IVF, from the past to the future: the inheritance of the Capri Workshop Group.

Fertility clinics have subsequently flourished. In the past, they used to offer hope exclusively to heterosexual couples who experienced problems conceiving. In more recent years, single women, same-sex couples and women following a professional vocation (or not having found a suitable partner yet) have also become part of the captive audience. Somehow, infertility treatments have overcome the boundaries of mere health care to expand into addressing societal issues of reproduction in the absence of infertility. The extraordinary achievements of this discipline have encouraged its global spread and attracted financial interests. Oocyte freezing in particular has turned fertility health care into a booming industry (The Economist, 8 August 2019: ‘Investors are pouring money into companies that promise to help people conceive’). The desire for offspring is universal. In some countries, more than 1 in 20 children are born following IVF, and fees of €10 000 and more per cycle are no exception. Hence the financial world is understandably thrilled. There is a long list of private investments and an increase in private equity firms involved in the global IVF sector in the last couple of years. The global fertility market, valued at €15 billion in 2016, is predicted to reach €28 billion per annum in 2023 ( Yunis and North, 2019 ). However, these exciting numbers should not hide some important and yet unsolved limitations of IVF.

In 2015, the United Nations (UN) launched a global campaign to ensure future sustainability for all. This initiative, aimed at addressing in a comprehensive manner the current global challenges, including those related to poverty, inequality, climate, environmental degradation, prosperity and peace and justice. It is planned to run for 15 years duration and is subdivided in 17 goals (sustainable development goals). Of utmost relevance is the concept that all these goals are interconnected and that all human beings and stakeholders should be primarily committed to these aims. Reproductive health in general, and infertility treatment in particular, do not represent an exception and have to be viewed as an integral part of this ambitious plan. Recently, a commendable global initiative (priority setting partnership for infertility) was undertaken, aimed at identifying research uncertainties in four main areas of reproductive medicine, i.e. male infertility, female and unexplained infertility, medically assisted reproduction and ethics, access and organization of care ( Duffy, 2019 ). In line with the modern UN commitment for sustainability, this effort was aimed at directing more efficiently the resources and energies to outcomes that matter to treating patients. Over 700 healthcare professionals and patients were brought together to disentangle the top 10 research priorities for the four areas using robust consensus development methods. Table III depicts those priorities as related to medically assisted reproduction. In addition, 11 out of the 30 priorities of the other three infertility-related areas are somehow related to IVF. Most of them are within the domains of ethics, access and organization of care, and thus in line with the UN commitment for sustainability. Top 10 research uncertainties in medically assisted reproduction. LBR, live birth rate. On the other hand, research is based on freedom. In spite of the above-mentioned commitment for rationality and sustainability, basic research in particular should go on freely to explore new avenues. There is currently particular enthusiasm for the possibility of gene editing of human gametes and embryos ( Church, 2017 ), reactivation of folliculogenesis through ovarian cortex fragmentation ( Suzuki et al. , 2015 ), mitochondria replacement ( Kang et al. , 2016 ), folliculogenesis in vitro ( Telfer, 2019 ), reconstitution of oocytes from stem cells in vitro ( Hikabe et al. , 2016 ; Morohaku et al. , 2016 ) and ovarian rejuvenation ( Labarta et al. , 2019 ). Of relevance here is that science should do everything possible to avoid false hopes and undue exposure of patients to risks. Researchers are called to resist financial pressures that can cause the premature and hazardous use of new technologies. Robust experimental evidence on effectiveness and safety are mandatory prior to fostering clinical application of any new technology ( Wilkinson et al. , 2019 ).

