George Daley, Harvard Medical School

Genome editing: Pathways to translation

Second international summit on human genome editing



Thank you very much. So, I'm giving a reprise on a session that I spoke in in 2015 where I was asked to outline the plausible clinical indications both for somatic gene editing as well as the prospects for editing away heritable diseases at the embryonic stage.

I would say that in 2015, that was a distinctly challenging task. There was widespread belief that the science was still incoate and undefined and many challenges were before us. Many preferred to not even consider the prospect of embryo editing.

And then of course, even in the case where you might accept that there was scientific feasibility, there is still dramatic ethical and moral debate that even if we can do it whether we should do it. If we believe that there are indications that warrant this application, what are they? And how do we engage in a discussion that leads to the conclusion that there is societal consensus about the use of this powerful technology for irradicating disease at the embryonic stage in a way that would be heritable.

I'm really going to be talking about the dramatic evolution that has occurred over the past three years. I want to make some provocative statements. I think after yesterday, with the revelations of the first gene edited twins, I think the tide was largely negative. The prospect of anticipating ethical uses in the future was almost set aside.

I think we do that with some risk, though. Just because the first steps into a new technology are missteps, doesn't mean we shouldn't step back, restart, and think about a plausible and responsible pathway for clinical translation. That's what I'll be trying to articulate today.

2015 statement

Going back into the statement from the organizing committee in 2015, we looked at the prospects for both basic and clinical research. There was a strong sense that there needed to be further scientific investigation of the techniques, of the fidelity, measures of off-target effects, and attempts to balance risks and benefits, not only for somatic applications but also human embryo editing.

There was a strong call, especially given the sensitivity around any work with human embryos, there was a strong call for rigorous ethical oversight. And at that time, it was very clear that prohibition was really called for. In 2015, there was no sense that we had any deep understanding of the science enough to make it reasonable to move forward with establishing a pregnancy.

And of course in speaking about clinical use, there was wide expectation- and I will review progress that has been made- but it was clear that genome editing for somatic applications was imminent and it should go forward with the existing regulatory frameworks which have already overseen various gene therapies.

At that time, given the risks both known and unknown, the organizing committee made a strong statement that it would be irresponsible to proceed with any clinical use for germline editing.

Genome editing clinical trials

In the past 3 years, it is clear that there are now many clinical trials that use both the zinc finger nuclease, TALENs, and more recently CRISPR. I have to thank Matt Portias who actually gathered all of the data on this slide.

What you can see is that Sangamo has launched a number of efforts. Interestingly, to target the CCR5 in somatic cells, with the goal of rendering cells resistant to HIV infection. Indeed, there has been a publication in the New England Journal of Medicine in 2014 that strongly suggested efficacy of this approach.

More recently, there has been more excitement and enthusiasm about the very exciting potential applications of gene editing for ameliorating sickle cell disease and beta-thalassemia.

TALENs has been deployed clinically and there is early data surfacing.

There are at least 12 clinical trials of CRISPR-Cas9-based gene editing for modification. A popular one is knockout of PD-1, the break or checkpoint on immune response based on antibody inhibition, which stimulates a potent anti-tumor response. Somatic cell gene editing might be the next step in that revolution.

As noted by Matt Portias, in doing this survey of clinical trials, there are very little pre-clinical publications of the experience of these labs or readiness for taking these projects forward. To date, there have been no clinical reports reporting progress.

So while there has been clear clinical engagement, there is still a question as to how much transparency and access to peer reviewed data. I think that's an important message as we take these trials forward.

Lessons learned from trials of somatic cell gene therapy

I think the lesson, though, is that we're comfortable-- the existing regulatory frameworks that have been in place, that have watched over the dramatic advances in gene therapy primarily retroviral and lentiviral gene therapy have worked, and do provide the kind of oversight that we know affords patients protections and rigorous clinical analysis.

