CHAIRMAN KASS: The first presentation will be by Dr. Collins. We also have as our second speaker of the morning Dr. Gerald Schatten, who is now the Director of the Pittsburgh Development Center, the Deputy Director of the Magee Women's Hospital Research Institute, and a Professor and Vice Chair of Obstetrics and Gynecology and Reproductive Sciences at the University of Pittsburgh's School of Medicine.

The procedure will be we'll hear from Dr. Collins. We will have a discussion of his presentation. We will break, and then after the break, we will hear from Dr. Schatten.

Thank you both very much for joining us this morning. We look forward to the presentations.

And, Dr. Collins, please.

DR. COLLINS: Thank you, Leon, very much, and it's a real pleasure to be here to speak to this distinguished group on a topic that I believe is of considerable importance.

I come to you as primarily a scientist. My efforts, I think, this morning are largely to try to provide a reality check on what, in fact, are realistic scenarios for enhancement and to try to contrast those with many of the ones that are floating out there in various formats, some of which are truly unrealistic now and probably unrealistic ever.

The challenge here is to try to sort through those, and one of the things I will try to do is to see amongst this very long menu of potential scenarios which are the ones that perhaps are most worthy of attention at the present time because they are both troubling and imminent.

And I think you will see that that is actually a fairly modest subset of the total list of scenarios that have been thrown around, and if I succeed this morning, maybe it will only be by trying to narrow the field a little bit here of perspectives that might be most in need of attention.

I know this group has already had distinguished presentations by others on the topic of sex selection; a talk yesterday, I gather, about aging. I will touch a little bit on those, but I will try not to duplicate information that's been previously presented.

So let me say as the Director of the Human Genome Project, I think there is a broad public perception that while this is an enormously exciting enterprise, reading our own instruction book as we commonly now are referring to this, but there are also concerns about where this leading and particularly in the areas of enhancement, and so it's not difficult to find cartoons like this one where the person carrying the briefcase says, "Sometimes I wish they had never mapped the human genome. These look like perhaps future members of the Congress, I suppose," in somewhat unusual garbs.

Cartoons are actually a wonderful window into sort of what the public is thinking about because cartoonists, after all, are both members of the public and reflect their perceptions of what the public's perceptions are about what's happening around them. So you can find lots of cartoons like this.

So even though those of us working on the human genome would argue passionately that we are engaged in a highly ethical, moral activity that has at its most fundamental basis the effort to identify the causes of disease and to alleviate human suffering.

It is clear there are other connected perceptions of where this might lead us. I don't think I have to remind this group that the Human Genome Project at the very beginning decided that these ethical, legal and social implications, many of which relate to nonmedical uses of the information, were so profoundly important that they deserved to be part and parcel of the genome project from the beginning.

And so rather than letting somebody else worry about those things at some future time, we have been investing initially three percent, now five percent of our research budget into studying those ethical, legal, and social issues, and I might say we have quite a portfolio of research that has been done on the topic of genetic enhancement, and I think out of that has come some thoughtful musings, but I would certainly not argue that we have done much more than scratch the surface.

I have looked over some of those musings, particularly from Eric Jungst, from Dan Brock, from Eric Parens, from Max Mellman, and I think there's much there that could be usefully looked at in terms of a foundation for the more philosophical aspects of these questions, but I would certainly not say that we've done that one.

In fact, I will argue that in our next phase of genomic research, enhancement ought to be one of our highest priorities for our ELSI program. Again, our program is aimed to do research in order to try to identify the issues and perhaps propose options for their solution. It is not a policy forming activity. We're not properly positioned to do that.

I guess one of the proverbs that I'm fond of, which often gets into these conversations, is this one. It's not good to have zeal without knowledge. There are certainly people out there who have a lot of zeal about science is the answer to all things. I think sometimes their knowledge is a little lacking in terms of some of the social aspects of what it means to be human.

On the other hand, there are certainly people out there with great zeal who would argue that all of this genetic manipulation is inherently immoral and dangerous, and I think oftentimes they don't have a terribly good grasp of the scientific facts.

So I guess this is an aphorism that can apply to all debates, but maybe particularly to the one we're having this morning.

Well, just quickly, let me say a word about the Human Genome Project, where we are, and particularly where it's going because I think it's relevant in providing some undergirding to our scientific discussion.

This project, now 12 years on, has basically come almost to the point of achieving all of the original goals and having done so well ahead of the original scheduled timetable, which is gratifying, indeed.

So, for instance, just a year and a half ago this initial draft of the human genome sequence was published. By next April we will have completed the sequence of the human genome in the highly accurate form that has been defined as the essentially completed target, and that will be a moment of some celebration coming as it does exactly at the moment of the 50th anniversary of Watson and Crick's paper on the double helix.

Just last week we published in the same journal with a somewhat similar motif on the cover, the sequence of the mouse genome, not the finished sequence, but an advanced draft with a thorough analysis of what can be learned from it, and many of the commentaries on that, including from some of my colleagues, referred to the Rosetta Stone kind of analogy here that have two mammalian genomes enables you by the comparisons between them to zero in on the parts that are most functionally important.

And I must say it is enormously exciting to be able to do this and to look and see what is happening in our mammalian neighbors, and to realize that there's a large fraction of the genome which is most strongly conserved between humans and mice, is of unknown function. Only about half of it is the stuff we expected, and the other half we didn't expect and presumably is involved in important things involving gene regulation and probably other important biological properties as well that we just have not had the tools to be able to appreciate.

So this is, I think, in many ways from the scientific perspective what we were able to say last week was even more exciting than what we said a year and a half ago, but less exciting than what we should be able to say in another two or three years when we have additional mammalian genomes and you can really do large scale comparisons.

Well, where is genomics going? We're in the midst of a very intense planning process to try to define the highest priorities for the next phase. Having seen the original goals of the genome project accomplished, I think we are appropriately at the end of the beginning, and we need now to build upon that foundation and move this enterprise forward as rapidly as possible for medical benefit, and that involves a host of activities.

There will not be a sort of single defining goal for the next phase the way the human genome sequence was at least perceived as the single defining goal for the first phase. And so we are focusing on a whole host of issues, and you can see some of them listed here.

I want to particularly say a word about the one related to medical genomics because we have been remarkably successful in the last dozen years identifying genes that play a role in Mendelian disorders. There are very few Mendelian disorders of any real frequency that have not yet had their genes identified because the ability of the tools of the genomic effort to make that process very streamlined. Things that used to take a decade or more can now be done in a couple of weeks by a good graduate student working all by themselves.

But the real frontier is those kinds of diseases that are not simply inherited, where we know there are inheritable influences, things like diabetes and cancer and heart disease and mental illness, but where the inheritance patterns are messy and undoubtedly are a combination of genetics and environment, and the genetic component is not just a single gene. It's multiple genes no single one of which has a very strong effect, a very difficult problem, orders of magnitude more difficult perhaps than finding the gene for cystic fibrosis or some other Mendelian disorder.

But I think you can anticipate that we are going to break that code, the genetics of common disease, in the next, oh, five to seven years. One of the ways we will do that is by understanding variations. So if this is a typical page of the instruction book, and it is -- this is a sequence from Chromosome 7 -- that was my sequence on one of my Chromosome 7s. That might be the sequence on the other Chromosome 7, or if that was yours, the other one might be mine. We differ only in about one position out of every 1,000.

Most of those variations have no particular consequences, but a small number of them play a role in risks of disease, and we would really like to understand those because each one will shine a bright light on the pathway that's involved in disease pathogenesis, a very important thing to learn about.

So we're building a catalogue of those variants, and we're also building an understanding of how they correlate with their neighbors, which they happen to do, something called haplotype map, and in the not too distant future, we will be in the position of being able to do a truly powerful set of experiments on virtually any disease you can think of, which is a very simple kind of experiment where you identify individuals affected with the disorder, and you identify controls who clearly don't have the disorder, and you test them for all of the variation in the human genome.

And you look for circumstances like Gene B here where the orange spelling of Gene B is associated with disease, and the blue spelling is not. Most of the time, of course, most of the 30,000 genes will not be associated with a particular disease, and you'll see something more like Gene A, and you can say, "Well, at least that variant in Gene A is not a predisposing factor to that disease."

So a pretty straightforward case control kind of analysis built upon this remarkable database of information about human variation, which is under construction right now, should enable us, I believe, in the course of the next five to seven years to find the major contributing genes for diabetes, heart disease, cancer, mental illness. I could go on.

Think of whatever disorder runs in your family. It's probably under the real likelihood of having its genetic components unraveled in the next five to seven years.

Some of that, of course, will allow an individualized form of preventive medicine where we can each figure out what our greatest risks are and adjust our preventive strategies accordingly, and that sounds like a good thing because it would probably be more effective than a one size fits all approach.

Some of this will enable us to figure out why drugs don't always have the same results in people who are given the same drug for the same disease and perhaps enable us to do a better job of choosing the right drug for the right person.

But of course, some of this will begin to tempt people to use this information in other ways that we're coming to because particularly while this strategy as I'm outlining it for you is intended for the uncovering of medical risks, the same strategy can, of course, be applied and probably will be by curious researchers to look at the genetic contributions to nonmedical traits, and those, of course, might include things that sound bland like handedness or a little less bland like baldness or not bland at all like intelligence.

And undoubtedly those kinds of studies are likely to be underway using these new tools, and I think actually that the studies themselves are probably valuable and appropriate, but the use to which that knowledge will be put obviously will challenge us in ways that the medical applications present perhaps somewhat fewer challenges.

So let's get on to enhancement then. A definition. I'm sure everybody has their favorite one. Maybe this group already has one that you're fond of. This one is borrowed from Eric Jungst, who as I said earlier I think has written quite usefully on this topic in a series of articles. "An intervention designed to improve human form or functioning beyond what is necessary to sustain or restore good health." We could tinker about that.

Of course, I am sure you have already discussed and no doubt we will again this morning the difficulty in drawing a bright line between medical improvements and enhancements, and I don't think we're going to be able to do that.

Yet at the same time, I think that's often the case when you're having a difficult moral conversation. Bright lines are hard to find, and I don't think that should be discouraging to an effort to try to set out general principles, recognizing that there are going to be gray zones where it's a little difficult to decide how the general principles apply.

