Francis Collins: 3 Scientific Breakthroughs Changing Medicine

John C. Reed, MD, PhD; Francis S. Collins, MD, PhD


February 15, 2012

This feature requires the newest version of Flash. You can download it here.


John C. Reed, MD, PhD: Hello, and welcome to Medscape One-on-One. I am Dr. John Reed, Professor and CEO of Sanford-Burnham Medical Research Institute. I'm joined today by Dr. Francis Collins, Director of the National Institutes of Health (NIH) in Bethesda, Maryland. Welcome, Francis.

Francis Collins, MD, PhD: It's great to be with you, John.

Sparking an Interest in Science and Medicine

Dr. Reed: Francis, I want to start off our questions on a personal level. You grew up on a farm in rural Virginia. What sparked your interest in science and medicine?

Dr. Collins: Funny you should ask that question, because you and I have a shared answer to it, which we only discovered after we were both fairly far along in our own professional careers. I did grow up on a farm. I was home-schooled by my mother until the sixth grade because she didn't think the county schools were up to her standards of educational rigor, but then we moved to a little town called Staunton, Virginia, and I went to public school. My interest in science, which started with the usual chemistry set and asking myself "What can I blow up?" really was sparked as a 10th-grader by a chemistry class taught by a very gifted teacher, Mr. John House. I later discovered that was also the way your interest in science was triggered -- by that same teacher in that little public school.

Dr. Reed: Indeed. What a coincidence. Who would have thought? Is there a lesson to be gained from your personal story that can help us encourage more students to pursue careers in science?

Dr. Collins: We really ought to value our teachers. We ought to be sure that teaching is valued by society as a career that our whole future depends upon. We all wring our hands about the fact that our science and math education in the United States is maybe not what it should be. But I'm not sure that we have made that kind of career path appealing to those who have real gifts, as Mr. House did, both in the grasp of the science and the ability to convey that excitement. That's one thing we should do.

I think also we need to be sure that the way science is taught actually conveys the excitement of the present time. Interestingly, my not-so-positive experience with science in high school was biology, which was taught in a rather rote fashion. It was about memorizing things, and I concluded that chemistry was really interesting and biology was really boring. It was only after many years that I realized I had kind of missed the boat on that and changed my direction from chemistry into medicine and into biology. So, we have to work hard to be sure that the science that's being taught keeps up with the science that's happening.

The excitement now could hardly be overstated. If we could just get that message across to that large number of the best and brightest who are out there looking for a way to spend their lives in a way that's going to be a grand adventure, then I don't think we'd have any trouble recruiting people to come and work with us.

Sequencing the Human Genome

Dr. Reed: The field of science is so fast-paced. It's interesting to consider that it was only a decade ago that the human genome was completed, a massive project. I think it took 13 years, $3 billion, and a team effort that you led. Just about 2 weeks ago, a US-based company announced an instrument that will sequence an entire genome in 2 hours for about $1000. Can you share with us your thoughts about what that means for medicine in the 21st century?

Dr. Collins: It is amazing that we can talk about such a thing as a reality. It was science fiction a few years ago when we all started talking about the "thousand-dollar genome." We thought that would be a good way to inspire people to think big, but I didn't dream we'd be there this soon. If someone can sequence all 3 billion letters of the human genome in 2 hours for $1000, it starts to sound like a pretty appealing thing to include in the medical evaluation of individuals, because there's information there that clearly will be valuable and will be actionable. Right now, for instance, we are learning an increasing amount about how to prescribe drugs in a more precise, individualized way. We're not all quite the same.

Physicians listening to this interview will no doubt agree. You have somebody who has a particular condition, you're pretty sure you have the diagnosis right, you write the prescription, you give the dose that's recommended, and it doesn't always turn out right. Some people don't get the benefit; some people even have a surprising toxic reaction. A lot of that variability is based on DNA, and at least for some drugs, now more than a dozen, we think we have a pretty good handle on what that's about. It's possible to make those predictions ahead of time so that you could write the prescription a little differently, knowing that person's particular makeup.

