Regulating reproductive genetic services: dealing with spiral-shaped processes and techno-scientific imaginaries

Abstract

Purpose

We have been inquiring into the diffusion process of reproductive genetic services (RGS) and the viability of geneticization in human reproduction.

Method

A 2-round modified-Delphi survey was applied amongst Israeli and Spanish experts to analyze regulatory attitudes and expectations about the future applications of RGS. We argue that an explanation of RGS diffusion based on a ‘technology-push’ impulse should be complemented by a ‘demandpull’ approach, which underscores the importance of regulatory frameworks and demand-inducing policies. The diffusion of RGS is advancing in a ‘spiralshaped’ process where technology acts as a cause and effect simultaneously, modulating social acceptance and redefining the notions of health and responsibility along the way.

Results

We suggest that there is a ‘grey-zone’ of RGS regulations regarding four procedures: the use of germline genome modification (GGM) for severe monogenic disorders, preimplantation genetic testing (PGT) for detection of chromosomal abnormalities, PGT for multifactorial diseases, and PGT with whole-exome screening.

Conclusions

Although far from the geneticization of human reproduction, our findings suggest that, since techno-scientific imaginaries tend to shape regulations and thus favor the diffusion of RGS, policymakers should pay attention to those procedures by focusing on good practices and equity while providing sound information on potential risks and expected success rates. A broad and inclusive societal debate is critical for overcoming the difficulty of drawing a clear line between medical and non-medical uses of genetic selection and engineering while searching for the right balance between allowing reproductive autonomy and protecting the public interest.

Introduction

Louise Brown’s birth in 1978 was the first successful in vitro fertilization (IVF) for human reproduction. Since then, assisted reproductive technologies (ART) have improved remarkably, thanks to advances in laboratory conditions and the emergence of new techniques such as ovarian hyperstimulation, Intracytoplasmic Sperm Injection (ICSI), and gametes cryopreservation [2, 18]. Nowadays, ART provide solutions to an increasing number of infertility etiologies and yield higher success rates [19, 38].

Since the 1990s, IVF has occasionally been complemented through preimplantation genetic testing (PGT). A biopsy is removed from a blastocyst (day five embryo) and analyzed for different purposes. Firstly, PGT for monogenic disorders (PGT-M) is mainly used to identify and exclude embryos with severe early-onset monogenic diseases, but also more increasingly for late-onset disorders with partial penetrance [4, 20, 22]. Secondly, PGT for chromosome structural rearrangements (PGT-SR) is used to test for hereditary chromosome abnormalities. Thirdly, PGT for predicting the risk of polygenic disorders (PGT-P) has recently become a validated test for diabetes and various cancerous and cardiovascular diseases [86]. Finally, the most frequently applied is PGT for aneuploidies (PGT-A), which tests for sporadic chromosomal abnormalities. It is applied to select euploid embryos in order to increase implantation rates, reduce the risk of abortion, and avoid the birth of children with chromosomal abnormalities. Overall, PGT is applied to a relatively small but increasing proportion of ART cycles worldwide. According to the latest ESHRE report, 2.4% of ART cycles in Europe involved PGT [23], while in the USA, it had grown from 5% in 2015 to 22% in 2016 [19]. However, the representation of PGT-M and PGT-SRT applied to ART cycles worldwide is small in comparison to PGT-A.

Whereas PGT is used to select from among a limited supply of embryos constrained by the parents’ genetic material, the latest and most controversial RGS technique is germline genome modification (GGM), which applies “Clustered Regularly Interspaced Short Palindromic Repeats” (CRISPR) to design human embryos. While it constitutes a breakthrough technology for intervening in the human genome [87], GGM still remains illegal or very tightly regulated in most countries [11, 58].

Potential uses of reproductive genetic services (RGS) have long been a matter of concern for the bioethics literature, typically addressing the attitudes towards embryo status including disability and human diversity; the advantages and risks of enhancement; and modern eugenics and inequalities. The possibility that, at some point, the qualities of genetically selected or engineered people might surpass those naturally conceived is often raised, which would motivate the use of RGS among the broad public [34, 46]. Lippman defined “geneticization” as “an ongoing process by which differences between individuals are reduced to their DNA codes, with most disorders, behaviors and physiological variations defined, at least in part, as genetic in origin” ([53]: p. 19). This concept has been framed as the interaction between medicine, genetics, society, and culture [85]. Along these lines, we refer to the geneticization of reproduction as the process of selecting or designing genetic traits of embryos in vitro which in turn may transform ART from a technical solution for infertility into a mainstream reproduction method.

In this paper, we have inquired into the diffusion process of RGS and the viability of a geneticization of reproduction by analyzing current regulatory attitudes and expectations about future applications of RGS, among two groups of experts from Israel and Spain through a 2-round modified Delphi survey. This analysis allows us to identify areas where regulation is most needed and anticipate the debate about how regulations should be set and should contribute to informing public policy in this area.

