Less than a decade ago, scientists gained the unprecedented ability to alter the genetic code of living organisms with the development of a tool called CRISPR-Cas9. In late 2015, recognizing the power of the CRISPR technology, a group of scientists held the first International Summit on Human Gene Editing. Led by Caltech's David Baltimore , president emeritus and Robert Andrews Millikan Professor of Biology, the group concluded that gene editing technology was far too underdeveloped to be used on humans.

This year, on the eve of the second international summit held in November in Hong Kong, a scientist announced that he had already edited the genomes of human embryos and inserted them into their mother's uterus—in spite of an international agreement not to carry out such an insertion—and that the twin babies had just been born. We sat down with Baltimore to discuss this report and the outcomes of the second international conference.

What was the motivation to hold this summit for the second time?

At the first summit, we made a clear distinction between somatic gene editing (where no edited genes can be passed down to the next generation) and germline gene editing (where the edited genes are passed down). We concluded that somatic editing was like any other medical intervention: once it is deemed safe and effective, it should become part of ordinary medicine. However, germline editing raises many questions, both practical and moral, and, we concluded, it needed much more study before it could be considered for human use. We did encourage further research to perfect the methods.

The second summit, three years after the first, was meant as an update—to take stock of the advances of the previous three years and to decide how perspectives had changed. We asked questions about how research was progressing on somatic editing and recognized that many investigators were initiating clinical trials focused on a variety of diseases. We saw that new, safer methods of germline editing had been developed, but we concluded that the moral and practical uncertainties remained to be resolved, and we continued to believe that it would be irresponsible to initiate trials in humans.

The day before the summit, a scientist announced that he had modified the genomes of two embryos, and that they had been successfully carried to term and born. What was the committee's response to this?

We were surprised, to say the least, when we heard just before the meeting began that somebody was going to announce that he had actually implanted gene-edited embryos back into a woman and that she had given birth to two children. What had been planned as a largely academic discussion became a media circus. This announcement was a very serious ethical challenge.

In a sense, the publicity that surrounded this basically irresponsible act had a plus side. It focused the attention of the general public on certain activities of modern science. Sometimes, it takes a very dramatic situation for nonscientists to pay attention long enough to recognize the advances of science; it produced a teaching moment. I think you'll find that many of the people who were at this meeting in Hong Kong are now back in their own countries and cities and laboratories, where they are being asked to talk to their local radio stations, talk to their local community organizations, and that's positive, although it does in no way justify the actions of Dr. He Jiankui, the scientist who carried out the human germline editing.

You and the committee described being disturbed by this research, but are you hopeful that one day germline editing could be conducted safely and responsibly?

I certainly hope that we will reach that point. It's part of my general belief that modern medicine will have the ability to ameliorate much of the burden of the diseases that we still suffer with as human beings, like cancer, inherited disease, heart disease. I'm hopeful that we will ameliorate those and that the world will become, in that sense, a better place because of modern biology.

Other than ameliorating disease, can genetic engineering solve other global problems?

Most global problems have a strong social dimension, and social problems are not solved by genetic tricks; they're endemic to the culture of our society. However, the general worry is that if we develop the ability to modify the germline, only wealthy people will be able to take advantage of that, and so it may exacerbate the difference between the opportunities available to the wealthy and to the impoverished. This concern applies to all medical advances and is not specific to gene editing. It's a social problem created by medical successes, and we have to think about how to make such successful treatments widely available.

What do you see as the future of the field of genome editing?

The science is advancing very rapidly. New ways of using the technology are being invented continually, so it will just get more and more effective and powerful over time. That's what I see happening, in particular in the area of somatic gene therapy.

We can do a lot with somatic gene editing. Some of it involves the direct modification of an inherited genetic problem. For example, sickle cell disease is caused by a single mutation in the hemoglobin beta gene. We could directly correct that or modify other genes to provide a replacement for the defective beta gene. That would majorly improve the lives of people who inherit sickle cell disease.

Another application of somatic gene editing is immunotherapy for cancer. We are today treating people who have cancer by modifying their immune cells and making them attack the cancer cells and kill them. You're actually getting the body to clean itself up, in a sense.

