NEW YORK – The US National Academy of Medicine (NAM), the US National Academy of Sciences (NAS), and the Royal Society of the UK this week convened the first meeting of the International Commission on the Clinical Use of Human Germline Genome Editing in Washington, DC, to obtain information on the state of genome editing technology and develop a framework for the scientific and regulatory communities on the appropriate use of germline genome editing.
The commission, which was announced in May, will identify the scientific, medical, and ethical requirements that should be considered before CRISPR research can be applied to the clinic, if society concludes that heritable human genome editing applications are acceptable.
Its formation came as a response to the work of He Jiankui, the Chinese scientist who shocked the world last November when he announced that he had edited the germlines of twin embryos. He was widely condemned by the global scientific community for violating scientific principles and ethical norms.
Indeed, the specter of He's work hung over the meeting in Washington — several of the presenters referred to his now-infamous experiments though few referred to him by name, instead showing pictures of him, or calling him "JK" and "a certain researcher from China."
Several of the assembled experts also noted worriedly that Denis Rebrikov from the Kulakov National Medical Research Center for Obstetrics, Gynecology, and Perinatology and the Pirogov Russian National Research Medical University has already signaled his intent to follow in He's footsteps.
In his introduction, NAM President Victor Dzau said that although germline genome editing is a technological breakthrough with transformative potential for medicine, it's controversial because the use of gene editing in embryos would produce permanent alterations that would be passed down to further generations.
Since the conference in Hong Kong where He unveiled his work, experts have been called on to produce expedited reports and release guidelines for germline editing, Dzau said. This commission has been tasked with developing such a framework for scientists and regulatory agencies, to govern the pathway of this technology from research to clinical use, "paving the way for specific criteria and standards before germline editing is considered permissible, if at all," he added.
He also noted that the scientific issues and ethical issues surrounding germline editing are linked, and that several problems would have to be solved before the technology could be considered usable in the clinic — for example, identifying appropriate protocols for assessing on-target and off-target effects, and long-term side effects; assessing mosaicism; assessing potential benefits and harms to a potential child; designing appropriate protocols to obtain consent from patients; and creating mechanisms for long-term monitoring of children who are edited.
Carrie Wolinetz, chief of staff and associate director for science policy at the National Institutes of Health, was in Hong Kong when He made his presentation. "It was an incredibly disturbing presentation and it hasn't improved with time," she said, adding that the NIH found the experiment to be highly unethical and irresponsible in many ways. The agency has called for a moratorium on germline editing until the international community comes to a consensus on standards for the technology's use, especially in the light of other scientists contemplating moving forward with germline editing experiments, Wolinetz noted.
She also said that all research involving human participants should harken back to the Belmont principles, which call for respect, beneficence, and justice for human subjects involved in scientific research. "Keeping those principles high in our thoughts as we have this conversation on germline editing is incredibly important," Wolinetz said.
The NIH is excited about the prospects for somatic editing, but there's a "well-worn regulatory framework for addressing ethics, regulation, and participant concerns" in the case of somatic editing, she added. The editing of genes that will be passed to the next generation should be held to a higher standard — the NIH currently thinks it's too risky and there's a lack of compelling medical applications.
"NIH supports the moratorium on germline editing and supports the commission's movement toward a framework and governance of this technology, to lead to a pathway forward that allows society to make decisions about what they'd like to do with this technology," Wolinetz said.
The commission's work isn't being done in a vacuum. The World Health Organization established its own global, multi-disciplinary panel of experts to examine the scientific, ethical, social, and legal implications of both somatic and germline genomic editing in response to He's work. The panel — whose members were appointed in February and who met for the first time on March 18 — are reviewing the current research and social attitudes towards genome editing technology in order to advise the WHO on appropriate oversight and governance mechanisms.
At this week's meeting, Margaret Hamburg, co-chair of the WHO Expert Advisory Committee and the NAM's foreign secretary, emphasized the need for the two commissions to collaborate to bring together the best information and ideas.
The WHO is looked to by many countries around the world as the first source on normative practice and standards for medicine and science, Hamburg said, making the agency a relevant source of standards and guidelines for the use of genome editing technology. At the WHO commission's first meeting, the experts discussed whether they wanted to focus only on germline editing or also on somatic, and decided to tackle both. In many parts of the world, regulatory oversight for somatic editing is well defined, according to Hamburg, but the commissioners felt that it was important to include somatic editing because some countries don't have such oversight, which could lead to exploitation of the technology.