ART industries in Israel and Spain are among the most active in the world. In 2015, Spain, the fourth largest ART provider globally and first in Europe, was practicing a third of the continent’s PGT procedures [23]. Meanwhile, Israel has the highest number of IVF cycles worldwide in relative terms [37]. Its public healthcare system funds almost all fertility treatments, including IVF cycles for women up to the age of 44, as well as most PGT-M and PGT-SR procedures (while in Spain, only about 20–25% of cycles are publicly covered, with an age limit of 40 years) [3, 79]. Both countries are characterized by relatively permissive RGS legislations, allowing to perform PGT for a wide range of severe conditions [3], which places them at the forefront of RGS worldwide in terms of scale and scope. Nevertheless, there are significant institutional and cultural differences between both countries, with religion, culture, and history greatly influencing today’s political and social norms, in turn affecting their perceptions of fertility issues and their ethical attitudes towards ART. Thus, Israel and Spain constitute a fertile ground for a comparative study concerning the potential geneticization of reproduction and the implications for policy.

Following this introduction, in Second section, we analyze the geneticization process in human reproduction in light of the diffusion of innovation theory [72], and identify its main ethical implications. Third section describes the method used to select the experts and retrieve information from them, and Fourth section summarizes the main results. Based on our findings, in Fifth section, we have developed a discussion about RGS regulations and ethics, their boundaries, and the conditions under which those may shift. Finally, Sixth section provides some conclusions.

Geneticization of reproduction: determinants and ethical implications

Geneticization of reproduction would involve a shift in ART application from primarily a solution to infertility towards genetic selection or engineering of human embryos. It would require extending the focus of genetic testing, from diagnosing severe early-onset monogenic disorders with high levels of penetrance (such as cystic fibrosis and sickle cell anemia), towards addressing less severe, late-onset polygenic and multifactorial disorders with partial penetrance (such as most metabolic, cardiovascular, cancerous, and neurological diseases) [12, 45].

How could such a scenario become a reality? RGS involves several technological innovations that have spread in a multi-cycle, two-way communication process between different agents in society, in an interaction between supply and demand [21, 61, 72]. Some authors have claimed that the diffusion of technology is led by a “technology-push” impulse [61], meaning that “a moment of breakthrough” [16], driven by scientific developments, makes new procedures possible. Hence, once the supply is created, demand follows. However, it must be considered that both technology and the social environment influence one another in a reciprocally extended process, shaped by the evolution of technological momentum and social values over time [16, 62].

The analysis of ART trajectories cannot focus exclusively on scientific developments but must take into account a “demand-pull” perspective, which underscores the importance of regulatory frameworks and demand-inducing policies [61]. Indeed, the extent to which the regulator allows and even funds different applications of RGS shapes social consent to these technologies and results in the establishment of socio-technical imaginaries, i.e., intersubjective beliefs regarding appropriate clinical applications, constraints for intervention in human reproduction, social and parental responsibilities, and desirable futures.

Therefore, we suggest that at least three conditions must be met for a broad diffusion of RGS. Firstly, visible technological developments should occur by making ART safer, more efficient, and more comfortable to bear. Secondly, in order to persuade a larger part of the public to substitute natural reproduction, RGS must introduce real or perceived medical or non-medical benefits and deliver enhanced children whose health is simpler and cheaper to maintain [34, 46]. Finally, regulations must be set in alignment with these developments and allow a broader range of RGS. We will focus on this last condition.

The role of regulations in the diffusion of RGS: the cases of Israel and Spain

Both Israel and Spain hold very pro-ART regulatory approaches [3], albeit, with stricter attitudes towards RGS. While regulations towards PGT-A are more permissive and guidelines for its application are usually set by clinics and physicians, PGT-M, PGT-SR, and PGT-P are restricted by the characteristics of genetic conditions under diagnosis, their severity, non-curability, onset, and level of penetrance.

Israel’s historical, cultural, and religious background results in strong pronatalism leading to the highest fertility rates in the OECD as well as yielding a favorable approach towards PGT and selective reproduction in general [70, 90, 91]. Therefore, Israel provides unprecedented public support for fertility treatment, which also covers PGT-M and PGT-SR cycles. RGS regulation is to some extent decentralized and rests upon expert committees in hospitals [80], where PGT is performed on a wide range of conditions, including some severe hereditary cancers, both late-onset and of partial penetrance and, even in some rare cases, PGT for sex selection [4, 91].

On the other hand, Spaniards are among the oldest parents globally and have the third lowest fertility rate among OECD countries [63]. The last decade saw a sharp increase in the demand for ART, which led to the emergence of the largest European ART industry involving some of the world’s leading ART corporations. The Spanish law on ART is comprehensive and quite permissive, with the National Committee of Human Assisted Reproduction (CNRHA) in charge of many regulatory aspects, including the approval of PG-M and PGT-SR on a case by case basis. Overall, RGS is practiced in Spain for a range of conditions similar to Israel, except for some late-onset diseases and sex selection [3, 66, 92].

Given their approaches, Israel and Spain belong to a small group of countries with a unusually high share of IVF births (5.1% in Israel in 2018 and 9.4% in Spain in 2018, in contrast with around 2% in the USA and Europe) [23, 37, 79]. Nevertheless, this rate could increase, given that 10–15% of all couples across the entire population may suffer from infertility [1, 7]. These figures might be growing due to environmental hazards and unhealthy lifestyles [40, 83], as well as the rising average age of parenthood [2, 65]. Additionally, a laissez-faire approach to the use of ART for social reasons is gaining public support, increasing the demand by women beyond the age of fertility (by egg donation or eggs cryopreservation), as well as by single women and same-sex couples ([13, 47, 79, 41]). That said, the shares of RGS cycles in these two countries are moderate but growing. Around 2.4% of ART cycles in Israel (2017) and 1–2% in Spain (2018) included PGT-M and PGT-SR. Additionally, more than 10% of Spain’s ART cycles included PGT-A1 [51, 79].

