The Future of Genomic Medicine – Anne Wojcicki, Richard Lifton and Eric Green

Eric Green:
So I now want to introduce to you two very special guests that I’m delighted were able
to come here and join us for this conversation. The first is Anne Wojcicki. And for those of you who don’t know Anne,
I’ll tell you a little bit about her. And then she’s going to tell about herself
a little in her own voice. Anne got a B.S. degree at Yale University,
and then worked in the health care investing area for about a decade, focusing on biotechnology
companies. But then she decided to found a company, a
company you’ve probably heard of called 23andMe, where she’s a co-founder and now
CEO. And 23andMe is really a key part of this personal
genomics revolution, one with really — focusing on amassing now a large database of individual
genomic information and doing all sorts of things with it that she’s going to tell
you about. And then in fact, in November of last year
she was named the most daring CEO by Fast Company Magazine. So this is what Anne had to say about herself. Anne Wojcicki:
Being on Wall Street and learning about finance and how health care works is really fundamental
to everything that I’ve done because I learned the mechanics, essentially, of the system. So I did it for about 10 years. And then after 10 years I said, look, like
I — “One, I don’t think I’m going to learn anything else; and two, like, this is
not the system that I really want to be part of.” Being surrounded by an environment where people
were really just pursuing money really started to conflict with my ethics and my values. And I spent a lot of time volunteering in
Bellevue Hospital in New York and then at San Francisco General out here. And I almost did that as a detox where I’d
be on Wall Street, and I’d spend all days, like, trying to monetize, you know, obesity. And then I’d go into San Francisco General
and it was tragic. The idea of just thinking of these people
coming in — sort of crisis situation, or being sick, and seeing people with their families
and thinking that — I’m just thinking about you as a dollar amount. Like how do I monetize the situation? And it became pretty disgusting to me. So again, I ended up having — feeling like
Wall Street did not reconcile with my values, and I had to leave. And 23andMe really evolved out of that frustration
that, I think, one of the ways that you can circumvent this — essentially, this whole
system is through the individual — is that the individual owns their data. And if I can empower you to make a difference
in your own health, we can potentially really change health care. Eric Green:
So please welcome the — please help me in welcoming Anne Wojcicki to the stage. Anne, come up, take a seat. [applause] Eric Green:
Middle seat. The hot seat. So the second — the second and other member
of this panel that will join me for this discussion is Rick Lifton. Rick graduated from Dartmouth with an undergraduate
degree, went on to get an M.D. and Ph.D. degree from Stanford University, and has just had
a meteoric rise, if you will, of a successful research career. He’s now a Sterling Professor and Chair
of the Department of Genetics at Yale University. He’s also an investigator, for actually
many years, in the Howard Hughes Medical Institute. And he is a recipient of many awards and accolades,
a small subset of which is shown here. And you can see the American Heart Association,
the American Society of Nephrology, and the American Society of Hypertension have all
honored him because of major advances he has made. In particular, doing the genetics of disorders
related to blood pressure, cardiovascular disease, and bone density with multiple, wonderful
successes credited to Rick and his laboratory. And this is what Rick had to say about his
circumstances. Rick Lifton:
I’m Rick Lifton. I’m Chair of the Genetics Department at
Yale and an investigator of the Howard Hughes Medical Institute, and we do human genetics
of cardiovascular, kidney, bone, and several other diseases. My favorite applications of genetics are to
reveal mechanisms of disease that we’ve known about the disease forever, but have
had almost no insight into their underlying, core biology. And when we think about where we make our
greatest advances in medical therapeutics, it typically starts with understanding what
the underlying biology is. And that allows you to target the, you know,
what is the real core disease process as opposed to the secondary pathways that revolve around
there. And so I think that’s what’s so beautiful
about genetics and the ability to understand complex traits — is the opportunity for the
first time to really get to the core, underlying, primary abnormalities, and understand those. Now, of course, those will not necessarily
tell you everything you need to know about the disease, but they are incredibly strong
starting points for unraveling the pathophysiology. And, of course, we expect that these will
define what the therapeutic opportunities will be. And we can’t predict with any certainty
what those therapies will be until we really understand the basic biology. Eric Green:
So please help me in welcoming Rick Lifton to the stage. [applause] Eric Green:
Thank you so much. Good to see you guys. Anne Wojcicki:
Good to see you. Eric Green:
So let me lay some ground rules here — what we’re going to do. We’re having a conversation. I’m going to kick it off with a question,
and both Anne and Rick prepared just a few slides to sort of help answer that opening
question and provide a little bit of background. We’re going to talk for a little bit, but
then we’re also going to take questions from the audience. And there will be index cards that will be
passed around at some point during this and you can feel free to write questions out for
one or both of them. If you have a really easy question, I’ll
take it. If it’s really hard, make sure it goes to
them. [laughter] Eric Green:
And then we will get — we probably won’t get to all the questions, but my staff over
there is going to sort through and try to find the best questions and pass them on to
me. And we will have some of the questions then
from the audience put before our two wonderful speakers. So what we want to start with — which, actually,
I think is a great way to think about it — I think about this a lot — is, you know, we
sort of in this decade of incredible advances; and we’re almost at halftime — 2015 will
be about halftime. And we knew when the decade began that this
was going to be an exciting one. And we sort of see 2020 now — really now
only about six years away or something. Soon it will be five years away. So I’m curious to hear from each of you
— we’re going to start with Anne — and sort of — in the year 2020 what is going
to be the role of genomics in medicine? I tantalized everyone with some of the early
examples. But one of the big questions is, how quick
will this happen? And over the next, you know, five, six years,
what is this going to look like when we’re celebrating the January 1 of 2020 or something
like that? So, Anne, you can advance your slides. Anne Wojcicki:
[unintelligible] it up? Eric Green:
Yep. Anne Wojcicki:
Okay. So I will start getting to that question towards
the end, but I will — I’ll go through a little bit on 23andMe and that will sort of
touch in here. How many of you actually have your genome
or have been sequenced or — okay. Eric Green:
About 10 or 12. Anne Wojcicki:
Yeah. Yeah. It’s less than I thought. So part of the whole purpose for me of 23andMe
is — I grew up as a child of a scientist. And that science in my opinion should be more
in touch with the consumer. And that you actually want to — like, you
should all be participating. There’s billions of dollars that come from
your institution and that we should all be following it. Like there’s nothing more exciting than
actually following science and watching it progress. So 23andMe was sort of born out of that idea
of, like, how can I actually engage you, the consumer, with this. And one of the things that we do is we, you
know, like we said, it started out as a $3 billion venture. And we’ve made it affordable and accessible. So it’s direct to consumer; it’s something
that you can actually afford. It’s $99. It’s something that you can get access to. And the idea is, like, then you should be
able to explore it. And what we have now is 750,000 people who
have actually done this. So what’s amazing here is in a pretty short
time period it’s now become accessible, and large numbers of people actually have
experimented and played with their own genome. And what we’ve also learned is that people
actually want this information. Is that science — when you actually empower
people with it — that it’s actually — people can understand it. And so when I think about the future — one,
there’s a big question out there — is how much can actually people understand the science. What we find is that the average individual
actually can be empowered with this information and they can make sense of it. And so here’s some of the areas, Dr. Green
already touched on these, you know, things that you can actually learn from your DNA:
you know, medication response, disease risks, inherited conditions like cystic fibrosis;
and then interesting traits, things like, you know, lactose intolerance or caffeine
metabolism. And then things like ancestry, and ancestry
includes other fun things like your Neanderthal status. Which again, is part of the things that we’ve
done because it’s science that’s coming out. It talks about you. That you’re not — just as much as we’re
talking here about disease and wellness, you’re not just about a disease. You know, you have your whole — I mean the
fascinating thing with your ancestry is that it’s, you know, it’s your whole history
of you. It’s your personal history. And that’s so cool. Like how much Neanderthal do you have? And then comparing that to other people. It tells your own personal story. So what we’ve also learned with this information
is that people get their genetic information, and that it does motivate them to change. And so this a study by Robert Green that — more
of it’s coming out soon — but people are getting their genetic information and they’re
actually showing up to their physician, and they actually want to make a change. And I think one of things that we’re seeing
genetic information do is that it gets people interested in their health before they’re
sick. And for all of you, if you raised your hand
and you say, well I’d actually rather, you know, treat my diabetes than prevent it, well
then you’re part of the old system. But for me, personally, like I’d rather
actually prevent my diabetes. And if I know that I’m at risk for something,
how is that I can actually, you know, prevent it? And I think more and more you’re starting
to see these trends like Walmart and your convenience stores getting involved in health
care, which is where you go to regularly. So more and more there’s going to be more
