DESIGN

Retrospective cohort study.

SETTING

University reproductive health clinic.

PATIENT(S)

A total of 1,152 women aged 19-42 years underwent 1,675 IVF cycles yielding 8,073 blastocysts for biopsy from 2010 to 2019.

INTERVENTIONS(S)

Preimplantation genetic testing for aneuploidy.

MAIN OUTCOME MEASURE(S)

Euploid rates, defined as the number of euploid blastocysts divided by the number of biopsied blastocysts per cycle.

RESULT(S)

A total of 225 women (20%) had DOR as infertility diagnosis per the Bologna criteria. Age was higher among the women with DOR (39.5 y vs. 37.0 y). Euploid rates were lower among women with vs. without DOR (29.0% vs. 44.9%). In generalized linear models controlling for age, women with DOR had 24% reduced odds of a biopsied blastocyst being euploid versus women without DOR. In a secondary analysis assigning DOR status to women producing the lowest quartile of age-adjusted mature oocyte yield, this relationship remained. No differences were identified in live birth rates between women with and without DOR after euploid single-embryo transfer independently from age (n = 944 transfers; 56.8% vs. 54.8%, respectively).

CONCLUSION(S)

Blastocysts from women with DOR are less likely to be euploid than those from women without DOR after adjustment for age. Given the concomitant reduction in euploid rates with quantity of oocytes observed in this study, quantitative ovarian reserve assessments (i.e., follicular machinery) may yield insight into relative ovarian aging.

Effect of culture conditions and method of conception on mouse live birth rate

OBJECTIVE

To understand in a mouse model whether there are differences in the decidua and live birth rate after transfer of blastocysts generated by in vitro fertilization (IVF) or by superovulation with spontaneous mating into unstimulated recipients.

DESIGN

Animal experiment.

SETTING

University-affiliated tertiary hospital.

ANIMAL(S)

Mice.

INTERVENTION(S)

IVF embryos were generated and cultured in either Whitten medium (WM, suboptimal conditions) and 20% O₂ or KSOM medium with amino acids (KAA, optimal conditions) and 5% O₂. The control blastocysts from superovulated mice were flushed out of the uterus 3.5 days (E3.5) after mating (FB group). The resulting blastocysts were transferred to nonsuperovulated CF1 recipients mated to vasectomized males. To understand whether anomalies of decidua were present, the expression of genes involved in decidual development and inflammation was analyzed at E7.5 and E18.5. Similarly, immunostaining was used to evaluate whether the pathways involved in activation of mTORC1 (p-S6) and Cox2 signaling (Cox 2 staining) were altered.

MAIN OUTCOME MEASURE(S)

Live birth rate, gene expression, and immunostaining of decidua.

RESULT(S)

Implantation rates at E7.5 were similar, but in vivo embryos (FB groups) were predicted to result in live births 3.3 times higher (2.2-5.1) and 6.6 times higher (4.7-9.3) compared with optimal and suboptimal cultures, respectively. Expression of genes involved in decidual development and inflammation or localization and intensity of staining for p-S6 (mTOR pathway), or inflammation (Cox 2 pathway) were not different among the groups.

CONCLUSION(S)

The predicted live birth rate was decreased in mouse embryos generated by IVF compared with embryos generated by mating, whereas the implantation rate was not different. Suboptimal culture conditions resulted in lower birth rate. We did not find evidence of abnormalities in decidualization that could explain these findings. These data indicate that blastocysts cultured in stressful conditions are less competent, suggesting that decreasing the number of embryonic manipulations may result in higher live birth rates.

Advanced Paternal Age Affects Sperm Count and Anogenital Distance in Mouse Offspring

In Western society, couples increasingly delay parenthood until later in life. Overall, studies have focused on the reproductive performance of older parents or the impact of advanced maternal age on pregnancy outcomes, but few studies have examined how advanced paternal age (APA) affects offspring health. The aim of this study was to investigate the impact of increasing paternal age on offspring reproductive performance and long-term metabolic health in a mouse model. Here, the same adult B6D2F1/J male mice were mated at 4, 12, and 18 months of age with 6- to 10-week-old naturally cycling CF1 females to generate 3 offspring cohorts conceived at increasing paternal ages PA4, PA12, and PA18. The offspring resulting from mating the same fathers at different ages (n = 20 per age; 10 males and 10 females) were maintained up to 20 weeks of age and morphometric parameters, growth curve, and glucose tolerance were measured. We found that increasing paternal age was associated with a trend toward longer time to conception. Litter sizes were not significantly different. Reassuringly, metabolic parameters and growth curve were not different in the 3 cohorts of offspring. Most importantly, increased paternal age (PA4 vs PA18) was associated with a statistically significant decrease in sperm concentration, sperm motility, and anogenital distance in offspring. These changes raise concerns about the potential impact of APA on the reproductive fitness in males of the next generation.

