For prospective parents known to be at risk of passing on a serious genetic disease to their children, heritable genome editing1 may offer a potential means of having genetically related children who are not affected by that disease—a desire shared by many such parents (e.g., Chan et al., 2016; Quinn et al., 2010). Thousands of genetically inherited diseases are caused by mutations in single genes.2 While individually, many of these genetically inherited diseases are rare, collectively they affect a sizable fraction of the population (about 5-7 percent). The emotional, financial, and other burdens on individual families that result from transmission of such serious genetic disease can be considerable, and for some families could potentially be alleviated by heritable editing. Recent advances in the development of genome-editing techniques have made it realistic to contemplate the eventual feasibility of applying these techniques to the human germline. As discussed elsewhere in this report, improvements in genome-editing techniques are driving increases in the efficiency and accuracy of genome editing while also decreasing the risk of off-target events. Because
1 “Germline editing” refers to all manipulations of germline cells (primordial germ cells [PGCs], gamete progenitors, gametes, zygotes, and embryos) (see Chapter 3). “Heritable genome editing,” a form of germline editing that includes transfer of edited material for gestation with the intent to generate a new human being possessing the potential to transmit the “edit” to future generations, is discussed in this chapter. The distinction turns on intent rather than on the technological intervention, which is highly similar in both cases.
germline genome edits would be heritable, however, their effects could be multigenerational. As a result, both the potential benefits and the potential harms could be multiplied. In addition, the notion of intentional germline genetic alteration has occasioned significant debate about the wisdom and appropriateness of this form of human intervention and speculation about possible cultural effects of the technology. As discussed below, these include concerns about diminishing the dignity of humans and respect for their variety, failing to appreciate the importance of the natural world, and a lack of humility about our wisdom and powers of control when altering that world or the people within it (Skerrett, 2015).
A similar debate is already under way regarding a related set of techniques—mitochondrial replacement—in which genetic disease carried by mitochondrial DNA is avoided by using healthy mitochondria from a donor. Because mitochondria in the egg are passed down maternally
through the generations, the effect of these techniques is to make a heritable genetic change, albeit one that does not change the DNA in the nucleus. Mitochondrial replacement has been used in Mexico (Hamzelou, 2016) and Ukraine (Coghlan, 2016) and has been authorized although not yet used in the United Kingdom (HFEA, 2016a,b). A recent National Academies study led to the recommendation that mitochondrial replacement be permitted to proceed to clinical trials in the United States provided it is subject to strict criteria and oversight (NASEM, 2016e) (see Box 5-1). In early 2017, there were reports of a child born after use of the technique in the Ukraine, in this case for an infertility-related condition that would not have met the criteria laid out by either the Human Fertilisation and Embryology Authority (HFEA) or the National Academies report (Coghlan, 2017).
This chapter begins by reviewing potential applications of and alternatives to heritable genome editing. It then describes in turn scientific and technical issues, ethical and social issues, and potential risks associated with these applications. The chapter then turns to the regulation of heritable genome editing. The final section presents conclusions and a recommendation.
Preventing Transmission of Inherited Genetic Disease
Opinions differ as to whether heritable genome editing should be used to prevent the transmission of inherited genetic diseases. Heritable genome editing is not the only way to accomplish this goal. Other options include deciding not to have children; adopting a baby; or using donated embryos, eggs, or sperm. These options, however, do not allow both parents to have a genetic connection to their children, which is of great importance to many people. Alternatively, in vitro fertilization (IVF) with preimplantation genetic diagnosis (PGD) of the embryos can be used to identify affected embryos so that parents can choose to implant only those embryos that are free of the diagnosed mutation. This option is not without potential risks and costs, however, and it also involves discarding affected embryos, which some find unacceptable. One can also avoid transmitting genetic mutations to the next generation by using prenatal genetic diagnosis of the fetus followed by selective abortion of affected fetuses. But as with PGD, some people find it unacceptable to terminate an ongoing pregnancy regardless of the predicted health of the future child.
In these situations, for those who are aware they are at risk of passing on such a mutation, the use of heritable genome editing offers a potential avenue to having genetically related children who are free of the mutation of concern. This form of editing could be done either in gametes (eggs, sperm), in gamete precursors, or in early embryos, but it is important to
note that IVF procedures would be required to generate embryos for subsequent genomic modification. In most cases, PGD could be used to identify unaffected embryos to implant.
However, there are some situations where all or a majority of embryos will be affected, rendering PGD difficult or impossible. For example, dominant late-onset genetic diseases, such as Huntington’s disease, can occur at high enough frequency in some isolated populations that one parent will be homozygous for the mutation. In such situations, all embryos would carry the dominant disease-causing allele that would cause the disease in the children, so PGD is not useful. In other populations, the frequency of particular disease-causing mutations may be high enough that there is a significant chance that both prospective parents will be carriers of mutations in the same gene. Examples include the tumor suppressor genes BRCA1 and BRCA2, which increase the risk of breast and ovarian cancer even when inherited in a single copy (because of loss of the unaffected copy of the gene) and Tay-Sachs disease and other early-onset lysosomal storage diseases that are caused by the inheritance of two copies of recessive mutations. In these situations, only one in four embryos would be free of a disease-causing mutation. Those unaffected embryos could be identified by PGD, but the number of embryos potentially available for implantation would be significantly reduced. There are also examples of diseases that are caused by pairing of two different mutations in a given gene, known as “co-dominance,” and combinations of specific alleles of two or more genes, in which PGD becomes more difficult.
As the survival of people with severe recessive diseases like cystic fibrosis, sickle-cell anemia, thalassemia, and lysosomal storage diseases improves with advances in medical treatments, the possibility cannot be dismissed that there will be an increase in the number of situations in which both prospective parents are homozygous for a mutation. The societal and medical pressures faced by these people often bring them together in social groups where they are more likely to interact and develop close relationships. Similar associations can develop among patients with autosomal dominant genetic diseases that allow development to reproductive age (e.g., achondroplasia, osteogenesis imperfecta), again increasing the likelihood of transmitting disease alleles. As our ability to treat children and adults with serious genetic diseases improves through both conventional and somatic genome-editing therapies, there may be a growing need to address concerns potential parents might have about passing along these diseases to their children. Such situations may well increase interest among carriers and affected individuals in using genome-editing techniques to avoid passing on deleterious genes to their children and subsequent generations.
