At A Glance
Name: Lawrence Lesko
Position: Chair of the FDA’s Pharmacogenomics Working Group; Director of the Office of Clinical Pharmacology and Biopharmaceuticals, CDER, FDA, since 1995
Background: After serving as an associate professor of pharmaceutics at the University of Maryland at Baltimore between 1981 and 1988, Lesko was named vice president of PharmaKinetics Labs, a position he held for three years.
When it comes to understanding pharmacogenomics, Lawrence Lesko, the FDA’s personalized-medicine go-to guy, thinks the agency is finally getting some traction.
After organizing a host of in-house seminars and lectures to help his staff get a handle on the role of pharmacogenomics in drug discovery and development, Lesko, who oversees the FDA’s 15-person Pharmacogenomics Working Group, expects the agency to roll out an industry guidance paper within three months.
In fact, Lesko may have an ally at the top: The FDA’s new commissioner, Mark McClellan, is a documented fan of new drug-development technology, including genomics and genotyping tools, and appears eager to put them and their data to use. Yet the biggest question for regulators and industry remains ‘How?’
SNPtech Reporter caught up with Lesko recently for a chat ...
It’s been a couple of years since the FDA created the Pharmacogenomics Working Group. What is the agency’s current position on existing genotyping and gene-expression technologies from the standpoint of drug discovery and development?
We’ve been attempting to encourage industry to adopt these technologies in drug development because we see the potential impact that they can have on regulatory decisions that influence efficacy and safety. And to the extent that new technologies, new tools, new study designs can reduce the drug-related adverse actions … we’re all for it.
But we’re also very realistic. I think we understand the stage of development of these technologies, and there’s been some promising things happening. We’re pretty excited about the Glaxo study … on [its AIDS drug] Ziagen and the use of SNP technology to identify markers that would predict toxicity. And it’s that type of advancement that I think has a lot of potential. …
Has your office begun looking at specific technologies and judging them based on their potential value in drug discovery and development?
We’re familiar with [the cur-rently marketed] technologies and we’ve set up a series of seminars to have these companies come in and talk about their technologies and tools so that we learn about them and hear where they are in their development stage and how they’re being used. … I think we have to be on top of that.
Does a SNP-genotyping or gene-expression technology exist today that you believe has an edge over competing technologies and would therefore have greater weight in the eyes of the FDA?
It’s hard to say yes. When we’ve seen, for example, microarray technology applied in an [Investigational New Drug application] there are no standards for submitting that data to the FDA. The submission ranges from pages and pages of numbers that we can’t really interpret to very brief three-page summaries that we also can’t interpret. So I don’t think we have seen, to be honest, a good submission of technology that deals with SNPs or microarray gene expression that would allow us to make an assessment of the technology and what provides advantages and disadvantages.
I think it’s too early in the game. This stuff is really changing rapidly … and I can’t really say where it’s all going to go. I think that we have to wait and see.
So at this point the FDA does not believe one technology or platform has an edge over another.
That’s right. And if anybody did say that, it would be strictly a personal opinion. However, we’d like to get to something like that. And one can imagine downstream a guidance for industry that maybe makes statements like that. But we have to see submission to the FDA. Quite frankly, companies involved in this research have been reluctant to submit this information to the FDA — or even to use the technology for fear of having to submit everything to the FDA.
So I think one of our challenges … is to create a process for companies to use and submit information that would give insight into the technology. And we actually are doing this. In fact, our safe harbor policy allows companies to submit exploratory genetic and genomic data that would not be part of the traditional regulatory decision-making on applications.
Do you foresee the FDA creating a guidance for industry?
We think a guidance for industry would be helpful, and we realize it is important to have a guidance that is general enough to allow the technology to grow and yet be able to say something to encourage firms to use and apply the information.
Can you say when the FDA might publish such a guidance?
I would say our plan is to have a document out for public discussion in the first half of this year. … I think the first milestone is to establish a draft policy for submitting genetic/genomic information into the safe harbor. We discussed safe harbor at the [2002 Workshop on Pharmacogenomics in Drug Development and Regula-tory Decision-Making, in May] as an idea and we’ve been meeting with pharma and internally to figure out what that means.
In other words, ‘What is safe harbor genetic/ genomic data? How would it be submitted? What would be the process for sharing the information publicly? What happens when we learn something from the data? How does that transition to a real regulatory tool?’
So I think the first goal of 2003 is to establish an FDA perspective on safe harbor as a draft position that could then be discussed publicly with the industry to arrive at a final position. I think running parallel with that is another milestone to develop a draft guidance for public discussion and comment by the end of the first quarter of 2003. It will be very interesting.
What are some of the hurdles that the SNP- and pharmacogenomics-research have to overcome?
I don’t think it’s so much hurdles. I don’t really believe there are technology challenges for companies to do what they want in drug development. I think the Glaxos, the Genaissances, the Variagenics’ [to be acquired by Hyseq] have the capacity to do the SNP analysis or haplotype analysis, and that doesn’t seem to be a big issue. There is a cost involved, obviously, in doing it but that cost is coming down very quickly over time.
I think the challenge is more in the area of study design. If you’re going to use genetics to predict efficacy or safety, what’s the appropriate randomization, numbers — that sort of thing. Or if you have a genetic or SNP test, what is the level of validation you need to apply it to make a claim related to safety or efficacy or drug dosing? I think that’s a challenge. I don’t think that’s well-defined.
We are really in a transitional phase where the challenges for industry to me are more questions about how regulatory agencies will see this technology when it’s applied. It’s just so new. There’s no standards to speak of. When you have gene-array data, how do you transition that to a functional analysis of what those upregulations or downregulations of genes mean?
Does the FDA plan to hire additional staff or specially train existing staff in order to help industry answer these questions?
I believe we need the strength in the review ranks to understand the information and not inappropriately apply it. Our approach has been multi. Number one, we’ve done a fair amount of continuing education here by having speakers who know this stuff come in and talk to the review staff in a seminar format. …
Secondly, we’ve conducted an internal course called Academics for CDER … where we brought in people from industry and academia to be lecturers. The third thing is, looking into the future, we have to begin to hire additional people who have the skill sets we want. The real action, I think, will happen when an actual submission comes in. You can only do so much preparing for something. I think we’re going to mobilize very quickly when something does finally come in. I know we have several good experts in the agency that have published in these areas, but if in fact this area takes off we’re going to need many more.
My position — and I think the public will demand it — is that we have to move away from the one-dose-fits-all paradigm. And the way to identify an optimal dose is to have a tool to identify the patient who should get it. And that’s on our doorstep.