Latest / Statecraft / How UK Biobank Was Built
Transcript
- 0:04SPEAKER_00: Hi,
- 0:05SPEAKER_00: I'm Santi Ruiz,
- 0:06SPEAKER_00: and this is Statecraft,
- 0:08SPEAKER_00: an interview series about how policymakers get things done.
- 0:12There are a lot of forces in policymaking in our lives
- 0:15generally that push us
- 0:16SPEAKER_00: towards the short term.
- 0:18Quarterly returns,
- 0:19the president's daily briefing,
- 0:21two-year election cycles.
- 0:23It's very hard for most politicians and most policymakers to
- 0:27break outside of that cycle.
- 0:29The pressures to get results and to show results on a tight
- 0:32turnaround in more and
- 0:33SPEAKER_00: more domains are incredible.
- 0:36SPEAKER_00: One of my interests on Statecraft for a while has been,
- 0:39SPEAKER_00: how do you build a machine to get long-term results?
- 0:42Whether it's a new agency or a new initiative or just somebody
- 0:47thinking long-term
- 0:48SPEAKER_00: in their own individual career.
- 0:50How do you set up a structure so that you can actually work
- 0:53with a goal that's 10
- 0:54years away or 20 years away or 50 years away and protect that
- 0:59thing from the
- 1:00SPEAKER_00: short-term political pressures that almost everybody's under?
- 1:03SPEAKER_00: Today's interviewee is Sir Rory Collins.
- 1:06He spent two decades building and leading one of the most
- 1:09important scientific
- 1:10SPEAKER_00: resources in the world,
- 1:11SPEAKER_00: the UK Biobank.
- 1:13SPEAKER_00: The Biobank represents a fascinating case study in long-term thinking.
- 1:17It's a database of half a million British participants whose
- 1:20health is being
- 1:21SPEAKER_00: tracked longitudinally,
- 1:23SPEAKER_00: that is,
- 1:23SPEAKER_00: for the next 30 years.
- 1:25The Biobank was established in the knowledge that the upfront
- 1:28work and the spending
- 1:29SPEAKER_00: required would only start to pay off 10 or 15 years later.
- 1:33SPEAKER_00: When Sorori went in for the 10-year review with funders,
- 1:36SPEAKER_00: they asked what had been achieved so far.
- 1:38SPEAKER_00: Sorori said, nothing yet.
- 1:41It's democratized access to population-scaled data for
- 1:43researchers worldwide,
- 1:45and it's already yielding amazing insights into the causes of
- 1:48and cures for a
- 1:49SPEAKER_00: disease.
- 1:50Sorori had this vision in the early 2000s of an institution
- 1:53that would pay off way
- 1:54SPEAKER_00: past the political sell-by date.
- 1:56SPEAKER_00: I wanted to understand how he built the UK Biobank and,
- 2:00SPEAKER_00: just as importantly,
- 2:01SPEAKER_00: how he managed to sustain it over a long period of time.
- 2:04SPEAKER_00: As a reminder,
- 2:05SPEAKER_00: you can find the transcript for this conversation and for many others at
- 2:08SPEAKER_00: www.statecraft.pub.
- 2:11SPEAKER_00: That's statecraft.pub.
- 2:13SPEAKER_00: If you like this episode,
- 2:15SPEAKER_00: or even if you don't like it,
- 2:16SPEAKER_00: leave us a review wherever you're listening to this podcast.
- 2:19SPEAKER_00: One final note,
- 2:20SPEAKER_00: I refer to Sir Rory Collins several times as Sir Collins,
- 2:24SPEAKER_00: but I've been informed that as a knight,
- 2:26SPEAKER_00: he's actually Sir Rory.
- 2:28SPEAKER_00: Whereas if he were a lord, he would be Lord Collins.
- 2:31SPEAKER_00: I regret the error.
- 2:33SPEAKER_00: Without further ado, Sir Rory Collins, welcome to the show.
- 2:37SPEAKER_00: Thank you very much for having me on.
- 2:39SPEAKER_00: I'm very excited to have this conversation.
- 2:42SPEAKER_00: I wanted to start with just some basics for those of us,
- 2:46SPEAKER_00: especially on the American side,
- 2:47SPEAKER_00: who may not be familiar with the biobank.
- 2:49SPEAKER_00: And I was wondering if you could put some numbers on it for me.
- 2:51SPEAKER_00: So there's 500,000 people in the biobank.
- 2:55SPEAKER_00: What has happened to them?
- 2:56SPEAKER_01: So UK Biobank was set up at the beginning of the century.
- 3:00SPEAKER_01: Initially came out of an idea from the Wellcome Trust charity,
- 3:05SPEAKER_01: which funds a lot of medical research.
- 3:07SPEAKER_01: And the Medical Research Council are UK equivalent,
- 3:10SPEAKER_01: essentially,
- 3:10SPEAKER_01: of the National Institute of Health to government funding.
- 3:13SPEAKER_01: Really,
- 3:14SPEAKER_01: the concept,
- 3:16SPEAKER_01: which I think came from a lot of epidemiologists,
- 3:18SPEAKER_01: geneticists,
- 3:19SPEAKER_01: health researchers,
- 3:20SPEAKER_01: was to set up a large,
- 3:23SPEAKER_01: what we call a prospective cohort.
- 3:25So we recruited half a million men and women who were aged 40
- 3:30to 69 between 2006 and 2010.
- 3:32SPEAKER_01: So they're from all across the United Kingdom,
- 3:37SPEAKER_01: from different socioeconomic status,
- 3:40SPEAKER_01: urban,
- 3:40SPEAKER_01: rural individuals who came along
- 3:44SPEAKER_01: Answered lots of questions about their lifestyle,
- 3:47SPEAKER_01: their environment,
- 3:48SPEAKER_01: their family,
- 3:49SPEAKER_01: and their own medical history.
- 3:51Agree to providers with biological samples and to make physical
- 3:54measurements.
- 3:55SPEAKER_01: And importantly,
- 3:56SPEAKER_01: to follow their health,
- 3:57SPEAKER_01: which we've been doing over the last 20 years,
- 4:00largely through linkage to the National Health Service health
- 4:04record systems,
- 4:06but also going back to participants themselves and asking them
- 4:09about different
- 4:10SPEAKER_01: aspects of their health.
- 4:11SPEAKER_00: And talk to me about the scientific value of the biobank.
- 4:16SPEAKER_00: How many studies, what have we learned so far?
- 4:19SPEAKER_00: What are we hoping to learn in the future?
- 4:21The beauty of studying people in middle age when they're
- 4:25relatively healthy is you
- 4:27SPEAKER_01: can find out about their exposures,
- 4:30SPEAKER_01: not just their genetic makeup,
- 4:32SPEAKER_01: but also the way in which they're living.
- 4:35SPEAKER_01: You can measure other things in their blood,
- 4:38SPEAKER_01: all of the things that predate the development of disease,
- 4:41SPEAKER_01: because obviously getting a disease can change your lifestyle.
- 4:44SPEAKER_01: It can change other measures in your blood.
- 4:47The big advantage of having a study even then and then
- 4:51following the health is we
- 4:53SPEAKER_01: can understand what are the causal determinants of disease.
- 4:57SPEAKER_01: So not only what are the genetic,
- 4:59SPEAKER_01: but also the lifestyle environment,
- 5:01SPEAKER_01: the proteins in their blood,
- 5:03the metabolites in their blood that subsequently lead to them
- 5:06developing some
- 5:07SPEAKER_01: particular condition.
- 5:09And the important thing that the Medical Research Council and
- 5:12the Wellcome Trust
- 5:13SPEAKER_01: required of us when we set it up
- 5:15was to make these data available to researchers around the
- 5:18world for any kind of
- 5:20SPEAKER_01: health-related research in the public interest,
- 5:22and that's the consent that these half-million altruistic
- 5:25participants have agreed
- 5:26SPEAKER_01: to.