There are tradeoffs with any regulatory system. It introduces a cumbersone, slow and fairly conservative approach to this work. We're constnatly debating those tradeoffs. The importance of regulation for ensuring rigor and patient safety, but sometimes coming at the compromise of innovation.

Statement of organizing committee in 2015

Now speaking directly to the very strong statement of the organizing committee in 2015, we used the word "irresponsible". It would be irresponsible at that time in 2015 to proceed with any clinical use of germline editing unless and until. There was a framework, a set of targeted goals that we felt had to be addressed, which had to include additional research so that we could have more insight into the safety and efficacy of various gene editing strategy.

There was an equally important and still critical goal for further discussion, debate, and an approach towards societal consensus around which indications would be most appropriate.

Of course, finally, and central to all investigation and all innovative medical therapy is the importance of appropriate regulatory oversight.

At that time in December 2015, just 3 years ago, there was wide consensus that these criteria had not been met for any proposed clinical use of germline genome editing.

And yet there was the call to realize that as the science evolved, and as our understanding of the ethical issues evolved, that the clinical use should be revisited on a regular basis. Arguably, this meeting is that chance to revisit this.

It was the call for this international forum that in fact led to this meeting today.

In the last 3 years, it's quite remarkable how many different groups have weighed in on these concepts. I'll highlight a few. I do think they speak to the rapidly evolving appreciation of the science and our now understanding I believe that we see gene editing not only as feasible but as very potentially safe, certainly safe within the risk-benefit analysis for the treatment of severe diseases.

Since 2016

In May 2016, I co-anchored a study group with international scientists and ethicists assembled by the International Society for Stem Cell research to look broadly at the conduct of both embryonic stem cell research and more broadly embryo research. It was clear to us that the debates about embryonic stem cell research were really founded upon the unique sensitivities of the human embryo. In whatever type of research was evolving- and embryo research was evolving far beyond the simple extraction of embryonic stem cells, but also embryo manipulation and gene editing. So we broadened the call for embryonic stem cell research oversight, to the more standard embryo research oversight, and really called for any type of research endeavor involving human embryos to be subject to this kind of rigorous and independent oversight.

The ISSCR clearly called for additional laboratory-based research including research into the modification of human embryos and human gametes. It strongly stated that without further understanding of the risks and benefits, and without a further ethical consensus, that any attempt to modify the nuclear genome of human embryos should be prohibited at that time.

Since 2017

A little more than a year later, in February 2017, the study that had been called for and set in motion from the last senate was taken up by the U.S. national academies and it led to a similar endorsement of the importance of basic research for both somatic and germline applications. It similarly said that clinical trials for somatic editing should continue subject to the existing regulatory frameworks. But interestingly, and I think the wording here was carefully chosen, and is quite powerful. It suggests a strong step forward from the perspective of the last summit.

Reading it directly: "Given both the technical and societal concerns, the committee concludes there is a need for caution in any move toward germline editing, but that caution does not mean prohibition. It recommends that germline editing research trials MIGHT be permitted, but only after much more research to meet appropriate risk/benefit standards for authorizing clinical trials."

This is a remarkably bold statement.

We heard from many different countries presenting their regulatory frameworks yesterday, which essentially prohibited clinical trials. And yet here we have a study which is saying that the committee believes that the science is feasible with further analysis, and that it is time to consider the moral principles that would call into possible clinical use of germline editing. They set forward a set of criteria in the document, criteria which adheres to universal principles such as the absence of alternatives, the conduct of the research with strict transparency and independent oversight and a number of others.

This set the stage for a possible translational pathway to germline editing.

Since 2018

In 2018, just a few months ago, the Nuffield Council on Bioethics concluded as well similarly that "We can, indeed, envisage circumstances in which heritable genome editing interventions SHOULD be permitted."

So we have an evolution from "might be" permitted, meaning the possibility of ethical permissibility, to a call to-- almost an imperative, for certain indications. That if we can solve the scientific challenges, then it might be a moral imperative to say it should be implemented.

