'Beyond Regulation: Shining a Light on Fertility Treatment', the third session of Progress Educational Trust annual conference 'Reality Check: A Realistic Look at Assisted Reproduction', was chaired by Alastair McLellan, the editor of the Health Service Journal. He opened by outlining how the session would provide insight into both current and future service provision in fertility healthcare.
Dr Cristina Hickman, Chief Scientific Officer at virtual fertility clinic Apricity, opened the session with her talk 'Medicine and Machines: Will Artificial Intelligence (AI) Take Over Assisted Reproductive Technology (ART)?' She explored the various ways in which AI could be used to reduce risks and improve safety in the IVF laboratory.
One example was the use of radio-frequency identification codes to identify embryos and gametes so that they could be placed in and removed from storage in an automated fashion. Other exciting AI applications discussed included a one-stop 'mini-lab' that could replace the need for large IVF laboratories, conventional incubators and costly specialist equipment.
Dr Hickman went on to explain how clinics could begin to use AI to capitalise on their vast data sets by using algorithms to improve embryo selection. She explained that during time-lapse culture, many images and data points are created as the embryo develops. By looking at this data it may be possible to create an algorithmic, standardised classification of embryo scoring and predictions of an embryo's implantation potential.
Dr Hickman said 'we are drowning in data, but thirsty for knowledge' to describe what she sees as the current situation in which AI, she believes, offers a solution. The message of her talk was that the goal of AI should be 'more babies, less money, safer'.
Machine learning can accept the diversity of populations to overcome the conflicting evidence often reported by small randomised control studies. The complete data sets can be used to augment current embryo selection, improving success rates and the provision of precision medicine.
Dr Hickman concluded her talk by emphasising that AI can improve all aspects of IVF treatment before and after embryo selection.
Furthermore, she called for international collaboration in the use of big data to help create effective evidence-based, data-driven algorithms. This global knowledge-sharing would operate within a secure architecture using blockchain technology, so that only the algorithms travel, and patients' data remains protected.
The second and final speaker, Dr Roger Sturmey, Senior Lecturer in Reproductive Medicine and principal investigator at Hull York Medical School, gave an enlightening talk titled, 'The Secret Recipes of Embryo Culture Media: Who Needs to Know?'
He began by highlighting that currently embryos are cultured in media containing components we think the embryo needs rather than one devised in accordance with studies of the natural early pre-implantation embryo environment in the fallopian tube.
Dr Sturmey shared graphs from various sources that indicate that one commercial culture media is better for embryo development than others. A Cochrane review, however, had concluded that there was 'not sufficient evidence to support or refute the use of any specific culture media'.
Dr Sturmey explained that this conclusion was drawn because different culture media had been used for the studies, so that data could not be pooled and, even more importantly, the studies did not follow up on the health of offspring.
One move towards rectifying this issue has been the regulatory CE marking of media. This has served to improve quality through validation of the composition and evidence base of the components. He highlighted that manufacturers tend to divulge only the percentage composition of the media and do not provide detailed information on the concentration of the media components.
His work has shown that the ratio of lactate/pyruvate can affect how an embryo uses nutrients metabolically, proving the importance of knowing component concentrations. Furthermore, analysis of media exposed other issues, such as how components sometimes change during storage and, more worryingly, the presence of proteins not declared by the media company.
Dr Sturmey said that, although an embryo has to choose its nutrients from those that are available in the culture dish, studies have shown that there are differences in gene expression and changes in metabolism for embryos cultured in different media.
Alarmingly, some changes have also been shown to have an effect at the level of the nucleus and the genetic code of the embryo. Dr Sturmey reassured the delegates that after 40 years of IVF and eight million children born worldwide, such changes, if they do exist, are likely to be subtle.
He concluded by stating that further studies in animals are necessary and urged companies to be transparent about the concentration of media components. Universal transparency would enable clinicians, researchers and regulators to investigate metabolic modifications in the embryo due to media composition differences and the subsequent health of babies and adults.
McLellan closed the session with a lively Q&A. Recurring questions included: What is the cost of development for AI?; Is AI just the 'next' thing to try?; How would the effectiveness of AI algorithms be evaluated without clinical studies?
Dr Hickman elaborated on the need for global collaboration to improve digital infrastructure so that big data can be used to improve outcomes with personalised care. She also highlighted that increased understanding of why a patient may not get pregnant is also valuable – we should not assume the patient only wants a baby.
When asked, 'Are all women the same incubators?', Dr Sturmey responded that every woman, embryo and incubator is different. However, the development of culture media optimised for individual patients is unlikely.
In summary, the participants in the discussion called for a better understanding of the oviduct environment, stratified by age and ethnicity, together with developing responsible relationships between commercial and clinical needs.
The Progress Educational Trust (PET) would like to thank the sponsors of its conference - the Anne McLaren Memorial Trust Fund, Edwards and Steptoe Research Trust Fund, CooperSurgical, the European Sperm Bank, Ferring Pharmaceuticals, the London Women's Clinic, NGA Law and the Institute of Medical Ethics.