- 5:26So the data are being used by scientists everywhere to try to
- 5:31understand why it is
- 5:32SPEAKER_01: one person that gets a particular disease and another doesn't,
- 5:35SPEAKER_01: Is it their genetics?
- 5:36SPEAKER_01: Is it their lifestyle?
- 5:38SPEAKER_01: And what are the pathways through which genes,
- 5:41SPEAKER_01: environment and lifestyle might lead to getting a disease?
- 5:44SPEAKER_01: So looking at proteins and metabolites.
- 5:47So there are thousands of researchers around the world working
- 5:50on the data.
- 5:51Last year alone,
- 5:52there were 5,000 peer-reviewed publications based on UK
- 5:56Biobank.
- 5:56SPEAKER_01: So it's unprecedented in terms of the scale of discovery and the range of
- 6:01SPEAKER_01: discoveries that are emerging.
- 6:03from UK Biobank,
- 6:04all down to the altruism of those half million participants.
- 6:08It's an interesting question as to exactly why there was a
- 6:12decision to set up UK
- 6:13SPEAKER_01: Biobank when it was made.
- 6:16SPEAKER_01: Now it looks prescient.
- 6:17SPEAKER_01: At the time, it was really a kind of risky decision.
- 6:22So if you're interested in genetics,
- 6:24the great thing about genetics is your genes don't change.
- 6:27SPEAKER_01: So in some respects,
- 6:28the most efficient way of studying the genetic determinants of
- 6:31disease is take
- 6:32SPEAKER_01: people with a particular disease and compare them with those who don't,
- 6:37SPEAKER_01: because the disease itself won't have changed your genetics.
- 6:41SPEAKER_01: And therefore,
- 6:42you can look to see what are the differences in the genetics of
- 6:45someone who has a
- 6:46SPEAKER_01: disease as compared to those who don't.
- 6:48SPEAKER_01: And that's,
- 6:48SPEAKER_01: of course,
- 6:49how we've learned so much about genetic mutations that are
- 6:52really strongly causal
- 6:55SPEAKER_01: of disease.
- 6:56SPEAKER_01: From an epidemiological perspective,
- 6:57SPEAKER_01: of course,
- 6:58it's much harder because the lifestyle or environment is
- 7:02influenced by disease.
- 7:03SPEAKER_01: So you want to study people when they're healthy,
- 7:06SPEAKER_01: find out how they're living when they're healthy,
- 7:08and then follow them for a long period of time to understand
- 7:12what it was that led
- 7:13SPEAKER_01: to the disease.
- 7:14SPEAKER_01: So, for example, you study people's smoking habits.
- 7:18SPEAKER_01: and then follow them and you find that smoking causes lung cancer,
- 7:21SPEAKER_01: many other cancers,
- 7:22SPEAKER_01: cardiovascular disease,
- 7:23SPEAKER_01: respiratory disease,
- 7:23SPEAKER_01: et cetera.
- 7:24SPEAKER_01: Of course,
- 7:25one of the best ways to get people to stop smoking is for them
- 7:28to have a hashtag,
- 7:29SPEAKER_01: then they stop smoking.
- 7:30SPEAKER_01: So they change their environmental risk factor.
- 7:35And there was a lot of controversy at the beginning of the
- 7:37century as to whether a
- 7:39SPEAKER_01: study like UK Biobank was the right thing to do,
- 7:42SPEAKER_01: if you were interested in genetics alone.
- 7:45SPEAKER_01: But we're not.
- 7:46SPEAKER_01: We're interested in how genes and environment and lifestyle interact
- 7:50SPEAKER_01: And for that, you need to set up a large prospective study.
- 7:54Large because you need enough people to develop any particular
- 7:58condition in order
- 8:00to be able to have the statistical power to determine what are
- 8:04the causal factors.
- 8:05And prospective with long-term follow-up so that you're
- 8:08actually studying the
- 8:10SPEAKER_01: causal associations.
- 8:11You can exclude events that occur quite soon where they may
- 8:16have influenced your
- 8:17SPEAKER_01: non-genetic risk factors.
- 8:19SPEAKER_01: It required a vision of the importance,
- 8:23SPEAKER_01: not just of genetics,
- 8:25but of all of the different drivers of disease for the Medical
- 8:29Research Council and
- 8:29the Wellcome Trust to decide they were going to set this up
- 8:33with a recognition that
- 8:34SPEAKER_01: for the first 10 or even 15 years,
- 8:37it was unlikely that it would actually produce anything of
- 8:41material value in terms
- 8:43SPEAKER_01: of discoveries.
- 8:44SPEAKER_01: Now, of course, it's a goldmine.
- 8:47SPEAKER_00: What made you think this was even a feasible project to undertake?
- 8:51SPEAKER_01: Well,
- 8:51there are a number of kind of epidemiologists who have been
- 8:54involved in setting up
- 8:55SPEAKER_01: large prospective studies.
- 8:58The kind of classic in Britain was a so-called British doctor's
- 9:01study that studied
- 9:02SPEAKER_01: doctors and their smoking habits and has followed them for a century.
- 9:07And so the ability to recruit large numbers of people is
- 9:11certainly there.
- 9:13SPEAKER_01: But when one looked at the studies that had been done,
- 9:17SPEAKER_01: they had largely looked at risk factors that were old risk factors.
- 9:21SPEAKER_01: So blood pressure,
- 9:23SPEAKER_01: body mass index,
- 9:24SPEAKER_01: so obesity,
- 9:26SPEAKER_01: cholesterol levels,
- 9:27SPEAKER_01: where cholesterol had been measured and the samples hadn't been scored.
- 9:31SPEAKER_01: A number of people around the world
- 9:33had seen the value in establishing a large prospectus with
- 9:37biological samples
- 9:38stored and then just waiting until it was possible to assay
- 9:43those samples at that
- 9:45SPEAKER_01: very large scale,
- 9:46SPEAKER_01: you know,
- 9:46hundreds of thousands of individuals with markers that hadn't
- 9:50been studied so far.
- 9:52SPEAKER_01: So we knew how to recruit large numbers of people.
- 9:55SPEAKER_01: In the UK,
- 9:56SPEAKER_01: we had been running really quite live,
- 9:58SPEAKER_01: randomised trials,
- 9:59working with the National Health Service to recruit directly
- 10:03into studies.
- 10:04SPEAKER_01: So it seemed to us that one could take that approach.
- 10:09of working with the Natural Health Service to invite people to
- 10:12join the study.
- 10:14SPEAKER_01: And then when they agreed to take part,
- 10:17to really look at how one could build a kind of factory
- 10:21production line where they
- 10:23SPEAKER_01: would go through answering lots of questions.
- 10:26SPEAKER_01: And we used touchscreen technology,
- 10:28SPEAKER_01: which was relatively novel at that time,
- 10:30SPEAKER_01: to get lots of questions answered.
- 10:33and have people moving through this production line of
- 10:36answering questions,
- 10:37SPEAKER_01: having physical measurements,
- 10:38SPEAKER_01: collecting biological samples.
- 10:40SPEAKER_01: And if you invited enough people and you had enough funding,
- 10:44SPEAKER_01: then you really could recruit very large numbers,
- 10:47SPEAKER_01: which is what we did.
- 10:49When you say that epidemiologists had put together very large
- 10:53prospective trials in
- 10:54SPEAKER_00: the past,
- 10:55SPEAKER_00: what's very large in that context?
- 10:57What became clear was that there have been a number of
- 11:02observational studies,
- 11:03SPEAKER_01: so long-term follow-up studies of feedback of pressure,
- 11:07SPEAKER_01: but cholesterol,
- 11:08SPEAKER_01: body mass index.
- 11:10SPEAKER_01: Tens of thousands,
- 11:12SPEAKER_01: 100,000,
- 11:13SPEAKER_01: perhaps the biggest was the so-called Mr.
- 11:17SPEAKER_01: Fit screening study in the US of a third of a million American men.