Well, pathways to enhancement are numerous, and they're covered nicely in the briefing paper that was put together for this meeting. I'm not going to say a lot about somatic enhancements, although I will give some examples and try to see where they fall on this sort of scale of concern because I think somatic enhancements obviously can be genetic, but they're often nongenetic. They can even be nonbiological in terms of enhancements, such as, for instance, education, which is an enhancement of functioning and one that we all find rather admirable.

I think most of the intention in inviting me here is to talk more about the germline possibilities, and there one can divide this up into two categories, although I don't think we should be slavishly adhering to them because there may be other sort of potential subsets here, but I'm primarily going to talk about these two categories, one being gamete or embryo selection, the other being germline manipulation.

Embryo selection, of course, is the opportunity already present now with pre-implantation genetic diagnosis to select which among a series of embryos are going to be reimplanted and allowed to come to term.

Gamete selection has perhaps less attention attached to it because I don't think there is general expectation that it will have broad utility, but obviously in the case of sex selection, it's a very interesting topic, and I imagine you may have already discussed that.

It does seem to create a somewhat different moral discussion when one is talking about the ethics of sorting sperm for X or Y chromosomes as opposed to applying pre-implantation genetic diagnosis to sort embryos from whether they are XX or XY, and that is, I think, often a useful discussion to have because the gamete selection strategy isolates you away from some of the other very compelling arguments about the moral status of the embryo and allows a sort of cleaner discussion about what are the social goods or evils associated with broad alterations in the sex ratio and inequities in access to that technology.

But, again, I think that's a topic that you have touched on before. I'm primarily going to talk about PGD for embryo selection and germline manipulation.

Let me start with germline manipulation, and let me say that this is also referred to in your document as directed genetic change, and it would probably be good if at some point the community arrived at some standard term so that we all knew what we were talking about, but there are various other kinds of terminologies. Germline gene therapy obviously is a subset of this, although many of these are not necessarily therapeutic when you get to the enhancement arena, of course.

Germline manipulation, as the scientists on this group know and many have led the effort to make this happen, is routinely carried out in experiment animals, especially mice. I would argue, and maybe there's a little controversy here, but I don't think there's actually that much, that current safety considerations are sufficient to rule out the application of germline manipulation in humans; that we would have to have very strong assurances of doing no harm not only to the individual who is being created by this particular strategy or having their biology altered anyway, and also that it would not create harms to future generations, and that is a very high standard that we are nowhere near meeting and I think we'll not be for the foreseeable future.

So one of the things I think we could probably talk less about in terms of using up the energies of the ethics community is whether or not it's appropriate to begin germline gene therapy right now. We clearly do not have the safety considerations anywhere near in hand to consider that as a justifiable activity.

Furthermore, I think when it comes to germline manipulation for medical benefits, I have a very hard time coming up with scenarios where this would actually be the treatment of choice. Even if we could do it safely, why would we want to? What would be the scenario where you'd argue that that's a good thing to do?

If one allows, I should say as a caveat, the possibility of pre-implantation genetic diagnosis as an acceptable mode of therapy, and obviously there are some who will argue against that, but if one allows that, in almost every situation where you might consider germline manipulation to avoid the birth of a child with a disease, PGD would be infinitely safer and cheaper and more likely to succeed.

So only in the circumstances, I suppose where someone has moral objections to PGD, but not to germline genetic manipulation does this seem to be an argument worth having, and even in that circumstance safety considerations at least for the foreseeable future make it totally unrealistic.

I think in the longer term though, if we can imagine those safety considerations ultimately being taken care of, and I think anybody who argues that safety considerations are always going to be a problem is probably on somewhat thin ice because science does move forward and we do figure out amazing things over the course of time.

Actually it's not the medical applications of germline gene intervention that would probably be the topic of major interest because, again, it's hard to come up with a medical scenario where one would necessarily want to do a lot of that. It's more the enhancement scenarios.

And so in that regard, it probably belongs on the screen as a general discussion, but not in the sense that it is any time realistic to be happening soon scientifically.

So germline manipulations come in several flavors, the simplest being the insertion of a transgene, a foreign piece of DNA often carrying a gene that you wish to have expressed, and we all do this in experimental animals all the time, and it's fantastically valuable from a scientific perspective.

There is work going on to try to do a more surgical job of this because, of course, when you insert a transgene, it's landing in some random place in the genome and potentially interfering with the function of the gene nearby and the risks have only been highlighted by the recent experience in the X-linked SCID where the child developed leukemia as a consequence of where the retrovirus seems to have inserted next to an oncogene. At least that is the current hypothesis based on the data that's present.

So insertion of transgenes into the germline when you don't know the consequences clearly does not seem like an ethically justifiable activity, but if you could come up with a way to go in and change that T to a C without leaving any footprints on either side of it, well, then that would potentially come much closer to an acceptable perspective on a safety basis.

There are experiments underway to try to do that. Obviously in mice we can do that using strategies called by sort of hit and run, knock-in, and other kinds of technologies, but I think we're a long way, again, from being able to argue that that's realistic in human applications, but it is potentially the pathway that people will want to go down to if we're going to seriously consider such germline manipulations for humans.

But the other pathway that I think does deserve a fair amount of attention has gotten some discussion in the various treatises on the use of germline manipulation, but perhaps not as much as it deserves because I think it's actually coming along in an interesting way, is the ability to construct artificial chromosomes.

Now, here the idea is not that you're going to try to insert a piece of DNA into one of the existing chromosomes, but you're going to build from scratch a chromosome that has all of the necessary materials, centromeres, telomeres, replicating sequences, and so on, and then use that as your vector to carry along maybe not just one gene, but maybe a bunch of genes that you think would be useful.

This is still very much, I think, an experimental enterprise, but I would say compared to where it was three or four years ago progress is certainly being made. The idea is pretty much an engineering effort here where you construct this artificial chromosome, and there are various ways to do this so that it has the appropriate components and then transfer that into the cell nucleus. If you were contemplating doing this in the germline, that would be an embryonic cell, and the artificial chromosome then takes up its home in the nucleus and gets replicated every time the cell replicates.

Now, notice, by the way, there's only one copy of it at least in this particular version that I show you here so that if you were to carry this experiment out, and this has actually been done in animals, only half the offspring would actually get that chromosome. The other half would not unless you went to some trouble to try to duplicate it.

It has been possible though to create such artificial chromosomes to show that they can pass through the germline, albeit with some difficulty, so that they can actually express proteins from the chromosome that has a particular gene imbedded into it as shown here by these red bars, a functional protein of interest.

A particular recent example comes from cystic fibrosis research where a human artificial chromosome carrying the CFTR gene was constructed by a somewhat unusual fashion that I won't go into and placed not into a human cell. This is actually a Chinese hamster ovary cell where you're actually looking now at the chromosomes stained, and you can see the big orange ones are the hamster chromosomes, and that little, tiny dot there is this artificial chromosome carrying the CFTR gene, which appears to be stable in this particular cell culture. So every time the cell divides, the artificial chromosome divides, and the copy number seems not to vary drastically from cell to cell. Obviously another interesting problem, and I'm sure Liz Blackburn could tell us much more about chromosome dynamics and the issues that reside here.

Again, I don't want to imply that I think artificial chromosomes for human applications are right around the corner, but I do think it is a pathway towards the potential insertion of foreign genetic material into a human cell that is going to get a lot of attention and a lot of scientific exploration.

Well, let's move then to PGD because, as I said, while the germline manipulations are interesting to talk about, I think they are some distance away, and for your purposes in terms of issues that are perhaps riper for consideration, particularly in terms of whether further oversight is needed, I think PGD is certainly very much on that list, and I know you've already talked about it to some extent.

This is a table that came from a presentation by Joe Leigh Simpson at a recent workshop we held in conjunction with our planning process for the future of our ELSI program. Joe Leigh is coming up with these numbers that approximately 6,000 cases of PGD have been carried out, most of them for chromosomal reasons, 20 percent for Mendelian disorders like Tay Sachs or cystic fibrosis, and at least his bullet says it's attractive to couples seeking to avoid clinical pregnancy termination, and I think that is, in fact, the major attraction for people who otherwise could perhaps achieve the same information by CVD or amniocentesis, but of course at a later step, a considerably later step in the pregnancy.

The process for doing this, I think, is probably familiar to most of you. It involves carrying out in vitro verification, then from the embryos produced, aspirating off a single cell. That single cell out of eight, of course, contains the entire genetic endowment of the embryo. Technologies using the polymerase chain reaction are able to make a diagnosis on a single cell, and then a decision is made about which to reimplant based upon the results of that.

So in this simple schematic the ones shown in red were found to have a Mendelian disorder. The other three are not affected and are, therefore, transferred to the uterus.

Again, this is a very technically demanding enterprise, and it still is the province of relatively few medical centers, but I think it is very likely that that will expand over the course of time, and I expect Dr. Schatten can speak more coherently about what that likely course of expansion of this technology will be when he comes to his presentation on ART in general.

I do think that this is an area where the opportunity for at least contemplation of enhancement strategies becomes much more immediate, and therefore, it's probably appropriate to spend a little time on it.

Some of those enhancement strategies, again, I think are based upon a deterministic view of what genes can do for the outcome that's not warranted by the facts, but it certainly makes for interesting discussions. And in that regard probably a very informative, albeit very science fiction, presentation of the notion of the application of PGD on a very broad scale is provided by the 1997 movie Gattaca, and I thought I would show you, assuming our technology works here, a clip from that, a clip which I actually think is extremely well done in a cinematic way and thought provoking, albeit based upon scientific premises that aren't quite right.

So let's see if we can make this work.

(Whereupon, a video was shown.)

DR. COLLINS: You can conceive 1,000 times and never get such a result. An interesting statement, and actually it raises a potential problem with this whole scenario because that implies the ability, I suppose, if this was going to happen to have 1,000 or more embryos to choose from, which is a bit of a biological quandary.

And, in fact, that is one of the many ways in which this scenario begins to fall apart scientifically because if you were, in fact, to attempt to try to optimize for ten or 20 phenotypes, as the smooth tongued counselor here was suggesting, and you consider that each of those phenotypes would probably be influenced by five or ten genes, each of which would have perhaps more than two alleles. You quickly get into a combinatorial problem where lacking a million or better embryos, the ability to actually do this in a fashion that gets you very far is pretty limited.

