We couldn’t begin in a better way than with our first presenter. Dr. Francis Collins is the Director of the National Institutes of Health. What we know today of the relationship between genes and disease is in no small part due to the results of the path-breaking research of Dr. Collins.
Dr. Collins is the former Director of the National Human Genome Research Institute. He led the Human Genome Project’s pioneering work that revolutionized our understanding of the building blocks of life.
A member of the Institutes of Health and the National Academy of Sciences, Dr. Collins is a recipient of the Presidential Medal of Freedom and the National Medal of Science, the highest honor bestowed on scientists by the United States Government.
Dr. Collins was one of the first people, indeed maybe the first person I met with, after my appointment as Chair. It is a pleasure to have you with us today. We look forward to hearing your views. Welcome, Francis.
Thank you very much, Dr. Gutmann, Dr. Wagner, distinguished members of this council. It’s a real pleasure to have a chance to address you at the beginning of what I think is going to be a really interesting day.
I congratulate you for identifying this pair of potential topics, genetics and neuroimaging, and the ethical consequences that are coming out of rapid advances in this field as an area that is ripe for this kind of deliberation. I wish you well in trying to figure out exactly which parts of this matrix you might decide would be most appropriate for a study and a report.
Before I get into that, though, I also want to say how much I appreciate the work that you’re doing on international human subjects issues. Obviously it’s an area of enormous interest at NIH, especially in the wake of the revelations about the studies in Guatemala on individuals who were infected intentionally with various pathogens often without their knowledge.
While that was 60 years ago it’s highly appropriate to have a close look now at the circumstances internationally to be certain that we have norms in place.
God willing that the Government doesn’t shut down, I will be on a plane tomorrow at 5:30 on the way to South Africa and I will spend the next nine days there as part of the initiation of two major projects in sub-Saharan Africa, one of them called Human Health and Heredity in Africa, or H3 Africa, which aims to look at both environmental and genetic contributions to common diseases in Africa, both communicable and noncommunicable.
And another called the Medical Education Partnership Initiative where we seek to provide capacity building opportunities to African universities trying to provide them the opportunity to be capable of doing much more research without having so much in the way of having this imposed upon them.
Of course, for all of that to be successful certainly human subjects issues are going to be critical so I urge you to do everything you can to investigate that, as I know you will, and look forward very much to seeing the outcome of those deliberations.
You have an amazing lineup of presenters for today so I think my task is to sort of whet your appetite and maybe provide a little bit of an appetizer and then all kinds of interesting entrees will be served up over the course of today. I guess you get to have the dessert when you get everybody around the table together and try to pin them down as to what exactly would make the most sense.
I apologize because of my own ridiculous schedule today that I won’t be able to come back for that discussion at the end of the afternoon but I certainly will be interested in hearing the outcome.
May I say we are delighted to have Val Bonham as your Executive Secretary, somebody that we are proud of having nurtured and encouraged and hopefully mentored a bit during her time at NIH.
You don’t get full credit but you do get part credit.
We’ll take whatever we can get.
So I’m going to try to go through a little bit of an overview about what NIH has been doing in the area of bioethics which is familiar to many of you but just a quick survey of that, and then dig into the topics, genetics and neuroimaging. I will spend most of my time on genetics because I think that is an area where I have more familiarity. Although as the NIH director of everything we are supposed to be doing in this amazing institution, I’ve also gotten very interested in the topics that will come up in terms of neuroimaging and I’ll say a little bit about those, too.
NIH has this mission statement which is science in pursuit of fundamental knowledge about the nature and behavior of living systems and the application of that knowledge to extend healthy lives. So this is basic and clinical. About 60 percent of our budget is in basic science. About 40 percent is in clinical applications.
I think it’s immediately obvious that there are ethical consequences, particularly the clinical applications. NIH has been in the area of bioethics for some time.
In a recent survey about exactly what kind of budgetary investments we make in bioethics, approximately $50 million per year is spent by the Institutes and Centers, of which there are 27, in bioethics research.
The largest of those is the program in the Genome Institute, the ELSI program, familiar to many of you, which I had the privilege of overseeing for some 17 years. That is currently budgeted out at about $18 million per year.
During the period of the ARRA funding there was a special opportunity with challenge grants to fund additional bioethics grants. Twenty-one of them were supported in a very hotly competitive arena. Some of those actually were supported by the director’s office and many of them by the institutes.
