Feature Story

Flunking the genetic test: Why we’re not ready for mass genetic screening and testing

Biotechnology forum for women

January 2003

Arthur Caplan, PhD, chair of the Department of Medical Ethics and Emmanuel and Robert Hart Professor of Bioethics at the University of Pennsylvania School of Medicine, addressed the Women in Biotechnology meeting in Baltimore Oct. 16, 2002. An edited version of his speech, “Are We Ready for Mass Genetic Testing? Ethical and Social Hurdles,” follows. Dr. Caplan wishes CAPTODAY readers to know that his comments do not pertain to genetic testing labs that meet CAP, CLIA, and other professional quality standards.

Are we ready for genetic testing and screening on a large scale? I won’t keep you in suspense: No. It’s not hopeless; it’s just that we’re not doing the things we ought to do.

Why? Well, when you look at large-scale genetic testing, a number of tough questions have to be asked: how to bring this to the public, who the correct audience is, what kind of values should guide a message about any genetic test at all.

Let me tell you a story that taught me something about genetic testing. Testing for the gene for Huntington’s disease has been developed with twin protections. First you have to go in and meet with a counselor. And the counselor explains there’s no cure for Huntington’s disease.

If you agree to the test, you have to come back again, so that you have time to think about whether you really want the test. The idea is that the information could be so disconcerting that some people might think about it and say, “I don’t want to know.” When you come back, you go through the same thing again. “Do you really want to know? This is what it means. There’s no cure.” If you then say yes, you get the test.

That’s a mode of counseling I want you to keep in mind: two visits; trained, master’s level genetic counselor; double opportunity to consent; time to reflect outside the test setting. There are a lot of hoops.

Three or four years ago, I got a call from our Huntington’s disease clinic. The person on the phone said, “We have a real problem. Somebody is here and tested negative for Huntington’s disease.” Nobody ever calls me up with that kind of problem. I’m thinking, what’s the issue? And they said, “Well, the reason there’s a problem here is that when we tested this guy and his dad, we learned that that can’t be his dad.”

The good news is he didn’t inherit the disease. The bad news is, we know, that given patterns of DNA inheritance, that man is not his father. So do we tell them that or don’t we? What would you do?

We didn’t tell them. We decided that he came to find out about a test. He got the test. And he left.

Of course, there could be side consequences to our decision. If he has brothers, they could come in for a test. Maybe they’re not at risk at all, and they could waste time and money to get tests they don’t need. He may someday find out the truth, realize we knew it, and sue us or something. But we decided it was in the best interests of the patient not to make trouble for him that he hadn’t asked for. So we didn’t.

But we did change the consent forms. They now say: “You could find out other things if we do this test. We could find out about paternity. We could find out about other diseases. And we’ll offer you the opportunity to decide if you want to know that.”

So we learned two important things about genetic testing. One, genetics is different because it tells you things about biological relatives. If you find a disease running in a family, it means that other members could be at risk. They may not have requested that information, but lo and behold, they now know.

The other thing we found out is that genetic information can be exquisitely sensitive. Most people aren’t going to worry too much about physiological measurements, like their respiratory volume output and pressure. But genetic information makes people feel that they are being told something about themselves. Oftentimes people will say, if they get an unfavorable genetic test or their children do, that they did something wrong. They feel guilty because they have a bad genetic outcome. They feel that they harmed their children, or that they themselves are flawed.

Most people don’t feel that way if their kidneys start working inefficiently. They don’t say, “I’m a flawed human being.” But they feel that way about their genes. Genes tend to be seen as somehow the blueprint of you, as opposed to just some inherited information that might carry mutations.

I constantly have to remind the medical students at Penn that genetic tests are something everyone fails. We have some knowledge now about certain risk factors or certain properties, but not all. So some people feel more put upon by genetic testing because their ethnic group or their family has information we can test for. But everybody ultimately flunks the genetic test, because everybody carries mutations and problems and risk factors. Something will kill you.

No one has perfect genes. But right now that isn’t how people react when they get unfavorable results. This is very sensitive information—not just because it sometimes leads to bad consequences, but because people identify with their genes more than they do their bones or internal organs. It’s very interesting to watch this phenomenon. Some of you ask why there was more attention paid to Jesse Gelsinger dying in a gene therapy experiment than Joe Blow dying in a pharmacology experiment. There are probably 10 deaths in drug studies for every gene therapy study death, but they don’t have the same symbolism and sense of personal connectedness that genes do.