support for a preventive kind of society that’s probably going to be outside of the existing
medical system. So again, one of the things that we’re doing
is we’re actually getting people excited about genetic information. And I think the more — one of the best things
I think that we can do is to help people like Dr. Lifton and all of NHGRI. Is if the entire world gets excited about
the genetics research coming out and participates, you’ll help clinical trials go faster. We’ll all be more excited about funding
going to this area. And I think we’ll uncover what the genome
means much faster. So as we see this — this is a slide that’s
probably shown all the time. That shows Moore’s Law and how, you know,
costs of computing is going down. And you can see that the costs of actually
getting their genetic information has dropped dramatically. And so this sort of leads to the conclusion
— oops, we have one slide I missed — is that at some point genetic information is
going to be free. And to me that’s actually one of the bigger
questions when we think about 2020 and beyond. Is that so much of population health today
is based on the fact that, you know, is it actually — is it cost-effective for you to
get this information? So you look at things like the Angelina Jolie
effect, you know, and the BROCA testing. Right now people are get — the BROCA test
if they have a history of breast cancer or if they’re Ashkenazi descent. But at some point, if the information’s
free, everyone’s just going to have this information. And so then how does that actually change
some of our population health guidelines? And how does that then change when people
are coming into you and they don’t necessarily have a history of, you know, something like
sudden cardiac death and they’re walking in saying I have this genetic variant, what
do you start to do? So I think that’s going to be actually one
of the interesting issues. One of the other areas — the obstacles is
going to be how is this actually all going to become regulated? So what is going to be the path forward to
actually get all this information out? And what you’re starting to see is there’s
all kinds of countries — you know, the U.K. has a massive genome — 100,000 Genomes Project. Beijing actually has the largest sequencing
shop in the world called Beijing Genome Institute. There’s tons of other countries. The whole Middle East is actually launching
these big sequencing initiatives. So how are we going to keep up? And how is that we’re going get through
all the ethical, legal, social debates about how to use genetic information? But because the genie’s out of the bottle,
the rest of the world is engaging in this, what is our role? And what is the role that the U.S. wants to
be playing? So with that I’ll pass it on. Eric Green:
Yes. Pass it on and — Richard Lifton:
Great. Thanks, Anne. So just to continue framing the discussion
a bit. To emphasize the pace at which this work is
progressing — if we think about humans as a species — a 500,000 year history of Homo
sapiens, give or take a hundred thousand years or so — it’s only in the last 50 years
that we’ve really begun having the faintest ideas about how the well-known principle of
like begets like, which is where genetics really starts, to understanding the fundamental
contributions of individual variation to health and disease. We only learned about the structure of DNA
in 1953. We unlocked the genetic code in the 1960s. In the 1970s, we just began to be able to
get our hands on individual pieces of individual genes. And then this led to the beginning of the
Human Genome Project in the late 1980s. I have to say, having started in this area
in 1975 as a graduate student — I think if you had asked anyone involved in genetics
and genomics in 1975 when would we be able to sequence human genomes essentially at will,
I don’t think anyone would have imagined that that would have happened within — certainly
within my lifetime or the lifetimes of just about anybody else who was engaged in the
activity now. So the pace at which this has happened is
really stunning. And if you think about the eras that we’ve
been through in human genetics, it was only in the late 1980s that we really began being
able to get our hands on individual disease genes and find these to begin with. And that’s been taken over in just the last
five years by this dramatic reduction in cost of DNA sequencing, which now allows us to
literally identify an individual with a particular disease and be able to think about the question,
can we figure out what that person’s — the cause of that person’s disease might be
just from studying his DNA and perhaps several other individuals. Now we’ve talked for — at great length
over the years about the impact of what this work is going to have on human health. And I’ll just give a couple of examples
of this. So in the case — Eric mentioned cancer as
a particular beneficiary of genomics research, because there we know that there are somatic
mutations in the cancers that are not present in the patient’s germline that occur in
single cells that initiate the cancer. And those single mutations with large effect
are terrific targets for new therapeutics. And in the case of cancer, we now have specific
treatments that are aimed at the specific genetic abnormalities in individual cancers. So if you have chronic myelogenous leukemia,
there’s a specific mutation that is found in nearly all patients with this disease;
and there’s a specific drug that has transformed the history of this disease from one that
is uniformly fatal to one that is now highly treatable with a specific drug that inhibits
the specific molecular mechanism that is mutated in this cancer. And there are a number of other such oncogenes
that cause malignant melanoma, lung cancer, and glioblastoma multiforme, the most lethal
form of brain cancer, that have specific treatments that are coming from this. Further, the study of rare patients frequently
gives us information about the pathogenesis of common disease. And these suggest particular targets that
will be beneficial for therapeutics. So for — one of my favorite examples of this
is if you’re missing a sodium channel that is in the dorsal root ganglia in your spine,
you’re completely impervious to sensory pain. Now those patients don’t need a drug for
pain relief. However, our current treatment of pain is
very ineffective and has a number of side effects. If you are able to develop a specific inhibitor
of that sodium channel, you wouldn’t even need that drug to get into the brain in order
to have effective treatment for pain. And there are many other examples where rare
mutations found in extremely outliers in the population have suggested therapeutic targets. And these are becoming the norm in the pharmaceutical
industry; to use human genetics to identify what are the best targets where we’re going
to have therapeutic efficacy. And this is rapidly emerging as these drugs
are coming on. And looking down this list, one that I’m
particularly enthusiastic about is the treatment of Alzheimer’s disease. Currently an intractable disease of the aging
population. And the best targets that we’re likely to
get for the next decade have come from rare mutations with early onset Alzheimer’s disease
that have pointed to a specific pathway for which drugs are currently in development. And we’ll have to see whether these will
prevent the development of Alzheimer’s disease in the population. But as both Anne and Eric have indicated,
the last several years have led to spectacular ability to rapidly and efficiently sequence
large numbers of human genomes. And this is unlocking problems that have heretofore
been unapproachable. And one reason that they’ve been unapproachable
is that there are some diseases that are caused almost exclusively by de novo mutations. Not necessarily exclusively, but there are
many patients with these diseases where mutations that are absent in the mother and father occur
and cause a severe disease in the offspring. One example of this is congenital heart disease;
where we’ve now sequenced a large number of unaffected parents with a severely affected
child, where the plumbing that is required for the oxygenation of blood and its distribution
to the tissues doesn’t work properly. And we’ve identified a number of mutations
that drive this process and have identified an underlying pathway involved in modification
of the proteins around which DNA in the nucleus is wrapped. And perturbation in that — in those pathways
are driving this disease in a significant fraction of cases. So this is one example of the kinds of discoveries
that are happening now. And I want to make a point that when we ask
what remains to be done — the answer is almost everything. So they’re 21,000 protein-coding genes in
the genome. We know what happens when humans have mutation
in about 3,000 of those genes. So you don’t need to put a very fine point
on it to say there’s much more discovery that lies ahead than lies behind. And that just encompasses the 1 percent of
the genome that codes for proteins. When we get outside the part of the genome
that codes for these 21,000 proteins, it’s really terra incognita. We don’t understand the language of genomes
at all. If I were to give everyone in this room the
sequence of the genome of a mouse, an elephant, and a human and said tell me which one is
which, we would be completely incapable without other knowledge of determining which is which. Our genomes are very similar to one another. We all share the — generally the same — when
we say we all — all in the animal kingdom share the same core set of genes. And it’s how those genes are used that make
the difference. We don’t understand that language hardly
at all. The last point that Eric alluded to was the
ability to use this technology in the clinic. And I want to give one recent, quite dramatic
example from our work at Yale. So we were presented with a case of a 15 day
old boy who had severe diarrhea and fever, who was progressing in an unexpected clinical
course that the physicians taking care of him in the intensive care unit were concerned
that he was not going to survive. He was developing coagulopathy, blood clotting,
and loss of red cells and white cells. And the physicians were very concerned about
his health. And he had been seen by every patient — every
consult service in the hospital, and nobody knew what the diagnosis was. So in five days we turned around the sequence
and analysis of all of the protein-coding genes in his genome and his parents. And surprisingly we did not find a new mutation
that described his disease. But we obtained the hint that there might
be a mutation causing his disease that caused an auto-inflammatory disorder. A disease in which the normal inflammatory
response pathway was activated in the absence of the normal stimuli for that. Unfortunately, the day before we obtained
our complete analysis, the boy died of pulmonary hemorrhage. However, it came as a matter of great surprise