Out of that, although those were just two-year grants as you know, that has provided some stresses for the system because science doesn’t generally operate on two-year cycles, nor does bioethics. Nonetheless, this was an opportunity to support some very interesting research.
Then in the budget that we were given in FY 10 there is a $5 million allocation to the Office of the Director. That is being used to support bioethics grants from the institutes who come to the OD and ask for shared support if they have something that seems to fit the bioethics research and training agenda.
A bioethics task force has been in operation since May of 2009 and was put in place to review in a more detailed way what NIH has been doing and to develop a strategic plan. One of your members is a co-chair of that task force, Christine Grady.
In addition, the other co-chairs are Amy Patterson, now the Director of our Office of Science Policy; and Grif Rodgers who is the Director of the Diabetes Institute. A number of other institute directors sit on this task force and it involves representation from no less than 25 of the institutes and centers.
They have gone through a first phase and it has been useful to look at what their results have been in terms of where the investments are. But the second phase is going to be more about the development of a strategic plan which is now under way.
The near-term priorities have certainly emphasized the importance for mission related bioethics initiatives and training initiatives. The longer-term goal is to try to do more to integrate bioethics into the full spectrum of biomedical research. Not having this as a separate discipline but one that is fully connected with what is going on in the laboratories and the clinics.
There is a process as part of this strategic planning effort to seek public input in diverse forums so this will be an important action for NIH going forward.
Of course, I will now have to mention something about the ELSI program because I think that has been such a major investment in bioethics. It also is familiar territory from my own background.
Of course, in 2001, just about exactly 10 years ago, we saw these publications on the draft sequence of the human genome. In that 10th anniversary, just a week ago, Nature published this follow-up special issue and very much putting forward a view of what we’ve learned in those 10 years and where we’re going, I think the general sense is that genomics has utterly revolutionized what’s going on in the laboratory.
Graduate students can’t imagine how you did anything in human biology without access to the genome sequence and many other databases and technologies. But its implications for the clinic are really only just beginning and a lot of what was being talked about in this issue of Nature related to that.
You can see the consequences in terms of discoveries of the causes of disease in terms of the molecular basis of conditions that have a Mendelian cause.
If you see the way this builds up over the course of the last 30 years, it is clear — 23 anyway — it is clear that there is a long list of successes here now reaching almost 4,000 conditions for which we have a molecular explanation. Most of those are rare diseases but if it’s your family, it doesn’t matter so much if it’s rare. You’re hoping for answers.
Today is actually Rare Disease Day, February 28th. We’re having a whole symposium at NIH about Rare Disease Day and it’s one of the reasons I have to run back there, to take part in that. I think there is a lot of excitement in the rare disease arena that this kind of research is leading to insights and ultimately to potential interventions for diseases that traditionally have not received much attention.
Not only rare diseases but common diseases are having a lot of molecular insights provided by genomic approaches. This is a diagram of the human karyotype and each one of those colored circles represents a well validated variant in the genome, in the human population that is associated with risk of a common disease.
On that chart, which looks rather daunting, and it is, are about a thousand of those related to risk factors for diabetes, for heart disease, for the common cancers, for hypertension, for asthma, you name it.
Most common diseases for which people have been able to assemble thousands of affected and unaffected individuals have been subjected to this genome-wide association approach and discoveries have been made about pathways that must be involved in disease.
While these variants are for the most part quite modest in their contribution to disease risk, and that has greatly limited the ability to use this information for prevention in terms of trying to identify individual risks for disease, still these are pointing at pathways that are of extreme interest in terms of identifying the molecular causes of disease and pointing towards new ideas for therapeutic interventions.
This is one of the motivations, in fact, why at NIH we are proposing the formation by next October of a National Center for Advancing Translational Sciences to try to take advantage of this deluge of discoveries about rare and common diseases.
On top of that, the technology for DNA sequencing has been advancing at a breathtaking pace. This curve shows you in the diamonds there, in the orange line, what’s happened to the cost per megabase, that is a million base pairs, of DNA sequencing over the course of the last 10 years.
That is a log scale over there. You will see the yellow line is what you get from Moore’s Law which we used to think was the gold standard for technologies advancing at a really breathtaking pace. Moore’s Law, of course, for computer capabilities. You can see that DNA sequencing is actually outstripping Moore’s Law substantially. This shows no sign of coming to an end.