Within the debate on the pros and cons of ‘add-ons’ in IVF, preimplantation genetic testing for aneuploidies (PGT-A) merits an independent and deepening discussion. In fact, genetic analysis of human preimplantation embryos before transfer to the uterus was reported in the late 1980s, thus in the early phase of IVF ( Handyside et al. , 1989 ). This soon was suggested to be used, not to prevent genetic risks but to increase the success rate of IVF, a procedure then called preimplantation genetic screening, but now referred to as PGT-A. The use of PGT-A, after the first reported pregnancy in 1995 and other enthusiastic initial reports ( Verlinsky et al. , 1995 ), found its way into routine practice despite the lack of robust evidence for its ability to increase LBR or other outcomes. Notwithstanding the detrimental impact on the chances of pregnancy reported in subsequent pivotal studies ( Staessen et al. , 2004 ; Twisk et al. , 2006 ; Mastenbroek et al. , 2007 ), PGT-A continued to be used. These studies, however, led to a reconsideration of the shortcomings of the first generation of PGT-A methods, and to adjustments being made ( Forman et al. , 2013 ; Rubio et al. , 2013 ; Scott et al. , 2013 ; Yang et al. , 2015 ; Griesinger, 2016 ; Rubio et al. , 2017 ). Despite substantial efforts to improve PGT-A, the evidence remains disappointing. To date, no single study provides high-level evidence of improved effectiveness of IVF with PGT-A. It has been suggested that potential benefits of the procedure are being limited to older women and to secondary outcomes (less transfers and possible shorter time to pregnancy) but high-level evidence is lacking here as well ( Munné et al. , 2019 ; Paulson, 2020 ; Pagliardini et al. , 2020 ). In addition, the rationale of PGT-A is also increasingly being doubted. To start with, the standard method for selecting embryos in IVF, namely morphological evaluation, seems quite capable of ranking embryos based on their implantation potential ( van Loendersloot et al. , 2014 ). More importantly, doubt has again been cast on the accuracy of the technical analysis methods used ( Popovic et al. , 2018 ; Lawrenz et al. , 2019 ). There is evidence that mosaicism, or the lack of understanding thereof, undermines the efficacy of PGT-A, since embryos labeled as mosaic and even aneuploid in PGT-A were demonstrated to implant and result in healthy live births, although perhaps with less efficiency as genuinely euploid embryos ( Scott et al. , 2012 ; Greco et al. , 2015 ; Bolton et al. , 2016 ; Patrizio et al. , 2019 ). Overall, based on the recent ‘Single Embryo TrAnsfeR of Euploid Embryo’ (STAR) RCT, the rate of live births lost as a consequence of misdiagnosis or blastocyst injury associated with PGT-A could be up to 30–40% ( Paulson, 2020 ; Pagliardini et al. , 2020 ).

After the initial pioneering period, safety in IVF has received more and more attention. To date, researchers are called to focus more than ever on the health of newborns, both in the short and in the long term. Prevention of multiple pregnancy with widespread uptake of elective single embryo transfer has been a cornerstone of the second and wiser phase of IVF development ( ESHRE Capri Workshop Group, 2000 ; McLernon et al. , 2010 ). But, still, huge efforts are required. Noteworthy, Louise Brown is only 42 years old now and the millions of IVF babies throughout the world are younger. Evidence on the risk of chronic conditions developing in the second half of human life is lacking. In this regard, the embryo and fetus can respond to an environmental challenge and develop into different phenotypes through an altered epigenetic regulation of genes, a situation that can increase the risk of chronic conditions such as cardiovascular diseases later in life ( Bateson et al. , 2014 ). In the last two decades, it has become more and more clear that insults during prenatal life are an important factor for the development of some diseases in adulthood. In a study using data of a large number of individuals born between 1930 and 1938 in the Ukraine of which 43 150 developed diabetes, Lumey et al. (2015) found a positive correlation between famine severity during prenatal development and the odds of type 2 diabetes in later life. Similarly, adults exposed during their intrauterine life to the Dutch famine in various phases of gestation (i.e. infants born between 7 January and 8 December 1945) had an over 50% excess rate of ischemic heart disease, more diabetes, a more atherogenic lipid profile, altered stress response and food preference ( Roseboom et al. , 2006 ), accelerated brain aging ( de Rooij et al , 2010 ) and shorter life span ( van Abeelen et al , 2012 ) than the comparison group of infants born before 7 January 1945 and conceived after 8 December 1945. Other stresses that affect the epigenetic regulation of genes, such as exposure to toxic agents and IVF, were shown in animal models to result in adult health consequences ( Feuer and Rinaudo, 2012 ). Factors in IVF that could theoretically have an impact on adult life are controlled ovarian hyperstimulation, cryopreservation, and in vitro culture of oocytes and embryos: concerning the latter, it has been shown that the type of culture medium does affect birthweight of the newborns ( Dumoulin et al. , 2010 ). The finding was confirmed by a multicenter, double-blind RCT comparing two culture media ( Kleijkers et al. , 2016 ). The effect becomes manifest as early as the second trimester of pregnancy and persists during at least the first 9 years of life ( Zandstra et al. , 2018 ).

