- il y a 2 jours
Accelerating Innovation in Drug Discovery
Catégorie
🤖
TechnologieTranscription
00:00Well, gentlemen, thank you. End of a long day, but delighted you could both join us.
00:05So let's jump right in, and Albert, maybe I can start with you.
00:10Maybe I can start with you. I think we're getting a bit of someone else's mic, but I'll persevere.
00:17Maybe I can start with you, Albert.
00:19You've had a fairly busy couple of years.
00:22But given now that we seem to have moved on quite a bit in June 2022 from where we were
00:30certainly two years ago,
00:31how are you now thinking when you sort of step back?
00:34How are you now thinking about what the future of healthcare actually looks like?
00:39Yes. First of all, it's a great opportunity that I'm having to speak about something different than COVID for the
00:46first time in three years.
00:47And I think that health in the next decades will continue to be one of the most important variables
00:54that define the political debate, the economic life, and the social cohesion.
01:00And the reasons why is because demographics are very intense right now.
01:07People are living longer.
01:09As a result, we are having new medical needs that before they were insignificant or absent.
01:19Now they are emerging as significant medical needs.
01:22On the other hand, science is going to deliver.
01:28I think we are entering into a scientific renaissance that is driven not only by breakthrough discoveries
01:36in the domain of biology, but also with significant advancements in the domain of technology.
01:43And technology is empowering right now significantly the knowledge and our ability to get inside from this knowledge.
01:52Now, so demand will be there and offer, I think, will be there.
01:56Now, not everything will be right because all of that, the demographics are coming with a huge cost to society.
02:03But people will live longer means that more people will get in for longer pensions
02:10and lesser people contribute to that.
02:13More people will be needing hospitalizations or they will not be self-dependent
02:19and they will need assisting living and all of that.
02:21But I hope that we will all see that medical innovation can be a solution to this problem
02:29rather than part of the problem, that we will be able to reduce the cost by providing people
02:35their ability to be self-dependent, not living their lives in hospitals, by bringing new innovation.
02:41So you mentioned a scientific renaissance.
02:43One of the people driving this is Yakov, who joins us today, both as a professor and also as an
02:49entrepreneur.
02:49Yakov, just sort of like building a little bit on what Albert says.
02:54So I've read that it costs on average $2.6 billion to bring a single drug to market.
02:59It takes on average 12 years.
03:02How do you think about the way that we might reduce the inherent risk involved in this?
03:07I mean, how sustainable is that model at the moment?
03:11So it doesn't look like it's a very sustainable model.
03:14It didn't used to cost that much money.
03:18If you take a look a decade or two ago, the cost was around half a billion dollars.
03:23Now we're talking about 2.5.
03:25It's just that a lot of the early drugs that we had access to, well, we developed.
03:30We took them out.
03:31Aspirin was known for 5,000 years.
03:34So now we're really trying to develop new molecules, new mRNAs, to break through into the new areas of disease
03:43that we are facing.
03:44People didn't used to live past 60 and are now living past 90 and 80.
03:49And the big problem when you're trying to innovate is that the models that we are using right now in
03:55the labs are mainly rodents.
03:58It's mice and rats, which are extremely different from humans.
04:02So rodents have different genetics, different metabolism, and different physiology from humans,
04:08and we're still heavily relying on them and means that simply we can't translate this to the clinic.
04:15We end up with 88% failure in clinical studies.
04:21That's what drives the cost up.
04:23And it's exactly this problem that makes everything else so much worse.
04:27And this is where new technology, specifically organoid technology that was developed by Hans Clevers and my lab over the
04:35last decade,
04:36allow us to take human cells with human genetics and human physiology and metabolism
04:43and essentially test and develop those drugs on those organoid models
04:47and get a much better prediction as to what is going to be finally effective in the clinic or not.
04:55If we manage to do that, and it's a lot of ifs, we can dramatically reduce the failure rate in
05:01clinical studies
05:02and the cost that is associated with the drug development as a whole.
05:06So I'd love to get a little bit into your start-up tissue dynamics later on,
05:11because I think this organ-on-a-chip technology is absolutely fascinating.
05:14But maybe, Albert, we can come back to you and let's just talk a little bit about mRNA as platform
05:22technology.
05:24What have we learned over the past couple of years when it comes to developing new use cases
05:31for this kind of technology and the ways that we might apply this in the future?
05:35What kind of conditions, what kind of treatments do you think we can kind of develop
05:39and think about when it comes to scaling this in different ways, using this in different ways?
05:46A few comments about mRNA to start.
05:48First of all, I think it is a very powerful technology.
05:53I think we have seen only scratching the surface of what the technology can do,
05:59but it's not going to be panacea.
06:00It's not going to be the technology that will resolve everything.
06:03There are a lot of things that cannot be resolved with mRNA,
06:06and there will be hundreds of other technologies that will be equally important.
06:10So just to set, let's say, the stage.
06:13Now, I think what we learned during the pandemic is not that we had a good technology that we were
06:20able to deliver.
06:21I think the most important was that we were able to do things differently in the clinical discovery and development
06:27because of the pressure of the pandemic, but we wouldn't do it before.