Now much of the handwringing is not about how to generate the sequence. It’s how to store it and how to analyze it which has become really quite a problem.
Biologists used to be considered by the computational experts as not having any really interesting problems. Not anymore. We are quickly arising at the top of the list here in terms of size of data sets and complexity of analysis.
This leads to the potential in the next three to five years of individual genome sequencing getting into the range of affordability as part of the standard of medical care if we decide that is a good thing. The cost may well be in the thousand dollar range within five years. Right now it’s about $8,000 roughly to do a whole genome sequence.
That has already led to some interesting examples where this application has shed light on diseases that were previously mysterious. The Milwaukee Sentinel has a front page story today about Nic. Nic, I have been getting to know a bit, who is six years old and had a very mysterious condition, severe inflammatory bowel disease, Crohn’s disease, and starting extremely early, just before his second birthday, with intestinal fistulas, unable to take solid food, 100 plus surgeries and no diagnosis, no understanding of what was going on.
So they sequenced in this case not his entire genome but the exome, that is the protein coding parts of the genomes, and found a mutation in a gene called XIAP on the X chromosome. This had previously been linked to a blood disorder.
In those patients who didn’t happen to have the intestinal problems bone marrow transplant had been curative. So recognizing it was a considerable risk, this resulted in a stem cell transplant in Nic from a healthy donor back last summer.
Today he is eating solid food and doing well and recovering quite nicely, the indications being that the transplant has both affected his blood problems, which were actually present but not the majority of the attention, and his intestinal problems. You might say in this case genome sequencing leading to a diagnosis leading to an intervention which may very well have been lifesaving.
Increasingly one sees more examples of this sort coming up. I might especially mention that is true in cancer where the ability to diagnose the specific mutations present in a tumor are allowing in some instances the choice of therapy in a way that is targeted specifically to what that tumor would be most likely to respond to.
Of course, this does raise issues about ethical, legal, and social implications and that has been the reason the ELSI program, I think, has been a flagship of trying to anticipate those issues. This paper, which now looks a little dated from eight years ago, was the paper that we published at the time the Human Genome Project essentially was completed by the finishing of a human genome sequence and all of the other original goals. This fanciful house here, which is intended to look a little bit like Falling Water, was Figure 1 in that article. You will see there were three levels, genomics to biology, health, and society. Then there were six cross-cutting areas that held it all together and ELSI being an important part of that touching on all of those floors.
That has now been updated as of a week ago by this paper, Charting a Course for Genomic Medicine Going Forward, by Eric Green and colleagues. In that is also a description of where this group felt would be — based upon much public consultation — areas of exceptional need as far as ethical, legal, and social issues.
Those would be in genomic research, in genomic medicine, legal and public policy circumstances, and broader societal issues. This might be a useful document to have a look at as part of your deliberations. So that has all been an important part of the program.
ELSI was founded in part because of the very difficult history that connects genetics and medicine, and particularly the eugenics effort which was not, as many in the public assumed, something that arose out of Nazi Germany but really arose primarily in the United States with the work of Charles Davenport and others, the science of improvement of the human race by better breeding, something that all now causes us to cringe.
ELSI was established as an integral part of the Genome Project in 1990. It’s about three to five percent of the annual budget of the Genome Project. It remains vigorously active. One of the major investments by ELSI now is in the Centers of Excellence in ELSI Research, six of them that you see listed here, which involve interdisciplinary effort by multiple investigators focused on specific themes.
Now let me turn to where I can sort of imagine you might be wanting to think about specific areas of interest. I will focus particularly on genetics but say something about neuroimaging in a bit.
First of all, although it may seem like a tired topic, and I’m not sure whether it is an appropriate one for this body to wrestle with, let’s be clear. Genetic discrimination, while we have made great headway, is not something that is completely over.
The great headway could be pointed particularly to the signing of the Genetic Information Nondiscrimination Act, GINA, in May of 2008. That took some 12 years of effort going back and forth with many examples where we thought progress was being made only to find that something else got in the way but it was ultimately signed into law.
Now, what GINA does is to make it illegal to use predictive genetic information in health insurance decision making or in the work place. Those are good but let me just say by way of why this is still on the list those are not the only areas where genetic discrimination could potentially rear its head.