There can be an additional problem in the case of mutations that compromise fertility, which is the case for women who carry mutations—
such as Fragile X, BRCA-1 (de la Noval, 2016; Oktay et al., 2015), and others—that cause the loss of oocytes during development or postnatally. Beyond inherited mutations, external factors like cancer treatments and environmental chemicals can also reduce ovarian reserve in women who wish to avoid transmission of a genetic disease. In these cases, women have fewer embryos available to screen from each cycle of superovulation, and the chance of establishing a pregnancy with an unaffected embryo (via IVF and PGD) is lower than it is for women without these mutations. As a result, affected women might require multiple superovulation cycles, with their attendant risks, discomforts, and costs, to identify an unaffected embryo.
In all of these situations, if it were safe and efficient to use heritable genome editing (e.g., in gamete progenitors) to correct the mutation, this alternative might be preferred by prospective parents who otherwise would be considering PGD. The number of people in situations like those outlined above might be small, but the concerns of people facing these difficult choices are real.
Treating Diseases That Affect Multiple Tissues
Some genetically inherited diseases affect specific cell types or tissues, such as particular types of blood cells. These diseases can be treated by somatic genome editing, and indeed, some of these treatments are already being used (see Chapter 4). However, somatic genome editing is less well suited to treating other genetic diseases that affect multiple tissues because it may be unable to target all aspects of the disease or may have difficulty reaching a sufficient number of cells in the affected tissues to ameliorate symptoms. Examples of conditions for which somatic genome editing is already being investigated but may not be fully effective include cystic fibrosis, which affects multiple epithelial tissues (tissues of the lungs, gut, and other organs), and muscular dystrophies, which can affect multiple muscle types, including heart muscle, as well as other tissues, such as brain. Couples with such diseases who want to have genetically related children might be future candidates for heritable genome editing because editing the defective gene in the germline could have therapeutic benefit in all tissues.
Duchenne muscular dystrophy (DMD) is an instructive example of the challenges faced by the application of germline genome editing. DMD is an X-linked disease that affects about 1 in 3,600 male births. Symptoms begin to appear within the first few years after birth and progressively worsen. The average life expectancy for a person with DMD is about 25 years. DMD is caused by mutations in one of the largest genes in the genome, dystrophin, which contains multiple repeating similar segments. Both the size of the dystrophin gene and its repeats predispose it to mutation, making this genetic disease relatively common. Somatic genome-editing approaches
are already being developed to remove the deleterious alterations in the dystrophin gene in muscle precursors. Such somatic genome-editing approaches will ameliorate the condition but are not expected to correct the symptoms in all tissues.
Once those somatic editing approaches have been tested clinically, one might imagine trying to use germline editing to correct the defect in all tissues. However, women who know they are carriers for a DMD mutation could use PGD to avoid having an affected child. Furthermore, one-third of DMD cases are due to de novo mutations, which would not be recognized until after birth and thus are not amenable to germline genome editing. Somatic editing approaches currently appear to be more useful than germline editing for this disease. However, the pace of advances in genome-editing methods and stem cell biology may alter that situation.
Practically speaking, considerable technical difficulties remain to be overcome in applying genome editing to zygotes and early embryos. Although the efficiency and accuracy of targeting can be extremely high, and there are sound reasons for believing that off-target effects can be greatly reduced (see Chapter 3 and Appendix A), there still would be a need to ensure that only embryos with correctly targeted alleles would be returned to the uterus to complete pregnancy.
If genome editing were performed in a zygote (fertilized egg) or an early embryo, there would be a significant chance that some of the cells in the resulting early embryo would not have the desired (or even any) edits. This situation is called “mosaicism,” and it presents a significant challenge to the application of heritable genome editing in zygotes or embryos. Screening of an edited embryo by PGD to test for mosaicism would not ensure correct editing of the implanted embryo because a single cell may not reflect the genotype of the other cells of the embryo, and removal of multiple cells for testing would destroy the embryo.
The impact of mosaicism depends to some extent on the gene being targeted. Mosaicism is a serious problem if the gene of interest encodes a required cellular function, but if the gene encodes a secreted factor (e.g., growth hormone or erythropoietin), or leads to the secretion of a required molecule (such as a blood clotting factor), then correcting the gene in only a subset of cells may be sufficient. Furthermore, because the germline in the resulting child may also be mosaic, editing only a subset of cells may not solve the problem for succeeding generations. But it may offer a better
chance of finding a disease-free embryo after PGD, or allow culture and selection of edited spermatogonial stem cells (see section on potential alternative routes to heritable edits below), thereby enabling those children to have unaffected offspring. Overall, the issue of mosaicism would currently present a serious impediment to the clinical application of heritable genome editing in zygotes or early embryos, although recent progress suggests that this impediment may eventually become surmountable (Hashimoto et al., 2016).
Potential Alternative Routes to Heritable Edits
Editing the embryo genome is not the only potential way to achieve heritable genome modification. Approaches that directly modify the genomes of the gametes (eggs and sperm) or their precursors before fertilization could overcome problems of mosaicism and would potentially allow preselection of appropriately targeted gametes before in vitro fertilization.
There are a number of potential routes to gamete genome editing, some of which are already in use in mice and others of which remain to be fully developed. For example, spermatogonial stem cells (which will give rise to sperm) could be isolated by biopsy from testes, edited in culture, tested for correct gene editing, and then reimplanted into the testes. Alternatively, generating sperm or oocyte progenitors via induced pluripotent stem (iPS) cells would allow genome editing to occur in the stem cell population. Correctly targeted clones could be identified and used to generate spermatids or perhaps sperm, either in vitro or in vivo, and used to fertilize donor eggs. Significant progress on such technologies is being made in mice and other mammals, including nonhuman primates (Hermann et al., 2012; Hikabe et al., 2016; Shetty et al., 2013; Zhou et al., 2016). A future in which this kind of approach could be extended to ensure precise and effective correction of a disease-causing variant in gametes is not unrealistic (see further discussion in Chapter 3 and Appendix A). The future prospects for heritable genome editing in humans will change dramatically if genome editing in progenitors of human eggs and sperm becomes a reality.