The WHO commission's first meeting produced three recommendations. First, the commissioners have suggested that the WHO create a more structured mechanism for planned and ongoing research by establishing a WHO registry of relevant genome editing research, Hamburg said. The commission believes that it's important to get support for such a registry from funders and publishers, to make it so that researchers would have to participate in the registry if they want to get their work published. A template for this registry will be presented at the next WHO commission meeting.
The second recommendation from the first meeting was a declaration that it would be irresponsible at this time for anyone to proceed with clinical applications of germline genome editing. The commission believes that the science is still underdeveloped, and that there's an incomplete understanding of negative findings, inconclusive findings, and successful efforts, Hamburg said.
And finally, the WHO commission is planning to seek input from the broadest range of stakeholders on the genome editing issue, and to explore opportunities for an open, online mechanism for seeking input.
In this vein, she emphasized the need for collaboration between the two commissions, noting that the International Commission could help the WHO group build a better understanding of the science, technical aspects, and related ethical aspects of genome editing, and define the universe of clinical interventions and experiment that should be included in the WHO registry. Hamburg also said the two groups could collaborate in each other's information gathering processes, help each other raise awareness of the issues and improve communication with the public.
If the groups fail to work together, there could be confusing or contradictory messaging, competitions for resources that could set back progress for everyone, gaps that would enable irresponsible research and result in harm, and long-term damage to the responsible application of germline editing technology, Hamburg concluded.
In response to a question from a commissioner on how the WHO's registry will overlap with existing registries, Hamburg said that the registry is currently being shaped and that part of the point of collaborating with this commission is to determine what should be included. Further, she added, there's no need to start from scratch — the registry can fold in information from existing registries like clinicaltrials.gov, and then fill in the gaps. "The idea is to give greater idea of the scope of work underway, create a resource to strengthen research, connect possible collaborators," and give a sense of responsibility to researchers doing genome editing work, she said.
However, Hamburg's presentation also brought up questions of how these standards and frameworks are likely to be enforced once these two commissions complete their work. "What authority does the WHO have over other countries in terms of governance?" one commissioner asked her.
"This is an area we are thinking about and working through," Hamburg said. "What tools does WHO have to implement a global governance structure? The majority of countries are member states of WHO, and participate in the governance framework. They look to WHO for normative standards and guidance in a number of critical areas. They use the regional and national offices of WHO. It's not an enforcement agency, but it does have a critical role, whether in identifying essential medicines, outbreak response, global surveillance."
The WHO commission's work is also about encouraging transparency, she added, and the commission wants to encourage the broadest level of information gathering possible through the formation of its registry. If members of the scientific community see unethical genome editing research happening around them, they may have a responsibility to report that research, Hamburg said. The commissions should explore that issue. But the WHO isn't in a position to enforce standards and guidelines like the NIH or the UK's National Health Service might be.
State of the science
The majority of the meeting was taken up with informational presentations on the current state of genome editing technology. Specifically, experts presented the commissioners with the latest knowledge in somatic genome editing in order to help them figure out whether or not germline editing is advanced enough to allow its use in the clinic.
Broad Institute President and Founder Eric Lander noted the importance of acknowledging the breadth of "what we know and what we don't know" about germline editing in order to decide if it's medically appropriate. And even if society decides it's appropriate, he added, researchers must be certain of how it works and whether it's safe to use in humans.
Aravinda Chakravarti, director of the Center for Human Genetics and Genomics at NYU School of Medicine, began with an overview of science's current understanding of how sequence variations influence the development of disease, emphasizing the complexity of the genome.
He pointed to a study published on the preprint server BioRxiv in March from the NHLBI's Trans-Omics for Precision Medicine (TOPMed) program, which sequenced 53,831 people and found 410 million variants. About 50 percent of the variation in each genome is unique to the individual, Chakravarti said.
This week, researchers led by the Broad's Daniel McArthur posted another study on BioRxiv analyzing sequence variation in 141,456 exomes and genomes. They found 443,769 loss-of-function variants in 16,694 genes and noted that genes vary considerably in their tolerance to these variants.