The problem has been that up until now, the logistics get in the way. You're a busy physician. You have a patient who needs a prescription, and you need this DNA what? You have to draw the blood and send it off, and who knows when you'll get the results back. It's just not practical to do this pharmacogenomics thing, which is what we call it, on a daily basis in the doctor's office. But if that patient and all your other patients already had that information determined because you did it once and for all, and the genome is not going to change constitutionally for that person, then it's a click of the mouse to say is this the right dose of the right drug for this person. It starts to get very appealing as a means of improving the way in which we take care of patients, taking full advantage of the individual differences that are often pretty important in determining whether the outcome is going to be good or not good.

Making the Thousand-Dollar Genome a Reality

Dr. Reed: Let's follow up on that notion of the click of the mouse. Now that the cost of generating the DNA sequence is potentially trivial, what advances are still needed in informatics and data information management to make this a reality in the day-to-day practice of healthcare?

Dr. Collins: That is a very important question, and we make jokes about having the thousand-dollar genome. For instance, do we have the half-million-dollar analysis or the half-million-dollar royalty fee for all of those genes that had a patent on them that you have now sequenced? We have to be sure we don't end up with those barriers. The analysis part is extremely challenging right now. Certainly, just putting the information together from what the machines tell you -- which is not all 3 billion letters all neatly laid out in order, but bits and pieces that have to be put together like a jigsaw puzzle. Those things are getting pretty good, though. The algorithms for being able to assemble a pretty good representation of your genome or mine are coming along.

It's more the interpretation, and that is really much of the goal of science right now. In biomedical research, we're trying to understand what it means if you have a T instead of a C in that particular position of that exon of that gene -- and how we build up that database of information, so that if you have the entire genome of one of your patients or yourself in front of you, you have some chance of saying what this means. How should you be better practicing prevention in an individualized way, or how should you use this for drug prescribing?

Another place where genomic sequencing is going to have a big impact is cancer. Certainly, we know that cancer is a disease of the genome. It comes about because of misspellings in DNA. Some are inherited. Most of them are somatically acquired during life. We are not far away; I think the cost of sequencing is coming down to the point of where every cancer, once a biopsy has been obtained, should have its genome sequenced to see what the drivers of the malignancy are in that situation and how that would influence your therapy. However, the assembly can be complicated because the cancerous cell will have a genome that's a bit scrambled, and putting it back together and figuring out exactly how it got there is not trivial. There is a lot of work that needs to be done there, but imagine what that's like.

I think we're probably within half a decade of a circumstance where any patient who has a cancer diagnosis will need to have a genomic analysis of that specific tumor in order to make a wise decision about what kind of targeted therapy might be most beneficial, as opposed to the one-size-fits-all chemotherapy approach, which is often what we've had to do because it was the best we had, but we can do better now.

Addressing the Challenges of Genome Analysis

Dr. Reed: What are some of the ways that NIH is trying to address this challenge of genome analysis, and what do you feel the relative roles are of the government and the private sector?

Dr. Collins: Certainly, a lot of the basic science of how we put algorithms together to try to assemble complicated DNA or RNA sequences is going on in universities supported by NIH. We are the largest supporter of biomedical research in the world, and most of our money goes to universities and medical centers across the country and even some outside the country. But when it comes to turning that into a product -- something that will be part of this genomic revolution, where potentially every one of us will want to have an assembled and reasonably well-annotated genome -- that's a great opportunity for the private sector. There are many companies looking at that as a business plan, expecting that we are maybe at an inflection point where the cost is coming down sufficiently to enable that business plan to be quite attractive.

The big question, of course, will be whether third-party reimbursements are going to be approved for this kind of information. The insurance companies will have to be convinced that this is paying off in terms of an ultimate medical benefit and maybe even reducing costs downstream by being able to do a better job of prevention and treatment. That will take a little time, as it always does, but I do think that if you're a small business trying to figure out what is needed for the genomic revolution, there is a lot of opportunity here in terms of not just the instruments -- although they're doing pretty well at that too -- but also in the analysis, the annotation, and the interpretation of genome sequencing.

Dr. Reed: I think it's an important message to share with our viewers, in this time of economic challenge and mantras about job creation, that public investment in NIH funding -- which is really at the foundation of these business opportunities -- plays a catalytic role in stimulating job creation and in making the country competitive economically on a global level.