Of course, that's only a small part of the scientific arguments because the concerns of trying to optimize offspring if they are predicated on this kind of scenario assume a degree of genetic determinism that we know is not correct, and it will not become correct just because we get smarter about genetics.

Just because we get to understand the nature part of the nature/nurture equation doesn't mean that the nature part becomes quantitatively more important. It just means we understand it better.

Just the same, I think that clip, in fact, that whole movie, is actually a thought providing exercise. It does to its credit basically point out the flawed nature of the premise because the scene you just saw where the embryo is being selected then gives rise to Vincent's brother.

By the way, Vincent is a kid who was born in the natural way, not benefitting from all of these particularly wonderful technologies.

His brother, on the other hand, did, as you saw about to happen there. The brother ends up committing murder and smoking and drinking to excess and having a generally rotten life, whereas the natural kid is the hero who can do almost anything because there is no gene for the human spirit.

So the movie discredits its premise, but it's still pretty interesting to watch and probably worth using in some of these discussions for people who don't quite get it.

I must say that scene was extremely well done scientifically right down to the four eight-celled embryos on the screen. They must have had good advice about some aspects of that.

Well, faced with all of this, to come to I guess what I'm hoping to help you with a bit is how to prioritize these enhancement concerns amongst all of the things that are out there. I tried this out on Leon and his staff a couple of months ago, and it seemed to be a useful way to organize things. So I'm going to try it out on all of you now.

Basically this is a proposal that one tried to formulate a two dimensional graph here. On the X axis we have level of concern, ranging from no concern. Actually it would be admirable to do this sort of thing. Over on the right to unacceptable.

The time table though, being the Y axis, is also important because I think you probably want to pay attention to things that are somewhat closer at hand and probably not to the things that may be never, and certainly the 100 years off are not as interesting as the ten years off.

So you could say maybe the zone of concern in my graph is the sort of lower right corner here, and things that fall into that category maybe are more deserving of attention than things that fall in other parts of the matrix.

So let's just for fun pick some particular types of enhancements and see where they fall, and I know as soon as I do this that you're all going to disagree with where I put the particular element, and that's interesting. And in fact, if we had more time, it might be fun to get everybody to fill out their own, and we'd compare and see how close were are, whether there's a large standard deviation in the placement of points.

And we'll also quickly see that it is hard to do this without getting fairly specific about what the application is for.

Well, to begin with something a little trivial, hair coloring. Well, I don't think it's unacceptable, but I don't think it's admirable. So I'm going to put this one sort of in the acceptable but maybe a little on the questionable side.

Music lessons. That's an enhancement of the somatic, nonbiological sort, and I happen to like music. So I think more people should know more about it and enjoy it. So we'll make that an admirable thing to do in terms of enhancing your children.

Cosmetic surgery. These are in no particular order, by the way. Well, okay. I have a little more trouble calling that admirable or even acceptable. It's a little more in the questionable direction, but people may disagree.

Exercise is a way of enhancing our physical performance. Ah, I think these days we ought to do more of that.

Fluoridated water, enhancing the health of our teeth. Well, although there's some arguments about that in terms of whether there might be untoward effects, I think most people would say it's on the admirable side of acceptable.

Immunizations. Obviously a biological form of enhancement. I think that's admirable unless, of course, there are some side effects that you didn't expect.

Not to be at all silly about it, prayer. I think that belongs on our list here. Prayer, if you believe in prayer, as I do, prayer is a way perhaps to enhance one's own existence and perhaps even try to influence our outcomes. That's what prayer is intended to accomplish in certain circumstances. So I'd say prayer is certainly an admirable form of enhancement.

I hope you don't find that too over the top.

Going from the sublime to the ridiculous, Viagra, on the other hand -- well, okay. I guess we now consider it acceptable, but a little on the questionable side of that.

You had a presentation about Ritalin, and here it immediately will come up. Well, Ritalin for who? Ritalin for a kid who has severe ADHD and is having major school problems is one thing. For a normal kid who's about to take their SATs, that's another thing. I'm sort of considering the former here by placing it in this part of the curve.

Mood altering drugs, well, more questionable. Recreational drugs, I think we'd all agree unacceptable. Maybe the same drug but used in a different way.

Now, you notice all of these are kind of things that are here now. HFH for normal kids, growth hormone. I think based on deliberations I've heard, there's pretty good consensus that for normal kids to try to achieve exceptional height in order to improve likelihood of basketball star status is an unacceptable application of that kind of enhancement.

Ditto erythropoietin for athletes, although erythropoietin for people with renal failure is a godsend.

Sex selection you've talked about. Maybe there's some deliberations about this. I'll tell you where I think it belongs.

IGF-1 to prevent aging. You had a presentation from Lee Sweeney, very interesting. Well, do we want to prevent aging? Is that really our ultimate goal? Is that part of the medical model?

Obviously it's easier when you take IGF-1 and apply it to athletes to enhance performance to say that's doping. That's not fair.

Individualized preventive medicine I guess is an enhancement because it enables us to avoid illnesses that we might otherwise suffer from, and I think that's a pretty admirable one, and many of us are hoping to see that come to pass as soon as possible.

Now, we get to PGD, and forgive me here because the first couple of these are actually not enhancements. They really are more along the medical model. PGD to try to identify an embryo with Tay Sachs disease and only reimplant those that do not carry that homozygous state.

I don't know exactly what the consensus would be about where that falls, but obviously it is going on in considerable sorts. There's been a recent survey by the Pew Foundation just issued a couple of days ago that says the general public seems to be teetering in the direction of finding that an acceptable use, but I think it's probably an overstatement to say that one can derive that completely from that response.

On the other hand, PGD for an adult onset disorder of incomplete penetrants like BRCA-1, much debate going on about that, and I think much more concern about whether that is an appropriate use of the technology.

Let's make it harder. We're going to discover genes involved in obesity, not just those rare forms of dramatic obesity that affect very few people, but the general genes that contribute to body mass index and people will be interested, I guarantee you, in applying that in a pre-implantation diagnosis setting.

And it could be that there's only a modest number of such genes, and it's imaginable that that might become feasible if one wants to recognize the fact that it's not going to be very effective. You're going to skew the odds a little bit, but you're certainly not going to guarantee a couple that they're not going to end up with an overweight child. Obviously environment is a huge aspect of this. That gets close to the unacceptable.

If we learn how to discover genes that are involved in whatever is measured by an IQ test, there will be pressure brought to bear to use that in a pre-implantation diagnosis setting even though the value of it in terms of how many IQ points you're going to skew the odds by is likely to be extremely modest.

Do we want to do that? I think that's much closer to the right end of the spectrum, and you know, people are already now studying genes involved in skin color. Is that something that we want to have utilized in this setting even further to the right?

We are going to discover in our abilities to understand pathways involved in obesity ways to interfere with those pathways, and we will probably in the course of time come up with pharmaceuticals that will not just treat morbid obesity, but will help normal people be more svelte than they otherwise would be.

Is that a good thing? Questionable.

Now we're getting into the zone of things that I think are much further off. As I said, human artificial chromosome transfer to human embryos are quite a distance off and, I think, carries with it a great deal of concern about what are you transferring and why, and many of those scenarios do involve enhancement, not just medical therapy.

Now we'll go really out there: enhancements that keep you from ever forgetting something? Well, I don't know. That's probably appealing in some ways, worrisome in others.

That keep you from ever having to sleep? Some of us seem to be pushing that boundary already, but we're not enjoying it.

To extend the life span to 200 years. Well, obviously that's a ways out there.

And I guess my final one here: designer babies with precisely predictable phenotype, the Gattaca scenario. Where does that fit? That's up there in the never. Let's be real here. Even if we have complete control of the genome, there are so many other features that fit into human phenotypes that relates to environment, that relates to free will choices those individuals make, and that relate to the part of us that is not really biologically definable, a part that Leon writes about, I think, very interestingly in the final chapter of his recent book.

So here's my matrix, and again, I guess having built it, I would say the things to pay perhaps most attention to are the ones that are over in this lower right quadrant that are both morally more questionable and also perhaps more imminent.

So what were those? Have you memorized all of the numbers? Well, I'll tell you again. Some of them clearly don't belong on your agenda: mood altering and recreational drugs. Obviously others are dealing with that, but some of the rest of these probably do, and I'm not going to reiterate them. We went through them already.

And, again, this is only a sampling of the kinds of issues that might well be deserving of current concerns and not putting them off for some far distant time.

Already you have touched on a number of these, which I think indicates that your planning process in reviewing the area of enhancement has been pretty effective in identifying the areas that are most in need of attention.

So let me conclude. First, I think defining the boundary between treatment and enhancement is difficult, but it's essential to attempt this, and I think those who argue, well, we can't get a sharp boundary so we'd better just not try are actually going to be defeating the opportunity to make some progress here.

Secondly, the moral boundary between acceptable and unacceptable uses of genetic technology does not map precisely to the treatment enhancement boundary. I think that's pretty clear from that matrix and the examples.

Certainly enhancement in the form of a vaccination is considered to be a highly moral activity. So you cannot simply say that enhancements are bad and treatments are good. It's much more complex than that.

There may be, in the mathematician's terms, a positive correlation function, but it is not a clean division.

Thirdly, many enhancement scenarios that are put out there are just not realistic, and it would be wise, therefore, to focus our energies not on those, but on the ones that are in my proposed zone of concern as defined by their presenting real challenges and being perhaps realistically possible in the next ten years.

PGD presents more immediate challenges than germline manipulation, and scientists can help provide a reality check on what's possible, but a broad societal dialogue is needed to define acceptable boundaries.

Again, I come back to the Pew survey that was just put forward a couple of days ago, which indicates the majority of those surveyed did feel that the government should be paying more attention to regulating these kinds of technologies.

So here we are looking out over the horizon here on a pathway which in many ways is going to be defined by our increasing knowledge of our own biology, which is in turn built upon our knowledge of the genome, and I think it's up to all of us to make sure that this particular prospecting team charts a course that leads us off there into that horizon in a fashion that has maximum benefit to human kind and avoids some of the potholes and pitfalls and chasms that we might otherwise fall into.