The report also highlighted the importance of the individual, and the wellfare of the individual, as well as societal concerns for social justice and social wellfare. Two important principles.

Those themes of both scientific analysis and ethical consensus are critical to the evolution in these various opinions over the past 3 years.

Clinical use might be or should be permissible

I point to a very interesting review by Brokowski in The CRISPR Journal, which recounts and has a table detailing the 60 odd reports on human genome editing that have occurred over the past 3 years (2015-2018). The vast majority of which agree with the proposition that it is premature at this time given the scientific uncertainties and the lack of ethical consensus that it is premature at this time to proceed with clinical use.

But there's a small number in the blue here, that raised the prospect that should the science be deemed feasible, that the clinical use might be or should be permissible.

So where has the science progressed? We have seen remarkable presentations yesterday from some of the leaders for defining and refining the technology. It's spectacular to see the array of different enzyme editing systems and the notion of base editing. It's clearly a transformative technology with the power for great medical use.

Genome editing in embryos (mice to monkeys)

  • One-step generation of mice carrying mutations in multiple genes by CRISPR/Cas-mediated genome engineering (2013)

  • ONe-step generation of mice carrying reporter and conditional alleles by CRISPR/CAS-mediated genome engineering (2013)

  • Generation of gene-modified cynomolgus monkey via Cas9/RNA-mediated gene targeting in one-cell embryos (2014)

There's also been quite a remarkable facility with being able to use this genome editing technology, to engineer whole organisms. Very soon after the first publications of the techniques, they were applied in mice. And soon thereafter, in monkeys.

  • CRISPR/Cas9-mediated disruption of SHANK3 in monkey leads to drug-treatable autism-like symptoms

  • No off-target mutations in functional genome regions of a CRISPR/Cas9-generated monkey model of muscular dystrophy (2018)

Over the last few years, we have seen a number of very interesting applications to primates. Indeed, there's a deep investment in using genome editing technology to establish primate models of human diseases, which I think is going to be profoundly powerful for creating more predictive and more useful animal models.

Human embryos

  • CRISPR/Cas9-mediated gene editing in human tripronuclear zygotes

It's also providing us with some of the technical feasibility for germline editing as applied to human embryos. The first reports, really happening just ahead of the last summit in 2015, were met with quite a bit of consternation in part because the message was how unfaithful the actual editing was. There was a significant burden of off-target effects, and there was ongoing concern about whether those issues could ever be resolved.

  • Introducing precise genetic modifications into human 3PN embryos by CRISPR/Cas-mediated genome editing (2016)

  • Correction of a pathogenic gene mutation in human embryos (2017)

  • Genome editing reveals a role for OCT4 in human embryogenesis (2017)

  • Correction of beta-thalassemia mutant by base editor in human embryos (2017)

And a number of groups have applied gene editing now to human embryos in the context of in vitro fertilization and attempting to determine variations of a protocol that would enhance the fidelity and reduce mosaicism. I think there has been an emerging consensus that the off-target problem is manageable, and in some cases even infinitesimal. There are some interesting proofs of principles, like diseases such as beta-thalassemia that could potentially be approached with this strategy.

Can embryos be effectively assessed for fidelity of gene editing safety?

In 2015, I laid out the prospect for how one could imagine assessing the safety of embryo interventions. We could use embryo biopsy, blastocyte biopsy, or trophectoderm biopsy. The facility now with deriving embryonic stem cells, even from single cells, is high enough that with great fidelity one would is able to generate cell lines that allow you to do a significant amount of functional characterization and sequencing. If one attempts genome editing at these early stages and there are still variations on when the variation occurs, and there's still much to learn to optimize that, the main concerns after doing all of this analysis are really concerns about mosaicism. That is, if the gene editing approach was not 100% efficient at the single genome stage, then you might end up with a mosaic-edited embryo and because of sampling issues you may or may not identify those cells that were effectively edited and it could lead to some misleading results.