Effect on the Human Gene Pool
Another consideration is that some genes that cause serious genetic diseases, like sickle-cell anemia, have been subject to positive selection to maintain the disease-causing allele in the population because it produces some protection against infectious disease when present in one copy (heterozygous) (see Chapter 4). The same might be true for some other disease-causing variants, and there is some evidence suggesting that might be the case for cystic fibrosis, although this has not yet been established
(Poolman and Galvani, 2007). Such examples have led some to question whether heritable genome editing to remove disease-causing variant genes might significantly alter the human gene pool. As discussed earlier, the numbers of cases of human germline editing to treat disease, if it were to be approved, would be very small, and there is little chance of any significant effects on the gene pool in the foreseeable future. It has also been proposed that any heritable genome editing should be restricted to making changes that occur naturally in the human population (i.e., converting deleterious disease-causing variant [mutant] alleles to a common nonpathogenic DNA sequence) to minimize the risk of unexpected effects of the modification in generations to come (see also Chapter 6 for discussion of genome editing for enhancement purposes). Changing a disease-causing mutation to a known existing nonpathogenic sequence would be the case in any currently envisioned therapeutic applications, and thus the effect of any such heritable genome-editing changes for therapeutic purposes is expected to have minimal effect on the human gene pool.
Ability to Select Appropriate Gene Targets
Finally, the issue arises of whether current knowledge of human genes, genomes, and genetic variation and the interactions between genes and the environment is sufficient to enable heritable genome editing to be performed safely. While current knowledge is arguably sufficient for some genes, in many cases it is not. There is uncertainty, for example, about why the APOE4 allele, which clearly correlates with increased risk of Alzheimer’s disease, is present in the human gene pool at such a high frequency. One theory is that it may confer an advantage in other respects, similar to the heterozygous advantage of sickle-cell mutations that confer protection against malaria (see Chapter 4). A gene such as APOE4 would not be a good candidate for heritable genome editing because it may confer some protection against liver damage by hepatitis C infection (Kuhlman et al., 2010; Wozniak et al., 2002), and also the fact that its deleterious effects are not fully penetrant. Knowledge of genome–environment interactions will improve over time as large-scale projects linking genomic sequences with details of health, environment, and lifestyle are carried out—such as the 100,000 Genomes Project in the United Kingdom and the Precision Medicine Initiative in the United States. As understanding of the genome progresses and genome-editing/stem cell technologies improve, future possibilities for editing the heritable germline to improve human health and well-being will need to be the subject of ongoing, careful consideration. Each potential target gene would need to be evaluated carefully on both scientific and ethical grounds, and only well-understood genes would be suitable candidates for heritable genome editing.
Nearly half a century ago, Bernard Davis published an essay that presciently outlined discussions still under way today about the promise, the risks, and the roadmaps for genetic research, including research on making heritable changes in the germline (Davis, 1970). He began his article with a call “to assess objectively the prospects for modifying the pattern of genes of a human being by various means” (p. 1279) and continued with a caution: that one must keep in mind that “the most interesting human traits—relating to intelligence, temperament, and physical structure—are highly polygenic” (p. 1280) and therefore depend upon large numbers of genes interacting in complex ways with the environment. This is still true, but more is being learned every year about the genetic regulatory circuits that control complex traits, and there is an ongoing need to consider the potential benefits and risks of heritable genome editing.
Balancing Individual-Level Benefits and Societal-Level Risks
One of the challenging characteristics of debates concerning heritable genome editing is that they require balancing possible benefits that accrue primarily to individuals (such as prospective parents and children) against not only risks to the individuals, but also against possible harms at a social and cultural level. This is a complicated ethical analysis, in no small part because the individual benefits and risks are more immediate and concrete, whereas concerns about social and cultural effects are necessarily more diffuse. In addition, although examination of past technological innovations can help in making predictions about social and cultural changes, these predictions remain necessarily speculative because any such changes resulting from a new technology take time to develop. Thus, the ethical debates become difficult because the arguments can fail to engage each other directly.
In the United States, appropriate consideration of social and cultural concerns is usually resolved within the context of civil rights jurisprudence and legal decisions, which compare the burdens on individual liberties or the discriminatory impact of those burdens to whether there is a rational or compelling need for these particular state restrictions. In these cases, the outcomes are often determined less by the substantive arguments for and against the technology or an individual’s choice and more by the level of justification required to uphold the restrictions. When ordinary liberties are restricted, a mere rational basis for the governmental restrictions will be upheld by the courts. If the liberties involved are fundamental, such as those specifically identified in the Bill of Rights or otherwise deemed fundamental by the courts, then a much more compelling and well-crafted justification
for the restrictions is required. The contours of the latter category can be uncertain, however, because of enduring debates surrounding the methodology and legitimacy of judicial determinations that some rights are fundamental despite their absence from the Bill of Rights. Procreative rights fall within this disputed area, making it more difficult to predict which level of justification will be needed should there be a challenge to government restrictions on one or more aspects of procreative activities (Murray and Luker, 2015).
The possible benefits of heritable genome editing accrue most immediately to individuals: the prospective parents who want to have an unaffected genetically related child (and that child) but fear passing along a disease. The desire for genetic relation is evidenced by the fact that many prospective parents, faced with the choice between foregoing genetically related children or risking the birth of a child with a genetic illness, will choose to risk having an affected child (Decruyenaere et al., 2007; Dudding et al., 2000; Krukenberg et al., 2013). If an edit is made in a gene that is well understood and the change is a conversion to a known, common, nonpathological sequence, heritable genome editing may represent an option that is at times more effective or acceptable than PGD. It would offer benefits to the parents and allow for the birth of a child who would enjoy better health. In the case of some disorders that are lethal at a young age, it would allow for the birth of a child with the prospect of a more ordinary lifespan.
Access to heritable genome editing would be consistent with the broadest legal and cultural interpretations of parental autonomy rights in the United States. The desire to have genetically related children may arise from a variety of factors, ranging from a wish to see one’s self or one’s ancestors reflected in the appearance of the children to a belief in the need for a biological linkage in order to satisfy a sense of lineage, continuity, or even some form of immortality (Rulli, 2014). Precluding access to this technology could be regarded as limiting parental autonomy, depending upon the country and the culture. Indeed, some people feel they have a religious or historical mandate to have genetically related children. There are others who do not share this view of parental autonomy, and see germline editing as a step toward seeing children as constructed products and an increasing intolerance of their inevitable imperfections and failures to live up to parental expectations (Sandel, 2004). And some would argue that satisfying the desire for genetically related children is not an unalloyed benefit, as it can be seen as reifying what some view as outdated notions of kinship and family precisely at a time when adoption, same-sex marriage, donor gametes, surrogacy, and stepparenting are being normalized (Franklin, 2013).
In the United States, procreative liberty is grounded in legal cases that relate to the right to have children at the time one wishes, and with considerable latitude in rearing practices.3 Relevant cases focus on a right to rear children and shape their characters largely to fit parental preferences, the right not to have the state involuntarily sterilize persons, the right to use contraception to avoid conceiving, the right to control one’s body even if it entails terminating a previable pregnancy, and the right to preserve one’s health even if it entails terminating a viable pregnancy. In a related case concerning statutory interpretation of the Americans with Disabilities Act, the U.S. Supreme Court acknowledged that procreation is a major life activity.4 The broad view of these cases would include methods of achieving pregnancy and a right to use the same technologies to reduce the risk of disease and disability in those children.
