Principle 1: the welfare of the future person

Gametes or embryos that have been subject to genome editing procedures (or that are derived from cells that have been subject to such procedures) should be used only where the procedure is carried out in a manner and for a purpose that is intended to secure the welfare of and is consistent with the welfare of a person who may be born as a consequence of treatment using those cells.

It is notable that in the formulation of this principle, the Nuffield Council has specifically referenced a general approach focused on the welfare of the future person rather than any particular distinction between different applications of HGE, such as ‘therapeutic’ versus ‘enhancement’. As the report itself notes, it is very difficult to draw a clear distinction between therapeutic and non-therapeutic applications and this approach sidesteps this difficulty. The report explicitly states that there is no a priori reason that applications beyond the prevention of disease would not also be consistent with the welfare of the future person (p. 76). This is a significant step in opening the door for a variety of traits and characteristics to be considered for HGE. For example, this raises the possibility of actively including certain welfare-improving characteristics, rather than limiting HGE to the removal of welfare-diminishing characteristics.

The report goes on to address the interests of society, noting that reproduction takes place in a social context. Reproductive behaviours already can change the composition of a society without deliberate coordination, but reproductive technologies allow for this with greater certainty (p. 78). Other people in society may be collaterally and indirectly affected by the use of HGE. The report highlights questions surrounding diversity, shifting norms, disability critiques, social virtues and equity and justice.

The report notes that it is possible that genome editing could lead to changes in the level of diversity within the population—whether more or less would depend strongly on prevailing social factors (pp. 79–80). If the use of HGE becomes standard, this could lead to a shift in norms and the expectation of its use, which could decrease freedom. This can be reflected in current concerns around prenatal screening—what people typically do becomes what people should do, and thus what people feel pressured to do (pp. 80–82). This follows on to disability critiques, such as the expressivist objection, which states that attempts to deselect or prevent disability expresses or presupposes a negative view of people with a disability. However, the Nuffield Council explicitly rejects this objection when it comes to genetic conditions that significantly affect both quality and length of life (p. 82). Another set of critiques is that HGE represents attempts to overcome fragility and weakness, which are in themselves a valuable part of the human condition and should not be removed. However, the difficulty with these approaches is that a person would have to value such fragility enough that they would be willing to impart it on their own children (pp. 82–85). Finally, the report considers the importance of equity and justice and concludes that HGE should be restricted so as not to result in unfair advantages for certain groups of people (pp. 85–86). This leads on to the second principle formulated in the report (p. 87).

Principle 2: social justice and solidarity

The use of gametes or embryos that have been subject to genome editing procedures (or that are derived from cells that have been subject to such procedures) should be permitted only in circumstances in which it cannot reasonably be expected to produce or exacerbate social division or the unmitigated marginalisation or disadvantage of groups within society.

The third set of interests that the report addresses is that of humanity and future generations. Potential adverse effects of HGE may only manifest themselves after several generations, and here the notion that we have responsibilities or moral obligations to future generations is key (p. 89). One perspective is that HGE could offer a means to remedy harms that have already been set in motion for future generations, such as runaway climate change (pp. 89–90). However, invoking the ‘precautionary principle’ would suggest that the uncertainty and possible negative consequences of HGE mean that it should not be applied. However, as the report notes, this constitutes no reason not to continue research and the development of the technology as a means of hedging bets against future events (pp. 90–91).

The report also addresses the relationship between HGE and transhumanism. Transhumanism is a concept that is closely linked to the emergence of technologies such as HGE. HGE could lead to the ‘self-overcoming’ of the human species to a grander, more capable species, and the report questions whether this constitutes a moral reason not to apply it. This may stem from a notion of a fundamental human dignity. HGE might be troubling because it threatens the integrity of human genetic inheritance (interference with the line of transmission that links the human family together) and the integrity of human genomic identity (distinguishing the human family from non-human beings) (p. 91). These concerns could be mitigated by replacing variants only with wild-type or typical variants, but of course the question of what is a wild-type or mutant variant is often value-laden in itself. The human genome is enormously varied, and many new variants are possible and likely to emerge. The report responds to this by stating that concerns about moving the human genome away from wild-type variants are prudential, rather than a categorical moral reason not to do so (p. 92). However, questions remain surrounding justice and equity, and the possibility of schisms between the ‘gene-rich’ and the ‘gene-poor’ (pp. 92–95).