when the next day we learned that his father had been hospitalized at the same outside
hospital with high fever and was intubated with respiratory distress. He turned out to have the same disease that
had been undiagnosed his entire life, despite the fact that he had been hospitalized in
a hospital his first month of life with a syndrome very similar to his deceased child. Because of this mutation that was identified,
we ultimately made the diagnosis of a new previously undescribed auto-inflammatory disorder
and he was put on high-dose immunosuppression and recovered. And he then revealed that throughout the course
of his life, he had had periodic fevers that were up to 104 to 106. They were always triggered by emotional or
physical stress. He recognized that with the stress of his
son’s death that he was kicking off another of these inflammatory episodes, but thought
he would ride it out the way he had every other one that he had had, but ended up with
this near fatal disorder in the hospital. So this is one dramatic example where sequencing
both defined a new disease and led to treatment of individuals in the family, and has suggested
preventive therapy that can be offered these individuals. I think that’s one of the last points that
I’ll add now. Is that frequently the discovery of the fundamental
pathogenesis of disease not only suggests the mutations and treatment, but frequently
will suggest prevention. Eric mentioned that we’ve had a long interest
in high blood pressure, a disease that affects a billion people worldwide and contributes
to 17 million deaths from cardiovascular disease around the world every year. Well, the mutations that cause this trait
converge on how the kidney handles salt, which immediately suggests an environmental interaction
with dietary salt intake and blood pressure. And this has led to a recognition that we
ought to be able to reduce blood pressure in the population by reducing dietary salt. And there are now 32 countries around the
world that have public health programs to reduce morbidity from cardiovascular disease
by reducing dietary salt intake with the scientific fundamentals coming from these rare patients
with extreme forms of high and low blood pressure. So if we look forward to 2020 and beyond,
where are we going to be and what are we going to be able to do? The things that seem most obvious are — we
are certainly going to be sequencing virtually every cancer in patients because those are
going to drive the therapies that we’re going to give these patients as we move forward. Similarly, we’re not going to waste too
much sleep wondering whether we ought to be sequencing patients who are in the intensive
care unit and critically ill in order to try to make what might be unexpected diagnoses. And there are multiple examples of this happening
now. The bigger questions, I think, going forward
will be to what extent will the sequence of every individual in the population contribute
to their health care? And that’s a research question at this point
in my view. There are relatively few examples where identification
of a specific mutation today we will be able to say, aye, we know what treatment you ought
to have as a consequence. The BRCA1 and 2 mutations are wonderful examples;
where if you have these disease-causing mutations, this has very strong implications for how
your diagnosis — your future diagnoses and susceptibility to disease, and we don’t
yet know how frequently that will, in fact, be the case. I think it will be relatively modest impact
if we tell our patients have your genome sequenced and we can tell you whether you have a 1 percent
or a 2 percent risk of developing schizophrenia. That’s not the kind of information that
we’re likely to be going to our physicians for. What we want to find are those mutations that
are going to be driving disease. And for that reason, I think quite conservatively
one can imagine that we’ll be starting in the clinic with diseases that — with patients
who have disease or are at high risk for disease. And then newborn screening, I think, is an
open question. We now screen in most states in the country
every newborn for a handful of diseases ranging from 20 to 40 different diseases. We might do much better than that with genome
sequencing rather than the current screening tests that we perform. But these are questions that we’ll have
to settle as we go forward. So I’ll stop there. Thanks. Eric Green:
So that was great. So what I want to drill down a little bit
because I think what you touched on, and actually each of you sort of put out something to think
about, really I think relates to prediction. Because, I mean — first I want to talk about
what’s truth, not that we know it, but let’s talk about what truth might be. And then we should think about some of these
implementation things because I think there’s relevant — I know both of you are very interested
in those. But let’s focus on truth because we don’t
know what truth is. And when we try to use genomics as a tool,
as a predictive tool, that’s where I know there’s disagreement in the scientific community. So Anne, you made a passing reference to you
want to know things to prevent your diabetes. And the question is whether it’s diabetes
or hypertension — something that Rick is passionate about. The question is what do we know now that is
predictive if handed somebody’s sequence or handed the kind of data that a 23andMe
test might reveal? And, you know, what do we know now? But also, where do you think we’re going
to be in 2020? Of course, we don’t know the answer. But where do you think we’re going to be? And does that tend to influence, sort of,
how we try to set up the system for — the medical system for dealing with this information? Because I think as Rick says, it’s going
to be a no-brainer for cancer, a no-brainer for rare diseases, probably for some examples
of pharmaceutical genomics. But what — I know it gets more, shall we
say, spicy and debate-y — Anne Wojcicki:
[laughs] Eric Green:
— out in the community is really when you get into these complicated diseases like diabetes
and hypertension — there’s an environmental component — and many genomic contributions. Will we have something predictive enough to
really tell us something that can be used for clinical care, especially in the prevention
realm? So Anne, what do you think? Anne Wojcicki:
Sure. So I would say for 23andMe one of the things
that we were doing — and again, just for disclosure, we’re not selling our health
reports today as we’re working with the FDA. But historically, what we did is we had two
sets of information; information that was what we call sort of the four-star reports
that was clinically actionable. So BROCA, for the BRCA1 and 2, cystic fibrosis,
associations with pseudo-cholinesterase inhibitors, for instance, about drug response. So we had that kind of information that — it
was sort of unrefuted — that was known to be — to have meaning in the clinic. And that if you walked into most physicians
that they should know what to do with it. And then there was sort of this grey area
of disease-risk prediction. And that’s where there’s been the most
controversy. And type-2 diabetes is a good example where
you can say there’s good data; there’s interesting science that is being done here
that — that again, I’ve always felt strongly that you, the consumer, the taxpayer, since
you’re paying for this research, you should have the ability to go and look at it yourself
since you’ve paid for it. And so can 23andMe engage you in the science? And we’re never going to know how to actually
predict disease risk unless we have massive sums of data. And so one of the reasons why 23andMe has
a huge research component is that what we’re looking for is how can we actually create
this community of tens of millions of individuals and understand which of them actually go on
to develop disease; what are the genetic risk factors that they have, and can we actually
really develop a risk prediction modeling system that’s based on tens and tens of
millions of individuals — the genetic information, taking in environmental information, and looking
at this longitudinally. Because I think that it is a grey science
right now, but in the future, there should be able to be pretty good risk predictions. And again, cholesterol is sort of an example
of that. Cholesterol doesn’t mean that you’re going
to get a heart attack. It means that you have a risk factor for it. And I think the genetic — this risk prediction
— you might have a thousand genes that put you at elevated risk. We might be able to calculate a score. And with that information, it’s another
risk factor that gets totaled up with your environmental factors, your family history,
et cetera, so that you know these are the areas that you’re higher risk for that you
might want to actually — there might be something that you can do about. And these are things that you’re — you
know, that you’re not as high risk for and potentially you don’t need to be as concerned
about. So it’s definitely one of those grey zones. But again, part of the reason why we’ve
had this massive research initiative is that you’re never going to get to that solution
about how — what actually is the meaning of all this information until you’ve actually
can do these kinds of massive, million types of person studies. Eric Green:
But 2020 — you think by the time we’re there, do you think for most of these diseases
that we’re talking about — hyperlipidemia, cholesterol, high cholesterol, hypertension,
diabetes — do you think we’re going to be mostly able to predict based on genomic
testing? Or you think we’re going to be just barely
up the curve in being able to predict? Don’t worry about it. We’ll save the video, so we want to make
sure don’t [inaudible] — [laughter] Anne Wojcicki:
[laughs] I would actually say it’s dependent upon how quickly we can grow. I mean 23andMe last year was on a mission
to get to a million individuals. If I could hit 10 million people I think in
five years, I think we can actually have pretty spectacular risk prediction algorithms. Eric Green:
So that’s — so your prediction is that that knowledge can actually be gained if the
right study is done? Anne Wojcicki:
I think — completely. I think if you can do — if we could get 10
to 20 million individuals engaged in research, filling out surveys, we could actually understand
genetics and what it’s going to mean for your disease risk. Eric Green:
Okay. Rick, I have a feeling you’re going to have
thoughts about this. Richard Lifton:
Well, so, I’m an optimist. And I want to see the experiments done. And the question is, what are the experiments
that one wants to do? There’s been a huge effort associating common
variants with common disease, and we clearly have associations. Sometimes they’ve been pivotal in giving
us insight into disease where we didn’t have any previously. But in general, these have small effect on