The ESHRE Capri workshop group is a long-lasting initiative, started in 1986, that annually brought together a varying small group of researchers from all over the world to discuss topics in the field of human reproduction. The Capri workshops were permanently discontinued in 2019. For the last workshop, held in October 2019, it was decided to discuss the achievements and challenges of IVF, the most outstanding accomplishment of modern reproductive medicine that evolved in parallel with the Capri workshop meetings. This article does not aim at a complete review of all IVF-related issues, it is a brief overview of selected arguments deemed to be particularly interesting for the scientific community to take forward in the future of reproductive medicine. After the birth of Louise Brown on 25 July 1978, Edwards and Steptoe, the two driving forces of her IVF conception, moved to Bourn Hall where, 2 years later, they started a new infertility clinic. Meanwhile, other groups also had initiated IVF programs. By 1982 roughly 300 IVF pregnancies had been reported and discussed by the group of interested pioneers: from one clinic in the UK, two in the USA and four in Australia. This same year, in California, USA, the first international meeting on IVF was organized ( Crosignani and Rubin, 1983 ). The conference attracted worldwide interest, and the key question of ethics was introduced, using the parable of Eve’s umbilicus, by R. V. Short, Professor of physiology at Monash University (Melbourne, Australia): ‘If Eve was really created from Adam’s rib she could not possibly have had an umbilicus, but all past painters chose to ignore this point at the risk of being branded as heretic’.

With the introduction of ICSI in the 1990s, ART provided broader coverage than had been possible with traditional IVF ( Rubino et al. , 2016 ). Higher fertilization and pregnancy rates were immediately realized as technology became more and more refined for a wide range of male infertility conditions. The profound impact of ICSI on the ART field cannot be understated despite persistent concerns being raised regarding possible long-term effects of both genetic and epigenetic determinants on offspring health ( Rubino et al. , 2016 ). In the course of refining ICSI as a core ART, methods evolved for sperm selection in order to obtain the most motile fraction of sperm after extensive washing and swim up processing. Only recently, more data has emerged to suggest that sperm carry important factors gained intrinsically during spermatogenesis or extrinsically during storage and ejaculation in the male reproductive tract ( Rubino et al. , 2016 ). In fact, the role of spermatozoa within ART has been neglected for years, but the time may have come to reorientate our thoughts and our scientific efforts. For too long, the systematic and blinded use of ICSI has shifted our attention away from fundamental biological processes that can ultimately reveal new means to improve IVF success. In our opinion, this new awareness on the fundamental role of the spermatozoa may open new fruitful areas of research and, therefore, deserves to be emphasized here. The fact that sperm carry more than a genome and centrosome has added more complexity to this process, especially since sperm factors other than DNA in animal studies indicate their role in the later stages of embryonic development. Among these factors are various species of RNA molecules that are acquired by sperm either in the testis or while transiting through the epididymis ( Sharma et al. , 2016 ; Burl et al. , 2018 ; Turner et al. , 2020 ). These RNAs are delivered to the oocyte at the time of fertilization, and appear to be directly involved with remodeling of the embryos maternal and paternal contributions to the newly formed zygote. Given that even normospermic ejaculates exhibit heterogeneity in sperm with respect to motility, degree of DNA fragmentation, and acrosome integrity, characterizing both the source and function of these sperm RNAs has become a future research priority in the area of sperm selection ( Turner et al. , 2020 ). Just downstream from fertilization itself is the process of syngamy during which integration of parental genomes is believed to occur. Mounting evidence now seems to suggest that the formation of a novel genetic entity, defined by merger of maternal and paternal nuclear DNA, appears to be far more dynamic and malleable than previously thought. Such core concepts derived from many years of research are being challenged on a regular basis as more sensitive and revealing technologies are brought to bear on the earliest stages of human development. A final example of our changing perception of fertilization relates to the blocks to polyspermy. High resolution live imaging studies have now confirmed a fast block to polyspermy in the human that must precede the biochemical block that had long been attributed to the exocytosis of cortical granules and their components ( Mio et al. , 2012 ). Moreover, the modulation of egg activation and concurrent influence of egg secretions on sperm motility constitute a much more robust defense against polyspermy than previously imagined. Specifically, divalent zinc ions stored in the oocyte are released in regular burst (‘sparks’) from a distinct population of cortical granules mediating both the removal of sperm receptors from the zona pellucida as well as zona hardening to effectively limit supernumerary sperm access at the time of sperm entry ( Kushnir et al. , 2017 ; Que et al. , 2017 ). These and other insights yet to be gleaned from ongoing investigations should be kept in mind as our current techniques in ART continue to be judged for efficacy and safety going forward.