At the current stage of RGS diffusion, PGT-M and PGT-SR are only generating a small share of the total demand for ART cycles, while PGT-A is applied as an add-on to ART to improve treatment outcome. However, as discussed by Beck-Gernsheim [14], RGS diffusion advances in a “spiral-shaped” process where technology acts simultaneously as both a cause and effect. Under such a process, ethical/moral values towards health and responsibility create social acceptance of genome analysis, while technological developments continue to redefine the concepts of health and responsibility and so on. In the beginning, PGT-M and PGT-SR are perceived as ethically justified for the clearest and most severe cases. Once results are positive and as technology improves, consent is extended, and new procedures are not only ethically justified but are in fact expected to be applied. Moreover, PGT-A may be applied as an add-on to ART, justified according to its effectiveness on the treatment. Although once a biopsy is already taken, there might be incentives to use it for extended testing.

Furthermore, RGS diffusion is shaped by a “dual-process” comprised of two cognitive routes [42]: firstly, “systematic processing” relating to the conscious and observable improvement in outcomes, i.e., better success rates and healthier babies, and secondly, “heuristic processing” which relates to the unconscious formation of social imaginaries, desires, and expectations [15, 16, 73, 84], i.e., a shift in the role of RGS due to new perceptions of social and parental responsibility and even new perceptions of health [14]. Socio-technical imaginaries shape social values and public opinion, which then shape regulations and thus favor the diffusion of RGS.

Ethical implications

ART has significantly contributed to the welfare of many families, as it is presently estimated that over 9 million people were born by ART worldwide [24]. Moreover, RGS has prevented a large burden of diseases and suffering. Nevertheless, these technologies raise many ethical controversies that need to be considered by their regulators.

In terms of clinical outcome, several authors have expressed high expectations that RGS will enable the prevention or eradication of specific disorders and even human enhancement, helping to cope with future societal challenges and provide remarkable advantages [36, 60]. However, doubts remain concerning undesirable collateral effects of IVF and ICSI, leading to epigenetic implications and health complications for both patients and children. Unfortunately, given the life expectancy of humans, it is still too early to judge how recent technological advancements in ART may influence offspring during late adulthood or impact future generations [25, 27, 35, 44, 62]. Some analysts have raised doubts concerning the increasing use of PGT-A as an add-on to ART, claiming that it could increase prices without contributing to success rates and could even reduce clinical outcome due to the “waste” of good embryos [18, 32, 64]. Regarding GGM, it has been argued that it carries significant uncertainties and unintended side effects [39, 58]; once a harmful gene edit is introduced, there is currently no method to remove it [26, 87].

Other controversial ethical dilemmas have arisen, mostly concerning potential future applications of RGS. Indeed, it has been argued that the potential geneticization of reproduction would reduce human diversity, emphasizing the differences between races, ethnic groups, and social classes, which might lead to modern eugenics, i.e., the desire to enhance society with stronger, smarter, and “better” people [5, 31, 53], resulting in increased social inequalities and discrimination [17, 29, 76]. Silver [81] has even suggested that in a distant future, a “geneticization arms race” could lead to the polarization of society. Privileged societal groups would evolve so far through genetic enhancement that, at a certain point, the “gen-rich” groups would lose interest in mixing or sharing anything with the “regular” people.

Conversely, Savulescu [77] argued that unlike eugenics, procreative beneficence “is an essentially private enterprise,” a fruit of individual choice. However, this choice could be limited to the extent that individual decisions are largely influenced by doctors’ counsel, social norms, and by fear of insufficient public support [68, 74]. A multifaceted diagnosis or engineering of embryos would mean that complex information must “be conveyed in a language of risk and probability” ([22], p. 1104), hence leaving many patients indecisive and strongly dependent on physicians’ expertise and interpretation based on their professional skills and personal values.

From a more philosophical stance, the ambition to design children has been criticized as weakening instead of empowering. By imposing our desires and beliefs on future generations, we would undermine their autonomy and deny their right for an open future [29, 43, 67, 76]. As anticipated by Lewis [52], p. 37–39): “The final stage is come when Man by eugenics, by pre-natal conditioning (…) has obtained full control over himself. Human nature will be the last part of nature to surrender to Man. (…) The battle will indeed be won. But who, precisely, will have won it?”.

Many of these concerns are usually downplayed by the claim that barriers such as cost and the difficulties associated with ART will limit the expansion of RGS [3, 44]. However, GGM has a greater potential for making breakthroughs, given its technical ease, low cost, and widespread application [62, 89] (and the fact that only a few embryos would be required while various DNA fragments could be edited). Nonetheless, GGM is not practiced clinically, although in the last few years there is an open discourse among the global scientific community regarding how to guide responsible research and its potential clinical use [58, 89]. Since 2015, two international summits on human genome editing have raised a certain consensus about the postponement of clinical applications, although this consensus was recently broken unexpectedly by one researcher in China, with uncertain consequences [48]. The way towards therapeutic uses of GGM is paved through clinical research, clinical trials, and post-approval distribution, in which each phase must have financial and regulatory checkpoints to ensure safety and ethical guidance [26].