Dr. Collins: Absolutely. There was a recent analysis done by a very well-respected economist that points out that NIH each year, through our grants program, is supporting about 490,000 jobs. If you then add into that all of the spin-offs in terms of the way in which this supports new start-up companies, biotechnology, and so on, it's well over a million jobs.

An interesting analysis was done by a different economist about the genome project, asking the question of whether the project has turned out to be a good thing for the economy. It was pretty stunning, the results they came up with. We spent about $4 billion in the United States getting the genome done, between NIH and the Department of Energy. We got that done in 2003. The research looked to see what the economic benefits were of the genome project, which continued up until 2010. They came up with a number that just blows you away: The genome project spurred $796 billion of economic growth in the United States alone. That's a 141:1 return on our investment for that original government plan.

That's a pretty amazing example of how the basic science, which the government supports through the NIH and the other science agencies, is an enormous foundation for building economic growth in our country. In fact, if you look back, since World War II, more than half of the economic growth in the United States has been on the basis of science and technology. That's true now more than ever.

How Do Physicians Benefit From This Science?

Dr. Reed: Let's think a little bit about what this means day to day for physicians. I think many of us who went into medicine did so because of a fascination with the life sciences, yet when it comes to the day-to-day care of patients and the challenges of actually getting paid for that care, I think sometimes physicians feel pretty detached from science. Yet we have this whole genomic health revolution right at our fingertips here. How do you see the physician today preparing him or herself for the genomic health era?

Dr. Collins: I have great sympathy for what the physician today is faced with on a daily basis, trying to practice medicine and trying to take care of patients in the best way they can with all the pressures that come down upon an MD.

My daughter is a practicing physician in North Carolina. She's an internist, a nephrologist, who has now turned her attention to the management of resistant hypertension. That's her practice, and she's incredibly dedicated to the service of these patients who have failed to be managed by first-line antihypertensives and need intense attention in terms of pharmacology and other kinds of support. Yet you don't get reimbursed for that, because that's mostly time and it's not attached to a procedure. So she does well in a given year to clear $50,000 for her own income. That's a pretty sad statement, isn't it, for somebody with this level of specialization focused on prevention that is obviously critical for preventing more cases of heart failure and kidney failure down the road.

Our system is not helping the regular physician, as exemplified by my daughter Margaret, to be able to actually do what they were trained to do, which is to pay attention to the new things that are coming along; incorporate them into their practice; and be able to practice medicine in an evidence-based fashion that feels like, "Yeah, I'm really doing the right thing." And that leads to so many other pressures that come down upon you.

Clearly, NIH sees part of our role as trying to get that information out there in a fashion that is easily discoverable through such things as MEDLINE and PubMed, which is a database of the literature that medical research is producing and is available for free and gives any physician who is quickly looking for information a chance to look at that primary literature. Many other sources of that information, such as WebMD, are also critical for people to try to stay up to date. It is challenging. Things are moving so fast. Even in a field where one is a relative expert, you can't necessarily claim to be an expert for more than a day or two because something changes. There you are.

What Role Does the Internet Play?

Dr. Reed: I'm glad you brought up this issue about the information that's available on the Internet today. Patients are more and more informed and more sophisticated than ever about their disease issues, or at least they think they are from what they read on the Internet. What implications does that have for practicing physicians and their struggle to stay current with the latest and greatest developments in science and medicine?

Dr. Collins: I think it's a good thing. A successful relationship between a physician and a patient is a partnership, and to the degree that patients are better informed -- that they have access to information that can empower them to come to the doctor with a more sophisticated understanding of what the questions are, and to be able to engage in a conversation that is really two-way and not one-way -- I think the outcome has got to be better. Just the same, it can be challenging and even threatening to a physician to have patients walking in armed with reams of information they've already derived from the Internet, some of which is accurate and some of which isn't, because we all know the Internet is not necessarily curated. You may then find yourself spending a lot of time trying to sort through all of that with a patient who is genuinely anxious for any information. Again, the way in which our medical care system operates, you may not feel like you have a lot of time to do that.

One of the critical things that NIH tries to do is to put ourselves forward as an authoritative, evidence-based source of information, and I think we've served that role pretty well. Our Website is one of the top sites in terms of government sites. People do know that if you go to an NIH Website and you see a fact about medicine, it's a fact that's backed up by evidence. It's not just somebody's opinion of the day.