- 11:22that showed this very lovely association between cholesterol
- 11:26level and the risk of
- 11:27SPEAKER_01: coronary artery disease that showed that throughout the range,
- 11:31SPEAKER_01: higher cholesterol was associated with higher risk.
- 11:34And there really was no normal level within the Western range
- 11:38whereby lower
- 11:39SPEAKER_01: cholesterol was not likely to be cardioprotective.
- 11:43SPEAKER_01: And so these very large prospective studies,
- 11:45SPEAKER_01: either individually or collectively through a meta-analysis,
- 11:49SPEAKER_01: a kind of combination of the results of different studies,
- 11:52SPEAKER_01: showed that if you had large numbers of individual,
- 11:55SPEAKER_01: hundreds of thousands of people in prospective studies,
- 11:57SPEAKER_01: either individual ones or combined,
- 11:59then you could get really very clear signals about the strength
- 12:04of the relationship
- 12:05SPEAKER_01: of certain risk factors,
- 12:06SPEAKER_01: the risk factors have been measured,
- 12:07SPEAKER_01: and disease.
- 12:09And so that really generated the hypothesis that if we could
- 12:12create studies of
- 12:13SPEAKER_01: hundreds of thousands of people,
- 12:15then we would be able to study many more risk factors if we
- 12:19stored the samples and
- 12:21SPEAKER_01: then just waited until we could analyze them at scale.
- 12:24As you and others and other advocates for UK Biobank began to
- 12:28shop the idea around
- 12:29SPEAKER_00: and build support,
- 12:31SPEAKER_00: did you run into trouble with
- 12:33SPEAKER_00: with the long-term nature of the project,
- 12:35that all the costs are upfront and all the benefits are out at
- 12:39least a decade?
- 12:40I know that a number of individual researchers put in grant
- 12:44applications which were
- 12:46SPEAKER_01: not supported.
- 12:47The International Cancer Research IARC in Lyon set up a study
- 12:53of the dietary terms
- 12:54SPEAKER_01: of cancer,
- 12:55SPEAKER_01: where they collected biological samples,
- 12:57SPEAKER_01: but it was really a study of studies.
- 12:59So in each country,
- 13:00they did a study and then IARC and Leon brought those together.
- 13:05They're all slightly different,
- 13:07but it did produce a large scale prospective study.
- 13:10But these were seen as studies that were of value for
- 13:14non-genetic risk factors.
- 13:17And when UK Biobank was being proposed by the Medical Research
- 13:21Council and Wellcome
- 13:22SPEAKER_01: Trust,
- 13:23I think that one of the drivers for it was the opportunity to
- 13:28study the genetic
- 13:30SPEAKER_01: determinants of disease.
- 13:31SPEAKER_01: And quite rightly,
- 13:32SPEAKER_01: people said,
- 13:33SPEAKER_01: well,
- 13:33SPEAKER_01: if you're interested in genetic determinants of disease,
- 13:35you can get quick answers relatively cost-effectively by just
- 13:39studying people with
- 13:40SPEAKER_01: disease and people without disease.
- 13:42And I think the messaging didn't really get across clearly
- 13:45enough that one wasn't
- 13:47SPEAKER_01: interested solely in genetics,
- 13:49SPEAKER_01: that one was interested in all of the different risk factors,
- 13:53SPEAKER_01: genetics,
- 13:53SPEAKER_01: environmental,
- 13:54SPEAKER_01: lifestyle,
- 13:55and that the beauty in the genetics along with the environment
- 14:00and lifestyle is
- 14:01that one can actually unpick the causal aspects of
- 14:07environmental and lifestyle risk
- 14:09SPEAKER_01: factors.
- 14:11SPEAKER_01: And so this so-called Mendelian randomization,
- 14:14SPEAKER_01: where you use genetic changes in,
- 14:17SPEAKER_01: say,
- 14:18body mass index to determine whether body mass index is
- 14:21actually causally related
- 14:24SPEAKER_01: to disease.
- 14:25SPEAKER_01: And so having all of these things in one study is incredibly powerful,
- 14:30SPEAKER_01: not only for determining the strength of an association,
- 14:33SPEAKER_01: but determining its causal nature.
- 14:36Will you tell us a little bit more about those initial years in
- 14:40the early 2000s
- 14:41SPEAKER_00: when you were building support politically and institutionally for it?
- 14:45SPEAKER_00: Were you going to Whitehall and convincing MPs?
- 14:48SPEAKER_00: What was the day-to-day as you tried to put together a coalition for it?
- 14:51SPEAKER_01: No, I wasn't seriously involved in it until really around 2005.
- 14:57The decision was made by the Medical Research Council of
- 15:00Wellcome Trust that they
- 15:01SPEAKER_01: wanted to fund a large prospective study.
- 15:05SPEAKER_01: So it really came from within those organizations.
- 15:08SPEAKER_01: It's quite difficult to understand exactly why those decisions were made.
- 15:14I think partly they were a result of epidemiologists having
- 15:19argued the case for the
- 15:21value of having large prospective studies that could study
- 15:25newer risk factors and
- 15:26SPEAKER_01: the value of having much bigger studies than we'd had before.
- 15:31I think part of it was the excitement about the ability to
- 15:35study genetics at scale.
- 15:37SPEAKER_01: And these kind of came together,
- 15:39SPEAKER_01: in my view,
- 15:39to help the funders decide that they wanted to set up a large
- 15:46prospective study
- 15:47SPEAKER_01: that would be able to look at the genetic determinants of risk,
- 15:51SPEAKER_01: but also to look at those in the context of lifestyle and environment.
- 15:55SPEAKER_01: And so they made an internal decision to fund this study.
- 15:59SPEAKER_01: They set up a small group led by Tom Mead,
- 16:03SPEAKER_01: an epidemiologist,
- 16:05to get input from scientists around the UK and beyond around
- 16:10what should that study
- 16:11SPEAKER_01: be in terms of its scale,
- 16:13looking at power calculations to decide what sort of
- 16:17differences in risk could be
- 16:18SPEAKER_01: detected with studies of 100,000,
- 16:19SPEAKER_01: 200,000.
- 16:19SPEAKER_01: They came up with half a million individuals.
- 16:24The age range was selected on the basis of young enough that
- 16:28you would be studying
- 16:29SPEAKER_01: people before disease was changing the risk factors,
- 16:32but old enough that the study would start to produce results
- 16:36within 10 or 15 years.
- 16:38SPEAKER_01: So there was always an expectation that this was a long-term commitment.
- 16:42SPEAKER_01: I think only organizations,
- 16:45institutional organizations like the Medical Research Council
- 16:48and Wellcome Trust
- 16:48SPEAKER_01: could make a decision whereby the first 10 years are really just waiting.
- 16:53I don't think they,
- 16:54or indeed any of us,
- 16:56thought it would become as productive as it has been.
- 16:59SPEAKER_01: And the reason it has, I think, is because of access.
- 17:04One of the things that the Wellcome Trust and the MRC decided
- 17:08at the very beginning
- 17:09was that this was not a resource that was being produced by
- 17:13researchers for
- 17:15SPEAKER_01: themselves alone.
- 17:17that this was going to be a resource that was going to be made
- 17:20available as widely
- 17:22SPEAKER_01: as possible in order that that investment
- 17:25both the financial investment and the altruistic investment of
- 17:29the participants,
- 17:30actually generated as much knowledge about how to prevent and
- 17:33treat as easy as
- 17:34SPEAKER_01: possible.
- 17:35SPEAKER_01: And that,
- 17:35SPEAKER_01: I think,
- 17:35SPEAKER_01: has been the transformative aspect of UK Biobank,
- 17:39SPEAKER_01: the kind of model of accessibility.
- 17:42When you say that you and others didn't expect that long-term
- 17:45value,
- 17:46why not?
- 17:46SPEAKER_00: And what changed that made you update?