I tell you these stories as setups for why I don’t think we’re ready for mass genetic screening and testing yet. There are five reasons why this is so.

I told you in the Huntington’s disease model that people are getting counseled. And right now some genetic counseling groups suggest that counseling always accompany testing. But no real mandate requires that. It might be urged as a morally good thing, but there’s no requirement. I have no doubt that people are going to need some form of counseling to understand genetic tests.

But if you talk about requiring counseling, who is supposed to do it? Large-scale testing for breast cancer or cystic fibrosis, for example, would take up all the person-hours of all the genetic counselors who exist. They couldn’t do anything else. And there are not that many of them.

So you might say, well, primary care physicians should do it. Or nurses or other people we train. But, as I think a lot of you know, there isn’t much training going on for most doctors and nurses about how to deal with discussion of genetic information. If you ask doctors whether they’re comfortable with talking abut genetics, the answer is no, they’re not.

So counseling is not required and we don’t have people to do it. Furthermore, we haven’t agreed on how to do that counseling. Should we use the two-visit method? Should we tell everybody not to get a breast cancer test because there’s really no cure? You could have a preventive mastectomy, you could take tamoxifen, but it’s not clear what you could do if you found you were in a high-risk group.

So should we say, yep, that’s disturbing enough that we should use the two-visit approach? Or should we say, “No, actually all you need is a brochure. Read it, have an 800 number to call if you have questions, and that’ll be adequate counseling”? I don’t know. It seems to me the discussion has to allow people to answer what style of counseling they think would be adequate. Or what the range of counseling options should be. It’s not really clear what ethics requires. Should we seek regulations that create counseling that’s minimally adequate or maximally adequate?

If we’re going to stand for informed consent—whatever that is—then I happen to think the two-visit model is probably not right. But just having a brochure is probably not right either. We’ve got to get that settled before we’re out there doing direct-to-consumer marketing for genetic testing.

Second problem: Genetic testing can be used to penalize a person. An employer might not want to take you on because you’re in a high-risk group. Your prospects for marriage might be diminished if you have a genetic disease. I’ve seen that a couple of times.

And if you’re trying to get insurance, what are we going to do about genetic risk information? Remember, we’re not talking about existing diseases. We’re talking about risk factors. As I read most insurance policies, you have to tell them if you have a preexisting condition, but historically that has always meant a disease, not a risk factor. So that is a completely unsettled area.

If I were to launch large-scale genetic testing for breast cancer, prostate cancer, juvenile diabetes, schizophrenia, autism—all of which I think will be doable in 10 years—I’d have to tell people: “Well, I can test you, but I can’t cure any of those things, and this information could cost you your job, your marriage, and your insurance. Want to do it?” Information is good, but it’s not always good.

At a minimum, you want to tell people what risks might result from this kind of information. You also want to set up a system of health care that mitigates the use of this information to penalize people. Information is good, and people should be encouraged to get it, but you don’t want them avoiding it because they’re worried about penalties.

A few states have tried to pass statutes on anti-discrimination. There’s no real federal statute. To be honest, even the states that have tried to pass anti-discrimination acts regarding genetic information haven’t done a very good job. And if you work at a company that self-insures, the regulations won’t affect you anyway. The current statutes apply to public employees or people with more commercial forms of insurance. But your benefits manager at Xerox or Aramark or some other self-insured place will have access to all your medical records and know exactly what’s going on.

The third problem is that there are no standards for the accuracy of genetic tests. As nuts as that sounds, you could announce tomorrow that you’re going to do genetic testing, dip a stick in somebody’s mouth, look at it, and say, “Your genetic test is fine.” There isn’t any standardization out there about how to do it, who should do it, or what level of sensitivity or specificity the test should deliver. And that is not good.

It’s one thing for Myriad to say, “We can test this and know what we’re doing,” but there are all kinds of other companies that could jump in the game. What do you think is the most common form of genetic testing in America today? Anybody know?

It’s paternity testing. It’s usually carried out by Jerry Springer or Montel Williams. And you’ve all seen these billboards all around the country offering paternity tests, the “1-800-WHO’S MY DADDY” kind of thing. Who are these people? Does anybody know if they’re accurate? What’s their false-positive rate? I have no idea. And nobody else does either. I’m astounded that no one in genetics has jumped up and said, “This is absurd.” Sensitive information, paternity testing, done by who knows what company on TV. No counseling for the guy getting the results in front of everybody else. That’s the context for large-scale genetic testing right now.

