Bespoke Babies: Genome Editing in Cystic Fibrosis Embryos

INTRODUCTION

Robert M. Sade, MD and Kyle B. Brothers, MD, PhD

Technologies capable of inserting new genetic sequences into mammalian genomes have been available for over a decade¹, but they were not promising for clinical applications owing to their inflexibility and lack of precision. Six years ago, however, clinical genome editing became far more likely. A cellular system that involved the combination of short DNA sequences called clustered regularly interspaced short palindromic repeats (CRISPR) and a family of enzymes called CRISPR-associated proteins (Cas) was found to be substantially more precise than previous technologies, and could easily be modified to insert sequences at virtually any site in the genome, including those of human cells.

Although the CRISPR/Cas system has already revolutionized laboratory science, its use in clinical contexts is controversial because it may create unintended modifications in human cells. Some recent findings have raised concerns that the risk for so-called off-target effects may be higher than originally thought².

Most controversial is the question of whether CRISPR technology should be used to edit the genome of an entire human being. Many of the most devastating genetic conditions permanently disrupt the development of important organ systems, and in order to treat such conditions using genome editing, it would be necessary to modify the genome at a very early stage of embryonic development, resulting in an individual composed entirely of genetically modified cells. This approach is referred to as germline genome editing, emphasizing that any genetic modifications can be passed on to subsequent generations.

The question of whether germline genome editing should be used to replace variants of the cystic fibrosis (CF) gene CFTR to prevent future offspring from suffering from this disease was debated at the 2019 Annual Meeting of the Society of Thoracic Surgeons by two experts in this technology. The debate focused on the following vignette.

Designer Babies via CRISPR editing

Harry Rose is 23 years old and at the age of 2 years was diagnosed with CF. He had a double lung transplant 3 years ago when chronic lung infections led to severely deteriorating pulmonary function. His new lungs are doing quite well, so well in fact that he was married last year to a young woman, Serena, whom he knew when they were both in their teens. She also has CF, but in a much milder form than Harry’s disease.

The young couple is considering starting a family but are very concerned about the certainty that their children will have CF, since they both have the required two variant copies of the autosomal recessive CF gene, CFTR. They have recently read articles in magazines and newspapers about new developments leading to the possibility of editing the human genome to replace abnormal alleles with normal genes in embryos. They especially like the idea that they can prevent their children from getting CF, and, by eliminating the variant gene in their children altogether, prevent generations of their descendants from having to worry about this condition.

After discussing this possibility with their doctors, Harry and Serena realize that the long-term effects of gene editing on future generations of their descendants are not fully understood, and that such procedures are controversial. They ask two experts for their advice about what they should do.

PRO

Kyle B. Brothers, MD, PhD

CRISPR technology should be used to prevent cystic fibrosis.

In the case presented here, a young couple, Harry and Serena, is interested in pursuing germline genome editing so that they can have a child who will not develop CF. CF is a recessive condition, so any child born to a man and woman who both have this condition would have, essentially, a 100% chance of developing CF. In theory, germline genome editing could prevent this, and no other reproductive technology currently in use would allow such a man and woman to have a child who does not have CF and who is also biologically related to both parents. Adoption and gamete donation are potential alternatives, but many couples are strongly motivated to have biologically-related children.

Any honest advice from an expert on germline genome editing would need to highlight to this couple that the safety of germline genome editing is, at this point, completely unknown. In 2017, the American Society of Human Genetics (ASHG) released a policy statement clarifying the unanswered scientific questions regarding safety and efficacy that would need to be answered before CRISPR technology could be used in clinical settings³. Although the laboratory science behind this technology has developed rapidly, clinical research necessarily progresses at a much slower rate. Optimistically, we are at least ten years from knowing whether CRISPR technology is safe enough for clinical use, and in fact the timeline could (and maybe should) progress much more slowly given that it may take years for adverse effects like cancer risk to become apparent.

In November 2018, scientist He Jiankui announced that he had modified the genomes of two embryos who had recently been born as twin girls⁴. The ethical layers of this case are so numerous and convoluted that it would be impossible to explore them in detail in this short discussion. But it will suffice here to observe simply that Dr. He’s actions in this case represent a grievous breach of ethical standards for research with humans. In addition to a number of concerns about the ethical review and consent process for this research, there is simply no evidence to indicate at this point that CRISPR technology is safe enough for clinical trials. The fact that Dr. He attempted such a trial does not mean that it is ethically acceptable to allow this to happen again anytime soon.