Applying Evidence-Based Medicine to the Physician Practice

Dr. Reed: This notion of evidence-based medicine is something that is talked a lot about in the context of trying to improve the efficiency of our healthcare system. What does that trend suggest in terms of physicians' ability to -- or their need to have the ability to -- be skilled at assessing in a rigorous way the scientific method with which studies are conducted, and to be able to interpret that information not only for themselves but also for their patients, who are going to come in and ask questions?

Dr. Collins: It's a heavy load to put on the back of a physician. You not only have to be up-to-date on the latest conclusions, you're also expected to know the details of the research studies that led to those conclusions. I think if it's something one is really critically involved in -- if you're an oncologist and you're contemplating giving a new therapeutic that's dependent on the most recent clinical trial -- it's good to be able to pull up that primary paper and read the methods section and have a good sense of what that Kaplan-Meier curve was all about.

Most of the time, it's probably not realistic for a physician to be able to incorporate that into an average interaction. Still, if we are, as physicians, really aiming to try to do the right thing for all of our patients, there are times where you want to go and look, especially if there is a controversy, and assess for yourself so that you can ask, "What did this study actually do, and what were the methods that were used?" Let's look at the tables and the figures and see whether the conclusions that we've just read about in The New York Times are actually justified by the work that was done, or whether maybe there's a nuance that got lost along the way.

Exciting Scientific Advancements Coming Down the Pike

Dr. Reed: I want to ask you a little bit about what, besides genomics, is coming down the pike. Certainly, the ability to sequence a human genome will provide a foundation for physicians caring for their patients, but we know that one's DNA sequence is not necessarily one's destiny. We know that there are a lot more layers and complexity to how cells work, how organs work, and how the body works. We need to think about proteomics and metabolomics and even new frontiers, such as stem cell technologies, which now make it possible to take an adult cell from a patient, synthetically convert it into a stem cell, and then program those cells to become whatever type of cell is relevant to the patient's disease, whether that's a neuron, a brain cell, or a heart cell. We can begin to envision an era where we'll be able to create in the laboratory cell culture models of each patient's disease and use that to then interrogate what medicines they might respond to or other issues. Could you share with us some of your visions as to where those kinds of technologies might take us in the next decade or two?

Dr. Collins: This is an enormously exciting area, although much of the promise is not yet fully explored and we have to be careful not to get overly excited until we know for sure what's going to work. The stem cell area, and particularly this ability to take a skin biopsy from you or me, grow those fibroblasts, add just 4 carefully chosen genes, and have those cells go back in time effectively to become pluripotent -- it's just breathtaking that that's possible. I think the standard scientific understanding of the process of differentiation 10 years ago would have said no way -- once you've differentiated, that's it; you're not going back to where you were. Well, we've learned otherwise, and maybe we should have guessed at that, since the genome of a differentiated cell for most purposes is the same as the genome of an embryo. So maybe the instructions were there, but we assumed that they were irreversibly shut down or turned on in a way that you couldn't erase. It is amazing that that is now possible, and it opens up so many windows.

You mentioned the ability to then use those cells, these so-called induced pluripotent stem (IPS) cells, to get insights into the molecular cause of disease. If you're studying Parkinson disease, for instance, and you want to know what's really going on in the neurons, you probably don't want to do that by performing brain biopsies in people with Parkinson disease, but you could take a skin biopsy, convert those cells to IPS cells, and then add the appropriate cocktail, and differentiate those into dopamine-producing neurons; then, you have a pretty good model of what's going on in the brain. You could then use those neurons, and people like you are doing this, to be able to assess among a library of possible drugs what would be the one that would be most likely to make those neurons happy. That is now being done for dozens of diseases, using this as a model of doing drug screens without having to depend on animal models (which sometimes have misled us), without having to just hope for the best and expose patients to drugs with less evidence. This is profound, with lots of potential there.

A technique with maybe even greater potential, but much less well-understood, is to use IPS cells themselves as the therapy -- again, for Parkinson disease, if you could take those IPS cells while they're growing in culture and put them in the patient and fix the genetic problem. For instance, if it's somebody who has an alpha-synuclein mutation, which is a cause of Parkinson disease in some families, you could fix the mutation, then differentiate those cells into neurons and give them back to the substantia nigra in somebody whose substantia nigra is really not doing what it needs to. That sounds like science fiction, but it has the appeal of being a potential reality.

