Aled Edwards

Drug discovery requires a new paradigm

Editor's Note: As drug development costs rapidly spiral upwards, the number of new drug approvals continues to drop. The pharmaceutical industry is closing many of its North American R&D clusters and moving to lower-cost jurisdictions. With the global cost of pharmaceutical R&D estimated at $250 billion ($50 billion industry, $200 billion government and academic), the current system for drug discovery is inefficient. In the face of mounting resistance to the cost of medicines and concern over the loss of R&D capacity, some are arguing for a new paradigm and totally new approach. Dr Aled Edwards is one of these people. The president and CEO of the international Structural Genomics Consortium spoke at The R&D Management Conference in Ottawa on June 18 and outlined a new public-private partnership for the discovery process. The following is an edited transcript.

By Dr Aled Edwards

The current economic structure for drugs does not work and it affects all of us. Society wants new medicines. The problem is they can't be made by outsourcing novel R&D mechanisms. It has to be science to focus on the problem. How do you go about tackling these kinds of problems. The other problem from left field is the genome. The human genome has 23,000 genes but most of industry is focused on 300. If you want to understand all that scientific complexity, we need to understand how the genes and the proteins that they encode work. You need to have an uber-understanding of all of this to be able to move on. Professors usually, up until the year 2000, worked on one gene at a time. And industry worked on this small sub-set of proteins in the human genome. We've developed our whole scientific enterprise to look at one protein at a time and ask brilliantly focused questions.

The genome age is a different way of asking the question. Instead of asking one question, you ask ‘how do I make that experiment work on 1,000 proteins'? You have massive parallelization of the research and you ask questions differently. The combined problem is, we want the medicines but the scale and scope is too large for any sector or any group to manage. The current way of making medicines is going to go the way of the dinosaur.

So what can we do? The first gut reaction by everybody is to blame somebody else for the problem. But we've got to stop whining. All participants are acting in the best interests. We set up this system to work this way. The drug discovery process has evolved this way where everybody knows their role and that doesn't work. We need to set up a new equilibrium. And the onus is on us – the public – to do something. We pay for most of the research anyway.

I can actually do something about this because I'm a scientist and I direct a very large organization in this area. We need to form research partnerships. Industry has a whole lot of experience in regulation and pharmacology that over the past 30 years moved away from academia and into the industrial sector. So we can't do it and industry doesn't have the money. Their budgets are smaller compared to ours and they've got to focus on staying alive for the next five years, let alone do R&D. If we call it a partnership, we've got to fund it and have skin in the game. And we've got to operate it in academic institutions.

The focus on intellectual property (IP) is killing us. We have a fragmented IP landscape and we cannot do experiments without freedom-to-operate searches up the wazoo. The concept is to tackle it as public-private partnerships that operate within these creative commons.

At the Structural Genomics Consortium, we put structures – three dimensional shapes of human proteins - in the public domain. When you're doing drug design, it can take 18 months off the process of developing a medicine. For a $1-billion drug, that's a lot of money. We're a charity and when our consortium members put in their money, they get to tell me what I have to do. The oversight is to a board of directors so it's really structured like a corporation. The stuff goes into the public domain without restriction. There's no prior access to anything. Our industrial sponsors don't now what we do until the day it's in the public domain and everybody sees it at the same time.

How did we do in the first four years? We got 5% of the world's structured biology budget and we produced about 20% of the world's human structures. We've dropped the cost per structure to about one third of what it is in academia and one 10th of what it is in industry. We've published a lot and we have a lot of collaboration.

We're a public-private partnership and we have another four years of funding to 2011 at $30 million a year. The funders put in their money and gave us a list of proteins to focus on. Deliver 560 structures on that list over the next four years. We're bang on target and will do that quite easily.

We write the papers long after the coordinates are available. Customarily between the time you know the answer and the time the world sees it is about 18 months. It's almost criminal. We are trying to show that by putting the data out there, you don't screw your chances of publishing in the top journals and we're continuing to push that story.

We're going to target to develop probes. For a human protein, you don't know what it does. You have to do an experiment. You need reagents to probe the question. Customarily these things are protected because people think they're of great value. We're going to take the opposite approach and we're going to develop small molecule inhibitors of these proteins and make them freely available.

And hopefully by putting these into the academic world — that $200 billion of research — the knowledge of these proteins will go up, we'll understand the physiology better and our attrition is going to go down and more drugs will make it through.

We'll develop these inhibitors by collaborating with industry and put them in the public domain. So what we're doing – and GSK (GlaxoSmithKline) is committing eight chemists to this – is make models for use in society and make them free. They realize that these molecules will do far more for them in the public domain than they would sequestered inside GSK that doesn't have the resources. It's a way of getting the academic world to work for industry or with industry or with all of us to create it.

The trouble with industry — the pharma sector — is it takes eight years before you know if it has commercial value whereas in the software sector it's very quick. Patents have gone the way of the dinosaur largely in that sector — instead it's mobility, quickness to market, marketing. We need to find a new way of doing it in the pharmaceutical sector. We have to get out of the idea that industry develops drugs. The academic world has to get involved. Even though we don't have the expertise, we'll need to create new ways.