However, such a time will come. The 2017 ASHG policy statement provides an extraordinarily helpful outline for the scientific steps that will need to be taken to make germline genome editing a clinically feasible intervention. If existing CRISPR/Cas technologies prove inadequate for this purpose, refinements of this approach are still likely to reach an appropriate threshold for clinical use in the future.

Assuming the technical barriers can and will be addressed, we will need to decide whether the use of this technology is ethically appropriate. Undoubtedly, there are important reasons to exercise caution. Modifying the human genome is not just dangerous from a medical perspective. It also raises profound questions: Is germline gene editing a subversion of natural law? What effects will uses and misuses of this technology have on society, including views on parenting, diversity, and disability? Does this technology threaten to reduce the human gene pool in a way that threatens the ecology of the human species?

Many scientists and other scholars have called for a deliberative process to explore the acceptability of germline genome editing with a broad range of stakeholders^5,6. Such a process is critical and should be undertaken with all due haste. However, the reality is that governments, professional societies, and other policymakers will block the use of a technology only if its harms far outweigh its benefits, or if it is highly unacceptable to stakeholders. This is particularly true for reproductive technologies, where concerns about autonomy and privacy have in the past significantly curtailed government intervention. The fact is there are already strong signs that germline genome editing for the prevention of serious medical conditions will be accepted by professionals. Professional organizations like ASHG and the National Academies of Sciences, Engineering, and Medicine have been careful not to make assumptions about the outcome of the societal deliberation activities they have recommended^1,7. At the same time, it is telling that they have called for deliberation rather than outright condemning the use of CRISPR for therapeutic applications. This implies that they would, in principle, be amenable to this practice in the future if other stakeholders agree.

Members of the public also seem open to the idea. In a 2017 survey of U.S. adults, 65% of respondents found germline genome editing to be an acceptable technology for therapeutic uses⁸. This is nowhere near the level of opposition that would prompt governments or other policymakers to interfere with reproductive decision making.

If therapeutic germline genome editing is acceptable in principle, the question remains whether CF would be an appropriate target. This is not at all a given. Recent advances in CF treatment have significantly improved the life expectancy for children born with this condition. A recent study in Canada, for example, found that the mean life expectancy for an adult with CF is 49.7 years⁹. With this increase in life expectancy, treatment often now focuses on improving patients’ quality of life. Although a number of factors can adversely impact the quality of life for persons living with CF, there is a potential for a relatively normal adult lifestyle¹⁰. CF is by no means the most severe heritable condition that could be prevented by germline genome editing. Sickle cell disease, Tay-Sachs disease, and osteogenesis imperfecta arguably create more suffering for affected children, and thus might seem to be better targets for this technology.

There are important reasons, however, to consider CF a high-priority target for germline genome editing. Although many patients survive into the fourth and fifth decades of life, some do not. Children still die from CF, even when they have access to appropriate medical resources¹¹. CF can also be an extremely burdensome condition. As in the case presented here, many patients need to undergo a lung transplant to survive, and extrapulmonary complications of this condition, such as CF-related diabetes, can create significant morbidity and mortality¹². It is telling that the parents in this case are willing to risk the potential pitfalls of germline genome editing to prevent their child from experiencing the burden of this condition. Clearly, their first-hand experience with CF has convinced them it is a condition worth avoiding, if possible.

The possibility of using germline genome editing to treat CF in a research context is particularly promising given the frequency of this condition. Most heritable conditions are quite rare, so it would likely take many years to accumulate sufficient experience with treating these conditions in a research context. CF, on the other hand, is the most common lethal genetic condition that affects children¹³. Although it is somewhat uncommon for two young adults with CF to decide to have children, the fact that CF is relatively common means that this situation is likely to arise more frequently than most other heritable conditions. This means that CF offers an excellent first application for building experience with germline genome editing in a research context. The experience acquired in this way could then be applied to other conditions that are less common.

Finally, CF is an excellent starting place for germline genome editing because couples with this condition already frequently require fertility treatments. 95% of men with CF experience infertility as a result of CF-related congenital bilateral absence of the vas deferens¹⁴. Most couples with CF will thus need to utilize in vitro fertilization (IVF) to become pregnant. Coincidentally, germline genome editing requires the generation of an embryo in the laboratory using IVF. This strategy is required not only because it allows the CRISPR technology to be applied to the embryo, but also to test one or more cells from the embryo to ensure that there are no detectable off-target effects. Because couples with CF already need to use IVF to become pregnant, they may be more willing to participate in a clinical trial for germline genome editing.