Common and specific transcriptional signatures in mouse embryos and adult tissues induced by in vitro procedures

Stressful environmental exposures incurred early in development can affect postnatal metabolic health and susceptibility to non-communicable diseases in adulthood, although the molecular mechanisms by which this occurs have yet to be elucidated. Here we use a mouse model to investigate how assorted in vitro exposures restricted exclusively to the preimplantation period affect transcription both acutely in embryos and long-term in subsequent offspring adult tissues, to determine if reliable transcriptional markers of in vitro stress are present at specific developmental time points and throughout development. Each in vitro fertilization or embryo culture environment led to a specific and unique blastocyst transcriptional profile, but we identified a common 18-gene and 9-pathway signature of preimplantation embryo manipulation that was present in all in vitro embryos irrespective of culture condition or method of fertilization. This fingerprint did not persist throughout development and there was no clear transcriptional cohesion between adult IVF offspring tissues or compared to their preceding embryos, indicating a tissue-specific impact of in vitro stress on gene expression. However, the transcriptional changes present in each IVF tissue were targeted by the same upstream transcriptional regulators, which provide insight as to how acute transcriptional responses to stressful environmental exposures might be preserved throughout development to influence adult gene expression.

In vitro fertilization affects growth and glucose metabolism in a sex-specific manner in an outbred mouse model

The preimplantation period is a time of reprogramming that may be vulnerable to disruption. This question has wide clinical relevance since the number of children conceived by in vitro fertilization (IVF) is rising. To examine this question, outbred mice (CF1 × B6D2F1) conceived by IVF and cultured using Whitten medium and 20% O2 (IVFWM group, less optimal) or K simplex optimized medium with amino acids and 5% O2 (IVFKAA group, more optimal and similar to conditions used in human IVF) were studied postnatally. We found that flushed blastocysts transferred to recipient mice provided the best control group (FB group), as this accounted for the effects of superovulation, embryo transfer, and litter size. We observed that many physiological parameters were normal. Reassuringly, IVFKAA offspring did not differ significantly from FB offspring. However, male IVFWM mice (but not females) were larger during the first 19 wk of life and exhibited glucose intolerance. Male IVFWM mice also showed enlarged left heart despite normal blood pressure. Expression of candidate imprinted genes (H19, Igf2, and Slc38a4) in multiple adult tissues did not show differences among the groups; only Slc38a4 was down-regulated following IVF (in both culture conditions) in female adipose tissue. These studies demonstrate that adult metabolism is affected by the type of conditions encountered during the preimplantation stage. Further, the postnatal growth trajectory and glucose homeostasis following ex vivo manipulation may be sexual dimorphic. Future work on the long-term effects of IVF offspring should focus on glucose metabolism and the cardiovascular system.

Impaired placental nutrient transport in mice generated by in vitro fertilization

More than 4.5 million children have been conceived by in vitro fertilization (IVF). Interestingly, singleton IVF offspring born at term have an increased incidence of low birth weight. The mechanism responsible for the lower birth weight is unknown, but alterations in placental function are possible. Hence, the goal of our study was to examine placental growth and function in mice generated in vivo or in vitro. To assess placental function, blastocysts were generated by IVF or produced by natural mating (control group); both IVF and control blastocysts were transferred to pseudopregnant recipients. Placental weights did not differ at embryonic d 15.5 (E15.5) but were increased at E18.5 in the IVF group (25.4%, P < 0.001) compared with control. Proliferation was increased in IVF placentae, whereas overall placental gross morphology and apoptosis were not affected. Both fetal weights (16.4% lower at E15.5 and 8.8% lower at E18.5, P < 0.05) and fetal to placental ratios were lower (P < 0.001) in the IVF compared with the control group at both time points, whereas birth weights did not differ. At E18.5, the mRNA for selected glucose, system A amino acid transporters, and imprinted genes were down-regulated in IVF placentae. GLUT3 protein level was decreased in the IVF group (P < 0.05). Importantly, intrajugular injections of (14)C-methyl-D-glucose or (14)C-MeAIB tracers (n = 6 litters per group) showed that placental transport of glucose and amino acids were 24.8% (not significant) and 58.1% (P < 0.05) lower in the IVF group. Fetal accumulation of glucose was not different, but amino acid accumulation was significantly (36 %) lower in IVF fetuses (P < 0.05). We conclude that IVF alters both fetal and placental growth and, importantly, decreases placental transport efficiency in mice conceived by IVF.