2019/01/29 by Julian Savulescu, Peter Singer · 1 citation
Biochemistry, Genetics and Molecular Biology · Medicine · #Biomedical Ethics and Regulation #CRISPR and Genetic Engineering #Ethics in Clinical Research
paper · pdf · doi:10.1111/bioe.12570
openalex publication_date 2019/01/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/29
Ethics is the study of what we ought to do; science is the study of how the world works. Ethics is essential to scientific research in defining the concepts we use (such as the concept of ‘medical need’), deciding which questions are worth addressing, and what we may do to sentient beings in research. The central importance of ethics to science is exquisitely illustrated by the recent gene editing of two healthy embryos by the Chinese biophysicist He Jiankui, resulting in the birth of baby girls born this month, Lulu and Nana. A second pregnancy is underway with a different couple. To make the babies resistant to human immunodeficiency virus (HIV), He edited out a gene (CCR5) that produces a protein which allows HIV to enter cells. One girl has both copies of the gene modified (and may be resistant to HIV), while the other has only one (making her still susceptible to HIV).1 He Jiankui invited couples to take part in this experiment where the father was HIV positive and the mother HIV negative. He offered free in vitro fertilization (IVF) with sperm washing to avoid transmission of HIV. He also offered medical insurance, expenses and treatment capped at 280,000 RMB/CNY, equivalent to around 40,000. The package includes health insurance for the baby for an unspecified period. Medical expenses and compensation arising from any harm caused by the research were capped at 50,000 RMB/CNY (7000 USD).2 He says this was from his own pocket.3 Although the parents were offered the choice of having either gene-edited or -unedited embryos transferred, it is not clear whether they understood that editing was not necessary to protect their child from HIV, nor what pressure they felt under. There has been valid criticism of the process of obtaining informed consent.4 The information was complex and probably unintelligible to lay people. The most basic ethical constraint on research involving humans is that it should not expose participants to unreasonable risk.5 Risks should be the minimum necessary to answer the scientific question, and the expected benefits should be proportionate to expected harms.6 While the Declaration of Helsinki states that research involving incompetent participants must be ‘minimal risk and minimal burden’,7 this is most plausibly interpreted as minimal overall risk or burden, where expected benefits match or outweigh expected harms. After all, children are exposed to trials of new toxic chemotherapeutic agents with significant risks. In deciding whether a risk is reasonable, it is important to evaluate not only the probability of achieving a benefit, but also the extent of the benefit in question. A greater expected benefit is worth greater risk than a smaller expected benefit. Avoiding HIV is certainly a benefit, but the probability that Lulu and Nana would have contracted HIV is low. In contrast, the unknown effects of the editing could cost them a normal life. Given our ignorance of the full ramifications of changing a gene, what could justify taking the risk of a gene-editing trial in humans? The answer is, if the embryo had a catastrophic single gene disorder. Several genetic disorders, such as BRAT 1, JAM3 and PHGDH, are lethal in the neonatal period, so for embryos with them, gene editing is potentially life-saving.8 There is a risk of off-target mutations, but the expected harm of such mutations is arguably no worse than the fate of the unedited embryos. The geneticist George Church has defended He's research on the grounds that HIV is a public health problem for which there is no cure or vaccine.9 Church is right to the extent that there is no problem in principle with editing out the CCR5 gene in the future. What he fails to take into account, however, is that Lulu and Nana are being used, at great risk, and without proportionate benefit, when there are more ethical experimental designs that would meet the need for greater knowledge of the effects of editing genes. At the conference at which He presented his experiment, George Daley, the Dean of Harvard Medical School, indicated that Huntington's disease or Tay–Sachs disease might be suitable targets for gene editing.10 It is not clear whether Daley is endorsing these as first-in-human trials. Huntington's disease is very different to Tay–Sachs disease. Babies with Tay–Sachs disease die in the first few years of life; people with Huntington's disease have around 40 good years. Hence Tay–Sachs disease is a better candidate for early trials, as babies with that condition have less to lose. This mirrors the rationale for experimenting with gene therapy on babies with a lethal form of ornithine transcarbamylase (OTC) deficiency rather than on adults with a mild form, such as Jesse Gelsinger who lost his life in a badly designed gene therapy trial in 1999.11 He Jiankui's trial was unethical, not because it involved gene editing, but because it failed to conform to the basic values and principles that govern all research involving human participants. Further into the future, if gene editing can be done without off-target mutations, it could be used to address genetic dispositions to common diseases, such as diabetes or cardiovascular disease. These involve tens or hundreds of genes. In principle, gene editing could be used to modify many genes accurately. Gene editing has been successfully employed to remove 62 porcine endogenous retroviruses from a kidney cell line.12 It is notable that the first human gene-edited babies were enhanced to have resistance to a disease, not to treat an existing disease. In future, perhaps gene editing will be used to engineer super-resistance to infectious threats. At the Second International Summit on Human Genome Editing, where He revealed his research, the National Academies of Science, Engineering and Medicine called for a ‘translational pathway to human germ line gene editing’. In our view, to be ethically justifiable, such a ‘translational pathway’ should be: catastrophic single gene disorders (like Tay–Sachs disease), then severe single gene disorders (like Huntington's disease), then reduction in the genetic contribution to common diseases (like diabetes and cardiovascular disease), then enhanced immunity and perhaps even delaying ageing. Should the translational pathway extend to enhancing normal traits, such as intelligence? This has been the subject of almost 20 years of debate.13 One approach to enhancement has been to ban it. Many jurisdictions, including most in Europe and Australia ban pre-implantation genetic diagnosis (PGD) for non-disease traits. However, one U.S. company recently announced the use of polygenic risk scores for low normal intelligence in PGD. They admitted the same techniques could be used to predict high normal intelligence and believe such a step is inevitable.14 Further into the future, gene editing could be used for enhancement of the genetic contribution to general intelligence. China is currently funding research that is trying to unravel the genetics of high intelligence.15 Perhaps the best we can hope for is harm reduction and a regulated market to make important enhancements, such as resistance to disease or the enhancement of intelligence (should it ever be possible), part of a basic healthcare plan so that the benefits of gene editing are distributed equally. Julian Savulescu, Uehiro Chair in Practical Ethics, Director, Oxford Uehiro Centre for Practical Ethics, University of Oxford, Visiting Professor in Biomedical Ethics, Murdoch Children's Research Institute Peter Singer, AC, Ira W. DeCamp Professor of Bioethics, University Center for Human Values, Princeton University, Laureate Professor, School of Historical and Philosophical Studies, University of Melbourne