Chakravarti further noted that even variants that are thought of as deleterious may prove beneficial in different ways. He pointed to the Duffy-null allele in people of African ancestry — some studies have shown that this allele is associated with a higher susceptibility to HIV infection. However, it also confers greater protection against malaria infection from the Plasmodium vivax parasite.
Chakravarti said that such examples show that editing a given genetic mutation in order to cure a disease may have unintended consequences, and that we don't yet know enough about how some mutations affect genes and gene networks to attempt to fix them. Further, most diseases are influenced by several genes or mutations at once, rather than one at a time, as well as by environmental and lifestyle factors.
Despite the progress science has made, it has not achieved complete genetic, epigenetic and environmental understanding of even one phenotype, he said, adding, "Gene editing [in the germline] is vastly premature" given what we don't know.
Importantly, Chakravarti also said that disease prevention of Mendelian disorders in the last few years has gotten more successful because of advances in genetic testing, education and counseling, and an increase in the number of couples at high risk for disease choosing prenatal testing.
Sarah Teichmann, a senior group leader at the Wellcome Sanger Institute and co-leader of the Human Cell Atlas project showed how her work could also eventually influence genome editing. She used cystic fibrosis as an example, noting that work done through the cell atlas program has shown the breadth of cells and tissues that are affected by the expression of the CFTR gene.
A map of gene expression in cells across a person's lifespan will provide a basis for the interpretation of rare disease genes and loci, non-coding variants, cell-type specific expression quantitative trait loci and single-cell quantitative trait loci, Teichmann said. In this way, the atlas will be critical for targeted editing, as well as gaining an understanding of gene-gene regulation — understanding how editing one gene will affect other genes in the network.
But there's still much research to be done, she added. She presented unpublished data on fetal lung development in which the researchers found new, rare ionocytes that are major CFTR-expressing cells. These cells had not been seen before in fetal lung tissue. Experiences like this show that further research and a lot of care is needed even before somatic editing is applied to humans, Teichmann said.
Editas Medicine Chief Technology Officer Vic Myer said there are also lessons to be learned from our knowledge of somatic gene editing. He showed the company's experience in editing a splice defect in the CEP290 gene to treat an inherited retinal degenerative disease called Leber congenital amaurosis type 10.
The first important lesson Editas learned was to understand the consequence of your editing strategy and the results of your editing, Myer said. The company's strategy is to cut out a section of CEP290, but Editas researchers learned that making multiple cuts in this manner requires multiple assays to analyze the on-target and off-target effects. They developed a single assay to scan the whole region after editing, but they then found that the individual assays they had previously been using had missed certain deletions and inversions that resulted from the editing. Had they stayed with the standard multiple assays they were using, they would likely have missed those edits altogether.
The experience showed that understanding the consequences of an editing strategy isn't a simple matter, Myer said.
Maybe not right now
Certainly, the presenters seemed to agree that right now, there is very little unmet medical need that would warrant a leap into germline genome editing. Harvard Stem Cell Institute Professor Chad Cowan noted that there are very few genetic mutations that he would be sure would result in a terrible disease. "The only one I can think of is cystic fibrosis," he said.
Stanford University Professor Matthew Porteus agreed. In laying out the technical criteria for germline editing, he said that there should be an absence of alternatives and that its use should be restricted to serious diseases or conditions, and to genes with clearly proven causality. Porteus also proposed that germline editing should be limited to editing sequences that are common and known to be healthy in the general human population. In other words, the goal should be to revert pathological sequences to normal and not to modify the normal to the bespoke.
By that token, he said, editing that results in heritable non-homologous end-joining should be prohibited because it would produce edits that are not normal or healthy. Only editing that results in homology-directed repair should be used.
Porteus also proposed that the commission might generate a list of sequences that would be acceptable for germline genome editing. That would limit the scope of possibilities to sequences that are normal and healthy in the general population, and would prevent anyone from using the technology for anything beyond the prevention of disease.
The commissioners noted that it might have a negative impact on future generations to know that their parents chose to have their genomes edited, either because their condition or their sequences had been put on a list. Porteus said that the UK currently has a similar program where people found to have certain conditions are eligible for preimplantation genetic diagnosis (PGD) before they conceive a child. The NHS is studying how the children interpret their parents' choices, whether they chose to use PGD or not, and how it has affected them as they grow up. The sociological implications could also affect the public's view of germline editing.
