Decoding the Human Genome: How Will it Impact Business?

Knowledge at Wharton: What kinds of opportunities will the mapping of the human genome offer pharmaceutical companies?

Danzon: It’s already changed the way some companies do research and development. [Last week’s announcement] was only the tail end of information that has been coming out for some time. People knew the announcement was coming and expectations had already been factored into the stock prices of companies involved in genetic research. Genetic information is already used in identifying drug targets. What remains to be done is to establish the connection between different genes and specific diseases. Having the gene sequences doesn’t tell us which genes are involved in which diseases, and that’s the huge gap that has to be closed before we see medications come out.

Research on the genome does give us a whole new way of identifying drug candidates. There’s a famous quote from the former head of R&D of one major company that pharmaceutical companies were previously going after 400 or 500 drug targets. Now there are thousands or tens of thousands. There is an abundance of riches as to what we try next. But there is still the bottleneck of figuring out which of these targets will work. That will involve costly trials. It won’t necessarily reduce costs in the short run.

Knowledge at Wharton: How much work remains to be done?

Danzon: Drug discovery is simply identifying potential drug targets. You then have to go through the process of drug development, which involves figuring out whether a drug candidate actually works. The gene helps you identify the receptor in the body. Then you have to come up with a molecule that you think will work. Then you have to test that in animals and humans to see if it does work. That part of the drug development process will always be something that must be done in addition to how we identify the various drug candidates.

Longer term, the hope is for true gene therapy, which would be therapy to modify a gene or to control the expression of a protein by a gene. These therapies would be incorporated into the body like a vaccine, a one-time fix for genetically-based diseases. This has enormous potential for diseases for which we have no treatment. There will be a lot of ups and downs in the process toward developing gene therapy – putting something into the body to change the expression of a protein is truly new. That’s still many years away.

Knowledge at Wharton: How soon will we see the development of products as a result of genome research?

Danzon: It’s a question of some starting soon and many more being developed over time. There’s already good information about genes associated with some diseases. That’s particularly easy for diseases where there is one gene associated with a disease. But for many common diseases, multiple genes are involved. And other factors like nutrition play a role. So what we really want to know is how does having a particular gene increase the probability that someone will get a disease. But determining that requires a lot of data and analysis that so far has not been gathered. But where gene therapies are already making progress is in the treatment of some tumors where genetic information enables doctors to figure out what drugs are likely to work. It’s an issue of progress at different speeds for different diseases.

Knowledge at Wharton: How much information from genome research will be proprietary and how much will be in the public domain?

Danzon: Information developed by the government consortium will be in the public domain. In addition, several major drug companies have formed a consortium whose purpose is to share research. Not only Celera but other genomics-based companies such as Millenium, Incyte and Human Genome Sciences are all interested in licensing or selling their information to companies that will turn it into products. Information doesn’t generate revenue till it’s turned into a product. There is a question about whether fees for this information will be such that it will deter research by other companies. But it’s in the interest of research companies to sell information.

One thing not yet resolved is what patents will hold up. Probably thousands of patent applications have been filed and there’s a huge backlog at the Patent Office. Even once patents are issued, there will probably be litigation. Since the life of a patent runs from the day you file the application, the life of the patent may expire before anyone has a patent. It seems fairly clear that genes will be able to be patented if the filer can, with some reasonable credibility, demonstrate what the use of the gene is. To obtain a patent you have to have something that is novel and has a utility; you have to say what you will use it for. That’s why patentability is so uncertain. People are making claims without really having evidence. Another issue is what happens when you request a patent for something less than the full gene: Can you patent the pieces?

Knowledge at Wharton: Will confusion over patents delay the development of products?

Danzon: There is a real concern that if a firm has to acquire multiple patents to do research on a particular drug — and that’s a realistic expectation given that we don’t know which genes are associated with each disease — then the so-called ’stacking’ of patents may become a significant added cost. Companies may only be willing to take on such costs if they have a chance of getting a significantly profitable product out of it. It would certainly deter investigation for minor diseases. But at the same time, since the patent issue has not yet been resolved — and since a lot of information is in the public domain — that so far doesn’t seem be likely.

Knowledge at Wharton: What role will pricing play in how people will be able to use drugs developed from gene research?

Danzon: There are two related areas where these types of drugs may raise special issues in pricing. First, let’s assume pharmacogenetics is used to identify the patient group that would benefit from a drug, and the drug is used on the condition of the patient having some diagnostic test to determine that he or she has the right genetic makeup for the drug. Well, these targeted drugs will treat a smaller subset of the population than traditional drugs, which means fewer ’customers.’ This means that such a drug would be much less attractive for companies to develop. (This is in contrast to the traditional shotgun approach where you give a drug to every patient, knowing that half will benefit and half won’t.)

