Session 20 - Future of Embryology: Q&A

Questions Marcos Meseguer:

Questions for Maurizio Poli:

In terms of proteins, the analysis of the blastocoel has the advantage of avoiding several contaminants present in the culture media, thus facilitating the identification of biomarkers expressed at much lower concentrations compared with proteins supplemented in the media (e.g., albumin).

It has been shown (and it is under our eyes every day) that embryos do have metabolic plasticity. This doesn’t mean that embryos thrive in it. Indeed, there are several arguments supporting the theory that if an embryo has to spend much of its metabolic resources to counterbalance the environment, then it has very little left for proper development. By providing an environment tailored for the embryo’s characteristics, it can focus on core developmental processes with minimal wastage of energy supplies.

In terms of genetic plasticity of the embryo, there is insufficient evidence of this.

        General Questions


I believe either of these technologies will be part of the future of IVF. However, they may develop into clinical applications at different speeds.

It mainly depends on what degree of AI and automation we are talking about. If we consider automated embryo grading, I believe this is just a couple of years away. However, if we are talking about AI based both on embryo images and on integrated analysis of data from multiple sources such as genetics, omics, patients medical details and laboratory parameters, then this will take considerably longer. Despite some of the data being already available and easily accessible (i.e., patients medical details, lab environment, embryology-related features such as procedures timings, etc.), no transversal digital infrastructure to record and collect all these parameters yet exists, thus no option of cross-evaluation, data mining and development of predictive algorithms is available. Unless a meaningful effort in this direction is taken in the next few years, results will keep coming scattered and inconsistent.

For this reason, a safer bid would be on genetics as it already has a solid technological background and infrastructure around it that, thanks to increasing adoption of genome sequencing in other medical fields, is growing independently from reproductive applications. Indeed, preconceptional genome sequencing may soon become the most employed genetic application in IVF.

For what it concerns full automation of the IVF lab, this objective requires several technological achievements, followed by extensive safety and technical validations. Therefore, I believe this complete reshaping of clinical embryology as we know it will take longer to reach clinical deployment compared to the other proposed here. Nonetheless, once available, lab automation will spread very quickly. At that point, theoretically, full process consistency in the IVF lab will be achieved and AI will have a strong foundation to build on.