However, the constitutional law cases in the United States do not directly address a right to destroy an ex utero embryo, nor do they address PGD or the legality of IVF. The expansive view above remains untested with respect to how broadly it construes procreative liberty, and related cases on parental rights and reproduction do not clearly support this interpretation (Nelson, 2013). Procreative liberty—like all liberties—can be viewed either as a negative right that protects parties from governmental prohibitions or as a positive right that obligates government to facilitate choices or provide services. In the United States, a negative-right analysis of procreation protects parents from government prohibitions on key aspects of reproductive choice (such as use of contraceptives) and parental discretion (such as choice of language of educational instruction). But reasonable regulation of a technology for the protection of the health and well-being of those affected is entirely permissible, even when claims of constitutional rights can be made. And the concept of procreative liberty has never been extended to a positive right to demand that government fund or even approve new reproductive technologies.
Balanced against the possible benefits of heritable genome editing are a variety of potential risks. As discussed earlier in the chapter, genome-editing technologies in their current state still face technical challenges that would need to be overcome before they could be applied to editing the human
3Meyer v. Nebraska, 262 U.S. 390 (1923); Pierce v. Society of Sisters, 268 U.S. 510 (1925); Farrington v. Tokushige, 273 U.S. 284 (1927); Prince v. Massachusetts, 321 U.S. 158 (1944); Wisconsin v. Yoder, 406 U.S. 205 (1972).
4Bragdon v. Abbott, 524 U.S. 624 (1998).
germline, necessitating caution and careful review of any proposals to proceed to clinical trials.
One concern is that human intervention may have unintended consequences. In the case of heritable genome editing, this concern has two distinct components. The first is the possibility of off-target effects of the editing process, as in the case of somatic genome editing. As discussed elsewhere in this report (in detail in Appendix A but also in other chapters), this technical question is receiving a great deal of scientific attention. Although improvements in genome-editing technology are reducing the incidence of off-target events, and methods for assessing their rate, some approaches already approved for somatic therapies, are being developed, they have not yet reached the point at which clinical trials of heritable genome editing could be authorized. Before any such clinical trials are approved, it will be necessary to demonstrate that the editing procedures will not lead to any significant increase in unintended variants. The required level of such variants will be necessarily lower than for somatic genome editing, but given that heritable genome editing may not go forward for some time, the technology for minimizing and assessing off-target events will undoubtedly have improved significantly. In addition, experience with somatic genome editing will have refined understanding of what might be considered acceptable or unacceptable rates of unintended variants.
Whatever standards are developed for somatic applications, there will be less tolerance for off-target effects in germline applications. By definition, those affected by the edits (future offspring) did not make the decision to be subjects of research or attempts at therapy, and adverse effects might be multiplied by reverberation across generations. Both factors lead to a more conservative approach to the risk/benefit balance.
A second concern is that the intended genome edits themselves might have unintended consequences, even in the absence of off-target effects. In the case of heritable genome editing to convert a well-understood disease-causing variant gene to a widely occurring nonpathological variant, the editing change would be to a version of the gene that is known not to have deleterious consequences. These variants are broadly present in the population already, so the chances that the edit would have some unexpected effects are small. On the other hand, the question of unintended consequences of a targeted edit would arise in the context of edits performed to make a change to a DNA sequence that is not already prevalent in the population, as would be the case for some so-called enhancements. In germline editing, the concern is magnified because the alteration could affect descendants, as discussed further in Chapter 6.
As with any new procedure, carefully monitored clinical trial protocols would be required for heritable genome editing, with attention to monitoring off-target events as well as the efficiency and correctness of the specific edit. Unlike conventional clinical trials, heritable genome-editing trials would likely require long-term prospective follow-up studies across subsequent generations. This follow-up would entail study of the future children affected by the intervention, none of whom would have been party to the initial decision to participate in a research trial. Data of this type would be important for determining whether the techniques had achieved their goals (Friedmann et al., 2015). Even those who have volunteered to be research subjects cannot be compelled to participate in long-term follow-up. Nonetheless, encouragement is permitted. Experience from xenotransplantation and some drug and device trials shows that this encouragement can be successful. And despite the particular challenge of studying offspring rather than those who consented to the research, experience with other reproductive technologies suggests that follow-up can be carried out in numbers sufficient to permit conclusions about many possible long-term effects (Lu et al., 2013).
Some have argued that germline editing can be justified on grounds that go beyond mere parental choice. There is a line of thought that germline modification could be used to create a level playing field for those whose traits now put their children and descendants at a disadvantage (Buchanan et al., 2001). Others see a public health benefit in access to heritable genome editing because it might somewhat reduce the prevalence of many devastating diseases, such as Tay-Sachs and Huntington’s disease. That said, it is important to note that the history of abusive and coercive eugenics (discussed in the section on human dignity and the fear of eugenics below) is intertwined with previous, undoubtedly well-intentioned public health and hygiene movements, which is one reason why discussions of public health benefits often engender some skepticism and unease.
Some contemporary transhumanists point out that the human body is flawed in that it easily becomes diseased, requires a great deal of sleep, has various cognitive limitations, and eventually dies. They suggest that it would make sense to improve the human species by making it more resistant to disease, more moral, and more intelligent (Hughes, 2004). Some, such as philosopher John Harris, say that in certain cases there is a moral obligation to enhance ourselves genetically (Harris, 2007). But these are
arguments about enhancement, not the restoration or maintenance of ordinary health. That topic is addressed in more detail in Chapter 6.
A “Natural” Human Genome and the Appropriate Degree of Human Intervention
Among the social and cultural arguments against heritable genome editing are positions that support a preference for a “natural” genome. Although there is wide acceptance of human intervention in agriculture and medicine, some hold the view that the human genome is different and should be free of intentional manipulation because of some aspect of its naturalness, whether defined as “normal,” “real,” or otherwise determined largely by forces other than human intervention (Nuffield Council, 2015), and some aspect of its “humanness” (Machalek, 2009; Pollard, 2016). However, the human genome is not entirely “human,” as it includes Neanderthal and Denisovan DNA (Fu et al., 2015; Pollard, 2016; Vernot et al., 2016). Nor does it exist in any single, static state. As reviewed in Chapter 4, each time a cell divides, numerous changes in the DNA sequence occur, and environmental insults such as radiation and chemicals (both natural and synthetic) also produce sequence changes. Moreover, meiosis combined with fertilization creates in each individual a novel assortment of gene variants. The result is significant variation in genomic DNA sequence among individuals (Kasowski et al., 2010; Zheng et al., 2010) and even within the cells of a single individual. Every human (other than mono-zygotic twins) begins with a unique genome—actually two genomes that subsequently diversify as cells divide, each of which is “natural.” There is no single human genome shared by all of humanity.
