Q&A: Dr. Felix Frueh, FDA Center for Drug Evaluation and Regulation

In April 2005, the FDA released a concept paper that outlined specific strategies for the codevelopment of drugs and diagnostic devices. In the first quarter of this year, the agency expects to release a codevelopment guidance document that incorporates feedback from the pharmaceutical industry. Recently, Executive Editor Randall C Willis spoke with one of the authors of the concept paper, Dr. Felix Frueh, associate director of genomics at the Center for Drug Evaluation and Regulation (CDER).

DDN: What prompted the creation of the concept paper?

Frueh: Thinking about pharmacogenomics and pharmacogenetics, it is pretty obvious to anyone that drug-diagnostic codevelopment is only going to happen if you have something to test for. For example, if you are defining a subpopulation that responds to treatment or is exposed to a risk, you need to have some means to identify and characterize that subpopulation. You need to find a test that allows you to group the people. If you want to have that associated with a drug, you need to co-develop the drug and the test. What we're really talking about are projects where a test is required for the use of the drug, as is the case for Herceptin.

DDN: What are the advantages of codevelopment?

Frueh: I believe that the market is changing. Companies are realizing that drugs really are effective only for maybe half of the people who are actually taking them. What happens next is a trial-and-error situation where you lower or raise the dose if it doesn't work, or you change the drug. But the treatment is definitely not as good as it could be.

The realization that genomics could be the critical factor in how people react to drugs is here, and I think that we're at the brink of this knowledge being used in drug development for a variety of reasons, including that you want your efficacy signal to be bigger and you want to avoid toxicity. These, in my mind, are the two major drivers that make it very obvious why I would want to use pharmacogenomics.

So, I think that the advantages of co-development in that sense are finding the best possible strategy to have a tool ready at the time that the drug is getting approved so that you don't have any delay. It's not all that special, in that sense. It's just the coordination, the strategic approach to development in a time when perhaps it's not 100% certain that you have the exact biomarker. During drug development, you're developing not only the drug but also the biomarker itself, so that poses a little bit of a challenge for developing the tool that ultimately is going to be used as a test.

DDN: The timelines involved in developing a drug (10-15 y) and an assay (1-3 y), however, are out of sync.

Frueh: We've received a lot of comments on this point. My view is that if we don't try to achieve was seems to be the distant future today, then we're not really going to change the way that we're doing business today.

I absolutely agree that there is a discrepancy between the development times. Someone once told me that the half-life of a device before it becomes updated and improved is 18 months. Naturally, that poses a challenge on the strategy of co-development. However, I think it also illustrates the flexibility of the device industry.

I feel it's very likely that because the device development time is much shorter than the drug development time, you're actually going to see device improvements over the period of time that the drug is being developed and very likely after the drug is on the market. We've seen that with Herceptin.

DDN: Will product regulation across FDA centers present any special challenges?

Frueh: It will pose a challenge, but we're preparing ourselves and are well prepared today. The interdisciplinary pharmacogenomics reading group, which was created to review voluntary genomic data submissions, is actually an FDA-wide organization with representatives from CDER, CBER, OCP, and CDHR, so we're already working very closely. In fact, the two people who have been the driving force in getting the concept paper written are Larry Lesko, director of the Office of Clinical Pharmacology and Biopharmaceutics at CDER, and Steve Gottman, director of the Office of In Vitro Diagnostics in CDHR. So you can see already the duality in the approach to actually getting these things done.

DDN: Do you expect the approval process for codeveloped products be more or less complicated than for individual products?

Frueh: I'm an optimist, so I think it gets easier simply because one thing you have to demonstrate for market approval for a device is clinical utility. Well, if you use the device in a clinical setting such as a clinical trial, you're done. There is nothing better than having that compelling evidence that you need for getting a drug on the market also associated with the device itself.

Vice versa, if the device works well, you're probably going to be able to generate better clinical data that is going to help you get the drug approved. So I see it as a synergistic effect—again, if you get the strategy right—and that's one of the reasons we wrote the concept paper and are in the process of writing it into a draft guidance document.

DDN: How has the industry responded to the concept paper to date?

The response was generally very positive, but it was also very critical. It wasn't our intention to something that we know is not the optimal situation. We don't want to encourage something that we believe can be improved. One thing that these documents do is stimulate very good discussion. What we really very much appreciate is the feedback that we get from the stakeholders because that is what tells us where we stand and that's critically important.

Of course, that has implications. For example, do we require that a test be approved before Phase 3 trials? So we're getting into more details about this, but the bottom line for me is that I feel that the time is right to get a guidance document out. It's not designed to guide existing practices, but rather to point to something new and hope that this is going to be adopted in a successful way. We don't have any control over whether or not the industry is actually going to use that information—it's not a binding document—rather it's what we think could help in the development of these products.