Thoughts and reflections from ASRM 2020

This year, online was the place to be when participating in the 76^th annual meeting of ASRM. In this blog post, thoughts and reflections from the Vitrolife team have been collected on some of the scientific content presented at the meeting.

Vitrolife’s scientific symposium "Reducing stress for embryos and embryologists: Time-lapse, AI and antioxidants working together" 

Prof. David Gardner gave a presentation on how to reduce stress on embryos and embryologists and his presentation demonstrated the improvements gained by adding antioxidants to gamete handling and culture media. Also discussed, was how purpose designed time-lapse incubators for human embryo culture permit almost constant visualisation of the embryo, which allows minimal disturbance and aids workflow efficiency gains. Prof. Gardner also talked about the use of time-lapse with morphokinetics and now artificial intelligence to provide more objective tools to support embryo selection.

Interest in using artificial intelligence in IVF is increasing exponentially

Applications for artificial intelligence (AI) range from predicting patient outcomes, predicting oocyte maturation, improved embryo evaluation of static and time-lapse images, and combining morphokinetics together with metabolomic analysis.

Berntsen et al. (O-181) presented the concept of AI “explainability”. An AI (iDAScore®) was trained on full embryo development time-lapse sequences with known transfer outcome. The AI automatically generates implantation potential score outputs for the entire embryo cohort. To look at explainability, scores for more than 17.000 embryos were compared to manually labelled grading of embryo morphokinetic and morphological scores. Reassuringly the iDAScore^® output matched well with currently accepted embryo evaluation assessments, which suggests that this AI for embryo selection is not a complete “black box” technology.

A new AI algorithm was presented by De los Ángeles Valeras et al. (P-84), combining morphokinetics and oxidative status. Spent culture media of embryos cultured in the EmbryoScope were measured using a Thermochemiluminescence (TCL) Analyzer. The AI showed an impressive ability to predict live birth and showed that a combination of morphokinetics and metabolomic measurements may be a useful tool to select embryos which have a high likelihood of resulting in live birth.  

In the interactive session “Benefits and Limitations of Time Lapse Imaging and AI in ART”, Eva Schenkman gave an excellent overview of the history and future of time-lapse technology. Benefits which were related to both workflow improvements and consistency as well as potential for better embryo selection were covered. The clearly improved workflow, quality control and knowledge building benefits of time-lapse technology were echoed in the interactive session “What We All Do, We Still Do Differently: The Continuing Debated Questions in the IVF Lab”. Here Dean Morbeck and Gerry Celia, discussed how procedures are performed differently between clinics.

In other talks, Bori et al. (P-79), examined the potential of two new morphokinetic markers of embryo development to predict embryo implantation. Novel morphodynamic parameters included distance and speed of pronuclear migration, blastocyst expanded diameter, inner cell mass area and trophectoderm cell cycle length. Performance of one of the artificial neural networks combining, known morphokinetic markers with new markers showed an AUC of 0.77.

Aycock et al. (P-98) examined the possibility of performing laser assisted hatching (AH) at the fertilisation check stage rather than at a later development stage. Embryos with abnormal fertilisation were donated for research and underwent laser AH at fertilisation check. Embryos were cultured undisturbed for 5 days in the EmbryoScope and the hatching rate was assessed. The study demonstrates that laser AH performed on day 1 at the fertilisation check can produce hatching blastocysts at the expected rate. This potentially allows culture of embryos in time-lapse incubators with completely uninterrupted continuous culture for 5+ days, until the desired hatching blastocyst stage has been achieved.

Significant number of studies presented on preimplantation genetic testing

Preimplantation genetic testing (PGT) featured prominently in general sessions, oral presentations and posters at this year’s ASRM.

McCulloh et al. (P-754) presented a compelling poster assessing the success of PGT using the largest data set to date. Evaluating data from over 303,000 IVF cycles (2014-2017) from the U.S. SART registry they compared outcomes using PGT versus no PGT. Live birth per cycle was significantly greater in all age groups with PGT-A; despite the transfer of fewer embryos compared to the no PGT group. Miscarriage, multiple pregnancy and preterm delivery rates were all significantly greater than when PGT was not performed.

The benefits of PGT-A were further supported by Tiegs et al. (O-73), who measured the predictive value of an aneuploid embryo identified by PGT-A. The group performed a prospective, blinded, multi-centre non-selection study in which all 402 participants underwent trophectoderm (TE) biopsy followed by vitrification. Patients underwent single embryo transfer of the best embryo selected by morphology. PGT-A was performed only after the clinical outcome was known. Outcomes were compared to PGT-A results and the predictive value of an aneuploid result for failure to deliver was calculated as 100%, demonstrating that the PGT-A assay does not result in the discard of embryos with significant reproductive potential.

Tiegs and colleagues also addressed the concerns that TE biopsy may have an adverse effect on embryo reproductive potential – an important question continually raised in the field. They compared sustained implantation rates (SIR) of the study group to a control group in which biopsy/PGT-A was not performed. When compared to the control group, the mean SIR of all patients in the study was not different, demonstrating no detectable detrimental effects of the TE biopsy. This study offers reassurance for those concerned about the impact of embryo biopsy.

With oocyte aneuploidy significantly associated with increasing maternal age, Katz-Jaffe et al. (P-792) presented a study where they evaluated the efficacy of single euploid embryo transfer. The cohort of 3773 single euploid transfers revealed that age-related increase in pregnancy loss associated with aneuploidy had been largely overcome with PGT-A and the selection of euploid embryos for transfer. They also reported that a small but significant decrease in live birth outcomes remains with women of advanced maternal age undergoing infertility treatment and represents the additional embryonic and maternal factors associated with reproductive success.