- 17:49There was quite a lot of uncertainty about the quality of
- 17:53findings that would
- 17:54SPEAKER_01: emerge if you just made data accessible to researchers around the world.
- 17:58There was also uncertainty about what kind of capacity there
- 18:02would be to turn the
- 18:04SPEAKER_01: samples into data.
- 18:06SPEAKER_01: To give you a sense of that,
- 18:08SPEAKER_01: when it was decided to move forward,
- 18:09SPEAKER_01: it was recognized with half a million samples,
- 18:12SPEAKER_01: and the idea was to sub-aliquot those samples into multiple aliquots,
- 18:16SPEAKER_01: collect multiple tubes,
- 18:18then sub-aliquot them in order that you could retrieve
- 18:20particular samples from
- 18:22SPEAKER_01: particular individuals.
- 18:23SPEAKER_01: It was decided to build an automated archive,
- 18:26SPEAKER_01: which was,
- 18:27SPEAKER_01: again,
- 18:27SPEAKER_01: state-of-the-art,
- 18:28SPEAKER_01: had never been done,
- 18:29SPEAKER_01: to keep the samples at minus 80,
- 18:31SPEAKER_01: but be able to track them,
- 18:33SPEAKER_01: retrieve them when you wanted to.
- 18:35And that archive was built with a single robot because it was
- 18:39expected that one
- 18:40would pull out samples from say 10,000 women who developed
- 18:44breast cancer after 10
- 18:45years and 10,000 controls to compare the people who do and the
- 18:50people who don't
- 18:51SPEAKER_01: develop a particular cancer.
- 18:53So-called case control approach or a case cohort approach where
- 18:57you might have the
- 18:58SPEAKER_01: same controls for a number of different diseases.
- 19:02SPEAKER_01: If what you're interested in is studying a particular disease,
- 19:05SPEAKER_01: that,
- 19:05SPEAKER_01: again,
- 19:06is very cost-effective because instead of analyzing all half a
- 19:09million people,
- 19:10you measure the people with the disease you're interested in
- 19:13and some match
- 19:13SPEAKER_01: controls.
- 19:14SPEAKER_01: But after we had recruited all the participants,
- 19:17SPEAKER_01: and we were discussing this with our scientific advisory board,
- 19:20SPEAKER_01: they said,
- 19:20SPEAKER_01: well,
- 19:21if you're trying to build a resource for researchers around the
- 19:24world to use for
- 19:25SPEAKER_01: all kinds of different research,
- 19:27then a strategy where you assay the samples as if you're doing
- 19:30it for a single
- 19:31SPEAKER_01: researcher isn't actually the best approach.
- 19:35So the thing to do is to wait until you can actually assay all
- 19:39of the samples for
- 19:40SPEAKER_01: some particular type of assay.
- 19:43So genotype,
- 19:44the whole cohort,
- 19:45until you can do that,
- 19:47then leave the samples alone.
- 19:49SPEAKER_01: And again,
- 19:49SPEAKER_01: until people have developed disease,
- 19:52SPEAKER_01: so during the first 10 years when they're relatively healthy,
- 19:55SPEAKER_01: there is no value in assaying the samples.
- 19:58So I think one of the things that we kind of really had as a
- 20:00watchword was to defer.
- 20:03SPEAKER_01: don't do anything now that you could do better later.
- 20:06Rather than doing case control,
- 20:09better to wait until you can genotype the whole cohort.
- 20:12And that then turned a start study of a particular disease into
- 20:17a resource for
- 20:19SPEAKER_01: researchers to do all kinds of different research.
- 20:22How did you manage to build an institution that has defer as a
- 20:27watchword?
- 20:28SPEAKER_00: Because I think many listeners will just be,
- 20:30SPEAKER_00: I think,
- 20:30surprised and impressed at an institution that was set up with
- 20:34the ability to pass
- 20:35SPEAKER_00: the marshmallow tests,
- 20:36SPEAKER_00: not to try and eat dessert first.
- 20:40SPEAKER_00: I'm curious,
- 20:40are there things operationally that were built into UK Biobank
- 20:43that helped make it
- 20:44easier to make those decisions to wait and defer and take more
- 20:47of the value down
- 20:48SPEAKER_00: the road?
- 20:49They say that the definition of an intelligent man or woman is
- 20:53one who can hold two
- 20:54SPEAKER_01: opposing views at the same time,
- 20:56SPEAKER_01: don't they?
- 20:56SPEAKER_01: In a way, that was a situation, I think, for all of us.
- 21:00SPEAKER_01: We all want quick wins.
- 21:01But on the other hand,
- 21:02we know there are no quick wins in a prospective study.
- 21:05SPEAKER_01: It was really kind of just reinforcing the messaging.
- 21:09SPEAKER_01: I mean,
- 21:09SPEAKER_01: I remember at the 10-year renewal application,
- 21:13SPEAKER_01: so every five years,
- 21:15SPEAKER_01: the funders say,
- 21:15SPEAKER_01: well,
- 21:16SPEAKER_01: what have you done?
- 21:16SPEAKER_01: And of course,
- 21:17SPEAKER_01: the first 10 years we were building,
- 21:19SPEAKER_01: we were recruiting,
- 21:20SPEAKER_01: and then we were starting to turn samples into data.
- 21:25SPEAKER_01: We were linking the people into their health record systems,
- 21:28SPEAKER_01: but it was a very slow buildup of use of the resource.
- 21:32SPEAKER_01: The 10-year review, I was asked, you know, what have UK Baibang?
- 21:35done for which it was set up it was a one-word answer nothing
- 21:39there was a kind of
- 21:40quiet pause at that point but that was the truth you know it
- 21:44was a resource that
- 21:46was in gestation i went on to say that it will be the next 5 10
- 21:5715 20 years where
- 21:52SPEAKER_01: it really starts to deliver on the investment that has been made
- 21:57SPEAKER_01: The thing that really,
- 21:58SPEAKER_01: I think,
- 21:58transformed UK Biobank that made it move from being a UK study
- 22:04to being an
- 22:04international study was when the UK government has some funding
- 22:09available and they
- 22:11SPEAKER_01: offered that funding to UK Biobank to get the genotyping done.
- 22:15SPEAKER_01: Athymetrics in California won the tender to do the genotyping.
- 22:19SPEAKER_01: We were able to tell researchers that
- 22:22SPEAKER_01: that there was genotyping data available on half a million participants.
- 22:26And that's where we were moving to a scale that had never been
- 22:29done before.
- 22:30Researchers had been combining studies of a few hundred or a
- 22:35few thousand people to
- 22:37SPEAKER_01: try to understand what were the genetic determinants of,
- 22:39SPEAKER_01: say,
- 22:39SPEAKER_01: blood pressure or height or cardiovascular disease.
- 22:42SPEAKER_01: And here now we're due to taping data from half a million people.
- 22:45SPEAKER_01: And that was really what put Biobank on the researchers' map.
- 22:49SPEAKER_01: And we started to see an increase in use of the resource internationally,
- 22:53SPEAKER_01: particularly moving from the UK,
- 22:56SPEAKER_01: where,
- 22:56SPEAKER_01: of course,
- 22:56SPEAKER_01: a lot of researchers have been involved in designing and building it,
- 22:59SPEAKER_01: to an increase in North America and in mainland Europe.
- 23:04As the Biobank got off the ground,
- 23:05especially in the second half of the 2000s,
- 23:09SPEAKER_00: There was a lot of public scrutiny,
- 23:11SPEAKER_00: a lot of public interest as half a million people were recruited,
- 23:15SPEAKER_00: obviously.
- 23:16Did that public scrutiny or parliamentary scrutiny lead to
- 23:19significant changes in
- 23:20SPEAKER_00: the operations of the biobank?
- 23:22SPEAKER_00: How much did you adjust in response to feedback?
- 23:25The thing that the Wellcome Trust and the MRC did before
- 23:29anything happened was
- 23:30SPEAKER_01: really widespread consultation.