For instance, life insurance is not covered by GINA. One might say that is okay because adverse selection could be a real issue if you knew you were at high risk of a disease and the insurance company didn’t and you bought a $20 million policy that would be quite destabilizing for the whole enterprise.
But should there be some floor, some basic policy level of life insurance for which genetic questions are off the table, certainly that’s worth asking about and some countries have started to deal with that.
I think in the U.S. because it was such a tough fight to get GINA passed people are still kind of breathing easily that that got done and a little reluctant to plunge into some of these other areas.
Long-term care insurance, disability insurance. I think about the whole issue of Alzheimer’s risk testing. If you test positive as being homozygous for APOE4, your risk of getting Alzheimer’s disease by the time you’re 80 is enormous. Should you be able to get the same kind of long-term care insurance as anybody else, or is the company allowed to ask whether you’ve done that and what the results were?
Educational opportunities have been worried about. I’m not aware that there has been a lot of activity in this area but would there be circumstances where somebody at risk, for instance, Huntington’s disease would be denied the opportunity to undergo expensive long-term professional training because of a limited expectation of their career duration.
Military service likewise. Even things like court decisions about child custody. Could that be brought to bear? There have been a few cases where that has happened where somebody is judged to be somewhat less appropriate as a parent because of their future risks of illness based upon some genetic test result. So there are unresolved risks there.
A second area, which I know you will be talking about this afternoon, which is a hot topic in every conversation about every genetic research protocol that I’m involved in, is what are you going to do with this vast amount of data that is now possible to collect as far as dealing with findings that you didn’t expect and do you tell patients or do you not?
The obvious cases come to light immediately in those conversations. Okay, you’re doing whole genome sequencing, or even whole exome sequencing, and your task is to look at people with diabetes, which is the research project in my lab right now. But you encounter a BRCA1 mutation in a woman who is still in her twenties. Are you obligated to share that information? Obviously that’s a tough question. A lot of it depends upon where you set your threshold for things that are truly actionable. Who decides what’s actionable?
I think in the research arena this has led to a lot of interesting discussions. There is a nice discussion about this in Science actually just two weeks ago that begins with a case and goes on to the general discussion about what would you do when you find these tough decisions over what information to give back.
You will have this afternoon a presentation from Susan Wolf who I have borrowed this particular algorithm from a paper that I think was in your packet in the Journal of Law, Medicine, and Ethics that had a very distinguished lineup of authors who are themselves major leaders in bioethics. This suggests, I think, a reasonable pathway if you do it prospectively because this starts with the consent form process and the opportunity to discuss with participants in research that this might happen and that is the ideal.
I guess the thing that I would lay out there is not necessarily as well dealt with, and I’ll be interested to hear what Susan Wolf is able to share with you, is how much have we gone down the road when it’s not a research project, when it’s part of the standard care of medicine.
If the $1,000 genome results in complete genome sequencing and millions of people in the next 10 years, there is going to be an unaccountable almost number of unexpected findings of uncertain significance that are not part of anybody’s informed consent IRB approved process. How should that be dealt with? What norms ought to be anticipated in a circumstance where you haven’t had the luxury of a prior conversation about what that individual would like you to do with the information? That is going to be, I think, an interesting conversation.
I will actually say of all the things I’m going to put in front of you this is the one that I think is probably most ripe for investigation and where I’m not aware that other bodies have been planning to delve into it at the level it needs to be and that is forensic applications of DNA analysis.
Obviously going back some 20 years the ability to use DNA analysis to identify perpetrators by DNA matching is an area that got a lot of attention. It has now become standard in forensics but it is expanding rapidly with the ability now to determine large amounts of genetic information and to be able to use that in ways that were not anticipated, I think, 10 or 20 years ago.
For instance, the whole issue of surreptitious collection, which is certainly increasingly possible with a discarded cigarette or even a glass of wine left at the table after the individual has departed the restaurant, this is in many cases documented as a means by which law enforcement is making a match.
Is there, in fact, some limit that should be placed upon this? Are we comfortable with that? Obviously if it’s for the identification of somebody who has committed a heinous act, then society may be well served by the ability to catch the perpetrator.
But if this is being done in a way where the evidence is much less compelling, or if it’s not even been done for criminal purposes but rather to sort of check out the DNA of somebody who is running for public office, which is entirely possible in the coming future, then you have a whole other set of issues in terms of invasions of privacy. I don’t think this has been appropriately mined for the potential that exists there because it’s not only the surreptitious collection for doing DNA matching, as one might want to do in a criminal case. It also may be done for the purposes of making predictions about that individual’s future.