Of course, the advantage here is that you're talking about that person's own cells. It's not like this is a transplant that's going to put you in a circumstance of immediate rejection. They are your cells. You've just reprogrammed them in a way that ought to provide what's missing. You could imagine doing that for other diseases, such as blood diseases, sickle cell anemia, or eye diseases. It's a very appealing possibility.

At NIH, we have just started a new center for regenerative medicine on the campus, which has the use of our clinical center with its 240 research beds as a real possibility of pursuing cell therapy using stem cells. I don't know for which diseases this is going to work, but I bet it's going to work for some, and I bet it will be pretty transformative in the array of opportunity we have to treat diseases that are pretty frustrating right now.

Combining Genomic and Preventative Medicine

Dr. Reed: It is an exciting time, particularly when you think about some of the recent advances in such things as genome editing that now allow us to change a single letter in that alphabet of the genome of a living cell and to be able to think about correcting a mutation.

I want to switch gears and talk a little about preventive medicine. There is a lot of interest in preventive medicine, because we think this is one of the best ways to drive down healthcare costs. In a genomic world, how do you see that changing? What sorts of conversations do you see a physician having with his or her patients, say a decade from now, when in theory everyone's genome sequence is available?

Dr. Collins: I think we should be focusing on prevention. We all say that, but we haven't necessarily been as successful as we might be in implementing that. A lot of the prevention we do right now is sort of one-size-fits-all. You're a person of a certain age and a certain environmental exposure, in terms of whether you smoke or not, and then there's a list of things that you're asked to do. People are not so responsive to that. I'm not sure physicians are all that excited about putting forward those generic prescriptions. Wouldn't it be better if we could do this in a more individualized way, because not everybody needs to have the same list of concerns? How could we do that more effectively?

Let's say, first of all, that there is a genetic test that is pretty good at predicting some degree of one's individual susceptibility. It's a genetic test that is free, and we don't use it very effectively. That's your family medical history. One of my pet peeves is when you read a particular medical record on a patient and you get to the family history and see that terrible hyphenated word, "non-contributory." I would submit that rarely is the family history non-contributory unless this is somebody who is adopted and has no information about any of their blood relatives, because there are clues and we should be using those more effectively. Part of the problem is that taking a family history takes time, and often the patient doesn't quite recall the details of what happened to aunts and uncles and so on.

There is a tool, which I think many hundreds of thousands of people have used, that the Surgeon General has put up on her department's Website. If you do a search for "surgeon general family history," it'll take you to that tool. This is something for patients to basically enter information about their own family members as far as their medical experiences. It ends up getting put into this software and then gets printed out as a standard pedigree. That gives the patient the chance to do all of the time-consuming work of getting the information; calling people up to be sure they have the right information; and then bringing that to the physician and saying, '"We're talking about prevention today. Let's look at this: Are there things in the history, such as that uncle who had a heart attack at age 42, that might change your view about what I should be doing?"

Those are useful adjuncts until the time that we have the whole genome on everybody. We will get there. It'll be hard to interpret. Right now, a lot of the information can skew the odds a little bit, but a lot of the heritability hasn't been sorted out. In a few years -- we could say 5, we could say 10 -- when that information is in hand, then I think prevention really will be empowered by this ability to have each of us focus on the things we need to pay the most attention to and less on the things for which we are not so much at risk. That's where we need to go.

What the Medical School of the Future Will Look Like

Dr. Reed: The importance of taking a good family history is something that you and I and every other medical student learned about in medical school. What do you think, in a genomic era, the medical school of the future will look like?

Dr. Collins: I wish I could predict. I know it shouldn't look like what it is now, and I know how hard it is to change medical school curricula. When I was a medical school professor in Ann Arbor a few years ago, I often scratched my head about how difficult changing just 1 hour of the curriculum could be, because there were all of these entrenched views about what the curriculum should look like. At one point, I think I said it was easier to sequence the human genome than to change an hour of medical school training, and it's still a challenge to do that.