The report concludes by stating that there are no categorical limits on the use of genome editing technologies, as long as (p. 97):

• They are not biologically reckless;

• They are consistent with the welfare of future people;

• They are not socially divisive;

• They are not initiated without prior societal debate.

We applaud the Nuffield Council’s approach to HGE and think it is an important step forward in the debate. A strength of the approach is that it outlines quite specific moral principles rather than merely appealing to broad concepts—such as earlier reports.

For example, the National Academiy of Sciences (NAS) report on genome editing endorses the principle of:

Promoting well-being: the principle of promoting well-being supports providing benefit and preventing harm to those affected, often referred to in the bioethics literature as the principles of beneficence and nonmaleficence.

It is left unspecified whose well-being we should be promoting through genome editing, what the components of well-being/harm are and exactly how that principle should apply to genome editing. Nearly all people would freely endorse this principle (and many of the other principles relied on in the National Academy of sciences report), but might draw radically different conclusions regarding genome editing. In contrast, the Nuffield Council’s first principle is much more specific and the scope and force is clear.

Nonetheless, there is room to improve and build from the Nuffield Council’s approach. In the next section, we show how an asymmetry in the structure between the two guiding principles leads to counterintuitive implications for an important set of possible uses for HGE, and suggest ways their second principle could be improved.

Modified principle 2: social harms

The use of gametes or embryos that have been subject to genome editing procedures (or that are derived from cells that have been subject to such procedures) should be permitted only in circumstances in which it cannot reasonably be expected to produce or exacerbate social harms, including increased social division or the unmitigated marginalisation or disadvantage of groups within society.

While this principle loses some of the advantageous of specificity, it has the resources to respond to concerns relating to collective action problems.

Single gene disorders

A mark of success of medical genetics has been the diagnosis of the disease phenylketonuria (PKU) at birth. This is an inherited metabolic disorder in which levels of the enzyme phenylalanine hydroxylase are lowered. This means individuals cannot metabolise the amino acid phenylalanine. In 1962, a test was devised that allowed PKU to be diagnosed through a blood test.17 The ‘heel prick test’ is now routinely given to infants as part of newborn screening. Those children who are identified as suffering from PKU are put on a low phenylalanine diet or else they will develop severe intellectual disability. This diet means no bread, pasta, soybeans, egg whites, meat, legumes, nuts, watercress and fish. Such an environmental intervention is demanding. There is always a risk that foods containing phenylalanine will be consumed by mistake. The ubiquitous sweetener, aspartame, can cause a crisis.

Imagine an artificial enzyme was developed to replace phenylalanine. If this was administered regularly it would allow sufferers of PKU to consume a normal diet. Such a cure would be hailed as a breakthrough. There would be a moral imperative to provide this cure, just as there is an imperative to provide blood transfusion for severe bleeding, and antibiotics for infection.

Now imagine that instead of getting a pharmaceutical company to manufacture the enzyme, we could get the body to manufacture it. By altering the DNA of someone with PKU, we could get a patient’s own cells to produce the missing enzyme, phenylalanine hydroxylase. There are many advantages to not relying on pharmaceutical companies. Production inside the body allows for a more targeted response and more accurate dosages. Furthermore, it removes all chance that a patient would be unable to access the treatment, such as when the company has supply chain problems.ⁱ

Just as there would be a moral imperative to provide a replacement enzyme therapy for PKU, there would be an imperative to make safe genome edits which prevent PKU. If it becomes possible for carriers of the PKU mutation to prevent PKU in their future children through HGE, they will have an obligation to use this technology, in the same way they would have an obligation to use an enzyme replacement therapy.