06:32The fact that we were able to develop a vaccine in eight months, and it usually takes eight years,
06:39this is something that's not because mRNA enabled us to do that.
06:42It was also a very quick technology, but save maybe two, three months.
06:48What was the critical determinant was that we did things in parallel, we ignored costs,
06:56and we had very good collaboration with regulators.
06:59When our people would spend endless nights developing their case for why a clinical trial should go this way,
07:09they would send it, and they know that they would go to bed,
07:12and then the FDA or the EMA people would start sleepless nights
07:17so that they can review it in one week rather than three months.
07:21Sure.
07:22And there are multiple stages like that.
07:25So I think for me that's the key lesson learned.
07:28Sure, and I guess that part of that was new applications of data-led approaches to drug discovery.
07:35Jakob, if I can come back to you.
07:37How important do you think, from your perspective,
07:40the application of technologies such as machine learning and deep learning,
07:44how important is that in terms of how we're going to move forward?
07:49So this is fascinating, and I have to strengthen what Albert said
07:54about the regulator playing a critical role in vaccine development.
07:57It was also the massive digitization of HMO data that allowed us to get the data
08:04and get the vaccine approved so quickly.
08:07When you're applying machine learning tools, data mining is extremely useful right now,
08:13as well as new ways to think about chemistry with machine learning and AI-based tools.
08:21AI, however, has its limitations right now.
08:24It's really, really good in pathology, for example, when everything is the same.
08:29So, you know, automated staining, automated microscopy.
08:33It's also not bad in radiology, right?
08:36Because it's essentially the similar images, even though there are differences.
08:40But when it comes to the bread and butter of drug development,
08:43when it comes to biology, there is massive variability.
08:48Massive variability between cells, between labs, between technicians.
08:54Sometimes it's even difficult to repeat an experiment and get the same results.
08:58And this amount of noise is just detrimental to any machine learning applications.
09:04You know, I call it the valley of death for machine learning.
09:09And the pharmaceutical industry has really been trying to solve that for many years.
09:15And it's really technologies based on robotics, based on sensors,
09:20that could allow us to automate biology to the point where we can actually apply machine learning tools effectively
09:29in doing the entire drug development cascade.
09:33So a vision of autonomous drug development is kind of like the next big thing.
09:40Yeah.
09:41But in Israel, during the pandemic, you were seeing results pretty much in real time, right?
09:47So how important was that in terms of kind of identifying particularly like new variants,
09:52that access to real-time data?
09:56Do you want to say it?
09:57Maybe I'll be happy to talk about it as well.
10:00I can say it was phenomenally important.
10:03I think Israel was selected for this reason to be the one that would be given at the beginning
10:10unlimited supply of the vaccine because it was a small country,
10:13but they had very good electronic data, American records.
10:16And we could get over there conclusions that can be used from the entire world.
10:22Don't forget that Israel had vaccinated already by months everybody.
10:25And then we had significant discoveries through the real world evidence.
10:29We discovered, for example, that the second dose was not enough.
10:33We didn't know that.
10:34Yeah.
10:34When we registered the vaccine, we didn't know if we'll go for six months or five months
10:39or six years.
10:40Yeah.
10:40In Israel, we realized that after the six months is waning and we were able very quickly
10:46to develop studies to prove if we need a third dose or not.
10:50The same was with variants.
10:51When Delta came and or now Omicron that came, in Israel again was the first place where very reliably
10:58we were able to see if the vaccines hold against the new variants.
11:03So extremely, extremely important.
11:05So I want to speak a little bit.
11:07I'm sorry, Jakob.
11:08I'm sorry.
11:09So this is extremely important because this digitization of patient data allows you to move extremely
11:16fast.
11:17When we just had a study that was published in Science and Inflational Medicine, from the
11:23development of the first human kidney model to the disease mechanism to the drug to the
11:30clinical validation was less than eight months.
11:34I mean, the availability of clinical data to validate your findings that simply are based
11:41on human tissues without going through animals can be transformative for this entire ecosystem.
11:48Sure.
11:48Sure.
11:49Now, obviously, during the pandemic, there was, and we had a panel earlier today here at
11:53VivaTech about disinformation and misinformation.
11:55There was a lot of that around COVID vaccines, vaccines, even mask wearing became politicized.
12:02First, I'm wondering, from both of your perspectives, and maybe you, Jakob, as a professor, and from
12:08your position as well, how do we do a better job of combating, how do we get, you know, do
12:15a better job of conveying the ideas of science and scientific discovery and combating this
12:21kind of misinformation?
12:24So, this is something that I've been facing with two fronts, right?
12:30I'm both on tissue dynamics when you are trying to communicate the findings, clinical findings,
12:37and I think exactly like Kasparov has said in a few hours ago, there is a lot of misinformation
12:45outside.
12:46That means we need to be very, very effective scientific communicators right now.
12:52Because unless we go and talk about the science and make the science available, other people
12:59are going to make something else up.
13:03Yeah.
13:04We have to be present, obviously with regulatory concerns, but we have to be present and we
13:11have to tell the most complete story that we can.
13:13My other hand is under Future Meat Technologies.