In conclusion, it seems likely that germline genome editing will be used in the future to prevent heritable conditions. Assuming the initial scientific barriers are addressed in animal models, it will be necessary to test this technology in a clinical research setting. CF offers a perfect opportunity for those future first-in-human trials. For the present, Harry and Serena Rose should follow the field carefully. There is a very real opportunity the treatment they seek will be available in a research setting during their reproductive years. They might even be the first couple to blaze this clinical path!

CON

Mary Devereaux, PhD

Germline gene editing should not be used to prevent cystic fibrosis.

Dr. Brothers argues in favor of germline genome editing in certain heritable conditions, CF, in particular. In the case under discussion, we are asked to consider two patients, Harry Rose and his wife Serena. Both have CF. Serena has a mild form of the disease, but Harry’s disease is severe. Chronic lung infections led to his double lung transplant three years ago. Harry is now doing well and the young couple wish to start a family. Unfortunately any biological children born to this couple have a 100% chance of inheriting CF, a condition widely recognized to carry a heavy disease burden. Germline genome editing offers them the possibility of CF-free children. Many would agree with Dr. Brothers recommendation that the couple stay abreast of developments in CRISPR technology in the hope of eventually qualifying for first-in-human clinical trials. There are, however, a number of ethical challenges that weigh against this recommendation.

First, there is little debate that using CRISPR (or other gene editing tools) to eliminate CF is not ready for prime time. The application of this technology to human oocytes, sperm, or embryos requires overcoming significant scientific hurdles. The animal work necessary to establish safety and efficacy, and the review necessary to approve clinical trials, is years if not decades away. There is broad scientific consensus that the human germline should not be scientifically modified at the present time and that rogue efforts to do so should be condemned. Evidence of this concurrence emerged in the near unanimous response to He Jiankui’s unauthorized and unethical embryo research and the resulting birth of Chinese twins in late 2018⁴.

But let us suppose that Dr. Brothers is correct in his optimistic projection that in a decade or so, while Harry and Serena are still of reproductive age, the technical challenges of making genome editing safe and effective will be overcome, thus making human genome editing a reproductive reality. Imagine we can offer Harry and Serena, and patients like them, the chance to have biologically related, CF-free children. The ethical question, of course, is not whether we can offer such intervention, but whether we should. Many argue that the particular heritable condition CF is an appropriate target for pre-implantation editing of embryos. As noted, it is a fairly common, but dreadful, genetic disease and CF males already require sperm aspiration and IVF for reproduction. Why then not use gene editing to eliminate mutations that cause CF, implanting only successfully modified embryos? Doing so improves the quality of life of the children and the family as a whole. It also enables the children themselves to have CF-free descendants. Indeed, one prospect is that over time CF will be eliminated from the human gene pool. Surely eliminating a fatal, highly burdensome, disease is an ethical good. Why hesitate to embrace this new and brave world?

Let’s begin with our couple. It is worth noting that pediatric pulmonology guidelines discourage CF patients from socializing, let alone living in close proximity to each other, due to the risk of serious cross-infection from their different bacterial strains. This particular couple thus risk frequent infections, and the prospect of recurrent hospitalizations. Of course love needn’t follow medical guidelines. But the risks this couple have assumed are not restricted to them. Their decision to have children — biological or adopted — requires taking into account the interests of those children. Both parents have a chronic, progressive, frequently fatal disease. Both have shortened life expectancies. Who will take responsibility for these children if and when their parents can’t? Even with improved drugs, on-going CF management and related symptoms may leave Harry and Serena struggling with parental responsibilities. In short, for this couple, even having children raises ethical questions. This is independent of any discussion of gene editing.

Admittedly, many couples choose to have children in less than ideal situations. Doing so does not — and should not — limit their reproductive liberty. The question before us is rather whether reproductive liberty should extend to altering the genome of future children to eliminate CF. Is ensuring that prospective parents are adequately informed and properly consented enough?

Notice that while it is ethically permissible to attempt to repair Harry’s phenotypic defect, e.g., with CFTR modulators, fixing Harry is not what we’re talking about doing. At issue is altering the genes of his potential children. In short, the parents are not the experimental subjects here. Their embryos are. The clinical research to which they would consent is research aimed at establishing the safety and efficacy of eliminating CF in their embryos. Testing for off-target effects prior to implantation, as Dr. Brothers describes, eliminates certain known risks. But early-stage research cannot guarantee protection against unknown, particularly long-term, health consequences. The point is that, however good the preclinical data, successful early clinical trials will unavoidably subject actual children to risks that cannot entirely be foreseen. Those accepting those risks, however thorough the process of informed consent, are not deciding for themselves. The risks, potential or actual, fall to the children born. It is their genes not those of their parents that technologies such as CRISPR will target. Nor is that modification limited to them (their somatic or body cells). The gene editing under discussion, germline modification, will introduce changes that are passed from one generation to the next. While all parents make decisions for their children, medical and otherwise, germline genome editing is different. It belongs to a unique category of modification: one that is irreversible and extends beyond the individual and the children of the individual. We have hence moved far beyond simply Harry and Serena or even their children.