disease risk and so our ability to predict who is going to get disease from these — the
assembly of common variants across the genome has been rather modest. And so I think the question going forward
is, will rare variation prove to be of sufficient importance to give us better ability to predict? And that’s an experiment that has not yet
really been done in a comprehensive way for any common disease. And this is — these are experiments that
are going to be done over the next five years for, I would expect, for virtually every common
disease as you go down the list. We’re going to be sequencing large numbers
of individuals, and the key question is how large is large? What is the number that we’re going to need
to — in order to actually believe that we have tested the hypothesis? And the current estimates are probably that
it’s going to require in the tens of thousands of individuals for each disease that we’re
interested in. And that’s going to require a lot of work
around the world to collect the cohorts, study them, and make conclusions. So I think it’s a very open question as
to where we’re going to be by 2020. Because, as you noted, that’s a short time
from now and we’re barely scratching the surface in most disease areas right now. And that presupposes that we’re sequencing
just the protein-coding parts of genomes. And we, as I indicated earlier, we really
don’t understand the language outside the protein-coding parts, so I’m not sure we’re
going to be able to digest that part [inaudible] — Eric Green:
Nor do we know the balance between. Of all the things that happen in the genome
that cause disease, we don’t even know is — what the fraction is happening in the genes
themselves versus those happening outside the genes. Richard Lifton:
Exactly so. And so for that reason, you know, I think
we don’t know much more than we do know at this point. And it is a long way to go before we are capable
of making confident predictions. Eric Green:
So what was your reaction to the INOVA health care ad that I showed in my talk? What was your immediate gut reaction? Richard Lifton:
Well, so, I think the field of genomics has always been fraught with optimism that frequently
is not tempered by reality. And I think my immediate reaction to that
ad fell squarely into that. I don’t think we’re ready to be making
those kinds of predictions for most people. I’ll give an example. So when exome sequencing became — came online
and suddenly people could do this, I must have reviewed one paper a week from a high-profile
journal that ran as follows. We sequenced 50 healthy year olds [spelled
phonetically] and identified variants in known disease genes. We then spent x number of dollars working
them up for those diseases, and were surprised to find that they in fact did not have those
diseases. And of course, this reflects a lack of knowledge
of Bayes’ theorem; that if the prior likelihood of a healthy 50 year old having a lethal genetic
disease is extremely low, if you find a variant in that gene, the likelihood that that’s
a disease-causing variant is probably very unlikely. And so you can spend a lot of money chasing
these kinds of diagnoses. And this, I think, is a major problem that
we have as we try to interpret genome-level sequence. Is — for example, in BRCA1 today, we know
a lot of mutations that clearly cause — are causally-related to the risk of breast cancer. There are other variants that change the protein-coding
sequence, but do not predispose to breast cancer. And the bane of our genetic existence today,
in many cases, is the variant of unknown significance. We find a variant that we’ve never seen
before in any of the people who have been sequenced around the world. And the geneticists are left with a puzzle. There’s a variant there; we haven’t seen
it before. How do we know whether it’s functionally
important or not? And there are approaches to try to address
all variation in every gene in the genome. And these are the kinds of things that we’re
going to need to have the kinds of diagnostic certainty that’s going to increase the power
of our tests ; so that we’ll know which variants have an effect on the encoded protein
and which — or the genome function — and which don’t. But we’re very, you know, the first papers
in this are just being published literally now. Eric Green:
Anne, what was your reaction to that ad? You’re the — you’re an optimist, I know. Anne Wojcicki:
I’m definitely an optimist. I mean, I think the ad was non-specific enough
that it can mean anything. Eric Green:
Yeah. Anne Wojcicki:
I mean, some of the things that I am excited about is that I do think that sort of this
general population health that — you know, I just turned — I’m 41, so should I get
a mammogram? So are all 40 year olds the same? And I look at other associations like — I’m
actually curious to ask you this question — there’s associations with macular degeneration
that actually are discovered out of Yale. And that falls in the grey zone. But if you were homozygous for that high-risk
variant, would you show up at your ophthalmologist sooner to check for macular degeneration? And — so would you? Richard Lifton:
So — [laughter] Eric Green:
You don’t have to tell your age. Just because she’s bragging — [laughter] Eric Green:
You don’t have to do that. Because we were both once 41, weren’t we? Richard Lifton:
So this is actually not a theoretical question. My father had retinitis pigmentosa. And he was from — his parents were from the
same small village in Russia. And it was no surprise that he was — probably
his parents had distant relationship to one another, and he was homozygous for one of
the genes causing retinal degeneration. And I was actually faced with that question
of do you want to know? And my father went — had a productive life. Had to retire early because of his blindness. It progressed to complete inability to see
in his 70s. And I made the conclusion that — drew the
conclusion that he had a productive life and I — would you — how differently would you
want to behave if you knew? And for me, the question was, is there something
that you would do about it today? And I’m very enthusiastic about things that
you have some ability to change. At that time, there was no treatment for the
disease other than laser zapping of the lesions to try to prevent them from progressing. And I decided that I didn’t want to have
eye exams at that point. So I think individuals will vary in how they
respond to that. APOE4 in Alzheimer’s disease today is a
very practical example of this. Allele frequency around the world is about
17 percent. And about 4 percent of the population will
have two mutant copies of this gene. And if you have one copy you’ll get Alzheimer’s
disease on average eight years earlier; and if you have two, 16 years earlier. And practical question — do you want to know? And currently there are no useful treatments
that we know of that will prevent progression if you are an APOE4 carrier. And I think individuals will vary in their
response to that. And in general, in my experience as a physician,
patients really want to know if there’s something that can be done. And if there’s not, it’s a very mixed
impression. Anne Wojcicki:
Right. But one of the things I think that will be
interesting is with population health guidelines. So things like macular degeneration or with
breast cancer, can I actually get more specific guidelines for me that are based on my genome? And that’s when I start to look at like
some of these new, you know, prevention diagnostics that are coming out that are very expensive. You know, if all women didn’t have to get
a mammogram — one, that’s great for me, but two, that’s a big cost savings. And so I look at things as well, like macular
degeneration. Where it’s in that grey zone now — still
of information — but I can see that actually being a really valuable risk prediction tool. To say this group of individuals — not everybody
actually has coverage with an ophthalmologist — but if I could actually send these people
at 50 to go and get screened, and there is now a treatment for delaying and actually
stopping the progression of macular degeneration — the cost to society of blindness is very
expensive. So can I find those individuals who are homozygous
and actually get them involved in front of a physician much sooner? And I think that is like — as much as that
information is the grey, that’s the kind of thing, I think that 23andMe could actually
start to validate by having this community-based research project. And then start doing these types of population
health studies to say yes, like, I can get these people in ahead of time and actually
try and prevent the disease before it actually really becomes costly to society. Eric Green:
One of the things I worry about a lot in these situations is managing expectations. And I, of course, like the idea of getting
community-based research involvement the way you describe it. But do you — are we convinced people see
the distinction between being involved in research and actually having what they perceive
as getting as truth? I mean, my reaction to the commercial when
I first saw it was at first was like, oh my gosh. They’re talking about genomics. Isn’t this fantastic? And then by the end of it, oh my gosh, they
are really subliminally implying some things that many people may not, sort of, appreciate
as not really here and now. It’s the promise, but it’s not reality. I worry a little bit about getting people
involved. Will they interpret, sort of, the excitement
of doing a study as being the new way that you should make life decisions? How do we — Anne Wojcicki:
So I think those are two — Eric Green:
— strike that balance? Anne Wojcicki:
I think it’s two different questions. So one thing for us is getting our customers
to understand how much we just tested them for. So for instance, if I just tested you on 50
different carrier statuses — so including cystic fibrosis and a whole host of others. How do I get you to appreciate the fact that
we actually just screened you for all of this and that we did not find any variance there? So one is like getting you to realize there’s
all this information. One of the things that we have done is we
include this whole section on traits. Is that — traits are actually really fun
for people. So if I can tell you what your likely eye
color is; or I can do things like caffeine metabolism. Every one of our customers gets something. And they love that. Like, they love thinking about lactose intolerance
or caffeine metabolism — Eric Green:
Everybody loves thinking about lactose intolerance. Anne Wojcicki:
— ancestry. [laughter] Anne Wojcicki:
Well, but it’s actually interesting. So like I look at things like my child, who,
you know, was having stomach pain and then I looked into the genome. I was like, oh, you’re genetically likely
to be lactose intolerant; like I’ll just try out that lactose-free milk. It’s a relatively — it’s cheaper than
going to the doctor. So like — and it seemed to work. [laughs] So I think those are things like — it just