I think much of the focus has really been now around enhancing the editing and optimization at the early stage. I think there's every prospect to imagine that even if we can't achieve 100% certainty of the lack of mosaicism, it could get to the point where it could be faithful.

Safe gene editing in embryos?

  • Generation of human oogonia from induced pluripotent stem cells in vitro (2018)

There has been another strategy with considerable progress, which is the ability to generate gametes in vitro. These can be derived from induced pluripotent stem cells, and then they are subjected to editing, and since they are immortalized cell lines they can be exhaustively characterized. These in vitro derived gametes have actually proven effective for birthing mice. There are now reports of very similar biology applied to human iPS cells.

I think we're not that far removed from technical competency for in vitro gametogenesis, which is yet another vehicle for ensuring safe and effective human genome editing.

A responsible pathway for clinical translation

I want to propose that it is time for us to at least consider a responsible pathway for clinical translation. What would such a pathway entail?

I would say that we're still as of this second international summit on human genome editing, lacking a deep scientific consensus for the optimal methodology for editing human embryos. There's still concerns about mosaicism. There's been relatively few publications that have exhaustively analyzed different strategies.

I would say we still lack a scientific consensus on the optimal methods for assessing the fidelity of gene editing, from many different techniques, but as we spoke about yesterday evolving towards a standard that could be applicable to clinical trials is still required.

Equally important are standards for assessing the technical competence of the practicioners that would be given the rights to go forward with human clinical trials. In the prior analysis of embryonic stem cell efforts, it was strongly stated by the ISSCR that one would have to ensure that the individuals given the right to derive embryonic stem cells and use the uniquely sensitive materials of human embryos, they had to be experts and they had to be trained. This is in part something that I think should be required for any clinical indication.

I want to suggest that I do think it's time to move forward from the prospects of ethical permissibility to start outlining what an actual pathway for clinical translation look like. What would be the regulatory standards that a group would be held to in order to bring this technology forward?

Equally important, and perhaps even more important, if we are ever going to practice this with a deep sense of social justice there needs to be additional definition of the distinctions between the permissible and impermissible.


I return to a slide that I made in 2015, which acknowledges the fact that there's a range of potential indications, some of which will curry much greater enthusiasm than others.

It would surprise me if the indications that we thought were the most feasible were not those indications that represent devastating life conditions and diseases such as Huntington's, Tay Sach's, cystic fibrosis, sickle cell anemia.

We know that preimplantation genetic diagnosis is frought with inefficiencies, and perhaps coupling IVF to genome editing will be more efficient way to irradicate these diseases instead of using preimplantation genetic diagnosis and embryo selection. Rarity alone is not an excuse to not pursue medical treatment; there are vanishingly few cases where obth are effected and the only option is genome editing.

Meeting in the middle, one option is to focus on modifying disease risk such as CCR5, PCSK9, Alzheimer's, BRCA1 for cancer, and resistance to global pandemics. As a species, we need to take control of our own heredity in response to future biosecurity threats. CCR5 is not completely curative but it is an example of modifying disease risk.

At the other end of the spectrum are the enhancements, such as muscularity (MSTN), height, skin color, learning and memory. Still not in any way plausible would be entertaining the use of this technology for enhancements.

If we're going to imagine a pathway for clinical translation for germline editing, solving and addressing these deep issues are essential.

Encouraging active engagement of the research and clinical community

I want to leave you with an important message and a cautionary tale. The fact that it's possible that the first instance of human germline editing came forward as a misstep should in no way I think leave us to stick our heads in the sand and not consider the very very positive aspects that could come forth by a more responsible pathway towards clinical translation.