The concern then devolves to the view that the human genome should be treated with a sense of humility and that humanity should recognize the limits of wisdom and science, and even that human intervention is more dangerous or more unpredictable than natural processes. This concern often is expressed by the term “playing God,” which captures the notion that humans lack a god-like omniscience that would be required to make any changes in the genome safely (and to predict that such changes would actually serve the intended purpose). Even those approaching the issue from a nontheological point of view may use the term to represent a more general notion about appropriate limits of intentional human control of the environment or of the human genome (President’s Commission, 1983).
To some extent, this argument implicitly accepts the thesis that the forces of nature and evolution are a better—or at least less problematic—source of genome alteration than human intervention. However, the natural variations in human genomes arise by chance and are selected for or against during evolution by founder effects and according to selection pressures
such as climate, nutrition, and infectious disease, some of which may no longer be relevant in the modern world. Accordingly, while it is important to recognize the limits of human understanding and proceed with all due care, this does not necessarily mean that society should forswear any human intervention at all.
Overall, from a scientific viewpoint, some conclusions about likely benefits and risks of heritable genome editing can be supported with a fair degree of certainty, while others remain uncertain and in need of further investigation and societal debate, calling for humility with respect to those conclusions. Assessing the probabilities of efficacy and risk is the focus of clinical trials, which can be viewed as a manifestation of the recognition of the limits of human knowledge.
Beyond the scientific assessment of risks and uncertainties, the question of the proper extent of human intervention in nature has long been discussed in spiritual and religious terms. In the contemporary West, where Christian traditions have had the most influence on what is today a more religiously diverse and often secular culture, these ideas are expressed in the debate about which tasks in improving or stewarding nature are the domain or obligation of humans and which are to be left to God (Cole-Turner, 1993; Vatican, 2015). This thinking reflects beliefs that are present across a variety of traditions and many centuries, including St. Francis’ Canticle of Creation and the belief systems among some Native American nations.
In the Jewish tradition, on the other hand, there is an explicit obligation to build and develop the world in any way that is beneficial to people, and such improvements are viewed as a positive collaboration between God and humans, not as an interference with creation (Steinberg, 2006). Similarly, many Muslims and Buddhists view genetic engineering as just one of many welcome interventions to reduce suffering from disease (HDC, 2016; Inhorn, 2012; Pfleiderer et al., 2010). The question will always be how much human-directed intervention in nature and in humans themselves is appropriate or even permissible. This is a spiritual and practical question asked by both religious and nonreligious people, although somewhat more often by the former (Akin et al., 2017). Even among the religious, adherents of different faiths will have varying degrees of interest or concern (Evans, 2010).
Human Dignity and the Fear of Eugenics
International covenants, treaties, and national constitutions, including European treaties focused on banning germline modification, typically invoke the concept of dignity (Hennette-Vauchez, 2011). While this term has many meanings, it is most often invoked in the debate about heritable genome editing to affirm that humans have value simply by virtue of being
human and not because of their capacities, and thus cannot be treated as instruments of another’s will (Andorno, 2005a; Sulmasy, 2008). Emmanuel Kant viewed human agency and free will as essential aspects of human dignity. The term also can signal a special regard for humans as opposed to other species, an appreciation of the intellectual capacities of humans, and a commitment to promoting autonomy and human flourishing. Since rights and other individualistic arguments cannot easily be used to address concerns about future generations or humanity, “dignity” has been invoked to “provide an ultimate theoretical reason to prevent a misuse of emerging biotechnological powers” (Andorno, 2005b, p. 74). Even if limited to preventing serious disease or disability, the prospect of using heritable genome editing triggers concern that purely voluntary, individual decisions can collectively change social norms regarding the acceptance of less serious disabilities (Sandel, 2004).
The disability rights community is not monolithic, and its attitudes toward genetic technologies such as prenatal screening can vary from supportive to skeptical (Chen and Schiffman, 2000; Saxton, 2000). There has been a long and ongoing debate among different parts of the disability community with regard to the use of screening technologies. These tensions are real, continuing, and unlikely to be resolved entirely. Still, disability activists have been among the most visible critics of using technology to screen for or determine the genetic qualities of children. Jackie Leach Scully writes of a fear that voluntary prenatal diagnostic techniques, which would also apply to genome editing, set us on a “slippery slope” (a concept discussed further below) toward intolerance of disability and even the risk of a return to the coercive practices of the past (Scully, 2009).
Others write that a policy of prevention by genetic screening (and by extension genome editing) “appears to reflect the judgment that lives with disabilities are so burdensome to the disabled child, her family, and society that their avoidance is a health care priority—a judgment that exaggerates and misattributes many or most of the difficulties associated with disability” (Wasserman and Asch, 2006, p. 54). The same observation has been made concerning the differing perceptions of disabled persons and medical professionals about the degree of distress caused by a particular condition (Longmore, 1995). Studies suggest that “many members of the health professions view childhood disability as predominantly negative for children and their families, in contrast to what research on the life satisfaction of people with disabilities and their families has actually shown” (Parens and Asch, 2000, p. 20).
Indeed, there is concern that the availability of the technology might actually lead to a judgment that parents who forego heritable genome editing are negligent, a theory seriously discussed (although ultimately abandoned) with respect to genetic screening when it became a common practice (Malek
and Daar, 2012; Sayres and Magnus, 2012; Wasserman and Asch, 2012). Others have cited fears that hard-won successes at developing laws and policies that make the world accessible to those with disabilities will lose support when the number of persons directly affected declines.
One can argue that these concerns reflect a false dichotomy, and that unconditional love for a disabled child once born and respect for all people who are born with or who develop disabilities are not incompatible with intervening to avert disease and disability prior to birth or conception. And the decades that saw the explosion of prenatal diagnosis (accompanied by selective abortion of affected fetuses) and preimplantation diagnostics (accompanied by selective implantation of nonaffected embryos) are the same as those in which public attitudes toward disability became far more accepting (Hernandez et al., 2004; Makas, 1988; Steinbach et al., 2016). It would appear that “encouraging attempts to reduce the incidence of a genetic disease is compatible with continuing respect for those born with the disease and providing support for their distinctive needs” (Kitcher, 1997, p. 85).