— it kind of just fits in with your life. And I think that’s part of it — is I think
part of the reason for the success of 23andMe is that we can give everyone some kind of
piece of information. And I think that the research — I think one
of the things we have found is that people — you know, you look at Susan G. Komen and
Livestrong. And there’s this, like, goodwill sense where
people want to contribute to the world. But we kind of make it hard. And in part, one of the insulting factors
that I found on Wall Street is how we treat people like a human subject. And again, you have the Henrietta Lacks up
here. Like, it’s kind of insulting how you are
in a clinical trial, we get as much from you as we can, and then you are deemed too complicated
to ever return any results for. There’s not a single federally-funded study
that returns the data back to the — the genetic data back to the subject. And I find it kind of insulting even just
to be treated as a subject. So part of what 23andMe has tried to do is
humanize it. And we do all kinds of research studies. Like, we did a big study on — we’ve done
now the largest study on human sexuality. And no one else — like, it’s going to be
tough for me to get an NIH grant for that. But it’s one of the things –consumers actually
wanted that. And if you don’t want to participate, you
just don’t take the survey. And if you want to participate, you take the
survey. And you look at the success of this ALS ice
bucket challenge. And you imagine, like, if you had a family
member with Parkinson’s disease and you could email out to all your friends and say,
“Take this Parkinson’s survey, you’re going to contribute to real research.” That’s like — then you can really make
a difference. And I think that’s part of what we’ve
done. So the customers like — they have their information. That’s their — like, their information
of their genetic data, what they should do, and then these surveys about, like, what is
it — like, how can I answer the questions? What are the genetic associations with type-2
diabetes? Why is it that it says I have this genetic
risk factor, but I don’t have a family history? How can we actually understand that? And that’s the responsibility of 23andMe
to keep innovating on that and make sure it’s clear to customers that we don’t know everything. There’s this fascinating world of gene and
environment. And the more environmental information we
can learn, the more we can understand of how your genes interact with your environment;
and then we should be able to give you better information about actually how to manage your
health. Eric Green:
So this is all about what truth is going to be, as I said earlier. What we’re really going to learn through
whatever means, whatever studies of the interaction of genomic variation, and diseases, and traits,
and the environment, and so forth. And I want to talk a little bit about implementation
— the real world. And I’m going to start with Rick. As a practicing physician, geneticist, and
someone who deals — in some cases with rare diseases, but much of your career has also
been looking very complicated common diseases such as hypertension. But you encounter patients and you don’t
have hours with them, but — you know, project to 2020. You’re seeing patients in the clinic. When they come in with their particular circumstance,
whether it be hypertension or whether it be some other renal disorder, and they come with
a lot of genomic information. What’s that encounter going to be like? How certain are you going to be to be able
to give an answer? And how certain are you going to be that they’re
going to understand what you’re about to tell them? Because it’s not going to be black and white. It’s going to be some shades of grey. Richard Lifton:
Yeah. I think in the long run both physicians and
patients are enormously practical. If things make a difference to patients’
individual health, patients and physicians will want the information. And I think BRCA1 and 2 are perfect examples
of that. When the tests became available, there was
a paternalistic strain in the community that said, “Well, we’re not ready for this
testing. We need time to figure everything out.” And patients and physicians in the community
said, “Wait a minute. My patient — or as a patient, I’ve got
a family history of early onset breast and/or ovarian cancer. I believe I’m at risk. I want this information now. And because I’m going to do things with
it.” And physicians, I think, were pretty quick
to pick that up. And a few more daring ones said, “Yeah. Of course we’re going to start” — Eric Green:
It was pretty black and white. Right? Those examples are pretty black and white. Richard Lifton:
So I think the — so I think the black and white examples are the ones that in the long
run — and it’s the getting from where we are now to the long run. Because in the long run I think it’s going
to distill down to a manageable number of things that matter. And physicians and patients will coalesce
around — here are things that really matter for your health in the long run, and these
are the things that we ought to be measuring in everybody. We have a rather shaky period going from where
we are today to that point where there is going to be a lot of uncertainty as to what
individual variants mean in individual patients. And this, of course, comes right back to Eric
because his job is to get as many people sequenced for as many diseases as rapidly as possible
to settle which variants actually matter so that Anne can put on her diagnostic list;
here are the things that we’re testing now because these are the things that we’re
certain really will matter to patients. And we’ll be able to hopefully use this
to improve public health. But I think we will have a very — a period
in the near term where we will have a lot of variants of uncertain significance where
— Eric Green:
Which is why I’m picking on 2020. Because I think, you know, I share a lot of
the optimism that will eventually understand a lot of this. But it will be an awkward phase. So in 2020 though, think about in our medical
colleagues, and our pharmacist colleagues, and our nursing colleagues, and genetic counselor
colleagues — you know, they’re either in training now or they’re out there in practice. And it’s just this tsunami of uncertainty. I think the BRCA1 is an easy example, but
I purposely pointed to hypertension, — Richard Lifton:
Yeah. Eric Green:
— diabetes. I know it’s going to be grey for a while,
and yet, some of it will start percolating in. And what is that going to look like — and
both from the patient point of view and for this busy health care professional point of
view? Richard Lifton:
Yeah. I think it’s going to be quite taxing in
many instances, and I think we need to be prepared for uncertainty. We’re going to be dealing with a fair degree
of uncertainty in many cases about what the significance. And for this reason, this is where I come
back to my earlier comment. I’m most enthusiastic about using this technology
for people who either — we have good reason to think they are at risk or they are presenting
with a particular disease. And the element that you and I have discussed
previously is — we’ve got a long list of genes to figure out what they do and what
they mean in the context of humans. And more rapidly we populate that space, the
better opportunity we’re going to have to understand when a patient comes to us, what
disease they actually have, if it is genetic; and if it’s genetic, which gene is mutated,
and how that’s driving disease. Another element that we haven’t touched
on yet is the potential impact of all the information coming from electronic medical
records and the ability to do very large amalgamations of that data with genomic data. And this again poses both opportunity as well
as enormous challenge for trying to make all of this coalesce onto — into new knowledge
pathways that will benefit public health. Eric Green:
As you know, we’re very interested in that, and research going on actively to try to see
what that future is going to bring. So Anne, I think you alluded to this. I mean, are the great discoveries the next
five years going to be in the United States? Or are we at risk of losing our lead on this? Anne Wojcicki:
See, I definitely — there’s two things. So one, I’ll answer that and then I wanted
to go back to your last question. Eric Green:
Sure. Anne Wojcicki:
So I do — I do really fear that the U.S. is going to be falling behind because there
are major initiatives going in a lot of other countries. So the U.K. has their 100,000 Genomes Project. Like I said, Beijing Genome Institute has
just a massive — it’s the world’s largest sequencing shop. You look at other countries like the Netherlands
where they actually have some of these massive, half-a-million person studies where the medical
records already are integrated. And I think that’s where we have this dream
and this fantasy of actually having all electronic medical records integrate. But I actually have — I mean, raise your
hand here if you’ve ever downloaded your medical record or if you’ve ever used — oh,
some of you. Anyone ever use Blue Button? Oh, we’ve been looking for you. [laughs] [laughter] Anne Wojcicki:
So I mean that’s part of it — is that trying to find people who are actually getting this
data is hard because medical records are not really integrated. And that’s part of the problem with not
having a single patient identifier in this country, and actually really being able to
get all this data. But there is massive potential for doing these
types of studies. And I look in the — Europe — I mean, the
U.K. they have, like, this million women study where it’s just like this one woman who
runs the study and they just have amazing amounts of data. So I do worry that at this time because we
don’t have a clear path for actually how we’re going to do these types of massive
research studies. And there has actually recently been a hold
on some of these research studies. Like, there’s BabySeq [spelled phonetically],
for instance. There’s this big sequencing program where
they want to, sort of, you know, understand the approval — the regulatory process before
they’re going to return all this data back to the individuals. So I think that right now, we’re really
kind of stuck because we don’t know the right way to deliver all this information
back. So the second thing I was going to say on
the variants of unknown significance — I do see that as actually one of the biggest
challenges. Is that you get tons of this data and you
don’t know what it means. And so this is actually a project that Rick
and I ironically ended up working on. Where somebody came to Rick where they had
three generations of pancreatic cancer. Rick did the sequencing. We found that there is a mutation in MLH1,
which is associated with hereditary colorectal cancer, and we thought this is the likely
mutation. So it was a variance of unknown significance