And in that regard, I think this Council is playing a very critical role in assessing those risks and trying to put forward reasoned, thoughtful recommendations about how we as a society might avoid those.

So thank you very much, and I'll be glad to engage in a conversation about all of this.

(Applause.)

CHAIRMAN KASS: Francis, thank you very, very much for what is your usual kind of performance. I mean clear, informative, and especially important to this group forthcoming about the seriousness of the need to assess the human and moral implications of where this is going.

I mean, I think it's for those qualities especially that the country is very fortunate to have you at the head of this project, and I mean that most sincerely. It's really a very, very wonderful thing that you've done in the past and done here again this morning.

Let me make a proposal. Dr. Schatten will be talking about PGD and its practice. I don't want to and I don't think we should raise the particular questions about the practice of PGD at this point. I think we should save that discussion for later and try to attend to the questions that emerge out of the selection qualities and the various assessments of what's possible and of what is morally admirable, acceptable, questionable, unacceptable in a way that Francis Collins has invited us to do.

So the floor is open for discussion. Rebecca Dresser.

PROF. DRESSER: Thank you for a very enlightening, comprehensible discussion.

A couple of questions. One is in my reading about speculation of germline interventions, I think most people think it would include PGD because you'd want to check to see whether the desired change had occurred.

And then my second question is once you move from PGD for a single gene disease and you're starting to look at enhancement interventions or enhancement screening, does the risk of false positives and negatives increase, that is, the chance that you would say, well, this embryo looks like it has the desired feature or the undesired feature, but it wouldn't show up, you know, if the child were actually born?

Do those concerns go up as you move toward enhancement screening?

DR. COLLINS: So, yeah, both good questions. I do think that if we were to undertake germline interventions, and again, I think that's quite a ways off, because of the extreme concerns about safety, ideally you would like to carry out a PGD after having done the intervention, and ideally you would like to sequence the entire genome of that cell that you had removed because that would be the only way of reassuring yourself that some untoward glitch hadn't occurred off there on some other chromosome that you didn't intend.

Obviously that's a very high standard to try to achieve that level of certainty, but I think the standard has to be very high, and that again underlines why it is so far away right now, but I think you make a good point. You wouldn't want to just sort of hope it turned out all right. You would really have to have a very strong scientific assurance of that being the case.

When it comes to your second question, the application of PGD to things which are non-Mendelian, I don't know if I'd call it a false positive so much as just the fact that the kinds of traits that people would be interested in in trying to optimize are not so neatly predictable on the basis of genotype.

And while you might be skewing the odds a little bit, you might still get not exactly what you counted on. I mean, I often sort of envision a well-heeled couple who have spent thousands or tens of thousands of dollars going through this to try to increase the likelihood that they will have a son who is a quarterback on the football team and gets an A plus in math and plays first violin in the high school orchestra, and instead he spends all of his time up in his room listening to heavy metal rock and smoking marijuana, you know. I sure he appreciates the music and all of that, but all of those aspects that perhaps parents would be most excited about enhancing are profoundly influenced by environmental aspects and by their own operating.

And if parents somehow imagine that they can just buy the phenotype instead of investing themselves on a daily basis in helping to create it by being good parents, then we've gone in the wrong direction. We'll end up with a society that's much worse behaved than it is now.

CHAIRMAN KASS: Janet Rowley.

DR. ROWLEY: Well, I, too, want to join my colleagues in thanking you, Francis, for coming and giving such a lucid presentation.

I was a bit surprised that you have within your diagram being able to do PGD for intelligence within a ten year period because I would have thought that the multiple components that go into intelligence are still so far in the distance in terms of our understanding them that that kind of enhancement is further off than ten years.

DR. COLLINS: Well, I'm glad you asked the question, and I should probably define what I meant by putting it in that timetable.

I don't mean that it would be very effective, but I do think there are people already, Robert Plomin most notably amongst them, who are by studying individuals who are at the extremes of the intelligence curves trying to identify variants that correlate with those who score highest on an IQ test.

It is clear from the study of identical twins versus dizygotic twins that how you score on an IQ test does have significant heritability, which says there are genes involved.

Now, are there 500 genes or 2,000 genes, or are there 15 genes that account for most of that heritability? We don't know. If it's hundreds or thousands, well, we've got a long way to go.

If it turns out that it is a modest number, a dozen or so, then it's very likely that some of those will get uncovered in the next decade because of this ability by having the catalogue of genetic variation to do things that are more powerful than what Plomin and others have had to do up until now, which was sort of linkage based.

This would move it into an association study. So I will be surprised if there are not reports in the next decade of specific alleles, of specific genes that are claimed to have a correlation with performance on an IQ test. Many of those will turn out to be wrong. They will not validate, but some of them will turn out to be right.

And then there will be at least pressure brought to bear on the PGD scenario by couples who would like to use that information when they find out that that particular testable allele on the average only changes IQ points by two. Many of them will be discouraged, but not all of them.

And again, by putting on the diagram, I did not mean to say that it's going to happen. I mean to say that the possibility of some couples wanting it to happen might very well be with us in ten years, and if we as a society think that that is scientifically unjustifiable because of the very small actual change in the outcome that it would predict and, more importantly from a society perspective deeply troubling, then it probably is appropriate in this next ten year interval to come up with a stance on that.

So I hope that helps clarify.

DR. ROWLEY: Can I just follow on?

So you brought up the use of what's called in our staff report mini chromosomes, so the artificial chromosomes, and so let's follow on with the intelligence. Let's say that within ten years' time we find ten genes. Maybe each one does only affect intelligence two or three points, but the combination of ten is a possibility of 20 or 30. It may be some of them may be complementary so that it would be not just additive.

So presumably these genes will be different chromosomes or different regions of the same chromosome, and the report does discuss the possibility of being -- because the mini chromosome can be used as effective for a large amount of DNA, you know the ten genes that you want to add on this mini chromosome.

What are the difficulties and the complexities of constructing a chromosome that gives you ten functional genes?

And of course, one has to understand that these genes are put together and added on the mini chromosome in a developing embryo that already has all of those genes derived from the normal fertilization, and I would think no way you're going to be able to cut out ten genes from both parental chromosomes.

But just deal with the complexity of making a mini chromosome, more than just the cystic fibrosis gene, which was the one that has been shown.

DR. COLLINS: Right. Well, the complexities are profound. Just the scenario you laid out there has many areas that would require vast amounts of work to be able to figure out just how difficult they're going to be, and they're probably going to be very difficult.

First of all, it's not clear to me that for a trait like intelligence that there's going to be a dose effect, that you add back additional copies of the gene that has the favorable spelling, and that actually gets a good result because you're going to end up with too many copies of that gene, and in general, you know, dosage is very carefully controlled within the human genome.

People with trisomies, like Down's Syndrome, there are consequences of having an extra copy of a gene, even if it's normal.

So it's not clear to me that that would be at all a safe activity, and again, how would you even do the experiment if you didn't know it was safe?

So actually for that application it's very hard for me to imagine how you could proceed with the uncertainties that would be associated with it.

The technical challenges of putting ten or 20 different genes onto an artificial chromosome and getting them properly regulated are going to be very large. Again, regulation we understand poorly. You probably would have to not just insert a small fragment, but the gene in its appropriate surroundings. Its appropriate surroundings would probably have some other genes in there because genes are packed next to each other in certain parts of the genome fairly densely. So you'd have a hard time creating that artificial chromosome and getting normal regulation without also bringing along some other neighbors that you didn't want to bring along.

This is a very big problem, and on top of that it is hard for me to see whether that would be an arguably more advantageous approach if what you were trying to do is optimize intelligence than the PGD approach, which carries with it sort of less in the way of wholesale reconstruction of the genome.

So I doubt that that scenario at least as applied to intelligence is going to be with us in less than many decades.

CHAIRMAN KASS: Could I just try to generalize Janet Rowley's question?

If I've understood the presentation today and also before, you seem to be leaving aside somatic interventions about which I hope we can ask you subsequently.

DR. COLLINS: Sure.

CHAIRMAN KASS: That genomic knowledge in relation to the prospects for enhancement into the foreseeable future really will function only in permitting selection rather than in encouraging positive manipulation directed genetic change by insertion of genes, and that's partly because of the answer to Janet's question. For any of the things that one would be really interested in, if one were interested, these things are much too complicated both to know what the genetics of those things are. The penetrance is unclear, the environmental, all of those things, and the safety questions of going to work either on egg and sperm are going to work on embryos.

So, in effect, you're saying genetic engineering of enhancement in a positive sense is not really in the cards, leaving aside the somatic interventions, but in talking about genetic interventions to somehow improve the offspring at the level of egg, sperm or embryos is science fiction.

DR. COLLINS: I think in a ten or 15 year window it definitely is science fiction. I would not want to say it will always be so because that's always a risky thing to say scientifically, and there are certainly people out there predicting that that will change.

But if your charge -- and I don't quite know what your charge actually is, but if your charge is to look at the most imminent concerns, the ones that may actually become real in a decade or a little more than that, I don't think germline intervention for enhancement purposes are going to be realistically on the table.

CHAIRMAN KASS: Okay. Good. I have Bill Hurlbut and Bill May and I should look to the right.

DR. HURLBUT: Francis, can you tell us what some of the implications are? I know we want to talk about PGD a little later, but just touching on this question of larger polygenic phenotypes, what are some of the HAP map and the possibility of selecting for not just genes, but for whole haplotypes?

Like let me give you a scenario. For example, you have an interracial couple that might say, "Well, we really want this ancestry emphasized in our offspring." Is that a realistic possibility?

DR. COLLINS: I --

CHAIRMAN KASS: Francis, would you say, because I'm sure not everybody understands the terminology?

DR. COLLINS: Right, right.

CHAIRMAN KASS: So you could educate us in the course of answering that.

DR. COLLINS: Sure. Yeah, thanks, Bill, for the question because I think there are some real relevances here to what's happening with this so-called HAP map.

So what is the hap. map? That stands for haplotype map. Geneticists have terms that only they could love, words like allele and haplotype, but you know, it's going to be hard to get them to stop using them. So let me try to define what a haplotype is.