Moreover, it has been suggested that clinical application not only requires medical justification and safety but also social acceptance and even consensus about the appropriateness of the proposed use [78, 89]. However, there is a distinction between safety and moral concerns [39], as the first can be seen as objective with clear reason for application, while the second is more subjective, divisive, and demands much broader public discussion [11, 58].

Ultimately, while the application of PGT to avoid severe disorders and improve success rates are gaining consent in Israel and Spain, it is particularly useful to focus on “gray areas” [22], i.e., the more challenging aspects of RGS regulations concerning applications of which the ethical debate has reached no consensus. These include applying PGT for less severe late-onset diseases, sex selection, and the right conditions to introduce clinical GGM, among other disputable applications.

Methods: the Delphi study

We have collected an expert’s opinion by conducting a 2-round modified Delphi survey, which we have designed that builds on a series of in-depth interviews which have enabled us to focus our study and design a structured questionnaire [75, 82]. Delphi is a widely used qualitative method for forecasting, assessing, and decision-making concerning complex problems. It builds on a panel of experts who contribute their insights based on knowledge and experience [50]. It is anonymous and interactive with controlled feedback. The second round allows experts to change their replies or add comments after learning the general views [57, 88]. Between October 2017 and January 2018, we conducted preliminary in-depth, semi-structured personal interviews with 19 Israelis and 10 Spaniards. These interviews not only guided the design of the Delphi questionnaire [75, 82] but also enriched the analysis with more qualitative insights and helped choose and recruit the final panel of experts. We aimed to assemble two Delphi panels which would serve as a bioethics committee for each country, as those traditionally accompany the legislation process. Members were chosen based on their skills, experience, and unique contribution to public discourse [54]. We searched for multidisciplined experts who were interested in ethical issues and whose careers were dedicated to ART from the fields of medicine, public health administration, law, ethics, philosophy, theology, sociology, economics, journalism, and psychology ([10, 57, 33]).

We began by consulting members’ lists of the Spanish bioethics committee and the latest (2012) government appointed Israeli “Mor-Yosef” committee (in Israel, advisory committees are occasionally appointed). We were assisted by a seasoned advisory board member of the Spanish Ministry of Health and by the coordinator of the “Mor-Yosef” committee, who guided us and provided some relevant contacts. Interviewees often recommended other experts, creating a “snowball” effect [71]. Each interview lasted on average one hour and dealt with different aspects of RGS according to the interviewee’s expertise, based on general open-ended questions, aimed at capturing a wide variety of views without conditioning the interviewees’ opinions. Gradually, we reached the saturation point where the marginal value brought by each additional interviewee in terms of new themes was virtually zero. By then, we had already drafted the first draft of a structured Delphi questionnaire which was tested and refined in the last interviews.

The final survey covered many aspects of ART. Five out of 29 interviewees decided not to participate, arguing that they lacked sufficient knowledge on some of the topics. We therefore recruited additional experts to balance out the panels of both countries and enhance the result’s fairness and comparability. The two final panels (see Table 1) included 18 Israelis (12 women and 6 men) and 18 Spaniards (9 women and 9 men), in which 15 Israelis and 9 Spaniards were previously interviewed and additional 3 Israelis and 9 Spaniards who had not been interviewed in the first phase.

Table 1.

Distribution of panel members, by country and area of specialization

Professional activity		Israel (18)		Spain (18)
Doctors and health departments directors	(5)	• Genetics (2) • Gynecology (3)	(7)	• Genetics and biology (2) • Gynecology (5)
Civil servants in health administration	(3)	• Health system Administration (2) • Jewish law (1)	(2)	• Medicine and health System administration • Law, bioethics
Academic researchers	(6)	• Philosophy • Law • Bioethics (2) • Economics • Epidemiology	(7)	• Law and philosophy • Law and bioethics • Bioethics (2) • Economics • Biology (2)
Psychologists and social workers	(2)	• Psychology • Social work	(1)	• Psychology
Others	(2)	• Rabbi • Journalist	(1)	• Law and bioethics (bioethics foundation)

The final Delphi questionnaire was based on 10-point scale questions, with room for open-ended comments. It dealt with general attitudes towards the RGS regulations across specific practices as shown in Tables 2 and 3 below. The questionnaire also included one open-ended question regarding the participants’ general views on the regulations of RGS in order to encourage respondents to provide more detailed comments.

Table 2.

General attitudes towards the regulation of PGT

From 1 (fully opposing) to 10 (fully supporting)	Israel (n = 18)		Spain (n = 18)
	Mean	SD	Mean	SD
1. The public sector should aim at reducing regulation regarding PGT to a minimum in order to allow the patients maximum free choice	3.4	2.83	3.7	2.79
2. When performing PGT for medical reasons, sex selection as an add-on service should be allowed. In other words, the physician may reveal the sex of the (clinically) selected embryos enabling the patients to choose	2.8	2.76	2.2	1.63
3. When performing IVF due to infertility, regulation should be more tolerant towards PGT. It should be allowed as an add-on service to IVF for some range of disorders	7.9	2.51	6.3	2.66
4. It is viable in terms of regulation to separate between the use of PGT-A and the use of PGT-M or PGT-SR for detecting disorders. In other words, in case PGT-A will eventually become a very common add-on for IVF cycles to increase the prospects of the treatment, regulation may still prevent the clinics from regularly using these biopsies for PGT-M or PGT-SR	4.8	2.25	5.6	2.37

Table 3.