Seeing that germline genome editing is not simply a matter of individual choice – or reproductive liberty – raises broader social considerations. Research on gene editing offers promise for many diseases. It is, after all, not only CF sufferers who may wish to avoid passing grave medical conditions to their children. Once demonstrably safe and effective, genome editing is unlikely to be confined to CF or even to the most serious of heritable conditions. Many who applaud the implementation of gene editing for devastating medical conditions may reasonably hesitate at the ethics of using this technology to enhance normal human traits such as height, athletic prowess, or mathematical ability. But where does the line get drawn between permissible, “good” uses of gene editing and impermissible or “unwise” uses? This is a slippery slope. Surgical techniques initially developed to restore facial function or appearance in wartime now routinely serve cosmetic ends, narrowing the scope of acceptable or “normal” human appearance. The ethical concern is that once the door to using germline gene editing opens, closing that door may prove difficult or impossible.

Defenders of opening this door reply that these reservations are overwrought or unfounded, like the early reactions to advances such as organ transplants and IVF. Certainly scientific advances will improve safety and diminish risk. But no scientific advance can itself answer the ethical, indeed philosophical, questions before us. Who will have access and for what? Who will decide and on what grounds? Is control over the use of germline gene editing even possible across nations and diverse cultures? And once developed, how will this technology alter our conception of normal reproduction, or impact acceptance of those with “uncorrected” disease or disability? It is not only bioethicists and philosophers who are asking these questions. The scientific community itself has articulated strong concerns about moving forward without careful deliberation. As the 2017 National Academies of Sciences report argues, now is the time to address questions not only about medical risks and benefits, but also how to govern the use of genome editing, what societal values to incorporate in clinical and policy decisions, and how to respect the diversity of values unavoidable in global research and medicine. We may, as some have argued, be crossing “an ethically inviolable line.”¹⁵

In conclusion, the problem with human germline gene editing is not primarily technical. Dr. Brothers is undoubtedly correct that the science will get there sooner or later. But getting the science right is only the first step. Even safe and effective alteration of the CF gene and healthy children for the Rose’s won’t suffice to make germ line editing ethically advisable. What we also need to get right are answers to the many significant philosophical and ethical issues outlined above. Even optimists about our scientific acumen may look in vain for evidence of the kind of public consensus and ethical judgment required to manage how far and in what direction we should go in altering the human genome. Until, and unless, we overcome the ethical hurdles, we should not attempt human germline gene editing.

CONCLUDING REMARKS

Robert M. Sade, MD

Our two essayists start from opposing positions: germline editing should vs should not be used to prevent CF. As their arguments develop, however, we find them in agreement on several important issues. They agree that technical aspects of achieving germline modification will be solved eventually, at least in animal models, and that a careful and thorough deliberative process to consider a broad range of benefits and harms is necessary as we move toward implementing this technology in human beings. The main difference in their viewpoints seems to be that Brothers believes on pragmatic grounds that preventing CF by modifying the CFTR gene makes sense and is going to happen, while Devereaux cautions that until and unless we resolve a number of philosophical and ethical issues, we should not move to embrace human germline modification.

An interesting question directly related to this problem is not addressed by either author: is there an epistemic obligation to undertake the clinical study of germline modification? The potential benefit to the Rose’s children is clear — living a healthy life without CF would be an enormous good — but we know little to nothing of the potential harms, great or small, that might result from such a clinical study. It seems that it would take a great deal of harm to outweigh the enormous good that might result, so do we have an obligation to undertake at a minimum a limited study of the effects of genome manipulation? Once that information is available, ethicists and policy makers could engage in a meaningful consideration of the balance of harms and benefits, which lies at the heart of ethical deliberation. Without the knowledge that could be generated by a limited investigation, both sides are debating from a position of inadequate information.

Many arguments could be adduced for and against the idea of an epistemic obligation to study germline editing, and I suspect our essayists would come out on opposite sides of that discussion. Although this idea is closely related to the present debate, however, our essayists did not raise it, so it remains a discussion for another time.