The disability community is characterized by a long and ongoing tension with regard to the use of screening technologies. The literature appears to support openly acknowledging that this tension is real, continuing, and unlikely to be resolved entirely, and that any step toward the use of genome editing to eliminate disabilities must be carried out with care and open discussion (Kitcher, 1997). The committee supports this call for continued public deliberation (see Chapter 7).
Public policy has shifted toward eliminating discrimination in employment or public services, and public investment in changing the social, physical, and employment environment to achieve this goal has increased, with measures ranging from accessible buildings to sign language presentations to aural signals for street crossings. The range of measures remains insufficient, however, and one cannot know whether this shift in attitude would have been even more dramatic if genetic screening and abortion laws had not made it easier to reduce the prevalence of birth defects. Nonetheless, this progress does to some extent address the concern that reducing the prevalence of disabilities will necessarily decrease empathy, acceptance, or integration of those who have them.
Economic and Social Justice
Recognizing that heritable genome-editing technology is unlikely to be used widely in the near future and that drastic transformation of the species or immediate changes in cultural norms are unlikely, some social justice arguments focus on the effects of the technology’s being accessible only to a few rather than to too many. In this framing, the technology is another example of a society’s allocating considerable resources to developing a
technology that will benefit only a relatively few wealthy people when this money could be used to relieve the suffering of millions of poor people through already existing technologies (Cahill, 2008). One counterargument is that the research phase may include those less well-off, or that even if treatments for rare but compelling diseases often start with the wealthy, they eventually become more affordable and available for the poor. Moreover, the research that will make heritable genome editing possible will likely provide insights that will lead to health care interventions for other disorders as well. More to the point, perhaps, is the reality that—at least in the United States—health care budgets are not set globally, and therefore the decision to refrain from spending in one area will not necessarily result in the funds being redirected to another area of need.
Another concern is that if heritable genome editing were to become prevalent among those who are wealthier or better insured, it could change the prevalence of avoidable diseases between advantaged and disadvantaged populations and could permanently establish what Harris (2007) calls “parallel populations.” While great inequality already exists, the argument continues, heritable genome editing would make a culturally determined inequality into one that is biological. While such a phenomenon already exists in the form of durable effects of better nutrition and use of vaccines among the advantaged populations of the world, some critics are concerned about adding yet another, more durable form of superior access to better health (Center for Genetics and Society, 2015). These concerns apply to a range of health advances, and are not limited to genome editing.
The Slippery Slope
Many scholars who support (or at least are not opposed to) germline modifications align the possible uses of genome editing along a continuum of acceptability. This continuum almost always starts with converting single-gene disorders to a common, nondeleterious sequence at the most-acceptable end, and moves toward enhancements that are unrelated to disease on the least-acceptable end. The slippery slope claim is that taking the first step with single-gene disorders is likely to lead, in some number of years, to the conduct of nondisease enhancements that many would rather see prohibited. As one group involved with somatic modification wrote in the journal Nature, “many oppose germline modification on the grounds that permitting even unambiguously therapeutic interventions could start us down a path towards non-therapeutic genetic enhancement” (Lanphier et al., 2015, p. 411).
The slippery slope arguments of most critics do not claim inevitability but are instead probabilistic. They are based on what could be described as predictive sociology about how societies actually function and rejec-
tion of the notion that placing barriers and speed bumps on the slippery slope will be a sufficient deterrent to less desirable uses (Volokh, 2003). Many scientific advances in the past—ranging from reconstructive surgery (which has led to plastic surgery for aesthetics) to prenatal screening for lethal disorders (which has led to screening of carriers for disease genes and preimplantation screening for nonlethal, even late-onset disorders)—have raised similar concerns about a slippery slope toward less compelling or even antisocial uses.
An opponent of editing the germline would not necessarily oppose on principle replacing a disease gene variant with a corresponding, common, nondisease variant, as such a change would give offspring no social advantage and is the type of instrumental action directed at future children that is currently part of modern medicine. Many opponents, however, do not believe genome editing would stop there and observe that a number of social processes make the slope more slippery. Parts of the medical profession might become invested in providing the service or patient groups in seeking the service, creating powerful interest groups favoring its maintenance or even expansion. IVF, for example, was originally developed to circumvent fallopian tube blockage. It soon was extended, however, to circumventing naturally age-related decline in fertility and even postmenopausal infertility, and later became an enabling technology for PGD. Likewise, PGD was originally designed to select against embryos with serious deleterious mutations but later was expanded to conditions that not all agree are diseases or disabilities, as well as to sex selection.
On the other hand, slippery slope arguments have their critics, who point to their inherent uncertainty and the fact that many such claims do not come to pass. Indeed, despite predictions to the contrary, neither IVF nor PGD has come to be used for convenience or for selection of trivial traits. Even artificial insemination, which offers an inexpensive way to “optimize” the male genetic contribution, has not become a widespread practice except when the male partner is absent, infertile, or at risk of passing on a seriously deleterious mutation. Furthermore, among those women who already needed to use donor gametes, almost none took advantage of the opportunity to obtain semen from the so-called Nobel Prize sperm bank (Plotz, 2006), although there has been evidence of a stronger tendency to “optimize” when it comes to egg donation (Klitzman, 2016). Those who reject slippery slope arguments often are less concerned than proponents about situations that might be viewed as the bottom of the slope.
Many of the attempts to introduce speed bumps or friction on the slippery slope in the evolution of genetic modification of humans have focused on the easily grasped linguistic/cognitive difference between a body/individual and offspring/society, thereby establishing the distinction between editing of somatic and germline cells. Critics would claim that the
current debate about crossing the cognitive barrier (i.e., crossing the germline) is proof of the existence of a slippery slope.
Overall, slippery slope arguments do not depend on universal condemnation of the initial, most compelling applications of heritable genome editing. But while many think that regulation could establish effective speed bumps, proponents of slippery slope arguments raise the question of whether and how society can develop regulations that are sufficiently robust to quell the fear of a progressive move toward less compelling and more controversial applications. Indeed, they would say that regulations would do little to stop the progression down a slippery slope because regulations are based on cultural views, and it is precisely the underlying change in cultural views that is the slippery slope.