Attitudes towards the regulation of specific RGS practices

From 1 (fully opposing) to 10 (fully supporting)		Israel (n = 18)		Spain (n = 18)
		Mean	SD	Mean	SD
Strong support
1. PGT-M for severe monogenic disorders for early-onset and of high levels of penetrance with no simple cure	Allow	9.6	0.62	9.3	2.20
	Fund	9.5	0.64	8.9	2.30
2. PGT-M for severe monogenic disorders for medium-late onset and of high levels of penetrance with no simple cure	Allow	9.4	0.81	8.9	2.19
	Fund	9.2	0.94	8.5	2.35
3. PGT-M for severe monogenic disorders for medium-late onset and of medium levels of penetrance with no simple cure	Allow	7.5	2.22	8.6	2.43
	Fund	6.9	2.29	8.1	2.60
Mild support
4. GGM (CRISPR/Cas) for severe monogenic disorders for early-onset and of high levels of penetrance with no simple cure in case PGT did not provide a solution	Allow	7.5	2.29	6.4	3.39
	Fund	6.5	2.79	4.8	3.51
5. PGT-A for detection of chromosomal abnormalities, in order to increase the prospects of an IVF treatment	Allow	6.9	2.79	6.2	3.29
	Fund	5.7	2.97	4.3	3.24
6. PGT for multifactorial diseases (cancerous/metabolic/cardiovascular/neurological) for medium-late onset and of medium levels of penetrance	Allow	6.4	2.60	6.0	3.28
	Fund	5.2	2.34	4.3	2.66
7. PGT for whole-exome screening	Allow	3.8	2.29	4.3	3.15
	Fund	2.6	2.22	3.1	2.82
Disapproval
8. PGT for sex selection	Allow	2.1	1.88	3.7	3.37
	Fund	1.8	1.51	1.4	1.42
9. PGT for cognitive characteristics selection	Allow	1.3	1.19	1.5	1.29
	Fund	1.4	1.22	1.3	0.96
10. PGT for physical traits selection	Allow	1.3	1.19	1.3	0.97
	Fund	1.4	1.22	1.3	0.96

The survey was delivered via two rounds of the same questionnaire, the first between April and September 2018 and the second between September and December 2018. In between both rounds, we delivered controlled feedback by highlighting those questions for which consensus, proxied by the standard deviation (SD), was not reached (SD > 2) [50, 82]. Each participant received feedback on their answers which were outside the interquartile range of their panel. We focused exclusively on inconsistencies or substantial deviations in order to reduce dropout rates in the second round while allowing experts to reach consensus or provide comments to explain dissensus [49]. Only one Spanish expert did not reply to the second round.

Participation was voluntary and without compensation. Therefore, to avoid “respondent fatigue” and respect the experts’ limited availability, the modified Delphi was restricted to two rounds regardless of the degree of consensus. Although the Delphi method typically aims to achieve a consensus, when this is not possible, the method is also useful to understand dissensus by exposing and debating differing positions on a topic, which has been proven relevant in previous studies to inform policymakers and regulators [49, 59, 88]. In our case, although there was only a mild convergence towards consensus during the second round, the comments from both rounds and interviews explained the dissensus and complemented the statistical analysis with rich insights. The comments collected from the survey, together with some of the statements from the interviews, were categorized according to their relevance to each of the 10-point scale questions, in order to facilitate their analysis and enable their presentation in "Results" section.

Finally, we applied the non-parametric Mann-Whitney U test for two independent groups to identify differences between subgroups of gender, country, and profession.

Results

The experts’ quantitative answers are presented in Tables 2 and 3, accompanied by some open comments in quotation marks, which represent their views.

General attitudes towards regulations

We introduced four different statements and asked the panels to mark their level of support (from 1, fully opposing, to 10, fully supporting). As seen in Table 2, there was no full consensus regarding these statements. Conversely, different approaches were introduced by some key comments.

The results concerning general attitudes towards the regulations of PGT may be summarized as follows:

1. The panels did not support the liberalization of RGS. Although most experts backed the promotion of PGT to reduce genetic diseases, they believed that regulation should control the practice of these services and reduce unexpected risks. An Israeli respondent stated in the first round: “In the long run, we may find that by trying to prevent cancer by PGT-M, we would have increased the incidence of other cancers or malformations (exposing the fertilized egg to laboratory conditions). Not enough years have passed to make us confident in the safety of these techniques.”

Some comments highlighted the importance of reducing asymmetric information, emphasizing that the technology is very sophisticated and the public is not familiar with all its implications: “Regulation should ensure safe and evidence-based services, making sure the offer of PGT is accompanied by appropriate counseling and that the important decision to perform IVF for the sake of PGT is fully informed and free from pressure” (as added by an Israeli expert). Additionally, according to a Spanish panelist: “We should prevent society from falling into the false belief that RGS assures 100% healthy offspring (…) we should not fall into genetic determinism when many other factors can influence people’s health.”

Although most experts supported regulating the field, some advocated for the consideration of personal autonomy as much as possible, urging (according to an Israeli respondent in the first-round) to “avoid heavy-handed regulation of PGT and leave the reproductive decision making to women and couples, based on a principle of reproductive autonomy.” However, a comment in the second round from a Spanish participant stated that “a patient’s choice and their consent for therapies should take place within a normative framework and a public health system. Therefore, the freedom of choice cannot be total.”