13:17It's a cultured meat company, and we are very much advanced, one of the fastest growing company
13:24in this sector.
13:25And again, we're going to be bringing a completely different type of food to people's mouths.
13:30We have to be completely transparent about the process.
13:34Otherwise, it's simply not going to be accepted in these days and age.
13:38Sure.
13:39How do we do a better job in communicating science?
13:42I think the problem of misinformation was paid dearly during the pandemic.
13:47I think millions of deaths could have been avoided.
13:51I think this misinformation created a significant wave of reluctancy to believe in science.
13:59It's not only a symptom about pandemic or the science.
14:02We see that in politics.
14:03We see that in everything.
14:05It's, in fact, the fact that the political life is so polarized finds a very fertile ground
14:11for misinformation to occur.
14:13I don't know how to handle this.
14:15From one hand, I don't think that we need to restrict free speech.
14:19From the other hand, I think this situation is costing dearly.
14:25I don't agree with many things that Elon Musk has said.
14:28But with that one, I agree.
14:30I think one of the issues it is that we are allowing to have non-verifiable accounts.
14:35I truly believe that everybody has the freedom to speak as long as he's human.
14:40If it is a bot, I don't think he's covered by our principles of freedom to speak.
14:47So I think that's a very important first step.
14:50So everyone has one opinion.
14:52And then let's see what happens.
14:55Jakob, let's continue to look forward a little bit about your work at Tissue Dynamics.
14:59You mentioned it earlier on.
15:00You're calling it organ-on-chip technology, which is fascinating.
15:04Just maybe you explained it to me earlier.
15:07It's fascinating.
15:09And the analogy you used was car sensors.
15:11Do you want to maybe just develop that for the audience?
15:13Because it really is quite fascinating.
15:16So organ-on-chip technology was developed.
15:19It sounds like science fiction.
15:21It's the idea that you can take human cells and make human tissues on a chip that mimics human physiology.
15:29Essentially, on a table like this, run a million experiments at the same time.
15:34But it's a technology that is about 30 years old.
15:38And one of the things that we've been doing over the last seven years is integrate sensors physically inside human
15:45tissues.
15:46So the previous talk, you heard Apple introducing sensors to our everyday life, like monitoring our health.
15:54So you know you're going to get a heart attack days before you do.
15:57Well, imagine doing this in the lab, essentially having sensors in every single organ in your body, telling you way
16:06before anything happened.
16:08The technology is akin to having sensors in your car, right?
16:13This is, again, a development over the last 20 years.
16:15When we, your car is not going to catch fire in the middle of the highway.
16:20You're going to have a sensor telling you you lost air pressure in your tires or oil pressure in your
16:25tank.
16:26And it will tell you that something is critically wrong.
16:29And more or less, what is it?
16:31So you'll stop driving before you crash.
16:34It's exactly this technology that allows us to develop drugs extremely fast in tissue dynamics.
16:40Because our sensors give us the mechanism.
16:44What does the drug exactly do?
16:47And what does the disease do?
16:51It makes the process of drug development iterative.
16:55Because when you fail, for the first time in history, you know why you failed.
17:00So you can fix it and go back.
17:03Yeah.
17:04So we've got 1 minute 46.
17:07By the way, that's fascinating.
17:08And I think that, you know, seeing that, seeing you work on that and continue to develop that could have
17:13really kind of significant impacts on, you know, human health in the coming years.
17:18But I do want to get one final question, and this is for you, Albert, which is, we think about
17:24the development of the vaccines over the past, you know, two or three years.
17:28And this was a coming together of science, of government, of the private sector, really achieving something pretty remarkable.
17:35Now, we face, you know, a very clear sort of threat, which is, you know, antimicrobial resistance.
17:40What have we learned over the past couple of years to take on that challenge, which is really potentially another
17:45real global health crisis?
17:46I think the microbial resistance is a ticking bomb.
17:51If you do the math in a couple of decades, we'll overtake deaths of cancer.
17:57More people will die because of a microbial resistance than from cancer.
18:03There are a lot that can be done.
18:05I think science is there.
18:07I think, though, we are facing a big issue that very few companies are working on it.
18:12Very few in the private sector.
18:14It's almost like a market failure.
18:16The fact that those antimicrobials take a lot of cost to develop, and there's not a market because people will
18:24not buy them unless if there is a need.
18:26If there is, let's say, a resistance that it is developed, creates uncertainty.
18:30So, really, we remain in anti-infectives as a result, but we see ourselves being among the very few companies
18:39that remain on that.
18:40I think it would be smart from governments to develop a set of incentives so that more companies could get
18:49into that.
18:49And the incentives could be in the form of either stockpiling the products or giving benefits for other products if
18:57you get a new product approved.
19:00I think that would be a good path forward.
19:04Okay, well, thank you.
19:05We're out of time, but thank you both so much for joining us today to talk about how we can
19:08accelerate drug discovery.
19:09Such an important topic, and it was fantastic at your perspectives.
19:12Thank you so much for joining us today.
19:14Thank you, and thank you to you for listening.
19:15Thank you.
19:16Thank you.
Commentaires