In the prize paper session Viotti et al. (O-3) presented their study on the features of chromosomal mosaicism detected in PGT-A and the impact on implantation rates and pregnancy. The group compiled data from the transfer of mosaic embryos (n = 1000) from multiple centres, with clinical outcomes and compared this to a euploid control group (n=5561). The results from this large dataset of transferred mosaic embryos confirmed that mosaic embryos have lower rates of implantation and higher likelihood of spontaneous abortion compared to euploid embryos. The data also revealed the level and type of mosaicism significantly effects positive outcomes and identified a ranking system for mosaic embryos in the clinic.

We have learnt a lot about PGT-A this year and the significant number of studies highlighted at ASRM and other meetings increase the support for the use of PGT-A.

Other embryo selection tools

Using time-lapse images of embryos or PGT-A are two ways to select embryos for transfer. Looking at oocyte/embryo metabolism could be an additional tool. One way to monitor metabolism under investigation is Fluorescent Life-Time Imaging Microscopy (FLIM). Using this technology, it is possible to get an indication of the metabolic state by measuring the autofluorescence of NADH and FAD. Venturas et al. (O-81) correlated metabolic activity of cumulus cells to oocyte maturation and fertilisation potential. The same author (P-44) also correlated metabolic activity of cumulus cells with clinical outcomes of resulting embryos. Using the same technology Shah et al. (O-185) demonstrated metabolic differences between euploid and aneuploid blastocysts. Metabolic differences were also measured using mass spectroscopy on spent culture media by Cabello et al. (O-13). It will be interesting to see how these technologies are further developed and how they perform in prospective studies.

Culture condition considerations

The O₂ concentration in the female reproductive tract drops from 5% in the fallopian tube to 2% in the uterus. For human IVF this would suggest that changing culture conditions on day 3 from 5% O₂ to 2% O₂ could be beneficial. Danis et al. (P-13) investigated how mouse embryos develop from the one-cell to the blastocyst stage after being exposed to either <1%, 1.5%, 2%, 3% or 5% O₂. Embryos cultured at 1.5, 2 or 3% O₂ showed the highest blastocyst formation. Further, when 5% O₂ was taken as the standard reference, the ODDS ratio for cleavage stage embryos to develop to the blastocyst stage was significantly higher at 2% and 3% O₂ compared to 1.5% and 5% O₂.

How does this relate to human IVF then? Brouillet et al (O-108) presented data from a retrospective in-patient control study. In a first IVF treatment cycle embryos from 120 couples were continuously cultured at 5% O₂ from day 0 to the blastocyst stage. In the following cycle, embryos from the same couples were cultured at 5% O₂ until day 3, followed by culture at 2% O₂ to the blastocyst stage. Culture at 2 % O₂ resulted in a significantly higher blastocyst conversion and blastocyst usability rate. Cumulative implantation, clinical pregnancy and live birth rates were higher in cycles with 2% O₂. The study was not designed for clinical outcome evaluation, but the data from both presentations suggests that a randomised controlled trial should be performed to investigate if differential O₂ culture can increase blastocyst rates and clinical outcomes.

Like O₂, pH shows a decrease when comparing the fallopian tube to the uterus. The effect of manipulating the pH for fertilisation and pronuclear (PN) formation was investigated in a sibling human oocyte study by Pham et al. (P-111). After ICSI oocytes were cultured to the blastocyst stage either at a continuous pH of 7.27 or at pH 7.40 until PN check, and then 7.27. Fertilisation and blastocyst formation rates were not different. Shifting the pH gave significantly more blastocysts for biopsy, but the euploidy rate per biopsied blastocyst was significantly higher in the continuous pH group. The authors concluded that the overall euploidy rate was not different. This study did not report clinical outcome data. As euploid embryos still show differences in their implantation potential, a follow-up of clinical outcomes from this study would be very interesting.

Bossi et al. (P-31) explored how osmolality could be optimised during embryo culture. No difference was found for osmolality in two types of culture media using mineral oil or paraffin oil (OVOIL) in a humid culture environment over a seven-day culture period. However, in the dry incubator the paraffin oil provided significantly better protection against evaporation for the single step medium compared to mineral oil. This study clearly shows that the type of the oil and the starting osmolality of the culture media are important to maintain good culture conditions.

A variety of insights on oocyte vitrification

Several papers on different aspects of vitrification were presented. Today, oocyte vitrification is normally not considered an experimental method and excellent results have been obtained. While it has been suggested that clinical outcome after oocyte vitrification is similar to fresh oocytes for oocyte donation, effects of oocyte vitrification on oocytes and subsequent embryo development is less known. Berteli et al. (P-186 and P-952), using a mouse model, showed that equilibration of oocytes in solutions for vitrification results in changes of oocyte membrane lipids, and that exposure time seems to have a substantial impact on this effect. Yang et al. (O-095), presented a study where they show that oocyte vitrification can influence embryo development. In a propensity score matching study they compared fresh and vitrified patient oocytes and found that fertilisation, embryo development as well as clinical outcome were better for fresh compared to vitrified oocytes. Analysis of oocytes donated for research showed that spindle-related aspects may be an explanation for the differences observed.

Porcu et al. (P-189) demonstrated that oocyte vitrification using a closed device is equally efficient as an open device, contributing to the increasing evidence that closed devices perform as well as open carrier devices. Labrado et al (P-131) showed that RapidWarm Omni can be successfully used to warm oocytes vitrified with media from another supplier. This is important information in view of changing supplier for vitrification and warming media. Finally, Martin et al. (P-133) presented a device for automated oocyte vitrification using microfluidics. Results indicate that both vitrification/warming of bovine oocytes was equally effective as compared to a manual procedure.

More information

If you would like to read more about some of the research presented at ASRM, you can download 'Vitrolife abstracts book ASRM virtual 2020' below.