Polygenic diseases

Most diseases are not the result of just a few genetic changes. They are the result of many, sometimes hundreds, of genes combining together with environmental effects. Such polygenic diseases are among the world’s biggest killers. Cardiovascular disease is emerging as the biggest cause of death in the low-income and middle-income world. Together deaths from chronic diseases in those under 70 years are responsible for approximately 30% of all deaths worldwide.21 In addition to causing pain and death to individuals, chronic diseases place a huge burden on national health systems, consuming resources that could be used elsewhere. One study found that the healthcare cost associated with treating cardiovascular disease totalled €104 billion annually, for countries within the European Union.22

We know that there are genetic contributions to chronic diseases. Genome-wide association studies have identified at least 44 genes involved in diabetes23; 35 genes involved in coronary artery disease24 and over 300 genes involved in common cancers.25

It is possible to differentiate between individuals based on their genetic risk of developing chronic diseases. Using next-generation sequencing technologies (like whole genome or whole exome sequencing), polymorphisms occurring across many genes can be tallied and weighted giving an individual a ‘polygenic risk score’ that reflects their genetic predispositions to develop particular diseases and traits. Individuals can then be stratified into different risk categories (such as high risk, medium risk and low risk) based on their polygenic risk score.26

As genome editing technologies can target many genes at one time, it may become possible to use them to alter an individual’s polygenic risk score at the embryonic stage,ⁱⁱ and shift individuals from a high-risk category to a low-risk category.^{iii, iv}

Alternatively, it will be possible for individuals who know they have a high polygenic risk to particular diseases to use HGE to alter their gametes to ensure they do not pass this high risk on to their children.

For example, by editing around 27 mutations associated with coronary heart disease, it would be possible to reduce an individual’s lifetime risk by 42%27; by editing 12 genetic variants one’s lifetime risk of bladder cancer could be reduced by almost 75%.28

This application cannot be achieved through current methods of genetic selection. Say a couple want to use PGT to select for 15 different genes in an embryo, to reduce their likelihood of cardiovascular disease. Then they would need to create thousands of embryos to make it sufficiently likely that one will have the right combination at all 15 loci. The chance of the couple having such an embryo would be <1% with traditional IVF and PGD.29

Given the massive disease burden caused by chronic diseases, we have strong moral reasons to develop technologies that reduce their incidence—whether these operate through genetic or environmental mechanisms. Imagine scientists develop a new technology which potentially could be incorporated into exhaust filters, and would drastically reduce the amount of air pollution cars emit. In cities where cars are fitted with the exhaust filter, the incidence of respiratory disease would be decreased by 40%.

There are clearly strong moral reasons to develop this technology and pursue its applications. Developing the exhaust filter is not merely something it would be permissible to do, but something that there is an imperative to do. The very same reasons apply to the development of HGE.

One might respond that the clear difference between this case and HGE, is that HGE makes heritable changes and will thus affect future generations. However, air pollution is a known epigenetic modifier,30 that is, it makes changes to gene expression which can be inherited by future generations.31 Hence, reducing air pollution could also affect future generations. Of course, we need to consider what the long-term effects of any changes will be. But if the likely effect of a genetic change in one generation is to reduce risk of disease in future generations, this seems only to strengthen the case in favour of those changes.

If HGE could make genetic changes which reduce risks of polygenic disease in current and future generations, there would be an imperative to use it. Obviously, this application is a long way away from being plausible, possibly decades. One major difficulty is that we do not understand polygenic scores well enough to accurately predict the effects of large-scale changes. Still, we have moral reasons to develop HGE with the intention of using them for this purpose. First, it will reduce rates of premature death and disability due to chronic disease. Second, the use of HGE to make the highest risk individuals the same as the lowest risk individuals will be equality-promoting. Third, using HGE to lower the incidence of chronic disease will also promote justice. As stated above, health systems spend billions in resources to treat and prevent chronic disease. Using HGE in germline cells will probably be a relatively cheap way (in the proximity of US$20 000) of reducing someone’s susceptibility to chronic diseases. In a world of limited resources, taking a more expensive therapy has the opportunity cost of preventing the treatment of someone else’s disease. Justice requires we choose the most cost-effective option, other things being equal. If we do not invest in the most cost-effective option, we harm others who could use these resources.