So, from an economic standpoint, the appropriate adjustment would be to raise prices to reflect the fact that each patient who takes the targeted drug is sure to benefit. Payers may not be willing to pay higher prices, yet that would be the appropriate economic response because the expected benefit of the targeted drug would be twice as great as the traditional drug which only benefits half the patients treated.

Second, if we get to the point of having true gene therapies — long-lived treatments — there’s the question of whether payers would be willing to pay for high-priced, once-a-year therapy. If once-a-year therapy replaces a daily therapy, would payers be willing to pay 365 times as much? They should be, but they may resist. If we get long-lived therapies, it may be better to treat them as therapies, like surgery, instead of drugs. The reimbursement framework will have to be adjusted to provide incentives for companies to develop these newer therapies.

Knowledge at Wharton: Would it be correct to assume that all or most drugs developed as a result of genetic research will be more expensive than traditional drugs?

Danzon: It depends. Compared to the old way of discovering drugs — going through lots and lots of compounds — today’s discoveries can be made by using computers, and that can be cheaper. Computer-assisted research generates more possible drug candidates, which is good. But that means there are more needles to find in the haystack, and you have to spend money to go through the haystack.

Knowledge at Wharton: What implications does the mapping of the human genome have for insurers?

Pauly: The insurance issue is this: If an insurer learns from genetic tests that certain people are high risks, will it charge high premiums and destroy the basis for health insurance? I think this is unlikely to happen. I say this based on research that my colleagues and I are doing on breast-cancer screening. We’re looking at whether the availability of current tests for breast cancer can disrupt the life insurance market, but I believe our preliminary findings apply even more strongly to health insurance. We’ve been doing modeling and surveys of breast-cancer patients. We’re not finished with our research and so we’re not sure what the exact answer is. But it seems that the extra information that the tests provide isn’t much more than what you’d get from knowing a patient’s family history.

The same probably will hold true for the mapping of the genome: The extra information provided by a genetic test may not be much more than what an insurer could get by knowing what your parents or grandparents died from. At the moment, genetic tests are quite costly. And the extra information such tests provide is sufficiently marginal that they don’t really tell the insurer that much more. Another way to say it is the genetic test will tell me I will incur health expenditures and die from cancer, but not when. And ’when’ is what the insurer wants to know.

Another reason why the mapping of the genome may not hold dire consequences for health insurers, as opposed to life insurers, is that more than 90 percent of people get health insurance as part of their jobs. They don’t apply individually and don’t have to supply information about their health because, under the current structure, most people get insurance as part of a large group. So I don’t think developments in genetic research would be terribly harmful to the health-insurance industry.

Knowledge at Wharton: What about the impact on life insurers?

Pauly: Here there is the potential for much more disruption. The issue insurers have to worry about is this: If a person tests positively for a disease, would that person go out and buy a whole lot of insurance? And, would a person with a negative test cancel his or her life insurance policy? With health insurance, you generally can’t buy much more insurance that pays you extra cash when you get sick; there’s an upper limit on how much people can use positive test results to their advantage. But with life insurance there is a greater danger. If a test shows I will die prematurely from cancer, there’s nothing to prevent me from buying much more life insurance coverage. Still, what we’ve found in our research on breast-cancer screening is that the extra information provided to insurers isn’t all that great, even in cases like this. So the mapping of the genome may not be a very serious problem even for life insurers.

Knowledge at Wharton: What privacy issues are raised by discoveries about the genome?

Pauly: From a public policy standpoint, one important issue, for example, is how genetic information could affect job opportunities. Would a company hire someone who is likely to die soon? As an economist, I have to say that the key will be whether genetic tests will be cheap. My hope is they won’t be. I hope they will be costly enough so that genetic information is not easily found out. One scary scenario that people talk about is somebody could pick up a water glass and get your DNA from it. If the test to analyze that DNA costs five cents, there may be privacy concerns, but not if the test costs a lot.

Knowledge at Wharton: What other issues have you been thinking about?

Pauly: It’s important to remember that, even though the announcement that the genome has largely been mapped was made only a few days ago, everybody knew it was coming. So, in some sense, it’s old news. The medical world has adjusted to this already. I think it’s still very uncertain what this will mean for the development of new treatments. Everybody is hopeful, but not all treatments developed from genetic discoveries work well. Also, even if you identify the genetic defect, to develop a treatment you have to have some vehicles for inserting the gene into a person. That’s proved very challenging.

The other issue I’ve been thinking about goes something like this: Suppose scientists discover a way to prevent a disease that at the moment is difficult to treat, say hemophilia. What would the seller plan to charge, and would most insurers pay that? That will be a tricky issue. Treatments may be developed—but if their price is too high, many insurers may not pay for them, and people may not even want to pay the extra premiums for the insurance plans that do cover it. There’s still a lot of confusion about what will happen along these lines. We often assume that if a discovery is made and the treatment is out there, people will buy it. But pricing will be a serious issue, and it is not at all clear how that will work out.