The New Genetics: Discovering Innovative Medicines for the 21st Century

Dr. Allen Roses -President and Worldwide Director, Genetic Directorate, Glaxo Wellcome Inc. NEW GENETICS: DISCOVERING INNOVATIVE MEDICINES FOR THE 21 ST CENTURY Chairman: Robert J. Dechert President, The Empire Club of Canada

Head Table Guests

David McFadden, QC, Partner, Smith Lyons and a Director, The Empire: Club of Canada; Julian Kirk-Elleker, OAC Student, Ursula Franklin Academy; Rabbi Baruch Frydman-Kohl, Senior Rabbi, Beth Tzedec Congregation; Dr. Heather Monroe-Blum, Vice-President, Research and International Relations, University of Toronto; Dr. David B. Shindler, President and CEO, Milestone Media Corporation; Dr. Manuel Buchwald, Chief of Research and Director, Sick Children's Institute; Ken Shaw, National Editor, CFTO Television and a Director, The Empire Club of Canada; Dr. Alan Bernstein, Director, Samuel Lunenfeldt Research Institute, University of Toronto; Dr. Michael Levy, Senior Vice-President, Research and Development, Glaxo Wellcome Inc.; Dr. Bruce Archibald, Assistant Deputy Minister, Ministry of Science and Technology; and Paul N. Lucas, President and CEO, Glaxo Wellcome Inc.

Introduction by Robert J. Dechert

Ladies and gentlemen, many of you will recall that a little more than a year ago the Chairman of Glaxo Wellcome Inc., Sir Richard Sykes, in his address to the Empire Club, described his company's involvement in the Human Genome Project and the many benefits that humankind could expect from genetic research.

Hardly a week passes without the announcement of the discovery of a gene linked to one of medical science's most perplexing diseases. With each new discovery comes the hope that those susceptible to a particular disease can be identified, treated and with the development of gene therapy potentially even cured. '

We are very privileged to have with us today one of the world's leading experts on genetic research.

Before joining Glaxo Wellcome in 1998, Dr. Allen Roses was associated with Duke University for more than 20 years as a senior research scientist. He was Chief of the Division of Neurology from 1977 to 1997, Director of the Joseph and Kathleen Bryan Alzheimer's Disease Research Centre from 1984 to 1997 and the Director of the Centre for Human Genetics from 1996 to 1997.

Dr. Roses was one of the first clinical neurologists to apply molecular genetic strategies to neurological diseases. During the early years of recombinant DNA research, Dr. Roses initiated several pioneering studies. His laboratory reported the chromosomal location for more than 15 diseases, including Lou Gehrig's disease. A multi-disciplinary team of Duke University geneticists led by Dr. Roses discovered a susceptibility gene which lowers the age of onset and increases the risk of Alzheimer's disease.

For his contribution to genetic research, Dr. Roses has received numerous prestigious awards including the Metropolitan-Life Award for Alzheimer's disease research and the American Academy of Neurology Potamkin Award for Alzheimer's disease research.

Dr. Roses is currently a member of the Glaxo Wellcome Research and Development Executive Committee and heads the Genetic Directorate. His current research involves genetic strategies for susceptibility gene discovery with respect to Alzheimer's disease and other neurological diseases.

Ladies and gentlemen, please help me in welcoming Dr. Allen Roses to the podium of The Empire Club of Canada.

Allen Roses

Thank you very much. It is an honour to be here. It is not often that I try to explain genetics to a primarily non-scientific audience.

What I would like to try to do today is explain some of the things that are going to happen in medicine in the next five to 10 years. Our primary customer is the patient and the most important part of our work is getting the right medicine to the right patient at the right time for an appropriate cost.

At Glaxo Wellcome I have had the opportunity to see things in a very different manner than I did when I was at the university for 27 years. The opportunity to actually affect medicine development and affect the way government and industry can work together to improve medicine development has been maximised by my being at Glaxo Wellcome.

Many of you probably read a couple of weeks ago (and maybe didn't quite understand its significance) an announcement about the SNP consortium which is a group that has as its founding members the Wellcome Trust, the largest charity in Great Britain for medical causes, and 10 contributing pharmaceutical companies sponsoring research in four genome centres to make a map that would be useful for pharmacogenetic purposes. I am going to explain more about that but it is probably the first time that companies that compete at the end product have declared that certain things like the human genome belong in the public domain. The human genome is not private property and we should treat it in an ethical way in order to be able to get the right drugs for the right patients.

One of the things you should know about the pharmaceutical industry (which I learned when I came into the industry and didn't really realise when I was an academic, even a seasoned academic) is that of 100 molecules that

go into development only 10 come out and of those 10 only three are profitable. Of those three only one pays for all of them. There is a blockbuster mentality in the industry and it is a real one because only by maintaining that output can an industry survive.

One of the things that genetics can do is find out what underlying genes are responsible for some of the diseases that we get; not cause a disease but give a propensity or a susceptibility for it.

Most of the genetics work that you have been hearing about or reading about in the newspapers for the past 20 years has identified genes causing monogenic diseases like cystic fibrosis or Huntington's disease. Most of the people who have these mutations actually get the disease sometime in their life.

There are a number of other genes called susceptibility genes. With these you don't necessarily get the disease but have more susceptibility or less susceptibility for the disease depending on which gene flavour you inherit from your father or your mother. Multiple genes contribute to the actual development of these diseases.