• 2.Most experts opposed revealing the gender upon performing PGT. Comments in both rounds emphasized the fear that allowing it would turn every PGT into a sex selection.
• 3.In contrast, PGT, as an add-on to IVF, was approved by the participants who again underlined priority in preventing medical disorders. The approval was significantly stronger among the Israeli panel in this matter, while no other statistically significant differences were found between the defined subgroups.

Several Israeli experts raised the “slippery-slope” argument in their comments in the first round: “The definition of ‘a range of disorders’ is very vague… where will it lead us? Where are the boundaries?”; “The process of selecting ‘perfect’ embryos occurs gradually and we have become accustomed to the idea so that later it will seem natural to prevent the birth of infants with treatable diseases or even traits that we have no reason to prevent”; in the second round, similar arguments included: “Why for a variety of genetic diseases rather than genetic anticipation or next-generation sequencing?”.

• 4.The experts were doubtful regarding the ability to draw a line between PGT-M/PGT-SR and PGT-A in case the latter becomes an official add-on to IVF, as stated by an Israeli respondent in the first round: “There is a regulatory weakness in this regard because the line is very loose. Where settings are not sharp, regulation will lose...”.

Regulation of specific practices of RGS

Table 3 presents the experts’ attitudes towards allowing and funding specific applications of RGS. After ranking the scores, we can distinguish three different levels of consent: strong support, mild support, and disapproval.

Strong support

The experts strongly supported (with a high level of consensus among the Israeli panel) applications of PGT-M for monogenic disorders of high levels of penetrance with either early or late onset. However, there was a lower consensus regarding PGT-M for disorders of medium levels of penetrance (including cancerous diseases), with the Spanish panel expressing stronger support than the Israeli one. Again, there were no other significant differences between subgroups. Despite the lack of consensus, both panels tended to support allowing and funding these procedures. As explained by one Israeli interviewee: “We are not talking here about curable diseases. Some genetic disorders of ‘medium levels of penetrance’ are in reality devastating cancers which may affect various family members for generations.”

Experts mostly supported the funding of PGT-M by the public health system in order to avoid health inequalities and guarantee fair and inclusive access.

Mild support

The panels displayed less consensus (higher SD) regarding four contested “gray-zone” categories for which the health benefits are not yet clear, thus highlighting the controversies of RGS regulation concerning these topics. Firstly, concerning CRISPR/Cas9 for GGM, we presented the panels with a specific case referring to a severe, early-onset disorder of high levels of penetrance where PGT cannot be delivered due to a “shortage” of eggs/embryos. The panels expressed reasonable support in this case, and the main concerns were around safety, beneficence principle, efficiency, and low cost. The “slippery-slope” argument was not raised.

Secondly, concerning PGT-A for detection of chromosomal abnormalities, some experts emphasized the undemonstrated usefulness or cost-effectiveness of the technique. For example, one Spanish participant argued in the first round that “PGT-A is contraindicated in the following cases: advanced maternal age, early ovarian failure, low response, poor embryo quality, severe male factor infertility and more.” An Israeli expert added that “it should be applied only in cases of repeated implantation failure and recurrent pregnancy loss.” Some experts (mainly Spaniards) expressed concern regarding the increased use of PGT-A, claiming that it is often offered as an add-on to IVF cycles, increasing the financial burden on patients for the benefit of private clinics.

Thirdly, concerning PGT for detection of multifactorial diseases, the experts distinguished between genetic diseases which can be reliably diagnosed, whereby PGT should be permitted and publicly funded and other multifactorial diseases which cannot yet be diagnosed by PGT and therefore should be strictly regulated. One Spanish expert claimed in the second round that “there are no immediate perspectives that PGT contributes anything significant regarding disorders involving more than a single gene, or when a larger number of factors cause a disease.”

Fourthly, the experts did not express high support for using PGT for whole-exome screening. As suggested by an Israeli interviewee: “The problem is that there will always be some suspicious mutations (although the emphasis will be given only to areas where there is a family history or where a gene for carriers has been identified), but if each finding was revealed to the patient, there would be no embryo remained to implant.” Moreover, a Spanish expert commented in the first round: “I reject the use of techniques that are not directly preventing or treating diseases, for which safety is not guaranteed or whose impact and consequences on the human species are yet unknown.”

Differences between “allowing” and “funding” the procedures were noticed, as explained by a Spanish respondent in the second round: “The public system, always with limited resources, should prioritize the financing of other health needs rather than PGT for multifactorial diseases, which could lead to an uncontrollable demand”… “Allowing these procedures would enable autonomy for couples in decisions concerning their children’s health. Another issue is financing these procedures with public funds when we have many other urgent priorities regarding health issues that are not subjected to probability.”

Disapproval

Finally, PGT for non-medical needs was mostly discarded by the experts. As argued by a Spanish expert in the first round: “The ethical boundary regarding the application of PGT should be between avoiding hazards and satisfying preference or choice of characters, simply according to the parent’s preference and not for the future benefit of an individual. Applying scientific knowledge to avoid suffering is ethically acceptable but not to satisfy whims.”