Regulation in the United States
In the United States, heritable genome editing would be subject to a complex landscape of state and federal laws and regulations (see Chapter 2). The legality of research, and perhaps even clinical application, would vary from state to state as a result of differing laws on fetal and embryo research. Federal funding for the research would likely be unavailable because of current legislative restrictions on funding research involving human embryos. Should heritable genome editing move into clinical investigations, the U.S. Food and Drug Administration (FDA) would have regulatory jurisdiction. The altered cells—whether gametes or the embryo—would need to be implanted for gestation, and this transfer would trigger the same FDA jurisdiction as that used by the agency in 2001 (FDA, 2009) when it determined that reproductive cloning could not proceed without authorization. This jurisdiction derives from the agency’s power to regulate tissue transplantation. While IVF and even PGD were developed before the FDA policy in this area was fully developed and therefore have not been regulated as closely as more recent products, genome editing would fall squarely within FDA jurisdiction.
A careful stepwise process (outlined in more detail in Chapter 2) would include consideration by the National Institutes of Health’s (NIH’s) Recombinant DNA Advisory Committee (RAC) (with public comment and transparent review), local institutional review board (IRB) and local institutional biosafety committee (IBC) review, and FDA review before any decision about whether to permit clinical trials could be made. If heritable genome editing succeeded in research trials and was approved for marketing, there would also be mechanisms for oversight in the postapproval context.
Because heritable genome editing would involve the use of other assisted reproductive technologies, oversight of its use would likely involve the same statutes and regulations that apply to IVF and PGD. Some of these regulations focus on donor material safety, transparency, and reporting requirements, as is the case with IVF, or on quality control of the laboratories (though not necessarily the actual diagnostics) used for PGD. Heritable genome editing would be performed in conjunction with IVF and PGD, and thus could involve statutes and regulations that apply to those technologies. For example, IVF itself is subject to rules that require registration of facilities, screening of donor gametes for communicable diseases, and compliance with good tissue-handling standards (FDA, 2001). Programs using IVF also must report pregnancy success rates to the Centers for Disease Control and Prevention (CDC).5
Following the publication of the National Academies report on mitochondrial replacement techniques, which can result in heritable changes in small amounts of mitochondrial (i.e., nonnuclear) DNA present in the egg, and the publication of genome-editing research in China using nonviable human embryos, NIH made a statement to the effect that it would not fund research involving genome editing of human embryos.6 Francis Collins, Director of NIH, stated that NIH “will not fund any use of gene-editing technologies in human embryos” (Collins, 2015). He noted that the “concept of altering the human germline in embryos for clinical purposes has been debated over many years from many different perspectives, and has been viewed almost universally as a line that should not be crossed” (Collins, 2015). But this research is already something NIH could not fund because of legal obstacles created by the Dickey-Wicker Amendment (forbidding U.S. Department of Health and Human Services [HHS] funding of such work) and the RAC policy of declining to review such work in accordance with the requirement for NIH-funded projects (see Chapter 2).
The NIH statement also highlighted the requirement for FDA approval of an Investigational New Drug (IND) application for any clinical trials involving transfer and gestation of an edited embryo. The FDA had never received or approved a proposal to modify the germline, but apparently alarmed by the direction of research, the U.S. Congress held hearings in
5 Assisted Reproductive Technology Programs, 42 U.S.C. § 263a-1 (current through Public Law 114-38).
6 The NIH statement can be accessed at https://www.nih.gov/about-nih/who-we-are/nihdirector/statements/statement-nih-funding-research-using-gene-editing-technologies-humanembryos (accessed January 30, 2017).
June 2015 on “The Science and Ethics of Engineered Human DNA.”7 These hearings were followed by the omnibus spending bill provision, discussed in Chapter 2, that prevents the FDA from using any of its resources to even consider an application to proceed with clinical trials involving germline modification.8 This limitation will last at least through the end of April 2017. Beyond that date, the prohibition may be extended or deleted, depending on the details of the next budget exercise. If the prohibition is lifted, the FDA will once again be permitted to entertain requests to initiate clinical trials in this area, although the restrictions on the use of federal funds for such research will remain.
Statements and Views from Other Bodies
Heritable genetic engineering has been the subject of public and academic discussion for decades. Salient instruments that have legal effect include the European Oviedo convention, which allows genetic engineering only for preventive, diagnostic, or therapeutic purposes and only when it is not aimed at changing the genetic makeup of a person’s descendants, thus precluding heritable genome editing. Although signed by 35 nations, this convention is binding only on those 29 that ratified it (6 nations did not ratify it in full), and even then requires implementation through domestic legislation (COE, 2016).
More recently, as discussed in Chapter 1, the organizers of the December 2015 International Summit convened by the science and medicine academies of the United States, the United Kingdom, and China called for a pause of some undefined duration in any attempt at heritable genome editing. Their statement read:
It would be irresponsible to proceed with any clinical use of germline editing unless and until
- the relevant safety and efficacy issues have been resolved, based on appropriate understanding and balancing of risks, potential benefits, and alternatives, and
- there is broad societal consensus about the appropriateness of the proposed application. Moreover, any clinical use should proceed only under appropriate regulatory oversight.
7 “The Science and Ethics of Engineered Human DNA.” Hearing before the Subcommittee on Research and Technology, of the House Committee on Science, Space and Technology, June 16, 2015. https://science.house.gov/legislation/hearings/subcommittee-research-and-technology-hearing-science-and-ethics-genetically (accessed January 30, 2017).
8 Consolidated Appropriations Act of 2016, HR 2029, 114 Cong., 1st sess. (January 6, 2015) (https://www.congress.gov/114/bills/hr2029/BILLS-114hr2029enr.pdf [accessed January 4, 2017]).
At present, these criteria have not been met for any proposed clinical use: the safety issues have not yet been adequately explored the cases of most compelling benefit are limited and many nations have legislative or regulatory bans on germline modification. However, as scientific knowledge advances and societal views evolve, the clinical use of germline editing should be revisited on a regular basis. (NASEM, 2016d, p. 7)
Similarly, in its 2016 professional guidelines for regenerative medicine research, the International Society for Stem Cell Research included the following statement: “Until further clarity emerges on both scientific and ethical fronts, the ISSCR holds that any attempt to modify the nuclear genome of human embryos for the purpose of human reproduction is premature and should be prohibited at this time” (ISSCR, 2016a, p. 8).
In 2015, a self-organized group of multinational experts called the Hinxton Group published a statement exploring the possibility that heritable genome editing might be acceptable, albeit with many caveats (Hinxton Group, 2015, p. 3). According to that statement, “[p]rior to any movement toward human reproductive applications, a number of crucial scientific challenges and questions must be addressed.” The statement proceeds to list a number of technical questions related to safety and efficacy and stresses the need to explore cultural attitudes and whether and how legal limits might be placed on particular uses.
