NEW YORK, April 13 - Recently, a number of genomics companies have announced bold new initiatives to use the genomics picks and shovels they market for their own drug discovery efforts.
But these companies may hit a drug discovery brick wall if they do not learn how to successfully negotiate the Food and Drug Administration approval phase, according to Attorney Edward Basile, a former FDA official who now heads the FDA practice group of Washington law firm King & Spalding.
On Friday, GenomeWeb spoke to Basile to find out how companies can learn how to deal with the FDA.
GW: You have said genomics companies need to move from "pure science to regulatory science." Can you explain what you mean?
Basile: People who come purely from an academic background may not appreciate the potential problems they will face in dealing with the FDA. The FDA will be more focused on particular endpoints and whether these endpoints were identified sufficiently in advance of a study. Often a company will start a study in order to find out whether a product will provide a particular benefit, and they will end up finding out that it provided a different benefit. From the FDA’s perspective, you have to start the study all over again.
GW: How can companies avoid getting clobbered by the FDA approval process for their products?
Basile: The FDA demands a very high degree of record keeping for studies that are conducted under FDA regulations. There is also an analysis that companies are going to have to do, for purposes of approval: 'Do I have a drug, do I have a biologic, do I have a device, or do I have a combination?' When you talk about personalized medicine, you might have a device and a drug combination. You might be dealing with two separate centers within the FDA, and that can influence the product approval process. So people ought to consider this when they are developing a plan.
GW: What should a company do first in navigating this regulatory process?
Basile : The first hurdle these companies need to face is whether the company wants to partner with a pharmaceutical company, and if they do, what would be the nature of that partnership? Will they give up all the rights to discovery in return for some percentage of the profits? Or will they retain control of the product and have some sort of joint partnership?
GW: What factors should companies weigh in making the decision of whether to go with pharma or enter the regulatory process themselves?
Basile : It’s really a question of ‘how are we going to generate the capital to complete this study and get it through the FDA process?’ And you’ve got to decide who is responsible for the FDA approval.
GW: Let’s say a company decides to go ahead and take responsibility for the FDA approval process themselves. What do they do next?
Basile: The next step is deciding how the product initially is going to be marketed, because often you have several potential different indications for a product and need to decide how marketable a particular indication is, as well as what you can prove most expeditiously and get through the FDA and get on the market. Another consideration is whether to initiate marketing in Europe, or [elsewhere] outside the US, where you can probably get the product on the market more quickly. This has to be thought through as part of an overall plan.
GW: Should companies hire an in-house FDA person to deal with this set of issues?
Basile: Different companies handle this differently. It’s a question of who is out there and is available to be hired. The other issue is economic. Is it cheaper to bring in someone who will perhaps take a lot of equity in the company, or is it less expensive to hire out the talent?
GW: Many genomics companies are banking on the concept of personalized medicine. This may allow doctors to use therapies that have an overall efficacy rate of say, 30 percent. But current regulations dictate that a company must prove “efficacy” for a therapy to get it approved by the FDA. How can a company with a diagnostic tool or personalized therapy overcome this hurdle?
Basile: If a company can show that the technology in its entirety could identify a patient population in which this therapy is, say 80 or 90 percent effective, I think that’s a therapy that could be approved. You have to show that your technology adequately identified those people in whom the product was effective at an acceptable rate, but there’s no reason you couldn’t do a study to prove that.
GW: A drug that is tailored for a small patient population might not be cost-effective to take through all of the regulatory hoops that the FDA requires, given that the normal drug development process costs about $500 million. Is there a way for companies to get around this cost factor?
Basile : There’s the Orphan Drug Provision, which requires a patient population below 200,000 Americans. If you qualify, you get certain benefits. One is exclusivity. In addition to patent protection, you have the entire market for some period of time.
GW: In February, Robert Oldham, CEO of Cancer Therapeutics, wrote in a Wall Street Journal editorial, “A major effort must be mounted now to rethink regulation for this promising new world of individualized therapies.” Do you agree?
Basile: If you have a process that is individualized, but that can be used to cure cancer in some populations, one is going to have to prove that. I don’t see how you can get around that requirement of doing a clinical study that shows that your personalized treatment works. It certainly is possible that some changes in the law would facilitate getting these products on the market more quickly, but it’s hard for me to see where the FDA is going to be willing to lower its standards.
GW: President Bush, as you know, is expected sometime soon to appoint a new head of the FDA. What should this person do about the genomics-based medicine issues?
Basile : I think the agency needs to be willing to work with these companies, to try to bring the technology to market as quick as possible. At the same time, the companies need to understand they are going to need to meet a certain standard of proof.