Keynote: A big pharma research platform, for all our partners
Wolf-Dieter Busse, SVP, Head of Biotechnology R&D, Bayer Corporation Keynote: A big Pharma research platform, for all our partners
By on May 17, 2001Wolf-Dieter Busse, SVP, Head of Biotechnology R&D, Bayer Corporation
Keynote: A big Pharma research platform, for all our partners
SPEAKER: We will start the session, before the panel we have the next speaker is Wolf-Dieter Busse, who is the Senior Vice President and Head of Biotechnology R&D in the Bayer Corporation.
WOLF: Good morning ladies and gentleman, it's a great pleasure to give a talk about how we build at Bayer and our technology platform during the last 3 years, which is moving us forward into this new millenium into the next decade of discovery. As you can see on the picture along the path that was really paved by the discoveries in .. project. I personally come from California, my base is California and I'sm responsible .. decide to do so in the 80's in Europe and particularly in Germany it was recumbent free zone type of approach, and now we have in California an energy free zone type ., and as you know Bayer has not been involved in a major , of businesses and since we are not ranking in the top 3 or 4 in the pharmaceutical business, we decided to be very aggressive, building a research platform, which today is regarded as one of the largest biggest and most aggressive ones, not just only by our own sane. Never mind, building a technology platform in research is not just a walk through a candy shop. Building a technology platform has to satisfy of course the basic business needs that we have been confronted to in the 80's, and which we are continuing to be a confronted tool now and in the next 10 years. In general terms, productivity, finance, innovation and time to market have been gaps where the expectation of boards of the pharmaceutical industry and stock holders were different than the reality. In terms of productivity the amount of innovative drugs that came out is far below expectation, the cost of producing innovative drugs is continuing to grow, despite the fact that we have containment in costs. In terms of innovation only 10 to 15% of all drugs being launched that are truly innovative, and time to market has been an issue, which has been addressed and has various aspects. One is the regulatory aspect and the FDA due to the FDA modernization act and producer has significantly reduced the time to market, and now it's back to the research shops. If you look at the numbers of drugs launched in the last years and in the last year, it is rather disappointing that the number of innovative drugs has not gone up. In fact in the last year it went even down to 37. If you would listen to the RD measures of the big pharmaceutical houses 10 years ago, they would predict by this year already 80 or more drugs on the market, but it simply hasn'st happened. So that's quite disappointing and when we build new technology platforms we have to consider the basic business needs being more productive. The sales have increased but only by just 5% every year, so it also was disappointing. On the other hand the costs for R&D have increased very significantly, 7% per year, so you can see the scissors is still there, as a gap between what you need in sales and profits and what you want to achieve in reduction of cost. With that in mind and I stop here with the general business ideas, we have started 3 to 5 years ago building a technology platform at Bayer, and the solution of some of the problems seem to come from the human genome project. As first published only recently, the human genome contains around 30 thousand genes. This was far different as expected, even the top bio enphamatics companies, like Inside and others projected only 1 year ago still 120 thousand genes. So it is very disappointing in a way, but on the other hand it pushes the rush forward. We believe today that biology is much more complicated as just a number of genes of course, and we believe that coming from these genes by different biological mechanisms, hundred thousands or million different proteins which act right at the specific site of the biological function could be the target of the next 5 to 10 years. In fact many companies are being formed dealing with protio mix as the next phase after genome mix. We believe there are still 5 to 10 thousand drug targets left in the world that give us the hope that we can move on. In fact coming from California it reminds me a little bit to the 49'sers, and not this football team today. But the real 49'sers who were coming to California to dig into the mountains and dig out the gold is very limited number of mother loads there, lots of excitement, lots of fun and lots of different enabling tools to find the wealth of the future. When we draw these lines it was about 5 years ago, nobody would believe the first seed curve, here in red, that by 2005 the genome would be there, in fact it was project 3 to 5 years later. Of course nobody would believe now that the prediction that the outcome of the genome in the so called post gnomic age is so close with another 5 years to go, but that doesn'st mean that all drugs are discovered in the next 5 years but the exploitation of the value of the human genome may be exploited very rapidly. You have to think about intellectual property management, you have to think about the discoveries, defining the individual differences of the diseases, individual difference amongst ourselves, and it is really like a gold rush, in some companies it turns into a mid life crisis, others are moving very strongly, I hope we belong to the second kind, and certainly we have made aggressive investments in the last 3 years, which will be in the ball park of 2 billion dollars over only 5 years. So we are moving and the biggest part of our new technology platform is an IS platform. What we wanted to achieve is to combine all the knowledge of the world to very pragmatic user friendly paradigms for all scientists. Ideally what we wanted to have is a scientist come in in the morning, have an alert servo, click on the computer and find everything that he needs, what has been discovered last night, what new information is out there, so a true customer orientated business in the research lab. It was clear that in our foray search operations that we have in Europe, in Connecticut, in