Enhancement

Just as polygenic scores could in theory be used to reduce rates of complex disease, they can target complex traits like intelligence.

General intelligence—the ability to learn, reason and solve problems—is the best known predictor of education and occupational outcomes.32

For decades, it has been known that around 50% of the observed variation in intelligence is due to genetic factors. A number of recent large studies have identified many polymorphisms, which help explain 20% of the heritable variation in intelligence.32

As with complex disease, using the polygenic scores it is possible to stratify the population into three board groups ‘high predisposition to high intelligence'; ‘medium predisposition to high intelligence’ and ‘low predisposition to high intelligence’. It will become theoretically possible to use HGE to shift individuals from the low or medium predisposition groups, into the high predisposition group.

Enhancing based on intelligence using polygenic scores would, in the words of the Nuffield report, be a form of enhancement that uses only ’wild-type’ variants (variants that already exist in the species) rather than a form of enhancing that goes beyond what currently exists in the species. In other words, it is a form of ‘normal range human enhancement’.33 While it may be possible in the future to enhance intelligence beyond levels that are currently observed in the species—such forms of enhancement are much less feasible at present.

Imagine a prenatal nutritional programme was developed, which was predicted to increase intelligence in children born with low innate predisposition to high intelligence. This would be seen as a breakthrough. We may soon be able to achieve the same with HGE.

One of the most intuitive concerns about technologies like germline engineering is the effect on equality. It is feared that germline engineering would only be available to the rich, and that it could widen the gap between rich and poor, adding biological advantages to already existing social ones. This is an important and complex issue, faced not just by genome editing but other goods like education. Ethically, we must take steps to ensure that the benefits and costs of HGE are evenly or fairly shared. As recognised by the Nuffield Council, this is not a reason to ban the technology, or fail to develop it, but a reason to ensure it is developed responsibly.

However, it is also possible to use HGE to directly improve equality, as the intelligence example shows. Nature is a biological lottery which has no mind to fairness. Some are born gifted and talented, others with short painful lives or severe disabilities. Currently, diet, education, special services and other social interventions are used to correct natural inequality. It may be that targeting combinations of genes is an effective means of promoting equality in education. For example, there are natural variations in people’s innate ability to learn how to read. This often matters little for people in higher socioeconomic groups, who can afford to spend extra time with their children teaching them how to read, or employ tutors, etc. However, for those in lower socioeconomic groups, this predisposition can leave them illiterate for life. While other measures could in theory no doubt remedy this inequity of outcomes, evening out the genetic starting point could prove the most effective way. This method would have the additional benefit of being passed to future generations. Genome editing could be used as a part of public healthcare for egalitarian reasons.

Boosting intelligence and other cognitive traits through HGE will be an ‘enhancement’, rather than disease prevention. As noted by the Nuffield report, this does not by itself reduce the moral reasons we have to pursue it. We have a moral imperative to use all reasonable means to produce equality in education.

Future generations and intergenerational justice

One of the key interests considered by the Nuffield Council report is that of future generations. It is crucial that the very long-term consequences of developing or failing to develop HGE be considered. Humans often exhibit a cognitive bias towards the near future and neglect then how our actions may affect the very far future. This can distort our appraisal of technologies.

The obligations we have to future generations are often described in terms of intergenerational justice. We owe future generations the same considerations that we owe our contemporaries. We should not unnecessarily deplete the ozone layer, for example, if this will greatly harm future persons at an only small benefit to ourselves.

