What do people think about genome editing? Part II

In a previous post, I discussed a survey on genome editing that the Royal Society had commissioned in the UK. The findings of the survey were summarised as “cautious optimism” about genome editing. Three-quarters of Brits seem to approve of genetic modifications that would be passed on to future generations, if the purpose is to prevent a severe disease.

My critique so far is that respondents were left in the dark about how such modifications would be done: they require artificial fertilisation in a dish and would be performed before implanting the embryo into the uterus. This particular procedure could not only seem less attractive than some vague “heritable modification”. In addition, the modification of embryos specifically involves potential risks that need to be made explicit.

In this current post, I am adding to my previous critique that media reporting about the survey has been selective, and examples that illustrated the survey questions for respondents have not been entirely appropriate from a technical and moral point of view.


The modification of embryos: Why things can look different with more information

The potential modification of embryos raises the question of whether there are alternatives, notably to a therapeutic genetic modification of an embryo. With pre-implantation genetic diagnosis (PGD), another alternative to genome editing exists, at least for many conditions that are involved here. PGD is a far-reaching intervention, but it gives less reason for concern than the genetic modification of embryos. So it would have been enlightening to ask if respondents find PGD a sufficient replacement of heritable genetic modifications. By contrast, for some of the questions that the survey did ask, it did not provide sufficient information about heritable changes.

Sebastiaan ter Burg,  “Workshop Gamestorming”, at flickr.com

In addition, news media reports (including the release by the Royal Society) did not mention exploratory workshops which the survey agency, Hopkins Van Mill, held in parallel to the survey. These workshops provided participants with in-depth information and encouraged discussion. Understandably, participants noted time and again how little they knew before the workshop, which highlights the need for detailed information. The workshops were not intended to document quantifiable, representative views, but they brought up interesting perspectives nonetheless. The survey agency mentions in the original report that workshop attendees had a notable concern about heritable genetic changes in humans. As these changes would be done in an embryo, there would be no way to offer a choice to the person directly concerned. This problem is not the same as the question of the management of potential accidents in future generations (the precautionary principle). Even if there were practically no risk of accidents or unwanted side-effects, informed consent remains an important issue. This is important notably as genetic changes in embryos raise the question of “improvements” that go beyond medical concerns – increased intelligence, blue eyes, etc. Such changes in an embryo would necessarily reflect the particular values and choices of someone else than the embryo. Yet workshop participants were concerned not only about enhancements, but even whether future generations would agree, in retrospect, with medical interventions done at the embryonic stage.

The potential of genome editing: The importance of realistic examples

A final point of contention concerns the choice of examples with which the survey illustrates interventions so that respondents know what they are responding to. As mentioned above, with the example of improved intelligence, the survey tended to oversell genome editing. Two further examples make good sense, by contrast: a particular heart condition, hypertrophic cardiomyopathy, and cystic fibrosis suitably illustrate life-threatening diseases that are currently incurable. If indeed genome editing were used to prevent these diseases, the required changes would probably be heritable. Both diseases are caused by identifiable genetic factors, and although there is yet no workable genome editing procedure for these cases, it is conceivable scientifically. Yet how about diseases that are not life-threatening – would respondents still approve of genome editing, apart from the question if changes would be heritable? The survey illustrates this question with arthritis, and the result is that 73% of respondents are in favour of genome editing to tackle such a disease.

Rheumatoid arthritis: National Library Of Medicine US at https://ghr.nlm.nih.gov/condition/rheumatoid-arthritis

The choice of arthritis as an illustration of diseases genome editing can treat is not necessarily helpful. On the whole, arthritis is fairly common in the west, where it is a well-known, if not particularly well understood ailment. Arthritis causes significant pain in the joints and reduces mobility, especially among women aged 50+. Many people will have a distant aunt, a grandmother or a parent with this condition. So to hold out a vague hope of a cure is likely to resonate strongly among respondents.

Things are more complicated, however, from a medical perspective. The term arthritis is a collection of about 100 different conditions, each with their own individual cause. Among them only few are clearly and identifiably associated with certain genetic factors. However, while certain genes can increase the susceptibility to particular kinds of arthritis, we cannot say that “arthritis” is “caused” by genes – the picture is more complicated. Even if genes are in some way involved, other factors like the lifestyle play a significant role.