that had a high probability of being this mutation. And what 23andMe did is we took that mutation
and we looked back in our database, and we said, oh, we have a 157 other people with
that same mutation. If this is really the highly penetrant mutation,
those 157 individuals also should have had hereditary cancers. And so we ran a survey. We got 12,000 people to respond in 12 hours. And we could see almost instantaneously that
this mutation had — that those individuals with that mutation had nothing above baseline. And so we could conclude than with a high
degree of probability that this variance of unknown significance was likely not the cause
of the mutation for this hereditary pancreatic cancer. And so this is something that 23andMe is looking
to do more and more; is that it’s not our job to mine all the data. We want to be, sort of, representing the consumer. So we are looking at ways that we can actually
make our entire database accessible to people like Rick, to you, to all researchers in the
world; where they can run a query in the database and they can instantaneously see is this variance
of unknown significance. What is being found in other individuals like
this? And that’s, I think, part of how you, the
individuals, are actually going to be able to contribute to research like us decoding
the genome really fast. And if we have that ability in five years
— if we have massive numbers of people, we really could decode a massive amount of this
genome. Richard Lifton:
So I want to completely agree with this point. If we sequenced our individual genomes today,
the best annotation tool that we could possibly have for understanding what is interesting
or potentially actionable in it, is to know the sequence of — not just everybody else
in this room, but a million other people. And today the largest publicly available database
— you could put together maybe 20,000 people from available databases. And I know NHGRI is passionate about trying
to get this data available. And, of course, it raises many issues with
regulatory authorities. And I think it is one of these — will be
one of these individual responsibilities where we will need individuals to be willing to
make their data in some way accessible. Because we all will benefit to the extent
that everyone makes their data accessible to be able to correlate genotypes with phenotypes. I think that’s critical. Getting back to your question about a — you
know, how the U.S. is fairing. It — we’re in a somewhat ironic time, when
having spent enormous taxpayer dollars since the end of World War II for basic biomedical
research that has led to this point where we are today. That we are now in the throes of cutting back
on our investment — on our public investment in research just at the time when the fruits
are most likely to benefit the public. And I think this is a poor decision to be
making at this time. Eric Green:
And we need to reverse it. Anne Wojcicki:
[affirmative] Eric Green:
So changing topics slightly, one of the things to be careful about is making sure then the
process of using genomics in a productive way to improve how we can practice medicine,
we don’t leave behind certain elements of the population. So health disparities — is — think genomics
is going to exacerbate or reduce health disparities in America and elsewhere in the world? Anne Wojcicki:
[affirmative] Eric Green:
Not an easy topic, I realize. Anne Wojcicki:
It’s not an easy topic. I mean, it’s a couple things that we’ve
done. I mean, part of the whole mission for 23andMe
was the accessibility and the democratization. So at $99, there’s a massive difference
between us and the entire genetic testing field. So we’ve tried very hard to make this information
accessible for individuals. And we’re trying very hard to make it so
it’s understandable, that it’s easy for people, that it’s — you know, you just
go online and you order it. I do think that one of the differentiating
factors is going to be when you get your information you show up to your physician and if your
physician doesn’t know what to do with it, then it’s a challenge. So I think there’s two things. One, it’s part of the reason why 23andMe
has a community — is that people are learning a lot because genetics is still so new, they
need to have a community resource to ask questions. And then we find those community members are
pointing each other to other resources. And we also have a list of resources for our
customers. I think second, I think you’re going to
have to have things like tele-medicine where you’re going to have — you know, we have
a partnership with InformedDNA, a group of genetic counselors that are trained on all
this genetic information. So it’s, again, relatively inexpensive,
but then someone from their home can then go and get this information. But I think it’s a challenge. And it’s part of the reason why we’ve
put in — we’re starting to put in significant resources into the education. Because most of the medical community is not
educated about genetics. And I do believe it’s part of our responsibility
to make sure that we are at least supporting the physician as much as possible. And I was an investor in the days of WebMD
when it first came out. And it was hated. You know, it was just wreaking havoc on the
world. And so I’ve at least learned from that in
that I want to be able to enable customers to get a report on Factor V Leiden printed
out and actually have the basic information on it where they can walk to a physician,
and the physician feels capable and competent at actually answering those questions. So that’s our goal and that’s going to
be the direction that we’re going. Eric Green:
Rick? Richard Lifton:
I’ll take a somewhat different tack on the question. When I was a medical student, the first time
I walked into a dialysis unit was in Palo Alto, California and I was astounded to see
that most of the patients in the dialysis unit were of African American ancestry. Because Palo Alto at that time — African
Americans constituted a small fraction of the population, and yet they dominated the
dialysis population. And I asked the attending physician, what’s
the explanation for this? “Yeah. Socio-economics, access to health care — don’t
know.” Wave of the hands and that was the end of
the conversation. But it was something that was persistent and
stuck with me over the years. And just recently in 2011, a really remarkable
study came out that provided an explanation that this major health disparity is actually
genetic in origin. And it turns out that if you have one copy
of the particular variant in a gene called APOL1, and you live in Africa, and are exposed
to trypanosomes, you are protected from development of trypanosomiasis. And that’s a very beneficial thing to have. But if you have two copies of the allele,
you — of that variant in APOL1 — you are likely to end up on dialysis at age 50 or
60. This was a mystery that was completely unknown
and people dealt with — what’s the origin of this health disparity — for a very long
time until this genetic explanation came forward. So I think there’s a path forward there
to actually address one of these health disparities. And, of course, the challenge now is to understand
the pathogenesis, how this variant APOL1 translates to this predisposition to end-stage kidney
failure. But is a nice example of — where in genetic
discoveries have the opportunity to reduce a health care disparity in a rather dramatic
way. Anne Wojcicki:
So do people get screened for that now? Richard Lifton:
Well, so right now it’s in that grey zone — Eric Green:
Grey zone. Richard Lifton:
— of we don’t really know what to do. And it’s a very recent finding. But it clearly has a very large — unlike
many common variant studies, this has quite a large effect on risk of disease and clearly
is of importance. Eric Green:
I just looked down. One of the questions — Anne, I don’t know
if you want to extend any more, but specifically related to this — where they wanted to ask
you what do you envision 23andMe can contribute to discussions on health disparities for various
populations? Anne Wojcicki:
What can we –? Eric Green:
What can you contribute to discussions on health disparities? In other words, is any — are there — I would
imagine some of the studies you’re doing or whether you’re doing enough population
collections to be able to answer some of the questions like what Rick was giving examples
about. Anne Wojcicki:
Yeah. So — so one of things that we have found
is in — I mean, you guys certainly know, is that most genetic studies are done on European
populations. So I think it was three or four years ago
we actually launched an initiative called Roots Into the Future, where we gave away
— we partnered with Skip Gates, and we gave away 10,000 free kits to African Americans
because we wanted to see — could we do sort of a massive replication study? So could we find, you know, this, you know,
this type 2 diabetes or Factor V — like all these genetic variants that are found in European
populations, do they replicate in the African-American populations? So 23andMe, I think, has — recognizes that
genetic studies are not done in all populations. And again, I think it’s part of where we
— we tried to rectify it at least a little bit there by having this, sort of, major initiative. And when we gave away 10,000 free kits, we
found then that each family member then goes and they get other family members to start
— to sign up as well. So then it actually expanded it quite a bit. But that it’s important for us to make sure
that the genetic information is relevant and meaningful to all populations. And today it is decidedly not. So it’s one of the things that’s definitely
on our mind. It’s some of the things that we think about. We’ve put — you know, as an early non-profitable
start-up, we’ve put pretty substantial resources into this already. And we try to advocate for this. Like Southeast Asians as well are routinely
not part of these big genomic consortiums. So that’s again something that we’re very
aware of, and that we think actually needs to be remedied. So we are thinking about that and we try to
recruit potentially certain populations. But it’s a problem in the industry. Eric Green:
So one of the members of the audience pointed out that we’re talking a lot about prediction. But let’s also maybe explore a little bit
about therapeutics and what is now possible. And they gave us an example, which is one
that I know Rick — you probably hear about frequently. You know, the examples where we’ve known
the genomic base of a disease like sickle cell for many years, and yet we really haven’t
been able to come up with good therapeutic or improved therapeutic options. So what — when you look in the crystal ball,
you know, what do you sort of see? Is genomics going to be mostly a diagnostic
and predictive tool? Or is this going to lead to new therapeutics
including, obviously, gene therapy, drug development, and so forth? Richard Lifton:
Yeah. So key question. I think the starting point from my perspective
is always the biology is going to tell you what your options are. And sickle cell anemia is a good example of
how difficult it can be to go from understanding biology to developing a new therapeutic. We’ve known the molecular cause of sickle
cell anemia since 1953, and yet we still don’t have effective therapy for that. And part of this gets to the nature of what
the gene is, and what the specific mutation is, and what would be the path forward. So some of the examples that I gave in cancer,
where we’ve gone in stunningly short time from identifying a mutation that’s driving
a particular form of cancer to a new therapeutic, has occurred specifically because the nature
of the mutation suggested an immediate path to therapy. The genes in cancer that I illustrated were
types of mutations that cause increased activity of a particular pathway and suggested immediately
that we could turn off that particular pathway and have a beneficial effect on the development
of that cancer. Much more difficult are situations where you
have lost the function of a particular gene or protein, and then are trying to figure
out how to reactivate that gene or some downstream pathway. And that’s frequently much more vexing. Similarly, structural proteins are very difficult
to figure out how to deal with their replacement. So thinking about potential magic bullets
going forward, there — gene therapy is always, you know, it’s one of those areas that the
future is always just around the corner. And I think that we haven’t really cracked
that nut yet. For trying to correct genetic mutations there
are, I think, quite promising technologies that, you know — as you know, just over the
last year the development of general approaches for either knocking out gene function or potentially
replacing — correcting specific mutations with technology called CRISPR technology,
which came from basic biomedical research identifying fundamental pathway used in bacteria
to ward off invaders has really quickly caught on in the biomedical community and is being
widely used. And there are potential therapeutic approaches
using this that clearly are not likely to happen immediately, but are of the sort that
are clearly getting a lot of interest in the public. So I think the bottom line is we don’t know
what we’re dealing with until we understand the biology. And as I always tell my students and fellows,
the surest path to therapy is understanding the biology. It’s just not the case that every time you
understand the biology, it’s going to immediately suggest a therapeutic. Eric Green:
So Anne, one of the people in the audience is asking questions about privacy issues. I am quite sure you’ve thought about this. Sort of, what are some of the — do you think
we’re properly situated in the United States to deal with a lot of genomic information
on a lot of patients in their electronic health records? Do you think this is something that’s going
to come and backfire? Or do you think we need a better framework
for putting that future together? Anne Wojcicki:
I mean, I think in electronic medical records it’s — I wouldn’t say as much of an expert
in that. But I think that in general, genetic information
is sensitive information. And I think it’s — in some ways the consumer
just needs to be really informed of the fact that it’s impossible to de-identify you
from your genetic information. So — so it’s part of where there’s a
responsibility; you need to understand what it means when you share your genetic information
and what those risks are. One of the things 23andMe has tried to do
is — what we have found, is that most consumers — actually most of our customers at least
are comfortable with sharing their genetic information, but they want to be in charge
of that. So they want to know, am I sharing this with
my mother? Am I sharing this with a stranger? You potentially want to share different levels
of information. I might want to share more information with
my physician, less information with my sister, less information with random people on Ancestry. So I think it’s important that the privacy
controls — you know, when we spent early — in the early days when we met with all
the privacy experts, one of the things that they said is that, like, look, the definition
of privacy is giving individuals that choice of maintaining whatever level of privacy they
want. And so 23andMe — you can join us. There’s no legal chain of custody that we
have. If you’ve ordered a kit, I don’t necessarily
know that that kit actually was spat in by you. So we don’t’ have that direct chain of
custody. And we do everything we can to protect that
privacy of your genome. At the same time, we give you those controls
if you want to share it with individuals that you can. If you don’t want to share it, you can’t. But I think it’s the reality for this country
is that we need to make sure that people understand that there is always a risk with this information
and when you’re sharing it. And I would advocate that there is a reasonable
risk-benefit reward. You know, what Rick was saying is that we’re
all going to learn from each other if we can all get over this. But we all need to understand what those risks
are. And I think that there does — like, GINA
is in place here. I think — GINA is the Genetic Information
Non-Discrimination Act — so there is federal legislation now that’s going to protect
you from having your employer and your insurance company from discriminating against you. It could be made stronger. It could include life insurance. And I think what we’re going to wait to
see is some severe penalties. If there are breaches of privacy, there needs
to be sort of that severe penalties coming after those individuals. Richard Lifton:
I do think — just to follow onto that — I think we have placed enormous weight on theoretical
harms that are rarely met, and have placed almost zero weight on practical benefits that
will come from data sharing of the sort that Anne is alluding to that is really necessary
to make full utility of the information that’s available. Anne Wojcicki:
I mean, it’s one of the things that when we get asked these questions, I look at all
of you in the room and — could I hack into your genome or would I rather hack into your
bank account? [laughter] Anne Wojcicki:
And the reality is, like, I’m sure your email or your bank account is much more interesting. Eric Green:
And a lot of these are federal employees. Our bank accounts are really boring. [laughter] Anne Wojcicki:
You never know. [laughs] So — so it’s still that there’s
more than incentive. And there is a lot of really great sophisticated
technology that’s evolved that we’re able to copy; and we’re able to, sort of, understand
what is it that they do in the banking industry with your mail, other things, that we can
then emulate with 23andMe. And actually the team that originally built
the infrastructure for 23andMe was a group that had come from PayPal that were specifically
from that banking industry and brought that kind of security obsession. So it’s clearly important, and again, I
would agree with Rick. Like, it’s that risk-benefit. We worry so much about the theoretical. But there’s a massive benefit to society
if we actually all can pool our data together. And I think that’s also part of the reason
why we have 750,000 people pretty quickly that have all come together, and 80 percent
of them are consenting to research and sharing information. Eric Green:
So Rick, another question came in. This is a person who is clearly going to be
recruited to be one of your reviewers next time you’re up for a Hughes evaluation [spelled
phonetically], which I know is a very stressful evaluation that comes in. So they said hypertension, which you’ve
spent much of your life studying, is a challenging example. But we can easily and cheaply see if a person
is suffering from it and there are large numbers of effective drugs. So why do we need genomics at all to study
hypertension? Richard Lifton:
Terrific. [laughter] Richard Lifton:
So when I started working on hypertension, I chose to work on it specifically because
it was one of the common diseases that, as I said, affects a billion people. And when we started our research on it, it
was passionately debated as to whether hypertension was a primary disease of the heart, brain,
kidney, adrenal gland, or vasculature. And as a consequence of that inexact understanding,
two-thirds of patients with hypertension are inadequately treated and we continue to have
17 and a half million deaths annually around the world from cardiovascular disease –remains
the leading cause of death worldwide. The rare outliers that we studied with extremely
high and extremely low blood pressure settled the question that probably only the salt lobby
really wanted not to be answered. [laughter] Richard Lifton:
Which is the kidney is a central regulator of long-term blood pressure by determining
how much salt is reabsorbed by the kidney. And the impact of this is it is immediately
suggested how you might want to try to attenuate the age-dependent rise in blood pressure that
occurs in western societies by modest reduction of salt intake. And as I mentioned, 32 countries now — as
an accumulation of knowledge implicating renal salt reabsorption — now have modest reductions
in salt intake as a uniform goal for the entire population, which is modeled to — and in
the U.K. there’s evidence — has reduced cardiovascular morbidity and mortality. Eric Green:
So if I combine a couple of questions related to education, I want to ask each of you — do
you think the bigger challenge is going to be in educating and preparing the health care
professionals for this future? Or is the bigger challenge going to be preparing
the patients and the general public for genomics and its use in medical care? Or is it going to be a tie? Anne, I’ll let you go first. Anne Wojcicki:
I think it’s going to be the physicians. Eric Green:
Are going to be harder? Anne Wojcicki:
Yeah. I know, I think the physicians are harder. And I had — Eric Green:
And other health care professionals — Anne Wojcicki:
And other health care professionals. And I think part of it is that, you know,
we had — at one of our early conferences we had a physician stand up and say, “Look,
the biggest problem with 23andMe is that you generate non-billable questions.” And I think that that’s actually true. It’s because there’s not, you know — we’ve
seen doctors make lots of changes when there’s a massive reimbursement incentive. So Lasik was one example. You know, when everyone was getting their
eyes cut, people made a ton of money off that so everyone learned that new technology. There’s not a lot of money to be made off
genetics. So it’s hard to justify with such a busy
world that you already have. Why is it that you’re going to learn all
this information? And I think the physicians have been generally,
you know — it just hasn’t been taught. It’s not really taught that much in medical