There are some medical schools, though, that have taken this on in a pretty thoughtful, creative way and tried to build medical school education around a holistic view of how biology works from the genome all the way up to the whole organism and to the environment, instead of having things pigeonholed by discipline.

Dr. Reed: Which raises a question: What training do you see physicians requiring in the future to practice medicine in a genomic world? How can we, particularly with our current base of physicians, prepare them to participate?

Dr. Collins: It is challenging if you were trained a few years ago in a medical school where genetics was considered an abstract thing that a few pediatricians worried about. To find out that now, all of a sudden, this is becoming mainstream and it's attached to its own jargon -- which may make it seem even scarier than it should be -- is off-putting. Frankly, the good news about genetics is that this is really straightforward stuff. This is not like learning neuroanatomy, which nearly broke my brain when I was a first-year medical school student trying to make sense of that. Genetics -- if you learn a few principles, you can figure out all the rest. It's very satisfying. It's very digital, but it is not necessarily in the current armamentarium of many practicing physicians.

There is an organization called the National Coalition for Health Professional Education in Genetics (NCHPEG). "NCHPEG" doesn't exactly trip off your tongue, but it's a very important organization, founded initially by the American Medical Association, the American Nurses Association, and the NIH. It is vigorously involved in the process of trying to provide accessible, case-based information for busy practitioners who are at that moment where they feel like they have to learn something. For example, a patient just asked me about that breast cancer DNA test, and I'm not quite sure what the story is. One can go to and see the materials that have been produced there; they're pretty valuable.

We clearly are going to need to have windows into this science, because things change much too quickly for anybody to assume, "Okay, now I've been trained in that field and I don't have to think about learning more." It's going to be a constant need for refreshing.

Dr. Collins' Science and Medicine Hall of Fame

Dr. Reed: Francis, we have just a couple of minutes left. What else would you like to tell us about today? Are there advancements in science or medicine that have made it to your hall of fame?

Dr. Collins: We've talked about a few of them. If I was thinking about moments of particularly amazing development in the basic science arena, sometimes you don't know them when they're happening and then later on you realize how profound they are. The work done by Elizabeth Blackburn and Carol Greider on telomeres, for which they won the Nobel Prize. Something that seemed so abstract, something that they studied in slime mold, of all places, and which now is profoundly important in our understanding of cancer and of aging, is just breathtaking as an example of a molecular insight into conditions that we never dreamed would have that particular connection.

My own laboratory works on progeria, which is the most dramatic form of premature aging. We have discovered the cause of that. We now have kids in a clinical trial for that, and there is a connection with telomeres that just recently has come to light through some of the work in the lab that I did not dream would be there. Again, it casts me back to what Liz and Carol did 20 or so years ago, and how amazing that adventure was that they went on. This is a lovely snapshot of medical research that's undirected in many instances in the basic level, but then reveals things that you could not have anticipated. So, that would be on my short list.

We talked about stem cell research. Shinya Yamanaka is a remarkable physician. He was a surgeon in Japan who decided that it would be better for his patients if he didn't do surgery because he didn't think he was that good at it. Instead, he just decided to get into the laboratory and took on this amazingly complex question in a fearless way and figured out this whole process of how you take a skin cell and turn it into an induced pluripotent stem cell -- amazing.

In the therapeutic area, I would pick Janet Rowley, who discovered the Philadelphia chromosome that characterizes chronic myelogenous (myeloid) leukemia (CML).

Dr. Reed: I think Peter Nowell helped her with that?

Dr. Collins: Peter Nowell, yes; I should have said that.

Dr. Reed: I trained with Peter. That's why I had to mention.

Dr. Collins: Thank you, John. I'm sorry, Peter. I really didn't mean to leave you out. But that led us down this pathway to the discovery of Gleevec® (imatinib mesylate), a remarkable drug that saves the lives of many people with CML and for which, a couple of weeks ago, Janet Rowley, Brian Druker, and Nick Lydon won the Japan Prize (having previously won the Lasker Award for some of that same work). That's a wonderful paradigm that they showed us could really work and could lead to great hope for similar advances in other kinds of cancer and other kinds of molecular disease.

Dr. Reed: Dr. Collins, thank you for sharing your time, your insights, and your vision with us today. For Medscape One-on-One, I'm John Reed.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.