Basically, I showed you that example of a couple thousand letters of DNA from Chromosome 7 and the two places in that DNA sequence where there's a common variant. So we call those alleles. Nowadays we call those positions single nucleotide polymorphisms, or SNPs. So SNPs are all the rage.

But I also mentioned that we're interested not only in the individual SNPs, of which there are about ten million in the genome that are common in the human population, but we're interested in how they correlate with their neighbors.

And you might not immediately think that they should. It might seem that they should be independent of each other, but that's not true. In fact, if you have two of these SNPs that are a few hundred base pairs apart and you know that you've got an A at this one and you might have a C or a G over there, it will be strongly predicted which you have by what your spelling was in the SNP next door.

That's because we humans are descended from a very recent common founder pool of only about 10,000 individuals roughly 100,000 users ago. There's only 5,000 generations separating us from that ancestral pool that lived in Africa, and there has been insufficient time for our genomes to get homogenized over that period, and so basically when you are looking at a segment of a chromosome, you're looking at that unbroken segment that came down from that common ancestor, and there hasn't been time for this variant and the one next to it to be separated by what we'd call crossing over and to travel off into different chromosomes.

That turns out to be very useful if you're trying to find a place where there is a gene conferring a risk for disease because it means that you have a real shortcut possibility here, that you don't really have to test all ten million of the variants that are common in the genome to find the place where there is a gene conferring a risk to diabetes. You can work with a much smaller subset because they're not independent of each other.

So you pick a gold standard set that basically represent the entire variation that's common in the genome, and you just work with those. That's what that haplotype map project is. A haplotype is a stretch of DNA where the SNPs are correlated with each other, and that is a project that's now underway and is going to have in the next roughly two years the ability to produce this gold standard set of SNPs.

That will greatly accelerate our ability to find genes involved in common diseases and also in non-medical traits. It will not give you the absolute end game because you will know where there's a chunk of DNA here that seems to carry a high risk for a particular disease, but you've still got to dig around in that chunk and figure out why.

But it will become almost immediately tempting to use the information anyway, even if you don't understand the biological basis, because having learned in a research study that this particular haplotype is correlated with this particular phenotype, you could then imagine applying that prospectively.

And that's another reason, I think, why PGD applications of traits will get talked about in ten years, because even though we won't understand the biology, we'll have the genetic ability to draw those particular associations, and I think that will perhaps tempt people who are interested in applying this in that setting to do so. So it's another part of the technology that I think speeds the enterprise.

CHAIRMAN KASS: To this point?

PROF. BLACKBURN: Just a quick. I wondered what is the estimate right now of the number as opposed to the ten to the seventh -- sorry -- ten million SNPs overall. What does the haplotype mapping reduce it down to? Is there a good estimate now?

DR. COLLINS: Yeah, about 300,000, maybe 400,000. So, yeah, you're saving yourself something like a factor of 20 to 30 in the amount of work that you have to do, which is well worth knowing about.

CHAIRMAN KASS: Bill May and then Michael.

DR. BLACKBURN: As I recall, there was a budgetary set-aside for moral reflection on this immensely important project. Would you say that the work in ethics has primarily focused on the issue of safety or has it dealt with the question of ends and purposes that you've sorted out here in your grid? That's my first question.

The second one is: has there been any intramural sharing across the work of your ethicists and the researchers in the genome mapping?

And my third and last question is: do you see any interest in those who have been engaging in privately funded research in this area? Interested in the work that has been done by your ethicists.

DR. COLLINS: So, yeah, those are wonderful questions. The ELSI portfolio is very broad. It certainly has addressed some of the things I've touched on, but many other things as well.

I think, in particular, there has been much research done on fair use of genetic information, and much of that has focused on the thorny questions of discrimination and invasions of privacy.

There has been research on how to make sure that genetic tests are applied in a fashion that achieves benefits, which involves questions of oversight. How do you know when a genetic test has actually been validated? And what kind of system is necessary from a policy perspective to make sure that that oversight is achieved, something that still has not been worked out?

There's been much research on human subjects' concerns, for participants in genetic research, a very complicated issue that's still, I think, a bit up in the air in terms of exactly what kind of consent is necessary for a study that connects genotypes with phenotypes and may have long term consequences.

And then there has been, I think, very rich philosophical studies on the consequences of knowing this kind of information about ourselves and how does that change our views of ourselves and of humanity at large. So it's really quite a broad array, and we have up on our Web site, which is genome.gov, a thorough enumeration of the kinds of things that the ELSI program has done for further investigation.

As far as your second question about sharing, certainly the ethics community has come together quite substantially. This is the largest investment in bioethics research on any topic ever, and it has, I think stimulated a whole cadre of scholars, some ethicists, some legal scholars, some philosophers, theologians, to put their careers very much into this enterprise, and many of them very effectively to interact with each other and share perspectives.

The sharing with the scientific perspective, I think, has also been good, but it could be better. Certainly many of our large centers do have ethics components attached to them, and it's very useful to have those conversations in the same room and not in separate rooms, but we are contemplating in the next phase other mechanisms to try to further stimulate that to make sure that the science and the ethics don't go off in separate directions because they clearly need to inform each other in the most vigorous way possible.

The interest in the private sector, your third question, I think, is variable. It depends on the particular entity and what they're attempting to pursue and their own sort of self-awareness of whether the ethical components of that are a major aspect of its likelihood of future success.

But I think in general there has been a pretty strong level of interest and certainly many of the people we've supported through our grants program get called upon regularly by private sector entities to come and consult or to make presentations about what their ethical perspective might bring to the table when a company is trying to make a decision about its research direction.

CHAIRMAN KASS: Mike Gazzaniga.

DR. GAZZANIGA: Again, a wonderful presentation. Picking up on Janet's point, Ellie somebody, whose last name I can't remember, at MIT has just shown that this past year between 500 and 1,000 genes are expressed during the propagation of one nerve impulse, and this takes me to your zone of concern and the issue of really understanding the polygenetic component of intelligence or something like that.

It seems horrendous to think that we could actually go to PGD to find out which genes are active in intelligence in the short term. I just can't comprehend how that could happen.

I was wondering if you could maybe just go a little bit more into your thinking on that.

And secondly, at the sociologic level I actually question whether people would do this. You know, I'm sitting here thinking of a couple of Yalies got married and they're sitting around thinking about their first baby, and they look at each other and they say, "Well, you know, we didn't get into Harvard."

(Laughter.)

DR. GAZZANIGA: "So you know, maybe we should do a little PGD."

I think such a --

DR. COLLINS: As a Yaley, I resent that.

DR. GAZZANIGA: Good. That's the idea.

I actually question with education and general public discussion whether this would be used much at all.

DR. COLLINS: Well, great questions because I have those same doubts, and I'm glad we're getting into this because I'm afraid maybe my initial presentation led you to think that I was contemplating that PGD for intelligence would actually be a valuable activity in the sense that people would really want a lot of this and it would really work.

I think they might temporarily think it a valuable activity, but would soon learn that it was very limited in its ability to make useful predictions.

So how would this actually happen? Again, I think the kinds of studies that are already ongoing would be to collect DNA samples from individuals with exceptional intelligence, those in the middle of the IQ scale, those on the lower end of the IQ scale who are not afflicted with some known syndrome, and once we have this haplotype map, run across the genome and see if you find areas where there is a skewing in the distribution of a particular haplotype between those three groups.

Already Plomin has made a claim based on a linkage study of a specific variant in a specific gene that on the average alters IQ by two points. I don't know if that will hold up, but undoubtedly other claims will get made and probably will have a higher chance of being right when we have the haplotype map to guide the enterprise.

So okay. Imagine then that in ten years there are four or five of those alleles that have been identified that at least one other has validated. Then that Yale couple might begin to wonder about whether that's something they want to use.

I actually think you're right, that perhaps after an initial burst of enthusiasm to think about this, driven by all of the forces that you can imagine, that people will realize that the value of this is so low in terms of its actual impact on its outcome that it's just not worth it, and it's a lot more fun to have kids the old fashioned way.

CHAIRMAN KASS: Michael Sandel.

PROF. SANDEL: Well, we ritualistically thank people for their presentations, but I think this was really a tour de force of clarity and clarity of a kind that helped draw upon science to focus ethical inquiry.

And what I would -- it also helped clarify for me the relation between discussions about enhancement in the ethical implications and discussions about genetic engineering and do these overlap or are they distinct, and PGD as a pathway to this as against germline intervention.

And this also would provide a way of addressing a worry that has come up from time to time when we've been engaged in the enhancement part of our project about how realistic is this. Are we just dealing with science fiction possibilities and spinning out ethical theories to address them?

But what it suggests to me, and this is directed not to Dr. Collins, but really to the group and to the Chair, would a natural follow-on or next step for the enhancement project, if as I take it we've more or less brought the enhancement project to completion with this meeting subject to the materials that are going to be prepared; wouldn't a logical next step be to take as a point of departure something like that lower right-hand quadrant on the two axes and focus on -- since it seems that germline interventions and manipulation are not the present or near term concern -- to focus on applications of PGD that fall into that lower right-hand quadrant?

We could elaborate the list of the ones that we take to fall there and to have now a more focused inquiry not on enhancement as such, but on some number of important practices that we agree fall within that quadrant that are here or could soon be here and really try to work out an ethical analysis with respect to those, say, four or five practices.

That could inform public debate in a more focused and practical way and yet still enable us to bring to bear the moral considerations we've been grappling with in the general enhancement project.

CHAIRMAN KASS: Thank you.

And I think at least the intention while not being exactly determined by a design of the quadrant is very much in the spirit of the comments. I mean, one of the reasons that, to speak simply for myself, that I was very eager to have this presentation by Francis this morning and several other members of the Council, Janet Rowley leading among them, since the very beginning, have expressed a great deal of skepticism about a lot of this hype about germline modification, eugenic selection.

And I think one of the services of this kind of a discussion is that for a change this Council is in a position to offer some reassurance rather than to foment and increase people's concern by saying, "Look. We've heard about this."

There are certain areas where genomic knowledge raises these kinds of concerns, but it's not the stuff that fills the headlines, and I think part of, I think, what we would like to do in whatever writing is, in fact, to sort out amongst the various things that have been talked about those that are reasonable to be concerned about and those are not.