- 23:33Consultation with the research community about what were the
- 23:37things that people wanted to study.
- 23:40So a lot of different working groups were set up to determine
- 23:44what questions or
- 23:45SPEAKER_01: measurements one would make.
- 23:47SPEAKER_01: to be able to understand, say, diet or activity.
- 23:51There were a whole set of pilot studies done by researchers to
- 23:57look at how you
- 23:58SPEAKER_01: collect samples and assay those samples.
- 24:03SPEAKER_01: in the future.
- 24:04So how would you collect and process a sample so that when the
- 24:08technology caught
- 24:10SPEAKER_01: up,
- 24:10SPEAKER_01: and you could do that at a scale of half a million people,
- 24:13SPEAKER_01: your sample would be the right kind of sample.
- 24:15So a whole set of different groups piloted how to collect and
- 24:19process samples,
- 24:21SPEAKER_01: as I say,
- 24:21SPEAKER_01: what questions to ask,
- 24:23SPEAKER_01: what measurements to make.
- 24:24There was a lot of consultation with patient groups around the
- 24:29issues of things
- 24:29SPEAKER_01: like feedback,
- 24:31SPEAKER_01: no feedback.
- 24:31So in UK Biobank,
- 24:33there is a policy of not feeding back information to
- 24:36participants.
- 24:37SPEAKER_01: Now,
- 24:37that was something that went through a lot of iterations
- 24:41because you imagine,
- 24:42SPEAKER_01: well,
- 24:42SPEAKER_01: you want me to join a study.
- 24:44SPEAKER_01: You need to make all these measurements.
- 24:45SPEAKER_01: Why aren't you going to tell me what those measurements are?
- 24:48SPEAKER_01: Then you go, well, we don't know what these measurements mean.
- 24:51Do you still want us to tell you what they are when they may or
- 24:55may not be relevant
- 24:56SPEAKER_01: to your disease?
- 24:57SPEAKER_01: They may be misleading, for example.
- 25:00So even things like single gene disorders that are identified
- 25:03in a population may
- 25:06SPEAKER_01: have quite different relevance in a free living population.
- 25:11SPEAKER_01: So they may be much less predictive of disease.
- 25:14SPEAKER_01: And we don't know these things.
- 25:15SPEAKER_01: And so the whole purpose of UK Biobank is to learn.
- 25:19SPEAKER_01: So the issue with feedback is it can actually cause harm.
- 25:23SPEAKER_01: It can mislead.
- 25:25SPEAKER_01: This was discussed in great depth about the feedback policy.
- 25:28SPEAKER_01: And I think there are different positions on this.
- 25:31SPEAKER_01: The one thing you can say is that if you don't have feedback,
- 25:34SPEAKER_01: you can guarantee you won't cause someone harm.
- 25:37SPEAKER_01: But it is one where there was a lot of ethical,
- 25:40SPEAKER_01: legal,
- 25:41SPEAKER_01: and participant engagement before that happened.
- 25:45SPEAKER_01: So there had been very widespread consultation in that respect.
- 25:48I think the areas where there was controversy was to set up a
- 25:53study like this is very expensive.
- 25:55SPEAKER_01: And so researchers were saying,
- 25:57SPEAKER_01: well,
- 25:58SPEAKER_01: that's been takeaway funding that could be used for my research now.
- 26:03SPEAKER_01: That is a perfectly reasonable concern.
- 26:07I think actually what happened was that Wellcome and the
- 26:11Medical Research Council
- 26:12SPEAKER_01: put additional funding in to set it up.
- 26:15SPEAKER_01: So I don't know that it really had that effect.
- 26:18SPEAKER_01: The other controversy was around,
- 26:19SPEAKER_01: is this what you would do if you were interested in genetics alone?
- 26:23That was allowed to rumble on for longer than perhaps it should
- 26:28have done.
- 26:28SPEAKER_01: It took a while to address that.
- 26:31SPEAKER_01: I remember when I was asked to take on UK Biobank in 2005,
- 26:36SPEAKER_01: the first thing I did was go and talk to
- 26:39geneticists who had been saying well this isn't what you do if
- 26:42you want to set up a
- 26:42SPEAKER_01: genetic study i said well it really isn't a pure genetic study this is an
- 26:46epidemiological study where we need to study people when
- 26:49they're healthy and we
- 26:50need to follow them up long term in order to look at the
- 26:53association between genes
- 26:55environment and lifestyle and they kind of went oh well why
- 26:58didn't someone say that
- 26:59the same with some of the people in government the chair of the
- 27:02health select
- 27:02SPEAKER_01: committee had also raised concerns about this
- 27:05SPEAKER_01: And again, I went and talked with him.
- 27:06So a lot of it was about addressing misunderstandings of what
- 27:10UK Biobank was about.
- 27:12SPEAKER_00: To go back to the point about communicating to participants,
- 27:17SPEAKER_00: these altruistic volunteers,
- 27:19imagine I was someone who was coming into the NHS in 2009 and
- 27:24signing up for my
- 27:25SPEAKER_00: data to be tracked anonymously in the UK Biobank.
- 27:29What information would I be told about my measurements and what
- 27:32would I not be told?
- 27:34When we wrote out through the NHS to people who were living
- 27:37within about 50 miles
- 27:39SPEAKER_01: of where we had recruitment centres,
- 27:41SPEAKER_01: they were invited to participate.
- 27:43They were sent an information leaflet telling them what the
- 27:45study was about and
- 27:47that there wouldn't be feedback of any of the subsequent
- 27:50analyses of the samples.
- 27:53They were told that when they came to the assessment center,
- 27:56they decided to come.
- 27:57SPEAKER_01: If anything was noticed,
- 27:59SPEAKER_01: particularly during the visit,
- 28:01any kind of issue that was an incidental finding and things
- 28:05like their smoky
- 28:06SPEAKER_01: history would be recorded,
- 28:07SPEAKER_01: their body mass index would be recorded.
- 28:09They got about half a dozen pieces of information given to them
- 28:13at the end of the
- 28:14SPEAKER_01: visit that told them about
- 28:16SPEAKER_01: the heel ultrasound density,
- 28:18SPEAKER_01: a test that was done during the visit,
- 28:21SPEAKER_01: which is a rather crude measure.
- 28:22SPEAKER_01: If you have a very low density, it can be suggested of osteoporosis.
- 28:26They were told their body mass index as to where they were on
- 28:29the scale in terms of overweight.
- 28:31SPEAKER_01: But that was about it.
- 28:33SPEAKER_01: But it was very,
- 28:33very clear in the consent that this was not a health check and
- 28:38that they would not
- 28:39SPEAKER_01: be getting feedback of their subsequent results,
- 28:42although we would be keeping them informed about what UK
- 28:45Biobank was doing in terms
- 28:47of the use of the data and research findings that emerged and
- 28:51how they were
- 28:51SPEAKER_01: contributing to helping to understand how to prevent and treat disease.
- 28:56SPEAKER_01: We also had their agreement to get access to all of their medical and
- 29:00SPEAKER_01: health-related records and to get back in touch with them to ask for more
- 29:05SPEAKER_01: information,
- 29:06SPEAKER_01: provided that did not constitute feedback itself.
- 29:09SPEAKER_01: So we wouldn't go back to somebody and say,
- 29:11SPEAKER_01: you know,
- 29:12you have a particular genetic abnormality and we want to get
- 29:15more information from
- 29:16SPEAKER_01: you.
- 29:16SPEAKER_00: And despite that strong limitation on information sharing,
- 29:20SPEAKER_00: it's my understanding you hit 500,000 participants earlier than expected,
- 29:24SPEAKER_00: that there was more interest in
- 29:26SPEAKER_01: We set up a machine that worked.
- 29:29The IT team that set up the systems for running these
- 29:32assessment centers did a really good job.
- 29:35SPEAKER_01: We were essentially, if you like, running an airline.
