Life science players have long complained about how big pharma’s bureaucracy hinders the pace of drug innovation. Last week, Duke University launched the Duke Drug Discovery Institute, a “new model designed to fill the void between academic drug discovery and translational medicine.”
The institute, funded by philanthropic support, private sponsors, federal grants, and research agreements with pharma and biotech companies, will work on projects focused on specific molecules. The institute will be administratively aligned with Duke’s Institute for Genome Sciences and Policy.
At the institute’s helm will be former GlaxoSmithKline pharmacogenomics maestro Allen Roses, who describes the institute as a “virtual pharma” that will harvest the best academic minds and industry experts to advance pharmacogenetic innovations.
“This first-of-its kind institute will be a new approach to the development of drugs by spanning early molecular discovery to proof-of-concept clinical trials,” said Roses, who after a long career at Duke left academia in 1997 to become senior VP of genetics research and pharmacogenetics at GlaxoSmithKline.
Roses, who left GSK last year to return to Duke to run the drug discovery institute, spoke to Pharmacogenomics Reporter recently about how this novel research and development model will advance molecularly-targeted medicines. Below is an edited transcript of this interview.
Why did you leave GSK?
There was a change in direction after Tachi [Tadataka] Yamada, [former GSK chairman of R&D,] left and the kinds of things I’ve been interested in doing became less of a focus for R&D. There is a lot of in-licensing. There is a set amount of money for R&D [at pharmaceutical companies] and how they use it varies. So, what GSK had decided to do is in-license molecules and projects that had a short timeline, a much shorter timeline than some of the things that were being done in the area of genetics. … They would like to get a molecule that starts making money in two to three years, as opposed to waiting for the five to seven years it takes to go down the pipeline. … It’s nothing sinister. It was a choice that was made. And they have a right to make that choice.
At your time in GSK, what kind of commitment was there at the company to pursue pharmacogenetic research?
GSK was more committed to pharmacogenetic projects in the kind of things they were doing before the decision to focus on later-stage projects. … The decision wasn’t so much to leave GSK. The decision was to go to something that was much more exciting.
So, what will you be in charge of at Duke?
I’m going to be director of the Duke Drug Discovery Institute. It’s essentially a virtual pharma with an executive team that will be made up of folks that have had industry experience across the pipeline from exploratory through proof-of-concept. The operation of it will be to outsource virtually every step, be it to a university or a biotech.
Can you elaborate on this idea of creating a ‘virtual pharma’?
The way a pharma R&D executive works, in a big pharma they have a lot of money and below them a huge bureaucracy. The way this is going to work is much more compact. The executive team is knowledgeable of the process from exploratory to discovery to proof-of-concept, and includes biologists, people with pre-clinical specialties, and chemists. We’ll discuss each project while the project might be temporarily in one of those groups, … basically the expert in that group determines where to outsource the next step in the project plan. The project timelines will be dependent on where we outsource it to get it finished in a reasonable timeline.
We would not have an expensive center of laboratories, or a center for bureaucracy. I believe that on the whole there is a need that is obvious to people familiar with the pharmaceutical industry. It is not being brash to say that on the whole the pharmaceutical industry has not been as productive as people expected that they would be at this time. Especially with the genome being available and the hype and things that went with it.
Academia does wonderful exploratory discovery, which is generally featured in magazines like Science or Nature. But then they get a grant with which they go on for years of basically studying that step but not going on to the next step. The reason academia didn’t go on to the next steps is that 10 years ago, there were no commercially available chemical libraries that could be screened, and the failure rate in early drug discovery is so high that no one could build an academic career that way.
So, there was no real reason to go on to the next step. They could expect that they would make a deal with a big pharma who would immediately embrace their ideas, or they could make a deal with a biotech company. In many cases, form a biotech with venture capital, which is time-based. They have two or three years to really do a seven-year project to get to its proof-of-concept. So that doesn’t work as well as it should.
So you have academia, biotech, and big pharma having difficulty. And it just seemed to me that if you take the brain power that’s at the universities and the know-how across the pipeline that’s present in big pharma, if you could put those together, then the only thing that would be missing is the money to do the projects.
Where is the money coming from?
Right now we have some star-tup money … to pursue some projects that we’re not able to discuss quite yet.
What disease areas will you be focusing on?
We’ll be focusing on major disease areas. One area of focus will be Alzheimer’s disease. Another area where we’d be interested in doing studies is with certain companies that have reasonably good molecules but they don’t know how to apply pharmacogenetics to accelerate their ability to get a proof-of-concept performed.
There are a lot of people who are consulting on pharmacogenetics. But what this group will be doing is bringing the experience of having done these experiments over the last six years and having results to be able to apply them to the molecule to come through this pipeline.
Will you be developing any diagnostics in your PGx efforts?
Part of this is to lay the groundwork for doing companion diagnostics. Now, companion diagnostics are really done during [drug] development. This is only going to go through proof-of-concept within the university.
The companion diagnostic piece, the way we envision this happening, and we’re a long way away from that at present, is that the molecules will be picked up by a big pharma. But not every big pharma is going to have a huge pharmacogenetics group. So, what we have established is a consultation company. In the university you are allowed 20 percent of your time to do consultations. So, the members of this team will also be involved for 20 percent of their time in a company. Their job will be to consult, assist, design, and manage specific projects during development for other companies.
At what stage is this consultation company?
Well, it has a name and it has one employee, president, CEO, secretary, dishwasher. What we’ve registered for this nascent company is the name Cabernet Pharmaceuticals. The institute is a drug-discovery institute. So, we’ve separated R&D. The consultation company will involve certain individuals from the institute and other high-profile individuals.
The way this would be ideally worked out is, a big company has a molecule and they want to see if they can use pharmacogenetics to develop a companion diagnostic in time for the registration of that molecule. They would work with us through the design, understand what they need. They would set it up but we would have project managers working with their pharmacogenetic internal person, and with the project team, so that our experience in doing this over the last five years can be transferred.
Are Rx/Dx partnerships the way to advance pharmacogenetic tests?
Rx/Dx partnerships are happening a lot. But there are ways to go about it. There are companies that can build you a test. There are companies that can give you lists of potential biomarkers. One of the things that is important is, how do you make the test simultaneous with the registration of a compound? If you look at all the examples out there, if you take [AstraZeneca’s oncologic] Iressa, it’s not a generalized third-line treatment. It’s a damn good treatment for a very specific, but small bunch of people.
Simultaneously, a year later, two academic groups found it. So, from a company’s point of view, if you get a third-line treatment approved, you get third-line pricing. When you get a pharmacogenetic test, it says, ‘Here is a great drug, for these specific people.’ And you can get premium pricing. But from AstraZeneca’s point of view, what they got was a smaller market for a third-line price. They can’t change the price.
So it’s very, very important to have your companion diagnostic be part of your development as opposed to being presented to you after development. It’s economically important. And that’s what companies should strive to do.