On the more specific suggestion, I think we should talk about it further rather than do this here: strategically how best to organize the case materials that we will bring forward. We've done our survey, and we have to now go home and figure out which of these case studies is most important to carry further.

So we're in sympathy, I think, with the approach that you offer. Is that okay for now?

To this, Paul, or do you want to join the queue?

Let me move to the -- I'll just add you to the list -- I have Gil Meilaender and then Frank and then Janet.

PROF. MEILAENDER: You're not concerned that parents give music lessons to their children or encourage them to play sports, that sort of thing, even though, as we all acknowledge, you can invest a lot of energy in that and the kid turns out not to care about Beethoven or unimaginably to like soccer better than baseball.

(Laughter.)

PROF. MEILAENDER: Even granting all of the caveats you've had about things in your zone of concern, take the possibility that there might be just even limited gains to be gotten from whether PGD or an artificial chromosome or something like that in areas like that, you know, and you would want to say to these people, you know, "Really, you're not going to get nearly as much as you're thinking you are. It's much more complicated," and so forth.

But there still may be limited gains, ten, 15, 20 years down the road. Why should that get into a zone of concern then? If I'm wary and attentive to your warning that I won't get as much out of it as I might have thought or some people think, but still, you know, if I can nudge my child over in the direction of Beethoven or whatever, I'm willing to do it. Why are you concerned about that when you wouldn't be concerned about my giving him music lessons or sports or whatever even though I might be just as attentive to the fact that that backfires sometimes and the child doesn't turn out the way that the parents hope?

Can you just say a little more about what it is about the fact that we're dealing with genes that makes it something that would fall into a zone of concern?

DR. COLLINS: Well, in a way this is not a scientific question. This is an ethical question. I mean, there's a scientific component to it, but we could probably deal with it fairly quickly. I mean, there is the question about safety and whether the practice of PGD is completely without risks, and ultimately I suppose we'll have better data than we do right now about that.

So obviously if there do turn out to be risks that increase the likelihood of some other kind of trouble, then this becomes an easy argument to say why music lessons are okay and PGD is not.

But let's assume, because I think it's the flavor of your question, that PGD turns out to be safe and that you're not exposing the embryos that you've diagnosed to other harms. Why is this not the same?

Again, I think this is not a question that science is in a very good position to answer. I think this is a question though that cuts to the heart of the matter of what is special about the germline and what is special about that kind of alteration, and I guess it's special in at least two ways.

One is, of course, that this is a permanent and heritable part of that individual. Unless you are Lamarckian giving your kid music lessons is not going to improve your grandchildren's performance on the piano.

On the other hand, if you're making a germline shift in the odds, well, you're shifting the odds for all of those future generations in that very moment, and that is, I think, a profound difference.

One might argue, oh, well, it's just a minor tweak in the throw of the dice. The dice might have come up that way anyway. But the fact is you sort of stacked the odds. You weighted the dice, and that is the kind of intervention that I think is in a fundamental way different than what one imagines through education or sports activities or music lessons.

I think there are more than that, the fact that it affects future generations being a pretty strong argument. It is also that there is something special about altering the biological inheritance, the instruction book; that this has a more profound significance for that individual's basic nature than an environmental alteration that attempts perhaps to enhance capabilities that were there and to refine them. This is an effort to try to provide capabilities that maybe otherwise wouldn't have been there.

And without, I think, being very articulate about it at the moment, I think if you ask the average person on the street, which I think I agree with Leon that that's not a bad thing to do when you're wrestling with a moral issue, that they would argue that this is very different.

Again, I keep coming back to this questionnaire that Pew Center for Genetics and Public Policy just posed to quite a bunch of people, and they certainly reflected that, that there is a general sense that this is much more at the heart of the matter in terms of altering our basic humanity or the nature of what that individual is going to have to work with than the mere fact of trying to improve your kid's performance on the piano or enjoyment of a Beethoven symphony.

CHAIRMAN KASS: Do you want to follow?

I have Frank.

PROF. FUKUYAMA: I am trying to think through some scenarios by which the germline stuff may get here a little bit quicker than you suggested, and it seems to me there are two basic baskets of reasons why you think it's very far off. The first has to do with complexity, and the second has to do with safety.

Now, on the complexity side, you know, you and various people around the table have argued that the characteristics we care most about are polygenic and very complex and dependent on interactions between genes. Is it obvious that there will never be a relatively simple, you know, single gene change that will have an effect that people will desire that will be simply discovered by accident?

I mean is there some theoretical reason for thinking that all interesting things have to be complex and polygenic? And I'll just give you a couple.

I mean, yesterday we heard, you know, in the discussion of aging that, you know, everybody thought that it was very complex, but, in fact, you know, it seemed that there is a single gene that actually does have an important effect. As I understand it, there's a species of, you know, ant where the entire social behavior is actually turned on or off by, you know, a single gene variant.

And so is there some assurance that we will never come across any of these relatively simple changes that will actually have a big effect? That's on the complexity side.

On the safety side, it seems to me that, you know, actually if you take our regulatory environment for granted, I think actually you could push a lot of this stuff into the never. I mean, just imagine what a clinical trial for, you know, some kind of germline intervention would look like. It's almost inconceivable.

But that presumes our regulatory environment. There are other countries around the world that have capabilities where you can take that for granted. There are rogue scientists. You know, there's lots of other situations in which something we would regard as unethical experimentation may take place.

And so it seems to me that if you're going to devise a scenario by which you could get to this stuff faster, it would have to be some combination of these two where you find a relatively simple genetic intervention that has a desirable effect, and then you carry out the experimentation in something other than our regulatory environment.

I mean, does that -- I'd just like you to respond to that.

DR. COLLINS: So with regard to the complexity, I think it's fair to say we're not going to find sort of monogenic aspects as you might in the ant and the social behavior that are going to explain significant features of human phenotypes that people might have the most interest in trying to enhance because if those were there, I think we would begin to see a lot more examples of Mendelizing characteristics like that, and we don't.

And certainly -- and this is a long tradition of genetics looking at identical twins and dizygotic twins and then further apart relatives -- one can look at distributions of various of these quantitative traits and say, I think, with great confidence that there are not single genes at work.

The real debate is is this oligogenic, that is, a few genes, or is this truly polygenic where it is gazillions of genes -- well, not gazillions; hundreds, let's say. There's only 30,000 altogether. so it's not going to be more than that.

And we don't have the data to know the answer to that for most of the traits that we've been talking about this morning, and that's where I think the next ten years are going to be very interesting, because we will begin to discover that.

If it's oligogenic, then amongst the genes that are discovered may be some that have a measurable effect. Again, going back to IQ, maybe there is a gene there where there's a variant which confers as much as a point spread on the IQ test of four to five. I doubt it. I don't think there's going to be any that big, but I couldn't say that I would be utterly shocked if that turned out to be the case.

But I don't think the scenario where you find a really big contributor to human behaviors is likely, given what we know from the more indirect studies of those traits.

When it comes to your question about the safety issues, of course, rogue scientists are always a possibility. Thank God they don't emerge very often. We're obviously all now very focused on that risk when it comes to human reproductive cloning, but I'm not sure that that should drive the debate on the Bioethics Council about how much time to put into an enterprise. If that is a possibility, well, that is certainly a task for governments and regulatory agencies to try to be sure they have the best control over, but it sort of ceases to be an ethical matter and becomes a policy matter.

PROF. FUKUYAMA: Well, but how about a country like China, for example, that has, you know, very substantial capabilities in this area that may, you know, be willing to do experimentation that we may not be willing to?

DR. COLLINS: So that would certainly be a concern, especially if the experimentation was based on shoddy science and was being carried out on unwitting and unwilling participants, and obviously that raises, I think a different set of ethical issues, which is the morality of conducting experimentation in a circumstance where it's truly unjustified and where the risks don't justify the benefits.

It's sort of a different question, I think than the question of whether germline interventions raise within themselves special ethical questions. They might some day, but I would say right now there are other questions that are more pressing.

CHAIRMAN KASS: Janet Rowley.

DR. ROWLEY: Francis, I'd like to follow on with some of the questions and part of my question, I think, is related to an aspect of what Gil Meilaender was asking.

In the real world, you're going to have a couple of whose genetic composition, if you will, for whatever genes they are interested in can be determined using the haplotype map, and then you would have a sense, say, for five to ten genes of whatever trait the alleles and the forms of those genes that each the husband and the wife carry, and then you would know from other information, mapping and such as you're already describing for intelligence, in the future what alleles are associated with either higher intelligence or taller stature or whatever feature is of interest to the couple.

But if those two individuals don't have the alleles that confer the highest level of height, beauty, whatever, then they either have to decide that they're going to have a child that will be less endowed in those features if they use their own genetic material or they have to then go and use somebody else's.

So that it seems to me that for doing, say, PGD, you can say this embryo has this collection of alleles and, therefore, the possibility. But those people unless they get different genetic material can only really sort or choose the assortment of the particular genes that they are themselves carrying.

So changing haplotypes is probably not going to be an easy thing to do in the future, but I'd appreciate your comments on this.

DR. COLLINS: No, I think that's very well said. You're basically in the PGD environment limited to the potential genetic contributions of the two parents. You're not creating new possibilities there.

You are, as it said in that Gattaca clip, just trying to give that couple the best possible outcome that they might have tried 1,000 times and not achieved by basically doing the picking and choosing prior to implantation.

But as you say, if what you're starting with is not what that couple is looking for, this is not going to create it unless they decide to go and seek genetic inputs from elsewhere. And you might argue that the widespread interest in PGD for enhancement might also lead to greater interest in donor gametes for people who decided that their own haplotypes were not sort of as good as they hoped for and were then motivated to try to further skew the odds by seeking out some other source of genetic material.

CHAIRMAN KASS: Dr. Schatten, one small comment on this?

DR. SCHATTEN: Just one small comment, and that is I hope we're not ignoring the whole dating game. I mean, people choose partners because they find, you know, a variety of features attractive or reject dates for other reasons, and you know, I think so many of the issues that you're referring to, you know, height or athletic ability or whatever, you know, happen over candlelight and nice meals.