- 29:37SPEAKER_01: We had 100 seats per day in each of the half dozen centers we had open.
- 29:43SPEAKER_01: We ran them seven days a week.
- 29:46SPEAKER_01: If we only got 90 in a day, we would run out of money.
- 29:48if we went to 110 a day the system crashed you people ended up
- 29:53having to kind of
- 29:54hang around waiting and they were unhappy the assistance were
- 29:57put in place to
- 29:58monitor the process of inviting people filling up the
- 30:03appointment slots not
- 30:05overfilling them monitoring to see whether there are any delays
- 30:08occurring in any of
- 30:10SPEAKER_01: the centers on any day as well as monitoring the quality of the data
- 30:14that was being collected they did a very good job in managing
- 30:17that so we were able
- 30:19to recruit on time a bit ahead of time and most importantly on
- 30:24budget i'm curious
- 30:26the biobank was established as an independent charitable
- 30:29company of all the forms
- 30:30that it could have taken was that ever in question were there
- 30:33ever other shapes
- 30:35SPEAKER_00: proposed for the uk biobank institutionally
- 30:38SPEAKER_01: Not really.
- 30:39SPEAKER_01: They didn't want it to be the study of a particular institution.
- 30:45SPEAKER_01: So they didn't want it to be,
- 30:46SPEAKER_01: you know,
- 30:46SPEAKER_01: I don't know,
- 30:47SPEAKER_01: an Oxford University or a Manchester University study.
- 30:50SPEAKER_01: The MRC and the Wellcome Trust were keen for it to be seen as being a UK
- 30:54SPEAKER_01: institution with no researcher ownership of it,
- 30:59SPEAKER_01: again,
- 30:59SPEAKER_01: to stress this accessibility.
- 31:02SPEAKER_01: And when I was made the principal investigator and chief executive,
- 31:06I kind of made the point that people could not collaborate with
- 31:11us on using the
- 31:14SPEAKER_01: data.
- 31:15SPEAKER_01: It was their data to use.
- 31:17So although I am the PI,
- 31:18I've probably published less on UK Barrington than almost
- 31:20anybody else.
- 31:21We wanted to kind of break the mold and make it clear that the
- 31:25data were for
- 31:27researchers anywhere in the world to use for health related
- 31:30research in the public
- 31:30SPEAKER_01: interest.
- 31:31SPEAKER_01: They didn't collaborate with us.
- 31:33SPEAKER_01: They made an access application.
- 31:35They complied with the material transfer agreement for
- 31:39accessing those data,
- 31:41and they applied their imaginations to learn more about how to
- 31:44prevent and treat
- 31:45SPEAKER_01: disease.
- 31:46I think the main driver from the Medical Research Council and
- 31:49Wellcome Trust's
- 31:49perspective was to reinforce this point that it was not a study
- 31:55that belonged to
- 31:56SPEAKER_01: any researcher.
- 31:57It was a study that belonged to the research community to do
- 32:01health-related
- 32:02SPEAKER_01: research in the public interest.
- 32:04I'd love to get your thoughts on the various lessons from
- 32:08setting up the biobank
- 32:10SPEAKER_00: and from being the principal investigator over the last 20 years.
- 32:14For those who are listening to this and who are looking to
- 32:17establish new research
- 32:18SPEAKER_00: institutions or new research enterprises,
- 32:21SPEAKER_00: especially large or more complex ones like the UK Biobank is,
- 32:25SPEAKER_00: maybe let's start with institutional design lessons.
- 32:28What would you say to people who are trying to stand up an
- 32:30institution that can
- 32:31SPEAKER_00: carry on complex research work in the long term?
- 32:35SPEAKER_01: I think the thing that we focused on was, what do we need to do now?
- 32:41What do we need to put in place so that we can do what we want
- 32:46to do in the future?
- 32:49SPEAKER_01: But don't try to do everything now.
- 32:51SPEAKER_01: At the beginning of UK Biobank,
- 32:52SPEAKER_01: there was a lot of pressure to say,
- 32:55SPEAKER_01: well,
- 32:55SPEAKER_01: what's your access policy?
- 32:57SPEAKER_01: And our point was, we haven't got anything to access.
- 33:02SPEAKER_01: What we're focusing on is getting something to access.
- 33:05We're doing that in such a way that we will be able to allow
- 33:10the data to be used in
- 33:11SPEAKER_01: the way that we want it to be used,
- 33:12SPEAKER_01: but we're not going to develop the access policy.
- 33:15SPEAKER_01: We want to focus on what do we can send the participants for?
- 33:18SPEAKER_01: Do we get the information from the participants?
- 33:21SPEAKER_01: Do we collect the samples in ways that can be used in 10 years time?
- 33:27One of the first things I did when I was made the CEO was to
- 33:31cancel an order for
- 33:33SPEAKER_01: biochemistry analyzers.
- 33:35SPEAKER_01: I said,
- 33:36SPEAKER_01: let's recruit the people,
- 33:37SPEAKER_01: let's collect the samples,
- 33:39SPEAKER_01: and we'll think about what we analyze when we've got the samples.
- 33:43It goes back to that thing about you defer what you don't need
- 33:46to do now.
- 33:47That means you can kind of focus your attention on the things
- 33:50that you do need to
- 33:51SPEAKER_01: do now,
- 33:51SPEAKER_01: and you can focus your resource on the things you can do now.
- 33:55SPEAKER_01: And hopefully in the future,
- 33:56SPEAKER_01: the things that you want to do will be easier to do and cheaper to do.
- 34:01SPEAKER_01: If we had genotyped all the samples,
- 34:04SPEAKER_01: as we were collecting them,
- 34:05SPEAKER_01: which some studies do,
- 34:07SPEAKER_01: it would have been a lot more expensive to do them when we did do them.
- 34:12SPEAKER_01: Perhaps more importantly,
- 34:13the quality would be worse because you would be collecting them
- 34:17in the order of the
- 34:19SPEAKER_01: participants you collected.
- 34:20SPEAKER_01: And if there were
- 34:21SPEAKER_01: shifts in the type of participant and shift in the assay methodology,
- 34:27SPEAKER_01: it's kind of very difficult to untangle those.
- 34:30Whereas if you wait until you've recruited everybody and then
- 34:32analyze the samples
- 34:35SPEAKER_01: in a kind of quasi random order,
- 34:37SPEAKER_01: you get rid of any systematic difference between the participants.
- 34:42SPEAKER_01: And, you know, you're using better technology because it's later.
- 34:45SPEAKER_01: You use a smaller volume.
- 34:47SPEAKER_01: It's at lower costs.
- 34:48I think you focus on what you have to do and just make sure
- 34:52that you're planning
- 34:53things in a way that allows you to do the things you want to do
- 34:56in the future,
- 34:56SPEAKER_01: but not necessarily do them.
- 34:59SPEAKER_01: People struggle with that sometimes.
- 35:01SPEAKER_01: You know, what's the quick win?
- 35:02SPEAKER_01: The quick win is not doing something.
- 35:05that principle of deferring which we've already talked about a
- 35:08little bit on the
- 35:09one hand i understand that it's relatively easy to keep
- 35:12insisting on that principle
- 35:14in a longitudinal study because the study is set up such that
- 35:17you won't get the
- 35:17value for a while but these things like choosing a data access
- 35:21policy all these
- 35:22other things that you could be doing how much political
- 35:25pressure was there either
- 35:26from parliament or just from other actors who wanted to be seen
- 35:29doing something
- 35:30SPEAKER_00: getting these quick wins
- 35:32SPEAKER_01: A lot.
- 35:33SPEAKER_01: It's often easier to do things than not do them.
- 35:36SPEAKER_01: The question is, what's the best strategy?
- 35:38SPEAKER_01: And if you try to do too many things at the same time,
- 35:43then you risk not doing the ones that need to be done as well
- 35:47as they could be
- 35:47SPEAKER_01: done.