For instance, if somebody is running for president, wouldn’t you like to know what is the APOE4 genotype for that individual so that you could make some guesses about whether you want to count on an eight-year successful presidency in somebody who is probably already in their 50s or 60s?
Tracking through relatives with the ability now to be able with increasing amounts of DNA information to make a very careful and accurate prediction down to the second and third degree level of relatedness. With DNA databases that are available in forensic situations, you’re increasingly going to have the chance of bumping into a circumstance where you don’t have a match but you know that the perpetrator must be the brother or maybe the cousin of an individual for whom you have a DNA sample.
What is the level of appropriateness of that and to what extent does that cause families to be dragged into criminal investigations in ways that violate their rights? Very much increasing opportunities there without clear guidance that I’m aware of.
Predictions. Taking a DNA sample from a criminal site. There are tests available, admittedly not with great accuracy, that claim to be able to predict the age of the individual based upon either looking at mitochondrial mutations or telomere length or, more commonly, by looking at T-cells to look at the amount of circular DNA which tends to decrease over time.
These are plus or minus maybe nine years but it still is an interesting new window that might actually be very attractive to law enforcement. May I say, with many of these part of the concern is whether these will be taken as more precise than they are. This will not be very precise but it might be taken as such and, therefore, used to include or exclude suspects in ways that went outside the science.
Certainly ancestry predictions which you can get, of course, direct-to-consumer as part of many different companies which will make a reasonable estimate if all of your grandparents happen to come from one part of the world about what part of the world that was. If you happen to be somebody with lots of different lineages, it gets a little more complicated but still probably a fairly good opportunity here to make some sense about geographic ancestry.
Again, great opportunities there but also great risks, especially if this feeds into racial prejudices which is an often possible thing to do. Just the same, this kind of ancestry prediction is now I gather being done in many criminal cases. Hundreds of such cases documented by Pamela Sankar.
Three states actually outlaw this process. The Netherlands has actually a standard legal approach to use this. Belgium and Germany outlaw it and say you can’t do it so there is a lot of confusion here in states and countries about whether this is an appropriate use of DNA and forensics or not. What is the right thing to do here?
Predicting physical appearance. Certainly there are those who would say they are getting pretty good at predicting skin color and eye color based on a DNA sample. There is a project called VisiGen that aims to actually get good enough to be able to draw a likeness of the individual based upon increasing knowledge about genes and how they affect craniofacial anatomy.
We are a long way from having that one yet but when you consider identical twins generally look a lot alike, that probably tells you that we ultimately will figure out how DNA is involved in even such subtleties and be able to do a better job of that.
Predicting recent travel. I mean, the Human Microbiome Project is underway now which is an effort to characterize in a very detailed way what microbes are living on your skin or in your GI tract and being able to do so without requiring that those be microbes you can culture in the laboratory.
Undoubtedly it will turn out that your GI microbiomes says something about where you’ve been. Will that turn out to be important? Just like what particular dust is on your shoes and where that came from, will your microbiome be part of the investigation to see whether you were there on the night of November 22nd or somewhere else?
Then assessing presence of disease. If you have a DNA sample in somebody and you do a genome sequence and, oh my gosh, they’ve got hemophilia, that tells you, okay, I’m looking for a hemophiliac who committed that crime.
Obviously that’s easier to do if you’re talking about a Mendelian condition where the DNA would be very strongly predictive. It would be much less useful in circumstances where you are simply making a guess about the likelihood of diabetes or the risk of somebody being a chain smoker.
Then there is the other side of this coin which is could people now begin to use DNA information in the courts as a defense against criminal activities as we learn increasingly some aspects of how DNA influences that?
The study, for instance, out of New Zealand that showed that individuals who were both abused as children and had a particular DNA variant were much more likely to end up in trouble with the law than those who had only one of those two factors.
So would such individuals be able to use that to ask the courts to let them off the hook, or at least to give a moderate sentence? Or would the court say, “Oh, no. You’re probably the one we want to put away because you’re predisposed to do this again.”
