At A Glance
Name: John Robertson
Title: Vinson & Elkins Chair in Law and Mark G. Yudof Endowed Research Professor in Law
Where: University of Texas School of Law, Houston
Pharmacogenetics can become a catalyst for better drug design; few experts will argue that point. What’s also clear is that the discipline, now entering adolescence, will encounter some legal and bioethical challenges as it crawls from the bench top to the bedside.
What’s not so clear is to what extent pharmaceutical companies, regulatory agencies, and instrument makers may be affected by debates between the clinic, the pharmacy, and payors.
"A perfect example of this is when the FDA last week approved a new AIDS drug [Fuzeon, made by Roche] … which costs $20,000 a year,” said John Robertson, a law professor at the University of Texas. “In this case, you can see a drug plan [and] say, ‘We’re not going to cover that drug for you unless you take a DNA test that shows that you can handle it.’ It’s in their interest too! Why should they pay for a drug that will not help someone?"
From a legal perspective, what are some of the biggest challenges facing companies that conduct pharmacogenetic research and development?
The main issues will be those that answer the question of what numbers will suffice in phase III trials to satisfy the [Food and Drug Administration]. And what should the labeling of the drug include, with regard to pharmacogenetic tests — that is, to what extent should the drug be approved on the condition that genetic testing occur prior to the prescription?
And there will also be issues about where that information should appear on the label; is a certain result in a genetic test a contraindication? Is it a warning? Is it a precaution? These will be issues that will be negotiated with the FDA and the company. This will also take into account the company’s concern with liability … about warning for genetic contraindications.
Considering the degree to which pharmacogenetics interacts with patients, and the myriad bioethical and societal components to genetics, what is your view on the current legal and regulatory landscape that governs pharmacogenetics? Are regulations too strict or not strict enough?
My sense is that, if the science is good, there’ll be easy regulatory acceptance of it. It will depend on how fast the science here progresses — for example, how strong the correlations are between certain SNPs and drug responsiveness.
Are there laws on the books today in any of the states or the federal government that touch on pharmacogenetics specifically, or is the relationship exclusively between the drug makers and the FDA?
That will be initially [the case], but then, once there are drugs that should be prescribed after a genetic test, there will be two other sets of important legal issues that arise: One is the need for physicians to become aware of these genetics tests — when to order them and what to do with the results. Failures in that area will raise issues of medical liability. That’s an important part of the legal [aspect]. This isn’t a big pharma issue — it’s just getting information to the troops out there in the field.
The second thing is, what do physicians have to tell patients before taking their DNA in order prescribe a drug?
Do you have a sense of how big that issue will be? Will it be something that the medical specialty societies will be able to sort out themselves, or will Congress have to step in write some laws?
It’s unclear at the moment; it’s still too unknown. The problem is, if every time a doctor sees a patient, and has to do a genetic test before prescribing a drug, that will lead to a whole level of routinization of genetic testing in the health-care system unlike we’ve ever seen. On the other hand, if the science works out only in certain classes of diseases or drugs, then it will be more limited and you’ll have specialists deciding this.
The question will be, ‘What does the doctor tell the patient?’ Would he say, for example, ‘I want to take a little bit of blood to see if we can prescribe this drug?’ Or should the doctor has to say ‘I want to do a DNA test?’ Here there are issues of protecting the samples and confidential results.
Are there really issues like this? I’d think it’s obvious that physicians have an obligation to inform their patients about any blood tests they run.
Sure. I agree, but here’s the question: It will be a question of ‘We want to draw some blood to see if we can prescribe a drug for you.’ Is that sufficient, or does a doctor also have to say ‘We want to do a DNA test?’ And the only reason there may be a difference is the sensitivity of genetic testing without consent. Plus, genetic tests can be informative of other conditions.
The hope of the pharmaceutical industry here is that you will be able to use pharmacogenetics to find SNPs that are not in any coding region, so there will not be information about mutations that could have independent significance — for example, not that a person has a gene for a disease, but that the SNP might show that the person isn’t responsive to drugs. For example, for the drug that a patient is being tested for, if he or she has … a mutation in the cytochrome P450 gene … that means this finding will not only affect the metabolization of the drug they’re being tested for, but it will also affect the metabolization of other drugs, too.
Considering the murkiness that you touched on, do you think there will be spillover into the upstream vendor market? Will vendors’ top line be impacted at all?
No. The pharmaceutical companies, the SNP Consortium, and the HapMap Project are all tools that are being produced by either the government or by industry that will help the development of pharmacogenetic markers.
Do you foresee more laws that regulate pharmacogenetics as the discipline gains momentum in more and more clinical specialties?
That regulate it specifically? No, I don’t, but we’ll have to see how this plays out. Take for example that maybe 20 states have laws against genetic discrimination in health insurance or employment. What about a drug-benefit plan that says ‘We will reimburse you for your drugs only when you pass a pharmacogenetics test that shows you are an appropriate recipient’?
A perfect example of this is when the FDA last week approved a new AIDS drug [Fuzeon, made by Roche] … which costs $20,000 a year. In this case, you can see that drug plan can say, ‘We’re not going to cover that drug for you unless you take a DNA tests that shows that you can handle it.’
Are things moving in that direction now? Are you aware of examples that show payors taking that stance?
They will as soon we have more science that supports it. It’s in their interest to! Why should they pay for a drug that will not help someone?
Clearly. But I can see patient-advocacy groups taking a ball-peen hammer to that argument.
Absolutely. There will be battles of that sort: Is the science good enough to justify requiring a genetic test to get the drug covered?
Will pharma companies ever feel the need to jump into this debate, or will this be an issue between payors and patient-advocacy groups?
I’m not sure how it will play out. If this will restrict purchases of drugs, [then] pharmaceutical companies may come out one way rather than another. They may have a label or a precaution as opposed to a contraindication, say, that genetic testing may be helpful here: It’s not a contraindication if you don’t have a certain genotype. And the drug-benefit programs are going to say a patient doesn’t get the drug at all. And so they could end up differing over that.
We know of drugs like Gleevec and Herceptin that are accompanied by genetic tests. In these cases, patient-advocacy groups haven’t issued a call to arms to Novartis or Genentech.
Because it seems to make good medical sense. But there could be issues with how precise the tests are. You know, for Herceptin, physicians can tell whether a woman overexpresses the her2 receptor gene, which, if so, she should get Herceptin. People aren’t arguing over whether someone is in the gray area [in terms of Herceptin treatment]. But you’re going to get a lot more cases like that with pharmacogenetics because the predictive powers for some of these tests will vary.