DR. COLLINS: Called assortative mating.

CHAIRMAN KASS: Let's see. I have Paul McHugh.

DR. McHUGH: I also want to thank you very much for your wonderful presentation because it opens up all of us to further questions to ask you about the role of the genome project in instructing our country about the matters that are in front of us, and I have three questions that I'd like you to develop a bit more for us, Dr. Collins.

The first one really relates to the issues of the genome project wishing and working to deal with diseases and recognizing, as you do, that some of the mental disorders are going to be at least oligogenic, if not polygenic.

But we also know that some of these mental disorders, that if we eliminated them, we might be as well eliminating certain aspects of human gift.

My colleague at Hopkins, Kay Jameson, of course, has talked about the element of manic depressive disorder amongst genius and our group talking about the process perhaps of eliminating the wonderful qualities that people might have that combine with in some close relationship to the diseases we're trying to eliminate and so making our population more uniform in a way that we would regret.

This brings me to my second question, and it's a question we've talked about before here. Much of our discussion about our offspring speak to what parents like or what parents want, and I'm sure my parents would have liked me to be taller. I'm just as glad they were satisfied with what they got because in the long run, our offspring don't just belong to us. We've talked about this before. They belong to the community, and surely our task is to speak to the diversity of our community in all kinds of the features that ultimately a family might want to affect, sex ratio being the most obvious one, but several others that might play a role.

And I'd like to be reassured that such considerations are coming forward in the teaching aspects of the genome project. I'm sure they are, but I'd like to hear more about it.

And then finally, of course, your wonderful contribution brought that picture, this moving picture to us that had this chilling, chilling ultimate genetic counselor there, and it was, you know, talk about smacking of the eugenicists right there.

After all, just last week we heard from the governor of Oregon that he was apologizing for what Oregon had done in sterilizing people with mental disorders.

Those of us who have the role of teaching medical students and teaching doctors and having the responsibility of teaching not only the material that they work with, but the ethical and characterological features of doctoring to people and appreciating the role you play in intervening in their lives this way; what are you doing to make sure that such glib and cold attitudes about humankind, human nature and its spirit are not only built into the issues that he's going beyond the science, but as well, that there might be an understanding of the spirit.

DR. COLLINS: Goodness, we could talk quite a long time about all three of those issues, and maybe the first two points you raise really do point to a potential tension, maybe even a collision between two principles that we hold fairly dear. One is the rights of parents to make decisions about their offspring without intervention from the state, and the other is the interest that society has in our future collectively being a future that encourages diversity and emphasizes fairness and access to goods.

And I think the whole discussion about enhancement collides in that way. Obviously a completely unregulated circumstance where parents with the greatest degree of resources could basically choose to do virtually anything they wanted to try to improve the characteristics of their offspring would potentially increase the divide between the haves and the have nots, would add perhaps to prejudice against those with handicaps and disabilities, and I must say the disability community is very concerned about many of these discussions about how we're going to improve ourselves, putting that in quotes, because of the impression that creates that those who have disabilities are less acceptable.

And clearly, we have to pay a lot of attention to that perspective. That is a major focus, I think, of the ELSI program, and some of the research that's been done on that is really worth looking at.

Your point about whether eliminating things like manic depressive illness is a societal good or not is a very tricky and important debate to have. I mean, clearly manic depressive illness in its full blown untreated state is devastating, often leading to suicide, enormous family trauma, but it is a treatable disease, and it is also very clear.

Kay Jameson has shown eloquently that many of our most talented artists, poets, writers, musicians have clear diagnosable bipolar illness, and how bland things might be if somehow such individuals were eliminated from society altogether.

So where does society's interests rest there? Is that a circumstance where parents perhaps, having been severely affected by the ravages of this disease and who wish to make sure it doesn't happen again would be told, no, you can't do that because you might be eliminating from us the next artist?

I can't imagine that scenario playing out. That doesn't seem like that would be an argument that could be won in terms of interfering with parents' perspectives, but it is a collision of intentions that I think is worthy of some more discussion than we probably have time for this morning.

In terms of preparing everybody for all of this, another aspect of the genome project is focused on education. A lot of our educational efforts are aimed at trying to help health care professionals understand how genetics is relevant to virtually all aspects of medicine, but also to understand the limits of that because we tend to go through these wild swings.

I think ten or 20 years ago most health care professionals would say, "Aw, genetics, irrelevant. Don't see any of those problems in my practice."

We may be heading towards a moment where it's quite the opposite. "Oh, genetics is everything," and we're sort of forgetting that environment and free will choices and the nature of being human is not going to get explained by having our three billion base pairs in front of us.

I think we have sort of a pathway underway to try to achieve that for health care professionals through a variety of partnerships with many organizations that represent them. The pathway towards the public is much more difficult.

I mean, I showed you that clip from Gattaca because I thought it was so powerful, and wouldn't it be wonderful if Hollywood would actually do a better movie about this that would actually convey some of the real realities of the situation in a fashion that would be thought provoking and not confusing. Because I think many people who probably saw Gattaca probably got scared and confused and probably didn't from that experience walk away with an idea about what to do.

In terms of public education, we depend so heavily on the media, and oftentimes the media by its very nature has to emphasize the things that are bad are really, really bad and things that are good are like incredibly wonderful. I'm going to save the world, and the idea of having a nuanced, balanced discussion rarely comes across in our public discourse, and we have, I think, a big problem in that regard as we face all of those entities in the lower right quadrant of my diagram because already there's so much distortion about expectations, and even more than that, there's a lot of things that aren't in that lower right quadrant that are perceived to be realities today or, at the most, tomorrow.

CHAIRMAN KASS: Thank you.

I have myself next in the queue, and I think I'll take my place.

A couple of comments and then a question. First of all, I would just make the observation that when one thinks about some of the desirable human qualities for which people might be interested in selecting at least if they could, when you get a bunch of professors and scientists together, intelligence is always the thing that gets on the list, perhaps the most complicated and difficult thing to talk about.

But it seems to me there are other matters more significant that eye color, and many of them even quite profound that I suspect we're going to discover the genetic contribution to, whether it has to do with things -- height is not a trivial matter, and especially in the athletic world. A certain premium placed on it might lead people to have an interest in that.

There are questions of sexual orientation and the complicated genetic questions about that. There are questions of memory. We learned yesterday that Paul McHugh inherited a certain feistiness from his ancestral line for which we are all very grateful.

But temperamental things, and as Bill leaned over and pointed out to me, a certain kind of selection against Marfan Syndrome might have robbed us of the 16th President of the United States or at least of a certain kind of bipolar disorder, whatever you want to say.

So it would seem to me that the list of things for the sake of actually stimulating a richer conversation about this would be to begin to fill out the whole range of traits that people might have an interest in and not simply focus on intelligence. That's an observation.

Second, in parceling out the various kinds of traits for which we would either find approval ranging to disapproval, the assessment was based simply on the merit or lack of merit of the thing chosen rather than on something which all of these things have in common, namely, the fact of selection itself.

And it's true that we have been practicing prenatal genetic diagnosis for 30 years and that there is a certain negative selective element there based upon limited information, mostly about Mendelian traits and chromosomal abnormalities, but the most that somebody can do there is to eliminate something undesirable. They don't have the responsibility of positively choosing something that they would like or that they think is right, and it seems to me that is something new.

Maybe we'll talk about this also in the next session, but I wonder whether or not the fact that even in a minority practice selection on the basis of some kind of genetic traits, whether for the traits that we find absolutely unexceptional or traits that we worry about doesn't produce a certain new kind of disposition where you now have, on the one hand, parental responsibility really for the genotype and in the community at large a kind of sense that, well, it really is somehow up to us whether people of this sort come into being or not and whether that isn't the kind of new moral dimension that's independent of the particular traits chosen.

And I guess the third thing I would say is I've been struck by the fact that especially all of the eugenic talk has, of course, disappeared. No one really talks about improving the species or improving the race.

And even these genetic choices that are being offered are being presented to people as, well, we're just going to give you the information and you can choose.

But the information which is being brought before the people on the basis of which they are to make their choices are not the things chosen by the people themselves, but they're chosen by us, by the scientific community. And I wonder whether there isn't -- and this is not a subject much discussed -- but whether there isn't insufficient attention to what might be in the aggregate something like at least a negative eugenic ideal here of what things we would be better off having removed from the population, to begin with, with various kinds of diseases.

There seems to be a tacit premise that we would be better off without certain kinds of conditions, and I'm not saying that we're wrong about that, and I'm not saying that the public and the scientific community don't agree with each other on this. I mean, I think there wouldn't be takers.

But this is also -- now that we have increasing genetic knowledge of the basis on which to make those at the moment mostly negative selections, isn't there a kind of implicit teaching that certain kinds of human beings ought not to be amongst us and it's no longer simply a matter of negation, but even to some extent positive choice?

I'm not sure I put that very well, Francis, but I do think that there are things that are new here that are quite independent of the particular choices made and having more to do with the kind of parental disposition to offspring. Admittedly if it's only PGD, it might be a very tiny practice. I mean, this is not something lots of people are going to run off to do I don't think, at least not for a while.

But would you comment?

DR. COLLINS: Yeah, I think actually this is a dynamic that's been going on for quite some time, and particularly as it's discussed around decisions about amniocentesis, for instance, for advanced maternal age.

Parents who have a child with Down's Syndrome are frequently asked, "Well, didn't you have an amnio? How did this happen?" And those may be parents who for their own very strong reasons were not interested in that because of their own feelings about the whole issue of abortion and who also in some instances did not view Down's Syndrome as an outcome that they would consider totally unacceptable and were willing to take that risk without prior warning, or in some instances did have the amnio, did know this was a child with Down's Syndrome and chose to proceed, and many couples do that.

And yet there is, I think, a societal dynamic driven by our broad use of amniocentesis for advanced maternal age that has this subtle aspect to it that you're alluding to that maybe that shouldn't have happened and something went wrong.

Now, already that has led to, I think, some debates about whether this kind of screening is doing favors for lots of people. I mean, you can also look at the experience with alpha-fetal protein screening for neural tube defects, the considerable anxieties that creates in couples that are found to have an abnormal result, which usually turns out not to be indicative of a problem, but certainly creates the sense that there's something wrong with my baby, a sense that may then linger for some time even after the baby is born.