Nevertheless, a few experts were tolerant of sex selection. An Israeli expert commented in the first round: “Sex selection in this area of the world is sometimes much more than merely ‘social’ and understandably allowed if having the ‘wrong’ sex baby might jeopardize either the mother or the baby.” In addition, a Spanish expert added in the second round: “It should be considered in some particular cases i.e. couples with few children of the same gender who are psychologically affected by the lack of offspring of the other gender. It is not necessary to criminalize them.”

Discussion

We conducted a Delphi survey among two panels of experts, selected from various fields, by attempting to simulate typical bioethics committees in Spain and Israel in order to match the practice of consultative and regulatory processes as closely as possible. We focused on regulation as one of three conditions that could lead to a broader diffusion of RGS, in turn leading to the geneticization of human reproduction, being the other two conditions real or perceived technological improvements that will make ART safer, more efficient, and more comfortable and enable RGS to deliver children with enhanced health. Both Delphi panels tended to approve medical applications of RGS that are safe and provide clear health benefits (as with most currently applied PGT-M and PGT-SR). Regarding the so-called strong support area, we found a correspondence between attitudes to “allowing” and “publicly funding.” What is allowed should be publicly financed to prevent inequality in access. The experts also outrightly rejected ART uses for non-medical reasons (selection of physical and cognitive traits), albeit with some level of dissensus in Spain regarding sex selection. PGT for sex selection is already allowed in Israel (subject to committee approval) for couples with four children of the same gender and none of the other.

Finally, we identified a “gray-zone” of ART regulations regarding four procedures: (1) the use of GGM for severe monogenic disorders for early-onset and high levels of penetrance when no therapeutic alternative is available; (2) the use of PGT-A for detection of chromosomal abnormalities; (3) the use of PGT for multifactorial diseases for medium-late-onset and medium levels of penetrance; and (4) the use of PGT with whole-exome screening.

A significant proportion of experts expressed their willingness to allow these procedures which, except for PGT-A, are not yet practiced today. Overall, their consent to RGS is founded upon applying these services to cases of severe disorders, i.e., “life and death,” in contrast with preventing disabilities or human enhancement. Their primary motive is around safety, followed by further health benefits that the techniques bring, followed by reproductive autonomy. However, as raised by some experts and as several authors have pointed out, in practice, the burden of proof regarding health benefits does not always apply whereby many times the technology is merely assumed to be beneficial, such as in the case of the growing application of PGT-A [18, 32, 64]. Interestingly, some differences between the attitudes to “allow” and “fund” arise in the “gray-zone.” Doubts and concerns about the benefits of these procedures may explain the reluctance of experts to support their public funding.

As new technological developments in ART and RGS materialize [2, 8], the regulators’ task will become more complicated since each “category” under the regulator lens (i.e., GGM) will continuously expand into numerous “sub-categories.” Such expansion of the “gray-zone” and the shifting of some sub-categories into the “strong support” area would represent a continuity of the previously described “spiral-shaped process” [14], as both civil society and policymakers develop larger consent to these services, which would also be influenced by real and perceived benefits. Following that process, this increasing consent would result in looser regulation that would further promote RGS diffusion. Thus, as we claimed in "Geneticization of reproduction: determinants and ethical implications" section, further geneticization of reproduction would not only result from a “moment of breakthrough” [15, 16, 62] but from self-reinforcing shifts in supply and demand. The first includes a steady increase in the share of ART out of total births, an ongoing approval of novel RGS, and the increasing introduction of PGT-A as an add-on. The second involves a growing perception of RGS as beneficial since, on the one hand, real outcomes are being produced, while, on the other, imagined or uncertain outcomes are being perceived. Socio-technical imaginaries would shape social values and public opinion, which then influence regulations.

It is particularly the case in the two analyzed countries in which current regulations consciously or unconsciously promote the medicalization of reproduction by generously funding IVF cycles to support fertility rates, like in Israel, or by facilitating a “laissez-faire” environment for the private industry, like in Spain [3]. The Spanish law leaves great room for expanding the future definitions of genetic diseases [65], whereas the CNRHA is subject to potential bias since the selection procedure of its members does not control areas of conflict of interest [3, 65, 66]. Israel also offers flexibility since crucial decisions have been decentralized and handled by local committees in hospitals and clinics [80].

Our findings reveal that experts from both countries have quite similar attitudes despite their different cultural backgrounds. Only two questions raised significant disparities between both countries. The Israeli experts showed greater consent to PGT as an add-on to ART cycles, while the Spaniards more strongly supported the use of PGT-M for medium-late-onset disorders of medium levels of penetrance. Additionally, the Israeli experts had a higher tendency to fund these “gray-zone” services. There is no straightforward explanation for these disparities, besides the fact that the key regulatory distinction between the two countries lies within their resource allocation and priorities of ART funding [3]. Notably, we did not find any other significant differences among the distinct stakeholders who participated in the survey (according to gender or professional background).