Some worry that by engaging in HGE we risk harming future generations by negatively altering our genome. There is no doubt that some application of HGE could harm future generations (e.g., see discussion of collective action problems in Section 3); however, such applications are not the inevitable consequences of the development of HGE, and can be mitigated or avoided.

Moreover, a deep engagement with the interests of future generations will show why there is strong moral imperative to develop HGE as a matter of intergenerational justice.34

Modern medicine is removing selection pressures that humans have historically been subjected to. This is increasing the rate of random mutations accumulating in the genome and poses a risk to future generations, as made clear by Michael Lynch in a 2016 article in the journal Genetics:

What is exceptional about humans is the recent detachment from the challenges of the natural environment and the ability to modify phenotypic traits in ways that mitigate the fitness effects of mutations, for example, precision and personalized medicine. This results in a relaxation of selection against mildly deleterious mutations, including those magnifying the mutation rate itself. The long-term consequence of such effects is an expected genetic deterioration in the baseline human condition, potentially measurable on the timescale of a few generations in westernized societies.35

As we develop effective and accessible treatments for disease, we all but guarantee that the incidence of those diseases will increase in future generations. This is because mutations which arise that contribute to those diseases are no longer selected against.

For example, short sightedness (myopia) has been historically very rare because it was selected against in hunter-gatherer societies.36 Modern technologies such as glasses, contact lenses and Lasik eye surgery help correct such vision problems. In modern societies, those with naturally poor eyesight have the same fitness as those who have naturally good eyesight. This allows deleterious mutations to occur in the genes which influence vision and not be selected against. Rates of myopia are now over 50% in many countries, making populations increasingly reliant on technology for this basic biological function. It is likely that reduced selection against poor vision has caused some of this increase. While it is easy to correct for myopia, the same process will allow mutations to accumulate in genes which influence other biological functions.

The percentage of people who require blood pressure medication,37 assisted reproductive technologies38 and have genetic predispositions to deafness, are all increasing. While social changes play a major role in these changes (eg, poor diet and sedentary lifestyle, delayed childbearing), biological factors also play an important part.39 40 In future generations, nearly all people may be reliant on technologies for these basic functions, as well as many others.

This will be bad for individuals, who become increasingly dependent on technologies for basic functions, and need to spend much of their time and money acquiring a range of therapeutic goods. Similarly, society will become burdened with spiralling healthcare costs. Furthermore, the consequence of natural disasters will become much more severe if people are reliant on a variety of complex technologies whose supply can be disrupted.

Fortunately, there is a way for our descendants to avoid such a medicalised future. Using HGE, we could edit out disease-causing mutations as they arise in our genome. This will allow our descendants to enjoy the same level of genetic health as we enjoy today.

Of course, many diseases have a lifestyle element—we have mentioned cardiovascular disease and infertility. Many resist using biological interventions to treat lifestyle problems. For example, it seems absurd to genetically modify human beings to be able to tolerate a diet consisting solely of foods with low nutritional value.

However, as we have argued, there are biological components to many contemporary diseases that are worthy of modification. Moreover, even if such diseases were entirely lifestyle or social in origin, which intervention we ought to choose—modifying the biological, psychological, social or natural environment—depends on the costs and benefits of the particular intervention, and relevant moral values. For example, it may be possible to prevent skin cancer by avoiding exposure to the sun, or by increasing the production of melanin, or by increasing the capacity of our immune cells to attack skin cancers. Which we should choose depends on the context.

A common assumption in environmental ethics is that we have obligations to members of future generations. According to one principle of intergenerational justice, ’existing generations ought not act so as to worsen the position of future generations by depleting non-renewable resources with no compensatory action or recompense’.41

It is clear that the use of modern medicine is worsening the position of members of future generation, by allowing random mutations to occur to our genome. Fortunately, there is a straightforward compensatory action—developing HGE. This is not something that is merely permissible—but a moral imperative.