- 35:48I think just focusing on what needs to be done ensures that
- 35:52it's done better and
- 35:54deferring the things that don't need to be done until they can
- 35:57be done better,
- 35:59SPEAKER_01: more efficiently,
- 36:00SPEAKER_01: lower cost.
- 36:01SPEAKER_01: I think it's an argument one has to make repeatedly,
- 36:04SPEAKER_01: but I think it is one that is worth considering.
- 36:07SPEAKER_00: If you could go back to 2005,
- 36:08SPEAKER_00: when you first became principal investigator and CEO,
- 36:12SPEAKER_00: would you do anything differently this time around?
- 36:16I'd try to work out how we could get a million people for the
- 36:18same money would be number one.
- 36:21SPEAKER_00: What would that give you statistically that 500,000 people will not?
- 36:26You'll get more people developing disease within a shorter
- 36:30period of time.
- 36:31SPEAKER_01: It's the number of people who develop a condition for actual power.
- 36:36SPEAKER_01: I think the age range was the right age range.
- 36:38SPEAKER_01: I think we got quite a lot of diversity for the UK.
- 36:44One of the arguments that was made against the UK Biobank is
- 36:47there aren't a lot of
- 36:48SPEAKER_01: people from Africa or from India or from Asia in the UK Biobank.
- 36:53SPEAKER_01: But at the time, they weren't in the UK.
- 36:56If you want to study a wider range of peoples,
- 36:59then you need to set up studies like UK Biobank.
- 37:01SPEAKER_01: You carefully selected different parts of the world.
- 37:05And in fact,
- 37:06we set up a parallel study in China,
- 37:08the China Kodori Biobank,
- 37:09SPEAKER_01: to UK Biobank at the same time in order to address that.
- 37:12And we also worked with the Mexican Ministry of Health to
- 37:15support a large scale study in Mexico.
- 37:17SPEAKER_01: So you can't get some kind of diversity in a population like the UK.
- 37:21But we have diversity in terms of socioeconomic status,
- 37:24in terms of rural,
- 37:25urban,
- 37:26etc.
- 37:27But larger numbers would give you greater power to study more
- 37:31conditions and rarer conditions.
- 37:34SPEAKER_01: I wouldn't have done the biochemistry assays when we did them.
- 37:37SPEAKER_01: We had to do hematology with fresh blood,
- 37:41SPEAKER_01: but with the biochemistry,
- 37:43SPEAKER_01: there was a lot of pressure to do the assays of things we know about,
- 37:48SPEAKER_01: like cholesterol and various cancer markers.
- 37:51SPEAKER_01: But I actually think it was a lot of work,
- 37:54SPEAKER_01: quite expensive at the time,
- 37:55SPEAKER_01: used quite a bit of sample,
- 37:57and probably one could do that now with some of the newer
- 38:01technologies with less
- 38:02SPEAKER_01: effort and less cost.
- 38:04SPEAKER_01: Again, that was something we could have deferred.
- 38:07SPEAKER_01: But in general, I think it's worked out well.
- 38:09SPEAKER_01: As I say,
- 38:10SPEAKER_01: the access model has resulted in very,
- 38:13very much wider spread use of the resource and many more
- 38:17findings than I have
- 38:19SPEAKER_01: anticipated at this pace.
- 38:21SPEAKER_00: Maybe this is a silly way of framing this question, but I'm curious.
- 38:24SPEAKER_00: If I was thinking of setting up a US biobank,
- 38:28SPEAKER_00: say 600,000 participants,
- 38:31what would the value of that American biobank be in a world
- 38:34where the UK biobank
- 38:35SPEAKER_00: already exists?
- 38:36SPEAKER_00: How much of the value has already been built by an existing biobank?
- 38:42SPEAKER_01: Well, of course, under President Obama, the All of Us study was set up.
- 38:47I was one of the people asked to advise on that study,
- 38:50given our experience in the UK Biobank.
- 38:52And I think it is going to be a very important study because it
- 38:56includes a much
- 38:57SPEAKER_01: wider range of people than UK Biobank.
- 39:01Very large numbers of people from various ethnic groups that
- 39:07will be interestingly different.
- 39:09So I think genetically and also in terms of lifestyle and
- 39:11environment,
- 39:12they will be different.
- 39:14I think that it's really important to have a few studies in
- 39:17different parts of the
- 39:18world in order to understand the full range of genetic
- 39:21lifestyle and environmental
- 39:23SPEAKER_01: risk factors and a wide range of different diseases.
- 39:27You go to different parts of the world,
- 39:28you have different rates of disease.
- 39:31And so studies that allow you to look at the widest range of
- 39:34exposures and the
- 39:35SPEAKER_01: widest range of diseases can't be achieved in a single country.
- 39:39SPEAKER_01: So I think what
- 39:40SPEAKER_01: UK Biobank does is it demonstrates the value of having such studies,
- 39:45and it demonstrates the value of making that data as widely
- 39:49available as possible.
- 39:50And I think our experience in UK Biobank of doing so has been
- 39:55positive in two ways.
- 39:57SPEAKER_01: One,
- 39:58SPEAKER_01: it showed that if you make data available to researchers,
- 40:01SPEAKER_01: a lot of research findings come out.
- 40:03The second thing is that there's been a lot of external
- 40:05investment in making UK
- 40:07SPEAKER_01: Biobank even more valuable.
- 40:09SPEAKER_01: The UK government funded the genotyping,
- 40:11SPEAKER_01: but it was actually industry that funded the XM sequencing,
- 40:15SPEAKER_01: a large part of the whole genome sequencing.
- 40:16SPEAKER_01: They're now funding proteomic assays on the data.
- 40:20SPEAKER_01: Government supported imaging in UK Biobank.
- 40:22SPEAKER_01: We've imaged 100,000 of the participants,
- 40:25but now industry and philanthropic organizations are funding
- 40:28repeat imaging.
- 40:30So every dollar of Wellcome Trust and Medical Research Council
- 40:35funding of UK
- 40:36Biobank has leveraged about $12 of external investment in
- 40:41enhancing the resource in
- 40:43SPEAKER_01: assays,
- 40:44SPEAKER_01: in imaging,
- 40:46and then those data be made available to that very wide range
- 40:50of researchers.
- 40:51That's allowed them to do even more research than otherwise
- 40:54would be possible.
- 40:55SPEAKER_01: So accessibility,
- 40:57SPEAKER_01: generated researchers using the data,
- 40:59SPEAKER_01: generated inward investment,
- 41:01SPEAKER_01: generated researchers being able to do even more with that enhanced data.
- 41:05SPEAKER_01: And I think that that experience has helped other similar studies,
- 41:09SPEAKER_01: like all of us,
- 41:11SPEAKER_01: consider the value of that kind of access model.
- 41:14SPEAKER_01: The Mexican Ministry of Health Study,
- 41:16SPEAKER_01: the Mexico City Prospective Study,
- 41:18SPEAKER_01: has just been made available to researchers around the world.
- 41:22SPEAKER_01: It was XM sequenced by industry.
- 41:25SPEAKER_01: Those data are now available to all researchers,
- 41:27SPEAKER_01: having been first made available to Mexican researchers.
- 41:31If you like,
- 41:32give them a head start for their contribution to having set up
- 41:35in the first place.
- 41:36One of the overarching lessons that I think you're pointing to
- 41:40here is just that we
- 41:42SPEAKER_00: underestimate the value of open scientific data infrastructure,
- 41:47SPEAKER_00: data that's available to a broad range of researchers,
- 41:49SPEAKER_00: because although the UK Biobank cost a lot of money to set up,
- 41:53SPEAKER_00: In relative terms, it's actually not that much.
- 41:56SPEAKER_00: And as science and markets evolve,
- 41:57SPEAKER_00: that infrastructure can become incredibly valuable in the future.