During interviews and comments in the survey, many ethical arguments were raised and discussed concerning social risks (such as those raised by Ramsey [67], Jonas [43], Silver [81], Buchanan et al. [17], Fukuyama [29], and Sandel [76]), eugenics and discriminative attitudes towards disability (as raised by Lippman [53], Asch and Barlevy [5], and Garland-Thomson [31]), medicalization, uncertainties, and freedom of choice (Ehrich et al. [22]; Rapp and Ginsburg [68]; Rothman [74]). The experts seemed to shape their consent to new forms of RGS with ambivalence, on the one hand showing caution about these ethical controversies and social risks but on the other conferring much higher importance to health and individual reproductive autonomy. Furthermore, the experts did not tend to base their approaches on possible radical scenarios of medicalization and geneticization since it is hard to identify how the general public would benefit from RGS. That is because the share of the human population affected by monogenetic or relatively simple polygenetic disorders is quite limited [6, 9] and the use of RGS for multifactorial diseases and traits would bring problems of gene pleiotropy (one gene controls the phenotype or expression of several unrelated traits). Aspiring to perform an expanded PGT would require a better understanding of the complex and flexible interaction between genomics, environmental factors, and epigenetic alterations [62] and might even reveal the falseness of genetic determinism [17, 65, 69]. Moreover, considering the shortage of human eggs required to produce a vast number of embryos for such a procedure, embryo selection would most probably stay limited. Only by designing embryos using GGM might it be possible to deliver those radical scenarios.

In this regard, both panels showed a tendency to approve GGM once it is safe and provides health benefits that PGT fails to deliver. The panels’ regulatory approach is compatible with the results from a previously conducted Delphi of similar design, for which two sets of Spanish and Israeli physicians were used, equally drawing a clear line between allowing RGS for medical benefit when the technology is safe and disallowing RGS for non-medical reasons [2]. Furthermore, the positive attitude towards RGS that found in the present study is consistent with Kalfoglou et al. [44] and Hollister [39], who also highlighted the distinction between safety and moral concerns.

Similar to previous studies [39, 58, 62, 87, 89], our findings may suggest that we are slowly approaching a global consensus for developing GGM. Nevertheless, even after its first applications, we expect consent to GGM to evolve gradually, as policymakers continue to consider health risks and outcomes [26]. Moreover, it was strongly implied by the experts (some of them actual members of national bioethics committees) that despite being leaders in ART usage, neither of the two analyzed countries will be the first to regulate therapeutic GGM, and both will follow the global scientific community.

Our analysis contributes to the design of future regulations by pointing out aspects that require more public control and supervision. That said, our results should only be taken as preliminary due to some methodological limitations. A Delphi panel selection always suffers from selection bias, even though the method is not aimed at statistical generalization but rather at capturing expert views [49]. Our sample is neither statistically representative nor exhaustive with different choices of experts expressing different attitudes. While the sample could draw some comparisons between Israelis and Spaniards, it was not large enough to detect any differences among other subgroups (gender, profession). A larger study involving experts from other countries would clearly enrich the discussion.

In addition, our Delphi design was based on interviews with experts, but not all interviewees finally participated in the survey. Instead, other experts were recruited to form the final panels. The survey was limited to two rounds to guarantee low dropout rates (with only one expert who did not reply), which did not favor consensus regarding most topics. Nevertheless, it has to be highlighted that reaching a consensus was not our primary aim, but to bring to light differing positions and to identify points of controversy [49, 59, 88].

We selected countries which are at an advanced stage of diffusion, have very pro-ART attitudes, and possess a comprehensive public healthcare system. Although these elements might bias our findings, they also suggest that Israel and Spain could be among the first countries to deal with the ethical implications of applying the said new RGS and could pave the way to others. State borders and national regulations are becoming increasingly easy to pass. Therefore, regulatory collaboration across countries will be of great importance [56]. Our results could also be useful for other, less pro-ART jurisdictions as they point out new possibilities that might soon be available for their citizens abroad despite being banned at home, particularly for some northern European countries, for which Spain is the leading destination for reproductive tourism.

Conclusions

Given the steady increase in the practice of ART, the growing use of PGT-A as an add-on, and the pace of development of GGM, regulations will increasingly have to engage not only with more complex safety measures but also with more challenging ethical dilemmas, requiring a broader public debate. As suggested elsewhere, a “tentative governance” approach building on new means of stakeholder engagement could be useful in this context, given that consensus is not always possible when dealing with complex emerging technologies [55].

Although far from the scenario of geneticization of reproduction, according to the results of our Delphi, regulators should aim to slow down the medicalization of human reproduction due to uncertainty about risks and benefits and the many ethical dilemmas regarding new procedures. It could be achieved by focusing on the three previously discussed conditions for geneticization. Firstly, more cohort studies should be applied to disperse some uncertainties concerning health risks for patients and children born by ART (none are yet to approximate elderhood). The importance of such studies will grow as further diffusion of RGS materializes, particularly concerning PGT-P and GGM.

Secondly, the public should be better informed regarding infertility determinants (mainly parenthood postponement and environmental factors) and real ART outcomes. As RGS becomes prevalent and as GGM is potentially introduced clinically, public awareness should be enhanced by disseminating clear and simplified information regarding the opportunities and risks of these technologies, considering that genomics may be an extremely complicated matter for laypersons and that as we claim in the paper, socio-technical imaginaries may shape perceptions of RGS. In this regard, while better knowledge distribution would help avoid excessive expectations, RGS advertisement by ART clinics should also be carefully regulated [3, 28, 30].

Finally, both regulators and the general public should avoid basing ideology on outdated and misleading contexts inspired by inflated hopes or dystopian theories. The regulatory approach must not be based on fear but authentic and updated scientific knowledge, despite the challenging pace of technological development and their growing complexity. Indeed, preventing RGS for non-medical reasons is not a total guarantee for stopping the geneticization of reproduction, and similarly, allowing patients to take autonomous decisions will not necessarily lead to social catastrophe.