The time for the Capri group workshops has drawn to a conclusion, after 33 years of activity, 56 workshops and 270 lectures. The initiative involved researchers from all over the world and from different generations. The meetings took place annually and consisted of grouping together top scientists for some days to discuss, in an open face manner, various topics in the field of reproduction. The ultimate aim was finding out shared views and providing inputs that could open new avenues of thinking and research. For the last meeting that took place in October 2019, and whose results are presented in this narrative review, a minority of the proud founders was still present. The time has come to pass the baton to the new generations and for the use of more modern and effective methodologies of consensus. As underlined in the present contribution, several new challenges are in front of us. They are complex but the scientific community has the intellectual armamentarium to take up the challenges and overcome them. We would finally like to express an immense gratitude to all the researchers who participated in the Capri Workshops over the years.

Fertility treatment is burdensome as, apart from the medical side effects, most infertile couples face emotional and financial strain ( Klitzman, 2017 ). To shape couples’ expectations, to allow self-empowerment and to prepare emotionally and financially, effective and transparent communication prior to embarking on their treatment journey is essential. Information on some fundamental aspects must be clearly provided to couples. The most suitable outcome measure is cumulative live birth rate (CLBR) per couple. This can be the CLBR per IVF treatment, including the fresh transfer and transfers after cryopreservation, or the CLBR per multiple IVF treatments. Both are calculated with a certain time window. Outcomes should not be reported per transfer ( Griesinger, 2016 ). The couple must be aware that IVF is a journey that rewards perseverance. In a prospective study based on the UK Human Fertilization and Embryo Authority (HFEA) data, 156 947 UK women received 257 398 IVF ovarian stimulation cycles between 2003 and 2010. These women were followed up until June 2012. The CLBR continued to increase up to the ninth cycle, with a cumulative prognosis-adjusted live birth rate (LBR) of 65.3% by the sixth cycle ( Smith et al. , 2015 ). This rate, however, is an over-estimation because of the high rates of couples who discontinue treatments. To note, dropouts may be viewed as the Achilles’ tendon of IVF ( Gameiro et al. , 2013 ). In addition, and most importantly, the proportion of low-prognosis couples who enter IVF programs is high: they cannot be neglected or omitted in the reports. Providing results of IVF programs should always take into consideration the main predictive factors of success such as age, ovarian response and number of previous cycles. Natural fertility as well as IVF success linearly decline with women’s increasing age after 35 years ( ESHRE Capri Workshop Group, 2017 ). A large national cohort study of all women initiating fertility treatments in 2007–2010 depicted well the effect of age: a long-term prognosis for live birth of 64% was achieved in women below 35 years of age, of 49% between 35 and 39 years, and of 16% over 40 years ( Malchau et al. , 2017 ). Apart from oocyte donation, no treatment has been proven to overcome the detrimental effects of aging on the capacity of embryos to implant. This is, to date, an insurmountable limitation of IVF. Up to recently, LBR per oocyte retrieval was believed to increase with the number of aspirated oocytes up to 15 and stabilize, or even decrease, thereafter ( Sunkara et al. , 2011 ; Briggs et al. , 2015 ; Drakopoulos et al. , 2015 ). However, a recent large multicenter study in women under 40 years of age showed that the CLBR per retrieval increased with the number of oocytes even beyond 15, reaching 70% when more than 25 oocytes were retrieved ( Polyzos et al. , 2018 ). Hence, ovarian stimulation may not have a detrimental effect on oocyte/embryo quality in good prognosis patients. A good response to hyper-stimulation is a good prognosis factor on its own. A Danish national cohort study including 30 486 women initiating ART treatment with their own oocytes showed that the number of aspirated oocytes in the first ART cycle was associated with an increasing CLBR in up to four subsequent cycles ( Malchau et al. , 2019 ). In women without a live birth in the first complete cycle, the number of aspirated oocytes predicted the outcome in the second and third cycle ( Malchau et al. , 2019 ). On the other hand, there is no evidence that increasing the number of oocytes by enhancing the dose of gonadotrophins could be of benefit ( Lensen et al. , 2018 ). The CLBR in poor prognosis patients is even more important for valuable counseling. In a Dutch multicenter cohort study based on the OPTIMIST trial, including 551 low-prognosis women aged <44 years and treated with a fixed FSH dose of 150 IU/day in the first treatment cycle ( Leijdekkers et al. , 2019 ), the CLBR of the low-prognosis women was satisfactory, being on average 56% over 18 months of IVF. In addition, the authors showed that the outcome in previous cycles can be used to estimate the prognosis during subsequent cycles. Providing an age-stratified prognosis for chances of live birth and risk of ovarian hyperstimulation syndrome, incorporating prior failed attempts and previous ovarian response, can aid couples in the decision of treatment (dis)continuation. However, even when prognostic factors are well known, a priori individual counseling remains challenging and imprecise. In general, improving our capacity to predict outcomes in IVF is fundamental in order to prevent useless exposure of women to the risks of the procedure, as well as wastage of resources, that could be employed in a more efficient manner. To note, IVF is unlikely to be cost-beneficial when the LBR per cycle is below 4–10% ( ESHRE Capri Workshop Group, 2015 ).