Again, there is a nice story about this whole issue of DNA and forensics in Science about a week ago. I’m coming to the end. Let me just say a word, though, about neuroimaging because I do think there are a lot of fascinating topics here. Although as your materials for the council point out, this is a field that is perhaps not as far along in terms of one’s ability to anticipate what kinds of inferences can be made.
Certainly incidental findings, though, is an enormous obvious overlap with the genetic situation. MRI scans often give things you didn’t expect and often they are things you don’t know what to do with. UBOs they are called, unidentified bright objects. Do you want to have one of those in your brain? Well, you might and what are you going to do with that? Very difficult to say.
Then getting again into forensics, lie detection. Certainly there are those who believe this will become a more accurate means of assessing when somebody is practicing deception.
And then other available methods. Personality prediction or profiling. Again, something that might be used in making decisions about who gets access to which particular kinds of jobs or in the military. Or also it could be used in courts of law.
Neuromarketing, a little scary concept there, trying to assess when people are responding effectively to certain kinds of advertising or branding, or even whose got the brain that is going to be most susceptible to a particular kind of manipulation.
The fascinating area of exploring spirituality and what happens in a PET scan or fMRI scan when someone is having a transcendental experience and what does that mean in a philosophical way about what those experiences are all about. More here, I think, of a philosophical/theological set of questions than perhaps a forensic one.
The utilization of this kind of neuroimaging to perhaps set a new standard for whether somebody in a coma is going to be able to recover or not. Obviously a topic of enormous interest where we all recognize that our best methods are not always absolutely perfect in making those predictions.
And, of course, prediction of future disease risk, especially if you combine it with genetics. Here is a place where the two would intersect. For instance, the ability to look, as was recently published in a JAMA article, at the deposition of beta amyloid using a particular PET scan in combination, perhaps, with APOE testing might give you a very precise statement about somebody’s likelihood of developing Alzheimer’s when they are still entirely without symptoms and what do you want to do with that information?
It might be extremely valuable for interventions if you want to start your clinical trial on people who have minimal disease. Of course, having that particular information provided to you if it turns out that clinical trial isn’t available or isn’t going to work could be obviously quite devastating.
So I will just conclude with this quote from Albert Schweitzer which I’m very fond of because certainly for me, somebody who had the privilege of leading the Genome Project, technology has been such an enormous advance for us in terms of our ability to understand how life works and how sometimes disease occurs. But we have to keep in mind that is not really the goal. The goal here is to benefit humanity. We must not allow our technology to exceed our humanity.
Thanks very much. I’ll be glad to answer your questions.
Thank you very much. What I would like to do is just very quickly go around and have the Commission members just introduce themselves and then I’m going to open it up to a few questions.
Barbara, would you begin? Put your mic on.
Thanks. Barbara Atkinson from the University of Kansas Medical Center.
Nelson Michael from the Walter Reed Army Institute of Research.
Christine Grady from the Department of Bioethics at the Clinical Center NIH.
John Arras from a place you might know something about, University of Virginia.
Jim Wagner from Emory.
Raju Kucherlapati, Harvard Medical School.
Nita Farahany from Vanderbilt University.
Steve Hauser, UC San Francisco.
Anita Allen, University of Pennsylvania.
Alex Garza, Department of Homeland Security.
Dan Sulmasy, the University of Chicago.
Great. Francis, let me just begin — just ask you a question based on an article in Lancet. This is about the challenges and the clinical application of whole genome sequencing. It’s expansive. I think it would be really interesting to me and other Commission members to know what your take on this is.
It just says in the view of the predicted frequency of recessive mutations in the population, every patient will learn that he or she is a heterozygous carrier of more than one serious or lethal autosomal recessive disease. This is the consequence of whole genome sequencing if we, as you have done, and have some members of the Commission have done, have their genome sequenced.
Now, it’s one thing when Dr. Francis Collins learns the results and finds out — I’m not going to ask you to tell us what you found out because we also have as one of our ethical issues the retention of privacy – but it’s one thing when you find out and you can get on the phone or have a meeting with any doctor you wish, or any expert you wish. What do you think are the potential consequences of the ability to, in the market, get the results for the average person who might be curious about this?
Yes, this will be an important aspect of this broad application of whole genome sequencing. I haven’t had my whole genome done but I did have the 23andMe and the Navigenics and the deCODEme analysis. In one of those they included carrier testing for a few of the common recessives so I know, and I will disclose because it’s in my book on personalized medicine, I’m a carrier for alpha-1 antitrypsin deficiency, a condition which I would like not to have. Since I’m a heterozygote I don’t expect will affect me, especially because I’m not a smoker.