Since this conference has occurred really as part of the intellectual tradition dating back to the Asimolar conference in 1975, I think that we in the scientific community embrace the tradition of self-regulation. It's one of our professional norms. But this requires transparency. It requires each of us in the research and clinical community to invite external scrutiny and peer review and insights.

It's against these standards which I think are widely accepted in the internationa land bioethical community that we judge the current revelation of the first clinical use of human germline editing.

I say that I believe with a community we can do better than that. Thank you very much, and I look forward to taking questions.


Q: Are you worried that with this first misstep that some authorities might have a reflex action and institute some tough regulations?

A: Yes, I do think that is a concern. It's part of the risk to society that the oversight process should actually consider when doing an evaluation of such an innovative clinical protocol. You're not only assessing the risk to the individual but you have to actually think about how a negative outcome could impact the future of a technology. It's a very compelling and interesting concept. If there's a technology that has potential for widespread good like relief of suffering and disease, it's terribly unfortunate if an initial misstep leads to an overall suppression of interest in that technology. I think there was a strong sense yesterday that the current standard is impermissibility. But once again, I hope we don't just stick our head in the sand.

Q: There have been a number of bioethics reports coming out and indicating that there might be a way forward for germline editing. There might be some consensus building up in the bioethics circles. I think the media response in the last day or so indicates that we are far far away from a social consensus. My question relates to your comments about responsibility. What will that mean for the scientific community? I'm very interested in this concept about responsibility and not just being a nice rhetorical word that we can tack on to things like innovation and research. What does it mean in terms of how we practice good science? In addition to the technical standards and research that you mentioned in your talk, could you reflect on what it would mean for scientists to practice social and moral responsibility in this area?

A: Excellent question. On the point of social consensus, I think it's true. I don't believe there's social consensus yet, either on the permissibility of any indication or any specific indication. It's another reason for ongoing discussions like the ones we have here. On the issue of responsibility, it's absolutely clear, and yet there's a tension. Scientists are by nature intrepid investigators. There's a reward for being a pioneer and for being first. But I think that we as a community have to accept the responsibility especially in the area where there is an intervention that might generate a certain degree of controversy that we be subject to external oversight. To be given the privilege of doing science, often at the expense of the public, we have to cede the right for this priviledge in exchange for independent oversight. When scientists go rogue, it carries a deep cost to society.

Q: You mentioned clinical applications should be permitted. My question is, we should distinguish between the research and therapeutic purposes. For therapeutic purpose, we should focus on solving severe genetic diseases. But for research purpose, it's focused on creating generalizable knowledge. We should distinguish these two goals, and we should have the different ethical principles or the rules aobut and how to use it. Thank you.

A: I think that point is very well taken. The difference between the motivation for general knowledge, versus the actual prevention or relief of suffering. They are two different goals.

Q: I want to raise some doubts about your suggestion that in vitro gametogenesis and the generation of embryos from in vitro gametes might provide a way forward for ensuring safety and accuracy. In theory, it sounds pretty sensible. But thinking about the translation pathway from this point in time forward, I am uncertain that in vitro gametogenesis will be able to do that. I think we're much further away from creating babies from in vitro gametes than we are from applying gene editing in the conventional way. I could not imagine a convincing reason for in vitro gametogenesis followed by IVF.

A: Infertility is a clear indication that would warrant that use. But I agree that these dual strategies of enhancing the fidelity of editing at the zygotic stage to ensure the uniformity of gene editing, versus doing it through an in vitro gametogenesis derived way. I agree that the science of in vitro gametogenesis is somewhat more futuristic, but the progress there is truly remarkable and it will be driven by the interest in clinical applications here.

Q: What about the assumption that the parents have an affirmative right to a biological child? There are other approaches to building families.

A: I can't say I'm an expert on that concept. I could say that speaking as a parent, there is a strong desire and drive to procreate on some level.

Q: I really appreciate you mentioning self-regulation and seeking consensus on this issue. If someone can just ignore the norms in our community... then what could be the response or what could we do in the academic community in response to this? There seems to be a positive response from the capital markets.