The indication for the first successful IVF leading to the birth of Louise Brown was bilateral tubal blockage ( Steptoe and Edwards, 1978 ). The whole process was actually indicated to replace the role of the tubes, i.e. the retrieval of the cumulus oocyte complex, the transport of the spermatozoa, the fertilization process and the final transport of the developing embryo into the endometrial cavity. Thereafter, indications have expanded widely, in most cases in the absence of evidence and rationale. While endometriosis, pelvic inflammatory disease and severe male factor are substantiated indications for IVF ( ESHRE Capri Workshop Group, 1994 ; ESHRE Capri Workshop Group, 1996 ), some other conditions are inappropriately taken for granted by most reproductive health specialists. The most controversial are unexplained infertility ( Crosignani et al. , 1993 ), repeated miscarriage, age-related infertility and low ovarian reserve ( ESHRE Capri Workshop Group, 2017 ; Annual Capri Workshop Group, 2019 ). The potential detrimental impact of the undue extension of the indications for IVF is generally under-estimated. Indication creep, i.e. the adoption of a technology shown to be effective in a clinical area to other patients group, may cause significant wastage of financial resources ( Bryan et al. , 2014 ). Technology adoption in the absence of documented benefits is an important driver of cost growth ( Bryan et al. , 2014 ). In this context, the application of IVF for unexplained infertility, repeated miscarriage, age-related infertility and low ovarian reserve is inevitably a main concern, in particular for the latter two situations where IVF is even less efficient per se . Well-designed studies are pressingly warranted to address this concern.

Thanks to ESHRE (with the European IVF-monitoring Consortium, EIM), international data have been available since 1997, and their reliability has constantly improved. The number of participating countries has increased from 18 to 39 ( De Geyter et al. , 2018 ). For IVF and ICSI, the number of treatment cycles has increased by 3.8 folds in this database, and the number of live births by 4.8. In the 34 countries with data on births, the number of IVF births in 2014 was 170 163, and the number of births following IUI was estimated at 22 767, resulting in a total of 192 930. This figure can be compared to the total number of births in the 34 countries in 2014: 7 465 421, giving a proportion of 2.6% for IVF births. If we assume that the number of IVF births in the other five countries is negligible, the proportion would be 2.3%, a minimum estimate. In fact, this rate varies widely among European countries, depending on factors such as the availability of public health coverage, whether registers are mandatory and whether registers have full coverage and proper follow-up. In Denmark, this rate reaches 6.6%, suggesting that there is yet a deficiency in recording and/or an unmet need throughout Europe ( De Geyter et al. , 2020 ). For all European countries with data on IVF births, we calculated the natural increase without IVF births and compared the two figures. In countries with a large natural increase (UK, Kazakhstan and France), the contribution of IVF is negligible, and in countries with a large natural deficit (Russia, Germany and Italy), IVF is far from significantly reducing the deficit. Its contribution is significant in a few countries with a moderate increase, such as Spain. For Europe as a whole (40 countries, Russia excluded), the natural increase in 2014 was +108 404 (a rate of +0.02%). The number of IVF births for the 33 countries in the ESHRE database (Russia excluded) was 145 129. Thus, without IVF, the natural increase of the European population would have been negative: 108 404 − 145 129 = −36 725. However, the general contribution of IVF to national fertility is limited: less than 0.1 additional children if we take into account the fact that not all IVF births are additional births, because many couples using IVF are only subfertile and may conceive on their own in the long term ( Leridon, 2017 ).