I think the challenge will be, first, for people to recognize that this information says that there are no perfect DNA specimens and that will be a come down for people who imagine that they’re the one. I think people will kind of get used to that.
There will be this list. It will not only be for things like alpha-1 antitrypsin deficiency where we know something about what the homozygous state leads to, but it will also be, you know, “You’ve got a stop codon in a gene for which we have no information. That means you’re a carrier but I don’t know what that means.”
You’ll sort of be left with that lack of knowledge. Where the real challenge will lie is the increasing likelihood that individuals will obtain this information prior to childbearing and they will want to know with their mate, “Okay, here’s my list. Are any of the things on your list the same as mine? In which case we then have a chance of a child, one out of four, who will be homozygous for a deficiency of this gene.”
Again, that will become extremely challenging when the list matches but the genes that match are not ones that we know anything about. How do you counsel a couple about their one in four chance of having null function of a gene whose functions is not known because these will be generally rare. Now, one thing that will help us there is having as many complete genomes on people who are already adults so that one could see are there, in fact, homozygous nulls in the population who seem to be doing fine? But often you won’t have that luxury if it happens to be a rare event.
There is an enormous educational challenge here of healthcare providers and of the public. I don’t think we can count on the roughly 1,500 genetic counselors to be able to handle the volume that is going to come out of this. This is a looming potential problem. It’s also an opportunity.
I might say one of the things that I think has not turned out so well as far as the way in which we do prenatal counseling is it generally appears after a pregnancy has already gotten underway.
Many couples really would have liked to have had the information about their risk for cystic fibrosis or Tay-Sachs disease or a long list now of recessive diseases prior to conception. In Tay-Sachs that is possible. In CF it’s generally not. If we get to this point of being able to give more couples that information sooner, that inherently should be a good thing even though it presents all kinds of counseling and decisional problems.
Thank you so much for your presentation and the perspective you bring on both of these.
I would like to probe you a little further on these two sets of ethical issues, Francis, one from neuroimaging and one for genetics. You pointed out a little bit of overlap between the two. In neuroimaging you said, of course, the incidental findings would be an area of overlap. You also mentioned the prediction of future disease risk would be something we would find common to both technologies.
You didn’t imagine, however, that personality prediction might be. I wonder also if another category that you just started to touch on, the age issue, the children — the impact and the ethics of — well, you just mentioned preconception and prenatal counseling on the genetic side, but also the issue of early imaging of children and the degree to which those sorts of findings might become self-fulfilling prophecies. What I’m really looking for, do you see a lot of overlap between these two?
I think you’ve just mentioned two important ones that I didn’t touch on explicitly. Certainly behavior. The ability of genetic factors to influence behavior whether you’re talking about depression or the ability to be a risk taker is an active area and there probably will be some findings that hold up over time, although this is obviously a challenging one to be sure that you are able to replicate results. Similarly a great area of interest in neuroimaging in terms of being able to make those predictions about who is more impulsive.
You rightly bring up the issue about early testing of children prior to the age of majority which in general in genetics we have discouraged unless there is an action required. We would like to preserve for individuals the opportunity not to know if that would be their decision once they became adults. I suspect many of those same arguments would play out in neuroimaging unless, again, there was a reason that the information could be valuable in terms of improving the outcome for that child. Those two also fit.
I do think you’re going to have an interesting time trying to figure out whether all these points of contact are sufficient to allow you to take on both of these areas which inherently have intense complexities unto themselves.
I want to follow up on Amy Gutmann’s question. It seems to me there is a conflict of values here with regard to access to all this genetic information by people out there in the market place.
On the one hand, it seems to me that experts — the more expert you get, often the more wary you are of information being given to people. In my experience experts in genetics don’t believe that more information is always better. On the other hand, you’ve got this sort of free market libertarian sensibility that it’s my body, it’s my stuff, and I should have access to it. Right?
You talked about the need for education. I’m just sort of wondering without knowing a whole lot about this whether there is room for regulation as well. What do we do with all this information that is floating around out there, a lot of which may be useless or even harmful?
Well, yes. It’s something we all worry about a lot, and should. Certainly what one would not want to have are direct-to-consumer efforts that market genetic tests that are actually scientifically invalid.