- 42:00SPEAKER_00: Beyond infrastructure assets like this,
- 42:02SPEAKER_00: like the UK Biobank,
- 42:03what other kinds of scientific infrastructure would you want to
- 42:07see that could
- 42:07learn from these kind of basic principles of very large sample
- 42:11size and open access
- 42:13SPEAKER_00: for researchers?
- 42:15SPEAKER_01: Well,
- 42:15SPEAKER_01: we've seen so much over the last 10 or 20 years of the value of data,
- 42:19SPEAKER_01: but the value of data are only if people can use the data.
- 42:24I think one of the issues is that people are often more
- 42:30concerned about the sins of
- 42:31SPEAKER_01: commission than they are the sins of omission.
- 42:35SPEAKER_01: What could go wrong if we make data accessible?
- 42:38SPEAKER_01: The risks are given great emphasis.
- 42:42Whereas the risks of not making data available are not
- 42:46considered seriously enough.
- 42:47And I think what UKBIN does is it really helps to redress that
- 42:515,000 peer-reviewed
- 42:54SPEAKER_01: publications last year alone by making the data available.
- 42:58SPEAKER_01: I mean, the benefit of that is enormous.
- 43:00SPEAKER_01: We're already seeing new targets for disease.
- 43:04new ways of identifying decades before disease develops who's
- 43:10at risk and therefore
- 43:12identifying ways in which we could target intervention to
- 43:15prevent people from
- 43:16developing those disease in the first place better ways of
- 43:19screening for disease
- 43:20than we do at the moment so you can imagine kind of precision
- 43:23public health using
- 43:25things like breast cancer screening or colorectal screening or
- 43:28prostate cancer
- 43:28SPEAKER_01: screening
- 43:29SPEAKER_01: in a much more precise way,
- 43:32SPEAKER_01: screening the people who really are at risk,
- 43:34SPEAKER_01: better ways of using cholesterol,
- 43:36lowering other cardiovascular protective treatments in people
- 43:39at high risk based on
- 43:41their genetic and other risk factor data that are coming out of
- 43:44the UK Biobank.
- 43:44So perhaps UK Biobank helps to redress that balance and
- 43:48understand the risks of not
- 43:50SPEAKER_01: allowing data to be used,
- 43:52SPEAKER_01: particularly in the health arena.
- 43:55SPEAKER_01: The ability now to have data on secure platforms,
- 44:01SPEAKER_01: which is where we've got the UK Biomank data now,
- 44:03SPEAKER_01: and having researchers come to the data also democratizes access.
- 44:08SPEAKER_01: So you don't need to have a big computer in your institution.
- 44:11SPEAKER_01: Because you can just come to the data,
- 44:14SPEAKER_01: which is sitting on a cloud-based platform,
- 44:16SPEAKER_01: and you can get access to the compute you need when you need it.
- 44:21SPEAKER_01: From wherever you are in the world, the data are more secure.
- 44:26SPEAKER_01: But I think even more importantly,
- 44:28the data are more accessible in terms of not only being able to
- 44:32get outcome,
- 44:33SPEAKER_01: but able to have the compute to analyze them wherever you are.
- 44:37SPEAKER_01: And one of the things that we've really been trying to do is
- 44:40SPEAKER_01: help support researchers in less well-resourced parts of the world,
- 44:44SPEAKER_01: in Africa and India,
- 44:46SPEAKER_01: but also East and Europe and South America,
- 44:48to take advantage of their brains by making the data accessible
- 44:53to them and
- 44:54SPEAKER_01: providing compute for them.
- 44:56Any other lessons that we should try and get to for an American
- 44:59policymaker audience?
- 45:01SPEAKER_00: I ask because there's a fair amount of folks in DC who listen to this.
- 45:04And if there's something we should say specifically to them,
- 45:08I would love to get it in.
- 45:10SPEAKER_01: I think that one has to have a very long view.
- 45:14SPEAKER_01: UK Biobank was set up 20 years ago, and it's now a mature resource.
- 45:19SPEAKER_01: It's now starting to generate important findings.
- 45:23SPEAKER_01: There's real value in setting up studies now.
- 45:27SPEAKER_01: They're not competing with UK Biobank.
- 45:30SPEAKER_01: They're complementary to UK Biobank.
- 45:32SPEAKER_01: They need to be interestingly different.
- 45:35SPEAKER_01: I think with the All of Us study,
- 45:37there's a very interestingly different population in the US
- 45:41that will really extend
- 45:43the range of genetic and lifestyle environmental risk factors
- 45:46that can be studied.
- 45:48SPEAKER_01: It will really start to come into its own in the next 10 years or so.
- 45:52SPEAKER_01: And the question is not where we want to be now.
- 45:55SPEAKER_01: The question is where we want to be in the future.
- 45:58Thinking about a strategy of having a series of these
- 46:00large-scale studies and
- 46:02SPEAKER_01: thinking about how they can be accessible and analyzed together to really
- 46:07SPEAKER_01: understand better how to prevent and treat disease.
- 46:10People kind of say to me sometimes,
- 46:12what's the sunset clause for UK Biomac?
- 46:14SPEAKER_01: And I said, this is a sun that never sets.
- 46:17SPEAKER_01: This is a resource that gets more and more valuable as time goes by.
- 46:21SPEAKER_01: And that will be true of studies that are set up now.
- 46:24SPEAKER_01: They will become more and more valuable as time goes by.
- 46:28SPEAKER_01: And the more such studies we have,
- 46:30SPEAKER_01: and the more we're able to assay in those samples,
- 46:33SPEAKER_01: we're just touching the surface at the moment.
- 46:36SPEAKER_01: Sequencing is just the beginning.
- 46:38The proteomic assays that we're doing in UK Biobank are just
- 46:41the beginning of the
- 46:42kinds of really detailed proteomic assays that will be possible
- 46:46in the next five or
- 46:47SPEAKER_01: 10 years.
- 46:48I think that what we're going to see is a set of these studies
- 46:52that over the next
- 46:5320 or 30 years will really help us to understand how to prevent
- 46:58and treat a huge
- 46:59SPEAKER_01: range of diseases.
- 47:01SPEAKER_01: And no single study is going to be sufficient.
- 47:04SPEAKER_00: I want to close with just a personal question.
- 47:07SPEAKER_00: You've spent two decades in this role as CEO of UK Biobank,
- 47:12SPEAKER_00: which is,
- 47:12SPEAKER_00: I think,
- 47:12SPEAKER_00: a long time for anybody to be in any one role.
- 47:14SPEAKER_00: I'm curious how, as you reflect on that two decades, how has it been?
- 47:18SPEAKER_00: Has it been as much fun as it seems like?
- 47:22SPEAKER_01: Yeah, it's been an absolute blast.
- 47:24SPEAKER_01: And I'm learning all the time.
- 47:25I find myself thinking not what have we done,
- 47:28but more what have we not done?
- 47:30SPEAKER_01: What are the opportunities that we haven't yet taken?
- 47:34SPEAKER_01: And how do we take advantage of those to make the resource better?
- 47:39SPEAKER_01: It's been a brilliant thing to be involved in.
- 47:41SPEAKER_01: I work with a fantastic team.
- 47:43SPEAKER_01: Just seeing how the research community has responded to...
- 47:48SPEAKER_01: having access to data has been really very interesting.
- 47:52SPEAKER_01: We kind of started out with this project where people said,
- 47:57SPEAKER_01: you mean really we can access the data?
- 48:00SPEAKER_01: Really we can access the samples?
- 48:02SPEAKER_01: And we would say yes.
- 48:03SPEAKER_01: And once they'd got past what the catch element of it is and started,
- 48:09SPEAKER_01: kind of found ourselves now turning into a utility.
- 48:12And we have to learn how to function as a utility and think
- 48:16about what are the
- 48:16SPEAKER_01: things that we can make.
- 48:18SPEAKER_01: How can we make UK Babac even more valuable for the research community?
- 48:22SPEAKER_01: And that's what we're focusing on now.
- 48:24SPEAKER_01: There's always the next thing we should be doing.