This is a different kind of genetics but lay people understand it much better than scientists or physicians. We know cancer runs in some families. We know Alzheimer's disease and heart disease runs in some families. These are not monogenic-inherited diseases but some families have a greater susceptibility for these diseases.

With the genome project sequencing the whole genome, we now have a chance to study the hundred thousand plus genes that are there. As a company we are interested in the 30 or 50 genes that are most likely to help us find targets that we can screen to develop medicines related to certain diseases. Rather than try to own the whole human genome we're trying to hasten the time taken to find what those particular genes are and work on them for disease targeting. At the present time targets are basically like hypotheses were at the university. People believe in them but their beliefs could be wrong. They can have great arguments for them but many of those arguments can be fairy tales. If we can get genetically relevant targets we can get a better hit rate and from that hit rate get more drugs that are going to help in more diseases.

In order to do that Glaxo Wellcome started a programme which is I think the only one in the industry. We took half the genetics budget and decided to place it in academia. We didn't think that companies should be owning DNA from people. We didn't think that companies should have access, direct access, to Mrs. Smith's medical record. There was a type of research that could be done in universities that we needed to be able to access but we didn't need to have the data sitting in the company. We formed what are called genetic networks, Glaxo Wellcome genetic networks.

We have three components to these networks. We got together the best academics worldwide that had an interest in accessing patients and their families. We got together a few laboratories set up as screening laboratories to do all the DNA screening for the networks. We took about 10 of the 15 people in the world who are really super at genetic epidemiology and made them non-exclusive consultants to do these projects for us. We provided support and we provided money to get this done, to centralise some of the things and to make sure that everybody was relatively happy and that all academic demands that ought to be in the system were adhered to.

From my point of view it's the perfect system that I would've wanted if I were still at Duke University. I have access to my own data. I have access to my own DNA. I can do anything I want with it except do a linkage study. The linkage study is what the company will do with this to narrow down the place on the human genome where the genes for a particular disease are found. We're using something called a SNP map. A SNP map is basically a single nucleotide polymorph, a variation, something that we can experimentally tell apart. This third generation genetic mapping promises a means of targeting the right medicine to the right patient, truly individualising treatment.

The big problem that people have once they get linkage information is they don't know how to narrow it down with current techniques. They use the candidate gene approach. They pick a gene. They think it is important and make up all the reasons in the world why they think it's important. They submit a grant, get paid for it, study the gene and it's not the gene. Or even worse they report it as the gene and somebody or a bunch of people say it isn't. Most of the reports you've read for instance about Alzheimer's disease recently are from genes that have not been confirmed and confirmation is the name of the game.

Once we get these linkage areas using a very fine, very dense map, we can find a small area that is easier to work with to find genes. We've done that. I'm not going to tell you about that but it works.

A SNP map is like a fingerprint or a profile. Using a SNP map you can measure people who have some sort of clinical response and compare them to people who don't get the clinical response and start targetting medicines. What do I mean by that?

There are really two things to look for--effective drugs and adverse events. We do about 20 per cent of our clinical studies here in Canada. It's a very important part of what we do. We had 20-per-cent responders in a phase 2 study, an early study where we just looked for response. One hundred people responded and 400 people didn't respond, We can take the DNA of each of those two groups, do a profile, compare the profile and see if we can get specific regions mapped in the people who responded so we can abstract those regions into a much smaller test and be able to pick out people who are most likely to respond to the medicine. Then when we do our clinical trials we would not subject as many people to the medicine. The trials would be smaller, faster and cheaper.

That's going to require some proof of concept and what. we're doing in proof of concept is something totally different. Other companies are doing the same thing. We have a drug that is an incredibly good anti-convulsant for epilepsy. We've got a problem with that drug; not a big problem but a small type of problem that keeps a very good drug from being a blockbuster. As a neurologist I can tell you that that drug is probably the best drug for certain forms of epilepsy that are commonly found in the population with fewer side effects than many of the other drugs. It is commonly used but why isn't it the blockbuster that everybody uses right off the bat?

It is because about 2 per cent of patients get an itchy skin rash. If you're a doctor and you see a lot of people with epilepsy you certainly don't want 2 per cent of them calling on the phone questioning what kind of doctor you are because they've got this itchy skin rash. The drug gets used in a dose escalation by experts but it doesn't necessarily get used by the general doctors who don't want to get involved in dose escalation or don't want to hear that they are not a good doctor for giving someone a rash. What are we going to do?

Using the 100 people who got the skin rash and the 200 people who took the drug but didn't get the skin rash, a SNP map from the SNP consortium is obtained and can be used to determine the profile of people who get the skin rash. From a blood sample it can be determined that you are not going to get the skin rash and can take the drug. You avoid dose escalation, you avoid the nasty telephone calls in the middle of the night, and you become a happy user of a very good and otherwise safe drug.

The consortium is working to develop and make the SNP map freely available. We will have a SNP map that's useful in less than two years. We are developing ways of reading the map in order to construct profiles. If we can bring individualised medicine, customised medicine, safely and efficaciously to the population the pharmaceutical industry is doing something good and I'm very very happy that Glaxo Wellcome is at the head of that movement.

Thank you very much.

The appreciation of the meeting was expressed by Ken Shaw, National Editor, CFTO Television and a Director, The Empire Club of Canada.