The French National Academy also appears to have taken the position that while heritable genome editing is unacceptable now, one can contemplate a time when it might be permitted, stating that “this research, including that on germline cells and human embryos, should be carried out provided that it is scientifically and medically justified” (ANM, 2016, p. 2).
These statements all recognize that issues of safety and efficacy associated with heritable genome editing are far from resolved and that attempts to apply this form of genome editing should not be made at this time. They all note, however, that the science is continuing to progress rapidly, and they avoid calling for permanent prohibitions. Indeed, the Hinxton Group recommends that “a detailed but flexible roadmap [be] produced to guide the development of standards for safety and efficacy” (Hinxton Group, 2015, p. 3).
In some situations, heritable genome editing would provide the only or the most acceptable option for parents who desire to have genetically related children while minimizing the risk of serious disease or disability in a prospective child. Yet while relief from inherited diseases could accrue from its use, there is significant public discomfort about heritable genome
editing, particularly for less serious conditions and for situations in which alternatives exist. These concerns range from a view that it is inappropriate for humans to intervene in their own evolution to anxiety about unintended consequences for the individuals affected and for society as a whole.
More research is needed before any germline intervention could meet the risk/benefit standard for authorizing clinical trials. But as the technical hurdles facing genome editing of progenitors of eggs and sperm are overcome, editing to prevent transmission of genetically inherited diseases may become a realistic possibility.
The primary U.S. entity with authority for the regulation of heritable genome editing—the FDA—does incorporate value judgments about risks and benefits in its decision making. A robust public discussion about the values to be placed on the benefits and risks of heritable genome editing is needed now so that these values can be incorporated as appropriate into the risk/benefit assessments that will precede any decision about whether to authorize clinical trials. But the FDA does not have a statutory mandate to consider public views on the intrinsic morality of a technology when deciding whether to authorize clinical trials. That level of discussion takes place at the RAC, in legislatures, and at other venues for public engagement (see Chapter 7).
Heritable genome-editing trials must be approached with caution, but caution does not mean they must be prohibited. If the technical challenges were overcome and potential benefits were reasonable in light of the risks, clinical trials could be initiated if limited to the most compelling circumstances, if subject to a comprehensive oversight framework that would protect the research subjects and their descendants, and if sufficient safeguards were in place to protect against inappropriate expansion to uses that are less compelling or less well understood.
RECOMMENDATION 5-1. Clinical trials using heritable genome editing should be permitted only within a robust and effective regulatory framework that encompasses
- the absence of reasonable alternatives;
- restriction to preventing a serious disease or condition;
- restriction to editing genes that have been convincingly demonstrated to cause or to strongly predispose to that disease or condition;
- restriction to converting such genes to versions that are prevalent in the population and are known to be associated with ordinary health with little or no evidence of adverse effects;
- the availability of credible preclinical and/or clinical data on risks and potential health benefits of the procedures;
- ongoing, rigorous oversight during clinical trials of the effects of the procedure on the health and safety of the research participants;
- comprehensive plans for long-term, multigenerational followup that still respect personal autonomy;
- maximum transparency consistent with patient privacy;
- continued reassessment of both health and societal benefits and risks, with broad ongoing participation and input by the public; and
- reliable oversight mechanisms to prevent extension to uses other than preventing a serious disease or condition.
Given how long modifying the germline has been at the center of debates about moral boundaries, as well as the pluralism of values in society, it would be surprising if everyone were to agree with this recommendation. Even for those who do agree, it would be surprising if they all shared identical reasoning for doing so. For some, the debate is about respecting parental desires for genetically related children. For others, it is primarily about allowing children to be born as healthy as possible. But as noted earlier in this chapter, some do not view heritable genome editing as a benefit to the resulting child, who otherwise might never have been conceived at all. And for others, the desire of parents who carry genetic disease to have a genetically related child through this technology, instead of having a genetically unrelated child, is not sufficient to outweigh the social concerns that have been raised. There are also those who think the final criterion of Recommendation 5-1 cannot be met, and that once germline modification had begun, the regulatory mechanisms instituted could not limit the technology to the uses identified in the recommendation. If it is indeed not possible to satisfy the criteria in the recommendation, the committee’s view is that heritable genome editing would not be permissible. The committee calls for continued public engagement and input (see Chapter 7) while the basic science evolves and regulatory safeguards are developed to satisfy the criteria set forth here.
Heritable genome editing raises concerns about premature or unproven uses of the technology, and it is possible that the criteria outlined here for responsible oversight would be achievable in some but not all jurisdictions. This possibility raises the concern that “regulatory havens” could emerge that would tempt providers or consumers to travel to jurisdictions with more lenient or nonexistent regulations to access the restricted procedures (Charo, 2016a). The result could be a “race to the bottom” that would encourage laxer standards in nations seeking revenues from medical tourism, as has happened with both stem cell therapy and mitochondrial replacement techniques (Abbott et al., 2010; Charo, 2016b; Turner and
Knoepfler, 2016; Zhang et al., 2016). The phenomenon of medical tourism, which encompasses the search for faster and cheaper therapeutic options, as well as newer or less regulated interventions, will be impossible to control completely if the technical capabilities exist in more permissive jurisdictions (Cohen, 2015; Lyon, 2017). Thus, it is important to highlight the need for comprehensive regulation.
As of late 2015, the United States was unable to consider whether to begin heritable genome-editing trials, regardless of whether the criteria laid out above could be met. As noted above, a provision (in effect until at least April 2017) included in a congressional budget bill9 contains the following language:
None of the funds made available by this Act may be used to notify a sponsor or otherwise acknowledge receipt of a submission for an exemption for investigational use of a drug or biological product under section 505(i) of the Federal Food, Drug, and Cosmetic Act (21 U.S.C. 355(i)) or section 351(a)(3) of the Public Health Service Act (42 U.S.C. 262(a)(3)) in research in which a human embryo is intentionally created or modified to include a heritable genetic modification. Any such submission shall be deemed to have not been received by the Secretary, and the exemption may not go into effect.
The current effect of this provision is to make it impossible for U.S. authorities to review proposals for clinical trials of heritable genome editing, and therefore to drive development of this technology to other jurisdictions, some regulated and others not.
9 Consolidated Appropriations Act of 2016, HR 2029, 114 Cong., 1st sess. (January 6, 2015) (https://www.congress.gov/114/bills/hr2029/BILLS-114hr2029enr.pdf [accessed January 4, 2017]).