A: You're asking about how does one balance the call for ethical oversight in a responsible way forward, with market forces and the possibility tha tthe industry can move forward. I note that in passing that the CRISPR trials, there wasn't a lot of publications associated with those trials. I think there's a need for more transparency in general for commercialization and commercial clinical trials, but especially for the first business use of highly innovative technology, especially if we apply it to the germline. I think that raises the bar for requiring transparency and willingness to be subjected to external oversight. Perhaps for that reason it is better done in an academic realm instead of the commercial realm.

Q: You are optimistic about the possibility of universal consensus. Immanuel Kant is famous for trying to come up with that universal ethics. I wonder if that is realistic, though. There are contested values even among us. Even yesterday we have seen all sorts of interesting moral values like from Islam, neoconfuscious ethics, and others. I am questioning your optimism about the possibility of consensus.

A: You have talked about the traditional tensions in authoritarian and other approaches. I think we aspire to universal principles. In every day life, I use utilitarian principles all the time. I think that's what's so exciting about the presentations yesterday where we saw many international principles including religious principles. This is part of the reason why we have to bring together communities like this to struggle with those issues so that we can get to the point of acceptance, which we can't do without finding common ground. Genetically modified organisms in its highest application could feed starving millions and yet it's hard because of missteps early on in their introduction and there wasn't broad societal consensus for those introductions, and it has caused a terrible cost for the world and I hope we don't go that way for human genome editing.

Q: What do you mean by consensus? In 1975, getting an unruly group of 100 scientists who arrived at an inevitable consensus at Asimolar took a lot of work and perhaps persuasion. There are asymmetries even within a vanishingly small scientific community, and a global community is much much more heterogeneous. In this domain, what would consensus look like? What sets the agenda for asking the questions that produces a consensus? What about broad societal consensus? Do you see this as following after the issue of scientific consensus? Are matters of scientific consensus achieved, and then broad societal consensus is an effect of what we have said? Or do the societal consensus inform what we ask ourselves as scientists?

A: Leave it to the philosophers to pose good questions... You're pointing out the obvious, which is that consensus is virtually impossible. But it's still a guidepost and it's still an important aspiration. Scientific consensus has to be sought contemporaneously with social and moral consensus. There has to be at least an attempt to come to some broader understanding, and it doesn't have to be universal. We do this all the time in lots of different interventions, whether in vitro fertilization or various treatments for a range of diseases. You might not reach consensus, but you will continue to strive for broad acceptance which can lead to greater embrace by the community especially for areas like that. I think we have seen this evolution in vitro fertilization. It's not clear that there was acceptance when it was first rolled out, and I think there's acceptance today. I think we can do better next time we do something like that as a society.

Q: .... We found that the precautionary principle was very important to us. Next was a self-regulation principle, which this community has been hyping for a long time. One of the challenges we had with genetically engineered plants and animals is that we have a scientific industry group that has been opposing regulation. But what I hear you saying is that despite this failure, that the national academy people here have not denounced this yet. The national academy of science is very... on gene editing for conservation purposes is much more pushy for strong regulations than I hear today. Why don't we need more stronger regulations? Why not a moratorium?

A: I can't think a single practicioner that goes against the norms of the fields represents a gross failure of self-regulation. We aspire to self-regulation in as many instances as possible. The fact is that one practicioner is unlikely to have as profound an effect as the entire community. As long as the entire community buys into these principles, just like the gene therapy community did early on, I think you can get as close to this ideal as possible. If you have widespread flaunting of the principles, then I think you need stronger regulations. I think we saw this with the proliferation of nonsense stem cell therapies. What happened is that there was a rise of this unregulated economy of stem cell treatments. Some have started to speak out against that, and in favor of enforcement of binding regulation. I do think that binding self-regulation is a principle that the community should continue to adhere to and aspire to.