So I think we do have to look at that experience and not draw the conclusion that, oh, this is a good model where everything turned out fine. I think obviously many people are awfully glad that that kind of prenatal diagnosis is available and have stories to tell you about how that in their own personal circumstances was a very positive aspect of being able to avoid a terrible outcome.

Certainly couples at risk for Tay Sachs disease, for instance would argue that there is nothing good about giving birth to a child who's going to die a terrible death over the course of a couple of years and having the ability by some other intervention to avoid that is from most of their perspectives, not all, but most, a very good thing.

But it does sort of start you into this zone of, well, that shouldn't happen, and for couples who make a different choice, societal pressure begins to bear.

Right now we have seen recently in the last year announcement by the American College of OB-GYN that all couples from populations where cystic fibrosis is fairly frequent should be offered CF carrier screening, and so lots of CF carrier screening is occurring in that first prenatal visit for couples to find out whether one or both of the partners are CF carriers, and I think many of them are walking into that without quite knowing exactly what they're getting into despite good efforts to try to achieve that sort of education.

And then some will be faced with a decision about a disease which is compatible with survival, average survival now age 31 with CF, and what to do with that information, and probably a part of that dynamic that's going to ensue because of the onset of this kind of screening is a general societal conclusion that, well, maybe we shouldn't have so many cases of CF anymore because of this.

Again, you can imagine how people with disabilities feel about this, and I think we should be paying very close attention to that.

So that experience, I think, does play very much into your question about, well, if it's already created some of that sense when it's an effort to try to prevent terrible diseases, how is that going to play out when the opportunity exists to even do a little enhancing? Will that also become an expectation? Will it be part of being a good parent?

I don't really think so. Again, I don't think it's going to work very well. I think that maybe we didn't emphasize enough this morning, although it came up a couple of times, the ability to use PGD anyway to do an enhancement is going to be so woefully disappointing that even though we may wring our hands and worry about the scenarios that might come about and think about how to put appropriate regulations and ethical principles in place, it's the simple, practical aspect of it not working very well that may be our best protection against an outcome that many of us would find very troubling.

I'm not sure of that, but I think that is an important part of the whole thing, but I do think you raise an important point. Of all of the many ripples that come out of this kind of discussion in terms of how society views imperfections, I think, I mean, this is getting us deeply into a philosophical or even theological sort of territory about the value of diversity, including imperfections of a medical sort in terms of our appreciation for what it means to be human. That is not a negligible feature.

I often think back to this story in the ninth chapter of John where there is a big debate going on between the disciples and Christ about the reasons for infirmaties, and in this case the disciples brought to Christ a child who was born blind and they said, "Why was this? There must be a reason?"

And their question was: who sinned here? Was it the child or the parents that this child was born blind?

So at that point in 32 A.D., the expectation was that it was one or the other. Somebody had done an evil thing in order for this consequence to have occurred.

And Christ's response is very interesting. He said neither of those. This happened so that the works of God might be manifest in this child, which is you were supposed to learn something from this. You were supposed to be able to appreciate something really important by the fact that we're not all perfect beings.

I think that's an interesting perspective, and while it collides immediately if it was my kid, I sure would not want to have them be the reason why other people would be learning important principles. That would trouble me greatly.

I think it does say in a general sense our headlong rush towards ultimate perfection of all of us in freedom from all diseases may not be on the broadest scale of things the most ideal outcome if what we're after is a little bit more than being machines, but actually being human beings.

CHAIRMAN KASS: Let's take a couple more minutes because there are a few people in the queue. Alfonso, Robby, and Dan.

DR. GÓMEZ-LOBO:: I have two questions, one of a theoretical nature and with broad implications, one a more narrow one having to do with policy and your project.

The broad one is this. It seems to me that you indicated that there is skepticism about the way to draw the dividing line between therapy and enhancement, right?

And I did like this, but you insisted that efforts have to be made to draw that line. Now, I'd like to know a little bit more about that because from what I've heard since I've been on the Council there seems to be areas where that dividing line is more or less clearly traceable. I imagine in the case of certain illnesses that's possible, but I'm interested in the broader issue basically because I think that there, on that line, hinges a lot of our moral thinking about this because, of course, clearly to improve the condition of someone in the sense of curing someone, in the sense of therapy is surely morally not only permissible, but probably obligatory in many cases.

Whereas about that we go into the enhancement area, and that's morally a very difficult area.

Now, the other question that I had is this. How do the people in the ELSI project get hired, are brought in?

And the reason, of course, is that moral philosophy is not a field like, I take it, like chemistry or biology where there are basic agreements across the board. In moral philosophy, the basic disagreements are above, are in the theories, and of course, that's a reason why people come down on bioethical problems on different sides of an issue.

So really that selection process I find intriguing for a public body that surely is going to have a tremendous influence, well, on what's going to happen down the line with some of these projects.

Thank you.

DR. COLLINS: So with regards to the effort to try to set boundaries between treatment and enhancement, I do think that's very difficult. I mean, take the example of obesity. Obviously obesity can be a disease, but I don't know how to draw a line between morbid obesity way over at one end of the spectrum and at the other end of the spectrum people who are actually of normal weight, but think that's not quite what they want to be, particularly given our society and the way they portray what's attractive.

And I can't really sort of say where between those extremes is the point at which I can say, well, intervening here would be a treatment and intervening there would be an enhancement. I just can't figure out where that boundary is.

Take drugs. You all talked yesterday about Ritalin. Ritalin in some circumstances seems like it can be a real advantage to a child and to their family. In other circumstances, it seems very much like an enhancement and one that causes a lot of concern.

But exactly where in the spectrum of people who might be given that drug can you say, "Oh, you just crossed the line. That person shouldn't have it, but it's okay to give to this one."

I have trouble seeing where that boundary lies, and I guess I'd be interested in examples where such moral boundaries are easy to draw because I must say I don't seem to bump into them very often. They often seem to be when you really look carefully a difficulty in drawing a precise line to say that things on this side are acceptable and things on that side are not.

End of life decisions it doesn't seem to me are any easier in that regard. So I guess my point was we should probably recognize that this is hard, but we should not be discouraged by that. While there are inevitably going to be middle ground scenarios where people disagree about whether this was an enhancement or a treatment, that should not prevent us from dealing with things that are a little further away from that gray zone in trying to come up with a consensus about how we feel about those.

And that is a lot better than simply throwing up our hands and saying it's too hard, although it is hard.

Your second question about where the ELSI scholars come from, again, we run a grants program. We at times issue requests for applications on a particular topic where we're soliciting investigators to come and study this issue.

Right now we just put out a number of these requests to study the implications of our understanding of genetic variation and race, for instance, a very thorny, difficult topic and one that we haven't talked about this morning, but which I think is going to be incredibly on people's minds in the next few years as we begin to unravel the genetic variants that characterize all of us and try to figure out what that means for this very muddy and mostly socially defined concept called race.

So we basically see who applies, although we do go out and encourage people that we think have good ideas and an interest to apply if they haven't already, and we run actually special sessions for investigators who aren't used to NIH and its particular schemes of getting grants approved to try to train them a little bit on how to be a successful applicant to the National Institutes of health.

It is a challenge sometimes to review those applications because if you have an ethicist who's coming in proposing to do a certain body of work and they're working on a framework that may be somewhat controversial, well, if you have a reviewer that doesn't like that framework, they might just sort of knock it down, and we have to be careful that our review panels have incredible balance in terms of perspectives, which sometimes tends to be an ethicist's Noah's Ark, but you know, it's very interesting to listen to those reviews as they go on.

And I think the outcome would demonstrate that it does, in fact, encourage many different perspectives to come in and seek funding and get it to carry out interesting projects.

If I have a concern about all of this, it's particularly about the dissemination. These result in publications, in books, a large bookcase full of information, but how is it actually getting out there is one of the challenges that we haven't yet, I think, completely solved.

Are we just having people, the experts, speak to each other or is this actually trickling out in a meaningful way to influence policy discussions and to raise the public's knowledge about the ethical consequences of these issues?

CHAIRMAN KASS: Last comment from Dan Foster, who has been in the queue.

Briefly Janet.

DR. ROWLEY: The only point I wanted to make to Gil was you used the term of public advisory bodies, and I think the point that should be emphasized about what Francis said is that these are individuals from a scientific and academic and ethical community, philosophy communities that are both putting in the requests for government funding, and their proposals are being reviewed so that the staff at NIH that is dealing with this really have more administrative functions and certainly no policy functions.

DR. COLLINS: Right. We don't choose the winners. We just make sure that the people apply, and if they're well reviewed, they get funded.

Thanks.

Janet is on our Advisory Council. She knows the --

CHAIRMAN KASS: You do choose the reviewers though.

DR. COLLINS: The reviewers are chosen, yes, right.

CHAIRMAN KASS: And I think Alfonso's point is worth emphasizing. There may be a certain relative homogeneity. There might be party squabbles, but it might be within a family.

Let Dan Foster have the last comment. Please.

DR. FOSTER: I'm going to pass in the interest of time, except I was going to comment on the fact if you look at some enhancement like intelligence, then it's likely that the part of our community that most needs help there would never get it; that the people who have the least need for intelligence enhancement, the rich and powerful, would be the ones who get it.

Dr. Collins spoke about the issue of widening of the gap, and I think that's a very important issue, but I'm not going to address it at this point. It concerns me very much.

I can't help but say that his quotation of 9 John ought to be matched with the mountain sermon where the teacher says that the sun shines on the righteous and the unrighteous and the rain falls on the just and the unjust. It's a good coupling.

Thank you very much. I'm going to pass.

CHAIRMAN KASS: That's better than a pass.

Thank you very, very much for a wonderful presentation and a wonderful response to the question and also for the biblical instruction, which -- this is an odd body, as everybody has discovered. We have proof here yesterday, the existence of God and a recommendation of strapping young boys instead of giving them Ritalin.

(Laughter.)

CHAIRMAN KASS: And you never know what's going to come next.

We're adjourned. Let's take ten minutes. I don't want to steal too much of Dr. Schatten's time, and people have to leave.