Kyoto and in Berkeley, that we could do it on our own. So we formed a joint venture with Lion Bio Sciences from Hydaberg in Germany, and we located this joint venture to Boston in the heart of gnomic if you wish, and this new unit would supply not only the pharmaceutical scientist, but on the right side you can see Agrikem, animal health care and diagnostics as customer of life sciences for this joint venture. There are about 180 internet based data pools in the world, speaking different languages, it's impossible to collect all this data in any company I believe. So you have to find ways to tie into a major technology provider and get all this information and then move them to the scientific labs. We are now in a major way, if you look at the middle part, dealing with detailed family analysis of genes with protio in functional prediction, prediction how key and locks fit together in way, and then SNP's, this is Single Nuclitite Pollen office and paradigms can be defined. So it think that has been the major platform at Bayer, taking advantage of the convergence of computers and genomic's in this age. Looking at the over all scheme of our discovery, this is only the IS platform, but of course we have to also really identify pills, and that starts generally with a composition of matter, a chemical substance or protein, and how do you find them, by screening. I showed you this slide which we developed a few years ago, which really shows you the driving force of pharmaceuticals in the last 5 years, which is the move from screening to high through put screening to ultra high through put screening, which has developed over the last years. The turning point was in 1990, when we used for the first time robots to do the work for us in the lab. At that time the chemist in our company got into trouble, because now it was no longer a limitation of how many substances you could test, but it was unlimited for any chemical you would find. In the meant time starting in the 90's where we could do 10 thousand tests per day it was a revolution, and today we are at 200 thousand per day, so for any given target we could be through in a few days. However nothing is that easy, the simple targets you can do that way, and then there are more complicated for which you do not find any answer of course, and I would estimate this as about 40%. So it's a mixed bag in a way, however this robot high through put system really pushes the paradigm into the future. So one would now be happy to know all these 30 thousand gene targets or 100 thousand gene targets, and this is of course the ways which is predictable to lead to innovative medicines in the near future. Now the chemists were challenged, and in 1990 also pushed by bio tech companies, for instance Pharma Copier and others, many companies developed now fast chemistries, so called combinatory chemistries, and again also here we are far beyond what was expectation only 5 or 10 years ago. A lab unit consisting of about 3 people, when I started at Bayer could make 300 compounds a year, combinatory chemistry they can now make 20 thousand compounds a year. Now given that all together the number of compounds made in pharmaceutical companies is extra ordinary. At Bayer 1996 we had in a bank 300 thousand individually purified compounds, we moved it up to 1 million compounds just last year, and in the mid term we believe we will have it in the ball park of 3 million compounds, it's not quite clear when we will achieve it, in 2 or 3 years in the estimate. So what you see here is a robot operated compound depository, our fort knox if you wish. This is the information, so chemistry has also turned from composition of matter to a composition of knowledge, in fact chemists are no longer seeing the chemist as compounds only, but as information feeding the discovery process, and that is really the biggest paradigm change that we have achieved, and with 3 million also we hope today in a ball park where we think that is enough, now we have to improve on diversity of chemical compounds and things like that. So these were the in . competencies that we built up and we thought chemistry is so important, first of all we are a chemical company and we should not let it take the butter off the bread of all our doing by the bio tech companies. So screening and chemistry are clearly our competencies. The question that we discussed 3 o 4 year s ago, could everything else be out sourced, and this was a question that was never really answered. We decided on a model in which we would tie in with a major bio tech company, in this case millennium to share the discovery process. As you can see here on the picture the front end producing qualified targets, which is DNA and protein, was the task of millennium, then the screening as I'sve just described, production of chemicals, the validation and then developing compounds is the task of Bayer. Now if you look at that everything looks simple, but the problem is still the validation, and the little mouse you see in the picture may become the most important future target of research, and understand that Silera which has published the human genome, is keeping all the mouse data in their own pockets, to discuss with pharmaceutical companies how one could share. There is a word of caution, whatever we share gives us a reduction of profit, but now there is the threat that with all these external collaborations, we could see up to 20% of our profit going away in terms of royalties and sharing innovation. So any model any deal that we do today is based on an intellectual property versus reduction of profit model, where we try to access what is the maximum that we could put lose. Now at that time this deal that we made was the largest ever in genomic's and I believe it still is and the biggest question was, what is a qualified target? Qualified target is a protein within a cell or outside of a cell, which lends itself for control of biological function with chemical proteins. But then the debate, and I was participating in the discussions, was always what does it mean qualified target. I'sd like to give you a little bit of a background of what we discussed. If you look into all the compounds that are moving into clinical trials today, you can classify the different pharmacologitry per actions and I know most of you are not scientists, and I don'st want to bore you with that. But just if