At the risk of sounding pedantic, it would be better to ask how people think about taking precautionary measures that would make a certain kind of arthritis less likely or ameliorate potential symptoms in advance, even before a person would have any symptoms. By contrast, the survey asked how people think about “treating” a disease like arthritis. This suggests that people who have arthritis could themselves profit from such treatment, but that is unrealistic.

Nevertheless, even taking precautionary measures could be helpful. One case where certain mutations do play a tangible role is the susceptibility to rheumatoid arthritis. This is a common form of arthritis that affects about 400,000 people in the UK. However, by far the most common form of arthritis is osteoarthritis, with 8 million Britons affecting about 20 times as many people as rheumatoid arthritis. Here genes also play a role, which is the focus of ongoing research. It looks like several genes contribute to the condition: the chances for someone with a whole bunch of particular genes to develop osteoarthritis can be 50% higher than for someone without those factors. By far not everyone will have most of the known genetic candidates, and nobody will have all of them that, taken together, lower the risk by 50%. Moreover, we do not know which genes contribute how much to susceptibility. So if we were to reduce the susceptibility to arthritis, it is difficult to say at what level we are starting and at what level we are ending up. Is that worth choosing artificial fertilisation and a genetic modification of an embryo, rather than having a child through sexual intercourse? Who knows. At any rate, promises about full-blown arthritis prevention, or treatment, need to be taken with a larger pinch of salt.

So while genome editing could possibly make a difference for future generations with a higher susceptibility, genome editing is not going to cure arthritis. In addition, rarer conditions like lupus or gout also cause symptoms that are called arthritis, and genes can contribute to these conditions as well. Yet again, these genes make only a contribution to a more complicated picture, and then knowledge about what role which genes really play is often patchy. Often symptoms can be prevented by lifestyle changes (in the case of gout) or treated with medication (lupus).

heart_care_medical_care_heart_health_medicine_symbol_health_care-843297.jpg!dClearly a genetic cure of arthritis would be desirable, yet the idea of a genome editing “cure of arthritis” is unrealistic: it would not be a cure, and there is no single thing as arthritis. Perhaps future research will be able to help prevent arthritis to a certain degree, but the survey suggests a golden bullet that no future research will be able to provide. In doing so, it peddles with the issue of genetic determinism, as if arthritis were due to an arthritis-gene” that we could eliminate. In one of the workshops held in conjunction with the survey, one group even produced a poster about genome editing, entitled “to eradicate all disease” – as if all diseases had one clear genetic cause, perhaps even including infectious diseases and mental illness! In sum, the question whether people think positively of genome editing, provided it could cure arthritis, is a leading, unhelpful question. Of course at this point most genome editing therapies involve some speculation and hypothesising. Yet we must be more careful not to create unrealistic expectations.


Having a representative survey that shows us what people really think about genome editing would be a good thing. The current survey, commissioned by the Royal Society, contains many interesting points in this regard. However, there are problems with the information the survey agency provided to respondents and with the examples with which they illustrate moral questions. If there were a technology that improved intelligence and cured arthritis, who would criticise it? However, as far as realistic prospects even beyond the near future are concerned, that technology is not genome editing. In addition, what a survey does is it not only tells us what people think who, as a group, are considered representative. It also tells others what views are acceptable in society, what the moral climate is like, and perhaps even what they themselves may in turn be expected to think. Here the reporting about the survey is just as important as the survey itself. Unfortunately, the news media has not at all picked up on the in-depths workshops that the survey agency held, with all the moral questions they raised, and some journalists even got the survey numbers wrong.



Hopkins Van Mill/Anita van Mil, Henrietta Hopkins, and Suzannah Kinsella, “Potential uses for genetic technologies: dialogue and engagement research conducted on behalf of the Royal Society”  (accessed 19 April 2018).

Hopkins Van Mill/Anita van Mil, Henrietta Hopkins, and Suzannah Kinsella, “Appendices to HVM report” (accessed 19 April 2018).

Kathryn Ashe, “UK public in favour of genome editing human embryos for disease”, BioNews 941, 12 March 2018 (accessed 19 April 2018).

The Royal Society, “UK public cautiously optimistic about genetic technologies”, 7 March 2018 (accessed 19 April 2018).

OMIM, “#180300 Rheumatioid Arthritis, RA” (accessed 19 April 2018).

NHS Choices, “Arthritis” (accessed 19 April 2018).