At a minimum the test ought to have been proven to be clinically valid, that it actually makes a prediction, even a weak one, for which there is unquestioned scientific evidence. In that regard if there are such efforts out there that are actually taking consumers’ money and giving them information that is incorrect, that’s a bad thing.
Regulation sort of rears its head at that point. The FDA after considering this whole question about laboratory based so-called home brew genetic tests for a decade is now engaged in moving in on this in a more active way.
One role that NIH has taken on is the establishment of a genetic testing registry which will be up and going by next fall which will be voluntary but which will encourage responsible companies to deposit in a standard format the information about a genetic test that they offer including what is the scientific evidence that this test actually makes a prediction that is clinically validated.
That, I think, will be a useful clearinghouse put together in a form that is user friendly for consumers to be able to go through it so at least you can find out what is the foundation of a particular claim before you decide whether to sign up or not.
My own personal view is that it’s hard to take a stance of being paternalistic here and saying, “There, there. It would be bad for you to have this information about yourself because I, the scientist [or physician] have decided that it’s just not something you can handle.” That doesn’t quite feel right.
But I do think we have responsibility as a society, and the government is part of that, to make sure that information is not being offered to people that’s actually bogus just as we do in other ways. That would be, I think, a more light touch way to approach the whole question of regulation but a touch that is needed.
There was a paper in the New England Journal, what was it, a month ago from Eric Topol and colleagues asking the question what happens to people who have gone through already this early stage of having their DNA analyzed? Do they all freak out? The answer didn’t seem to be no. In fact, it had very little effect.
We are running a study at Henry Ford of thousands of people who are given predictive information and it’s actually been pretty reassuring that it looks as if people do understand that this is not yes/no. This is statistical up or down kind of risk factors.
They retain it reasonably well. Some of them act upon it in terms of health behaviors. Others don’t but it didn’t seem as if we were starting some sort of psychological disaster scenario which some people had worried about.
It’s reassuring to see that evidence. We are actually over time but I wanted to go into the audience and see if there is somebody who would like to ask Dr. Collins a question. Anybody? If not, I’m going to ask Nita to ask the last question.
I wanted to focus on the area that you thought was the most salient one to take up now which is the issue of forensic collection. In particular you identified two things. One is that there is a difference potentially between identification uses of this technology and prediction uses.
And the extent to which you think that is actually something that is feasible to do to really maintain a strong distinction between the two for technological purposes.
Second, you mentioned, but don’t really focus very much, on the overlap between the collection of neurological and genetic information for forensic use. Of course, both are being used quite extensively and both of them have significant investigative potential. I was hoping you could speak to the extent to which you actually think they overlap or really need to be treated as distinct concepts.
In terms of distinguishing identification potential ad prediction potential, it does get a little blurry. I mean, is predicting age identification or is that prediction? It sort of seems like it’s on the border there somewhere.
Of course, the identification issues, as I tried to point out, themselves are complicated when you start drawing in family members as your hook to get at the potential perpetrator.
In terms of the way in which the courts are utilizing this information both in neuroimaging and in DNA analysis, I think it’s still early days to see how those are going to fit together.
Certainly the neuroimaging part of that in my view and, again, you’ll probably hear from others who are better informed about this during the course of the day, it’s still pretty exploratory in terms of what the actual value of that information is in trying to assess responsibility for actions and whether or not that ought to be part of the process of a legal proceeding or not.
I think we have to be very careful that we don’t over interpret what we can learn from these particular kinds of analyses, be they genetic or imaging based and use that information to deny what is a cardinal principle in humanity; namely, personal responsibility for your actions.
There are, after all, in this room roughly half the people have a 16-fold higher likelihood of ending up in trouble with the law than the other half. Those are the males. Yet, we don’t, I think, have people in a court of law saying, “I shouldn’t really be put in jail for robbing the 7-Eleven because my Y chromosome made me do it.”
Whatever we learn about criminality potential from other kinds of genetic information or imaging information will in general be a much weaker effect than the Y chromosome. If we’ve learned that the Y chromosome is not an excuse for criminal behavior, then we certainly shouldn’t take other factors that are probably less connected and elevate them to that status.
On that note, which I think is very good for us to keep in mind moving forward, I want to thank you, Francis. You got us off to a great start. Thanks very much.
Thank you. It’s great to be here.