school, or it’s taught just about things like cystic fibrosis. So this — it’s a whole new class of information
to learn. And what we find is that no one cares, you
know, from the consumer perspective — no one cares more about your health than you. And if your genetic information is all about
you — and you’re worried about how you can stay as healthy as possible, and you see
your family has a history of one thing or a history of other, you’re really motivated
to learn as much as you can about you. And that’s — that’s where I think what
we’re seeing is. Like, our customers are really — get educated. You know, we have these videos that — called
Genetics 101. It’s a YouTube series. And we have over a million views on these
videos collectively. So people are — people want to understand
this. And so I do think that you have more of an
incentive because it’s about you; versus the physician, where it’s part of your workload
and there’s not a reimbursement structure to really support that right now. Eric Green:
So Rick, are all the Yale first-year medical students watching these videos and getting
educated about — [laughter] Eric Green:
No? Seriously — Anne Wojcicki:
[laughs] I can trace them back. Eric Green:
I’m curious what you think about what’s going to be — where is the bigger challenge
going to be: the public or health care professionals? Richard Lifton:
I think they evolve just about in parallel in my experience. And I go back to BRCA1 where I think patients
and physicians — patients were a little bit ahead because of exactly Anne’s point — that
it’s me, it’s my family, and I really want to know. But physicians were pretty quick to pick up
that this was practical and important and made a difference. And surely, physicians currently in medical
care don’t have huge amounts of time to spend with each individual patient. And so they’re going to be focusing on main
effects. Right? What are the things that actually are going
to matter most to long-term health of patients? And I think to the extent that things matter,
people will be paying appropriate attention to that. Certainly, in the first-year Yale medical
school curriculum, we teach a lot of genetics now. Genetics is — you know, there is no better
framework for understanding human biology today than focusing on the diseases of every
organ system from what happens when you have a specific mutation. And you take one, you know, essentially one
piece out of this working organ, what happens? It’s a wonderful way to organize a curriculum. Eric Green:
And how much has that changed in the last decade — that curriculum? Richard Lifton:
So I think in medical schools around the country, it’s quite common now to have aspects of
genetics and genomics tie into almost every organ system; because it provides such a special
way of understanding the biology of what happens when you mutate this gene and why does, you
know, all the secondary consequences evolve from this primary abnormality. Eric Green:
Okay. We’re winding down and we want to end at
an appropriate time to have a break before the next panel. So I’m going to ask each of you one more
question. The question is, when thinking about everything
we’ve discussed, what keeps you up at night? And Anne, your answer can’t be the FDA. [laughter] Eric Green:
You have to come up with something other than the FDA. [laughter] Eric Green:
So we’ll let Rick start and then we’ll come to you. Anne Wojcicki:
Yeah, I have to think about it. [laughs] Richard Lifton:
Well, so I think we’ve already touched on the part that keeps me up at night, which
is how do we get from here to there? You know, we can see over the event horizon
to see to — once we’ve got a million people sequenced, maybe more than that million genomes,
we’re going to have ideas of what variants are highly — most strongly predictive of
particular diseases or their outcomes. And trying to get from where we are today
to that point I think is going to be — have a lot of challenges because a lot of genomes
are going to be sequenced. And just as Anne articulated, people are going
to show up and say, I’ve got this variant. What does it mean? And right now we’re not going to know with
nearly the precision that we want in many cases what those variants actually mean. And I’ll add the FDA part for [laughs] — for
Anne so she doesn’t have to answer that. I think there’s huge opportunity for innovation
in health care. And I do have serious concerns that if we
over regulate — I think the FDA has a completely appropriate role in making sure that tests
are being done well and appropriately and the information is accurate. But I do have reservations that — about if
we go down a path where we insist that these are the only tests that people can offer at
any one time, that that will really stifle development and innovation in this area. Eric Green:
Anne. Although now I realize your answer’s going
to be your young children keep you up at night. [laughter] Anne Wojcicki:
My young children do keep me up — they wake me up early rather. I mean, there’s things that I lose the most
sleep about is that the most rewarding part of 23andMe is — I always wanted to be a doctor
and help people. And I get an email almost on a weekly basis
of somebody who says, “Like, oh, you know, you saved my life because I learned this information.” Or, you know, there’s a man who came up
to me recently who said, “You know, my son hated sugar. I couldn’t figure out why. We took him to all these specialists. And then my sister was pregnant and she did
23andMe. She found out she’s a carrier for fructose
intolerance. Lo and behold, I go to another five physicians. I get my son tested and he’s homozygous
for fructose intolerance. He can’t absorb fructose and that’s’
why he hates sugar.” It’s normally only detected in their twenties,
once they actually have severe disease. And so this little boy, like, we’ve just
impacted his life. We prevented him from actually having this
disease. And, you know, information, you know, significant
percent of the 23andMe customers really learn something that influences their life. And it could be that, you know, it could be
on their ancestry. It could be that they find a relative. It could be that they, you know, actually
have Jewish ancestry and they didn’t know about it, or they didn’t know whether they
were adopted. Or it could be on the medical side. And so it keeps me up at night that people
can’t get this information right now. And I think more than that, I’m impatient
like I am — I’m 41, which is young, but it’s also sort of old — Eric Green:
It’s actually disgusting, but that’s okay. [laughter] Anne Wojcicki:
[laughs] But I really — like I feel this, you know, this pressure. And I was looking back for this talk — I
was actually looking back at where were cellphones 10 years ago — just to look at that delta. And, you know, it was like the old flip phone. Like that was really hot, like having one
of those StarTACs — Eric Green:
I just got rid of my flip phone like six months ago. Anne Wojcicki:
[laughs] I know. Don’t admit that. Eric Green:
[laughs] Anne Wojcicki:
So — and then I compare that to the iPhone 6, which just came out. And then I think about, well, we want that
same kind of evolution to happen in genetics. And so how are we going to get there? And at this current pace of funding individual
little studies, it is not going to happen, frankly. And so we need a massive community of individuals
to help us decipher what this genome means. And I am definitely the big believer. I mean, I spend every single day doing this
because I believe that if we can have the world’s information combined with genetic
data, you really will be — have a much better path for making drug discovery, you’ll have
much better diagnostics, you’ll have just a much better way of living your life. So how can we get there faster? And I don’t want 23andMe being the only
one sitting on this data. I want to enable every researcher in the world
to get access to our data. So when I think about what keeps me at night,
like, I want to empower this revolution. It’s so — like it is so exciting right
now. Like it’s the most exciting time in science. And I just — like, we’re on the cusp. We’re just slowly getting there. And I want to see it happen faster because
I’m impatient. Eric Green:
So — [applause] Anne Wojcicki:
Thanks. [laughs] Eric Green:
So I would say, in case anybody’s interested, what would keep me up at night — actually
in some ways is exactly what you were just saying but brings a different dimension. Is that — this incredible, enthusiastic opportunity
— but also knowing that not only we have to do a lot of things right to make this happen,
here in the U.S. in particular. We also have to convince the next generation
that being part of this revolution, at many different levels whether it’s as a health
care professional, as a researcher, whether it’s a data science expert, we need them
to come into this and embrace this as part of their professional career. And I’m just — what I lose sleep about
is then seeing them being discouraged because of a decaying support for biomedical research
in the United States, and seeing other opportunities that seem much more exciting to them. I really worry about the workforce and the
next generation, because I think without the — well, some of us on this panel aren’t
as young as you, and the next generation are going to be needed to really see us through
some of the things we talked about today. Anne Wojcicki:
Yeah. Eric Green:
So we need to change the course and make sure they get enthusiastic. And some of your infectious enthusiasm needs
to be part of what they bring into their profession, so. So we will — please thank Anne and Rick for
a wonderful conversation. [applause] Eric Green:
And thank the audience for some terrific questions. Please — you get a 15 minute break now. Do not wander far because the next panel’s
going to be just as good as this one. I promise. Thank you very much. [end of transcript]


  1. On St Elsewhere and Chicago Hope tackle this subject of monetary system trumps medical considerations. Those programs were 20 sum years ago! Nothing has changed.

  2. Let us get the FDA out of most health care such as companies that help with DNA.
    In the case of 23andME FDA overstepped their authority,

  3. The panel have not talked about the side-effects of pharmicuticals on the genes. Also,The is a great site that can combine genomics and medical records, A person can download different DNA results from say and 23andME.

  4. This is so dumb, why don't they release medical info for internationals. That way it would be easy to set up an outside service that you send your dna to and have them collect the data and send it to you ………….

  5. The cost of sequencing a person's genome is now $1000 and dropping fast. 23andMe's less comprehensive service is $100. In 5 or 10 years, genome sequencing will surely be standard medical procedure.

Leave a Reply

(*) Required, Your email will not be published