Both of those things are in clinical trials today, and we expect this will become a part of medicine within the next few years.

I suspect that as time passes, we will want to rethink whether gene editing ought to be used in modifying the germline.

There are thousands of single-gene defects in humans. We're going to see, I think, some pressure to use germline technology in people where the medical need is great, such as sickle cell hemoglobin disease, Huntington's disease, and others. I think we'll find situations in which the benefit-risk ratio is very much in favor of the benefit. At that point, I think there's a moral argument to be made that we have to use gene editing because we can improve the lives of people.

What are the challenges in doing germline modification? Where do we need to be cautious?

There are two kinds of practical challenges in using the technology. One is an off-target effect: you want to modify a gene at a certain position, and you inadvertently cause a change somewhere else in the genome. These are accidental errors that would be passed on to later generations and need to be carefully avoided.

The other problem is that, if the edit is done as the cells are dividing in the embryo, you could have a situation in which the embryo becomes what we call mosaic: some of its cells are edited, some of its cells are not edited.

I think people who work with this technology are reasonably comfortable that off-target effects can be assayed and minimized. But there are real questions about whether we know how to handle the mosaicism.

What are some of the moral and ethical boundaries surrounding gene editing?

Because germline editing involves making alterations in the genome that would be passed down through the generations, it should only be done when we have a clear idea of what the consequences of the gene alterations will be. Right now, I believe that limits the use of germline alteration to genes with predictable behaviors, like that which causes Huntington's disease.

Some people oppose gene alteration on basically religious grounds. They would say there should never be gene modification. For instance, if you believe that humans are perfect, then you may not want to modify them even if they're not healthy.

Then there are, I think, lots of other people who believe that if there is a way to make the lives of people better, we should do it. Those people now have to make another distinction: that distinction is between a modification that is only in your own body and a modification that is inherited by your offspring. That's a fundamental difference, not because of the mechanics of it but because of the moral status of the individual; by modifying the genes, have you modified some essence of the individual? Again, there are people on both sides of that question.

We're going to debate these questions over the next years and decades, and there are always going to be people on both sides of the issue. We will have to decide to go one way or another. I think it's pretty clear where I would go, but I don't have any more important status than anybody else in this discussion, and so it really will come down to what the majority of people think is the right way to behave.

It seems that your camp of researchers must also determine what modifications will actually improve people's lives and what modifications are for aesthetic preferences or maybe superfluous.

Yes, that is a fundamental distinction that is very hard to make. When is a gene alteration a way of improving an individual's health and when is it an aesthetic preference or a socially desirable characteristic? That's a conversation that's going on with the whole world today. I have emphasized the easy case, which is where an individual has genes that are in some way driving ill health. But how about genes that people would just like to see in their children? Blue eyes, or intelligence, or the like. I think the general feeling is that we shouldn't be doing that, but there is a concern that once we perfect the methods for improving health, the same methods could be used for other purposes. That is a "slippery-slope" argument, and people are even saying we should not use the methods for dealing with serious diseases because it opens up the slippery-slope concern.

Predicting all the consequences of a gene alteration is difficult. For instance, in the U.S., sickle cell disease is clearly something we would want to avoid if possible—but in Africa, the sickle cell trait protects an individual against malaria and therefore has a positive consequence as well as a negative one. So there is a risk/benefit calculus to consider for any gene alteration, and we simply may not know enough to make the judgment confidently. So we must ask whether we know enough to make a judgment, or would we be best off taking a humble stance in the face of uncertainty. Thus, our advances in science face us with a mixture of practical and moral questions, and opportunities that are not easily resolved.

The First International Summit on Human Gene Editing was convened by the Chinese Academy of Sciences, the Royal Society, the U.S. National Academy of Sciences, and the U.S. National Academy of Medicine. The Second International Summit on Human Genome Editing was convened by the Academy of Sciences of Hong Kong, the Royal Society, the U.S. National Academy of Sciences, and the U.S. National Academy of Medicine.