you look at it more than 60% are known pharmacological targets which you could easily make drugs with. So in general those you would call a drugable targets, and then there are several other classes. The objective for the deal was to identify these drugable targets with also new mechanisms, but for which you could do very fast high through put screen. If it's not possible within in 4 weeks to develop a high through put screen compounds go back to millenium. Mark Leven is always proud that he gets 80% back, because he thinks that with a little bit more effort he could get very valuable things back. We believe through put, high through put and speed is everything that counts, so we move on. Well it sounds easy but 40% of all the targets that we have today which are highly promising and which are innovative, which have never been proposed before, all coming from the 30 thousand genes, are so called orphans, what does it mean? It means we have no idea what this receptor is for, we have no idea what this protein is for, and we don'st know how they work together. So we had to design a strategy in which we can de-orphanize the information, and it's like a completely new language, you have to found codes. And what was the basic driver here, was chemistry again, trying to use expression profiling, defining what cells produce and how disease mechanisms influence expression levels, and then developing with this high through put mode already chemicals to validate what's happening, for instance in animal models. So it's a whole new strategy which worth's any knowledge, should say worth's any knowledge, but it can live without any knowledge and the chemical validation to the end is possible. So what was the background? We have asked Millenium to give us is 6 therapeutic areas, 225 targets, and the duration of our deal was 5 years or is 5 years, and the total value was 565 million US dollars including 11% of equity participation, which turned out to be a very good, very wise investment in the meantime. This was and is the largest deal in the industry and we where very much put in question whether this all makes sense, and now it's 1 years into the deal and it's time to put everything on the table, and say where are we today. Well the collaboration is on track, we have more than expected turn out all together, we have received so many interesting targets that we have in fact increased this deal, and it is clear that in an explanation growth of information you have to push that part that is limited. So that's the genomic information, so we have been increasing this deal quite a bit. We have identified 2 projects for the development, so called strategic projects, I don'st want to go into the details of exactly what that means, but it means it's a valuable logical target for which we are improving the chemistry. One of the 2 so called eye cast project has made it into the clinical development stage in only 18 months, which it think is a record for this type of research, and we believe depends always how you define it that this eye cast molecule is the first one truly discovered along the path of gnomic's. How do you find it? Well you find it by screening clinical samples, and that is also done by Millenium to evaluate these targets, 50% is done by Bayer, 50% by Millenium, and you find that the expression is up to a thousand fold higher in clinical tumor tissues than in normal corresponding tissues. Then the molecule was converted by the technologies into a drug, and by using this new chemical we found that it stops growth of tumor tissue. The development of this compound is on going and we hope to also break all the records to bring it to the market. This is an example how genomic's translated into big pharma can be working and we are quite happy with this collaboration. Moving to a different field that will now be the target of future platforms, is pharma co genomic's and toxic co genomic's. Pharma co genomic's is of course a very hype issue at the moment and the belief is that in an ideal world you would find the right medicine for the right person at the right time, because today it is believed that most medicines do not work and respondarate of any medicine may be in the ball park of 30% average, so why treat so many people the wrong medicine if you can do it based on the gnomic profile. Well it is a very intriguing idea, and is supported by the FDA, by patients, by physicians. However it will of course break down the classical markets, because now the markets could shrink to a third or tenth depending on the genomic's pattern of an individual indication. So why should this be of the interest of the industry? Well if we don'st do it somebody will and so there's no choice, rather than to go along these routes and then also reduce significantly the costs of clinical trials, and getting faster and better drugs applications moving along these lines. Okay, theory in the next 2 years will know for sure. Toxic co genomic's is another issue, 30% of all the drugs die from entering the development stage to the clinic. So what if we could predict and compare human genes, mouse, red genes, dog genes, and predict what are toxicological problems early, very early in the discovery process so we could select and weave out the ones that we don'st have. So toxic co genomic's is today a major important information source for the scientist and it's just beginning. So we have decided this year to do another line which is in bio tech dollars 1,3 billion dollars, but I don'st want to talk about how you split it up and who's sharing what. In fact we are working with Curagen. out of Brand Fort Connecticut on 80 targets, but we are developing this new pharma co genomic and toxic co genomic paradigm for our company. In fact we believe that we have so many drugs coming out that we can'st even develop all together, so we are sharing with Curagen the cost down the road to develop drugs. So this deal and I'sve been and my colleagues have been approach by many other bio tech companies who have money in the pocket from last years run, who would suggest to us this is a good idea, why not we as bio tech companies can forward integrate faster if we tie up with the pharmaceutical company. So I think that is a model of this year where several companies in the bio tech field genomic's, will team up with companies like us to share the cost of development. My area in