The commissioners were also concerned about the potential for off-target effects resulting from germline editing, and asked what the long-term effects could be on a population. Chakravarti said that such scenarios must be studied because there's no doubt that they could present a problem, but that the real effects are unknown at this time.
What are the rules?
Another issue the presenters all seemed to agree on was that there's currently no legal, ethical, or regulatory framework that would support the use of germline genome editing, and that such considerations are just as important as the scientific and technical aspects.
University of Pennsylvania Professor Bruce Levine said that in his experience, regulatory bodies don't currently have the proper expertise to review proposals for research and INDs related to somatic gene editing, and that regulators without the proper understanding of what they're looking at could easily approve unethical or substandard research. That would set back the field as a whole, whether it's somatic or germline editing.
Porteus noted that an international oversight body of some kind is a necessity before germline genome editing is approved in humans. If a researcher were to change the genes of a person in another country, he argued, everyone would be affected since we all share the same gene pool.
"No one should get to change the gene pool for humanity without some oversight," Porteus said.
Levine and Porteus also agreed that it will likely take several years, if not decades, after the first somatic genome editing products hit the market to accrue the necessary amount of long-term patient safety data to determine what the secondary effects may be. Porteus estimated that thousands of patient years would be needed to analyze the effects of a product that eventually disappears from the body, and that tens of thousands of patient years would be needed for a long-term product.
"I think we're nowhere near ready for germline editing in the next decades," Levine added.
Denise Gavin, a team leader from the US Food and Drug Administration's Office of Tissues and Advanced Therapies was on hand to explain the current state of somatic genome editing regulation in the US. She noted that gene editing products are being regulated as gene therapies, and said that the FDA is now receiving three to four applications per month for such products and clinical trials.
There are several considerations that the FDA takes into account, Gavin said. The agency looks at what components are being used in a genome editing product — the nuclease, targeting elements, the donor template, and so on. An IND must provide details on how the components were designed, manufactured, and tested. If the components are modified during the product life cycle, the FDA may require comparability studies.
Further, basic safety concerns need to be addressed. Gavin cautioned researchers to look at off-target genome editing as well as the unintended biological consequences of on-target editing such as mutagenesis or misfolded proteins. She also said the FDA looks for signs of additional adverse effects due to genomic cleavage of DNA at on and off target sites, such as chromosomal translocations or inversions; immunogenicity; possible adverse impacts from the delivery system; and off-target cell or tissue editing, in the case of in vivo editing.
In Gavin's experience, none of this has been easy for researchers and companies looking to file INDs for genome editing products. The most notable challenges have involved selecting appropriate methods for predicting and identifying intra-chromosomal off-target events and inter-chromosomal genomic modifications; accounting for genomic variation between individual subjects; differentiating between off-targets that will lead to adverse biological effects and benign off-targets; and dealing with the possible limitations of animal models for evaluating safety and efficacy.
One commissioner asked Gavin how the FDA might regulate germline editing, if it was deemed acceptable. She noted that because germline editing involves the use of reproductive technology, it might require some rejiggering of the current FDA protocols, since the agency doesn't regulate reproductive technology. Most likely, germline editing technology would be regulated similarly to somatic editing technology, but with the addition of consultation with reproductive experts and a larger volume of preclinical studies.
The NAM's Dzau said that the commission will have its second meeting and an international workshop in November, in London. The third meeting will be held in January 2020, after which the commissioners will begin writing a report. The commission is aiming to release its report to the public in the spring of 2020.
The WHO's Hamburg said that organization's advisory panel will have its own second meeting at the end of August, at which it is planning to discuss existing regulatory schemes and a possible global governance framework for genome editing technology. The panel will take a look at existing global governance frameworks from other areas to see what could be useful in this area, she added.
The panel's third meeting is planned for early 2020 and its fourth meeting will be held in the summer of 2020. If the international commission's report does indeed come out in the spring of 2020 as planned, Hamburg said, the WHO panel will take it into account for its own final recommendations, to ensure synthesis of work rather than duplication, and to delve deeply into any gaps that may arise.