Berkeley California is protein therapeutics, where as the genes and the proteins you find are so called targets and these targets lend themselves for competition with chemical drugs, and it takes a while to optimize it. Proteins could be immediate drugs. So we are very much focused on producing these new therapy's, we are engineering proteins and we are making mono prone anti body's to proteins. So this has been a very interesting paradigm and we have built also a technology platform that I call when we built it, fit technology platform, getting fit for the new Millenium, getting fit for discovery and it meant flexible innovation technologies. So we have the same basic driver which is the genomic's information powered by Lion, our joint venture in Boston, and also powered by a big company inside who not only delivers us information on the genome, but also delivers us all the information and clones that is heart wear directly. So we receive all the clones that inside has and has patterned it directly, you can work them to proteins in our own shop, because that's out core competency, and we are screening them with ephmetrics. We have developed specific chips based on Lion, based on inside, so we have Bayer customer made chips, so we are sure that we are a little ahead. Then we are now focusing on protiomic's, that means genome independent discovery picking proteins out of the different cellular components. We also formed a joint venture with Mophosis, which has a combinator anti body library, 10 to the 10 anti body's in which we can pick individual human anti body's against these proteins to develop new drugs. So it's a fantastic concert, like the symphony orchestra of technologies, and I think we are making extremely great progress. Now I described to the chemical small molecule paradigm and the protein therapeutics paradigm. For the small molecule Bayer has 3 research centers in Germany, in Connecticut and Kyoto, and for proteins we have 2 areas, the world center in Berkeley California, and another one for production technologies in Germany. Looking again at the small molecules, you see here how this whole paradigm evolved. If you look at the technology area, genomic's, function genomic's, protio mix, combinator chemistry, then the development of drug candidates test and then moving into human clinical trials, we have a major support platform of various companies that help us. You can see here that the basic area of pharma co genomic's leading into the clinic is also covered by Curagen. So we believe exploiting the human genome, we are well established now, we'sve made tremendous investments and we are moving on. Every of these functions is high through put, if not ultra high through put, that has really been the major change of everything. In the bio pharmaceutical part, proteins, therapy's, these light yellow orange ones are different paradigms, you can see here the expression profiling, the ephmetrics technology protein engineering and bio tech process development are differing, and the platform members are some what different. But as the saying is true, high through put, ultra high through put, rapid development. So these are the 2 new evolving drug discovery engines in our world. If we look back into the general strategy, what did we have 5 years ago? Five years ago we had a discovery process which took about 5 years from the idea of an area to getting the first lead compound. We had very few targets, we had to spend a lot of time in target validation, and then sequentially on SA development screening it and developing lead structure, all sequentially at that time we thought it's high through put but now we know it better. Today we do it differently, today we believe we can do the same drug discovery process with many more drugs, with many more targets at the same time in 3 to 3 years in average. We believe that we can manage in the next 5 years up to 100 thousand protein targets and we believe that this will lead to a productivity increase of up to 50% if you look at the out put per person in our discovery organization. So we move from a sequential process to a parallel process with the idea to make everything faster and more efficient and with a higher quality. This translate into new business objectives that any board of any company in this world loves to hear. Let me remind you that what I said at the beginning that if you ask the pharmaceutical company's 5 years ago, we should have 80 drugs every year into the market and not 40, so how can we be so confident we can make it. So we drew a line and said okay to our board we promise we will deliver and we are now in the year 2001, and in fact we could deliver, in fact do one more drug every year than we promised, and we will make also the 13th in this year, because we have 10 already. So we are on the path I believe where significant investment, the right strategies, and high through put every where has moved us forward into a higher productive research world. However we make so many drugs that it will now cost a fortune to develop them. This is of course what we are presently working on. Just a few figures for those of you who are interested, R&D has a percent of sales, the industry average is in the ball park from 18 to 20% depends where you make the cut top 20 companies or so. Research has a percent of R&D at Bayer today in the pharmaceutical areas 40%, it used to be 25%, so in the last 5 years it has increased quite substantially. But most importantly we have increased the external portion of our research, which was only 12% 5 years ago, to 26%, so the process of sharing technologies, building technologies, out sourcing the up stream of our research has become reality. The bio R&D strategy has significantly changed the company, it has significantly increased productivity, making drug discovery faster, getting not only a jump start on new technologies, but be a leader in this field, developing greater potential for truly innovative drugs, being first in class, and the possibility to making better decisions in the future based on better knowledge and information that we have through our integrated IS platform. With that I'sd like to stop here, and thank you very much for your attention, not only we but also Leman Brothers say this year Bayer is one of the best research platforms in the world and certainly we are proud of that. Thank you for your attention.