The fourth problem with genetic testing is how to pay for it. The reason we don’t see more large-scale genetic testing is that it’s not covered by insurance. We still have a system that doesn’t do a lot of reimbursement for preventive services. If this field is going to move, there’s going to have to be some reimbursement. And there isn’t any. If you’ve been through amniocentesis or chorionic villi sampling, you know it’s hard to get reimbursed even for that.

So far we’ve got: no counseling, no standards, the possibility of discrimination, and no financial coverage. And that leads me to the last problem: What constitutes adequate, informed consent? What can you put in an advertisement? There’s been a lot of discussion recently about direct-to-consumer advertising, and genetic testing fits smack in the middle of this.

For example, it’s clear that genetic forms of breast cancer make up about 10 percent of all forms of breast cancer. You could have a genetic risk factor, but if you don’t have a family history, it’s probably not likely you’ll get the disease. On the other hand, you could have a genetic risk factor that no one knows about yet. So how do we describe that in an advertisement for breast cancer genetic testing? Do we target it to people who are at high risk, or do we just say, “Hey, any woman should get this”?

And don’t think there won’t be people running in to get their kids tested. There absolutely will be. Whose consent is required? Does the kid need to know? Would we test children or adolescents at all? I have no idea.

So we’re in good shape, as far as I can see, for large-scale testing. We have no standards, no counseling, no informed consent, adverse discrimination, and no way to pay for it. Other than that, it’s ready to go. And what makes me nervous right now are three letters: GMO, genetically modified organisms.

I watched that technology shoot itself in the head by moving forward without any answers to these questions: Do you label the food? Are there standards for how you plant your crops? What are the safety standards? What’s the rationale for using genetically modified food? Is it a way to increase pesticide availability? Are we going to use it to put nutritious things or medical things into foods? None of that got answered, and for the most part, the industry is in dire straits because the public doesn’t trust it.

People place a symbolic value on food. I think the first genetically engineered product out there was milk. I wouldn’t have started there. If they had brought me in as their ethicist, I would have started with something like asparagus. Milk is supposed to be pure and unadulterated, and it represents your mother, and that’s just a bad opener. I don’t like that on symbolic grounds. I don’t care what the market for it is. If you think the technology is guaranteed to advance just because it’s useful and a lot of scientists like it, you’re nuts.

Another technology in the same boat is gene therapy. I know about this firsthand from Penn. I watched our gene therapy program implode over human subjects’ protection problems. About once a week, a reporter asks me what’s going on with gene therapy. Nothing. There’s some experiments puttering along, but it’s going to take 20 years to get back to a clinically effective form of gene therapy trials.

Inadequate attention to ethics can kill a technology faster than anything. Those two areas are dead because they did not attend to their ethical infrastructure. And if mass genetic testing doesn’t attend to its ethical infrastructure, it will go the same way.

I’m not against genetic testing. I think it should be done, and done responsibly, and I want insurance companies to pay for it. I think information is valuable. People can make better choices about their lifestyles and careers. But if you don’t have these issues addressed, I’m here to tell you that this field is standing on the edge of problems. When I ask what Myriad plans to do with large-scale breast cancer testing, given what I’ve just told you, I don’t get good answers. Now, maybe they have answers and they’re just not giving them to me. But I’m worried.

Do you know what the Framingham Heart Study is? It’s probably the most powerful epidemiological study ever done. In 1948, the National Heart, Lung, and Blood Institute decided to select a town that was demographically typical of the United States, study smoking behavior closely in that cohort, and do annual physicals on everybody in the study as long as the study went. It began in ’48 and it still goes on. It’s in the third generation of people. I trust the Framingham Heart Study completely because it was well done, standardized, and extends out over this 50-year period. You can actually start to look for the heredity risk in the third cohort.

They got a good high correlation between smoking behavior and heart disease and later cancer. But a lot of other information is in the Framingham Heart Study because they were doing these physicals all the time. Most of the information is still in boxes. It’s never been transferred into a computerized database.

Well, a company, Framingham Genomics, realized that it might be able to create quite a database if it could get genetic samples from the current cohort of people, look at tissue samples from previous generations of people, and create a kind of massive DNA database. So they got venture capitalist and equity people together and began to talk. They found investors and got incorporated and off they went.

Lo and behold, the people in the town of Framingham said, “We don’t want a private company involved in this. We’ve participated in this study for three generations for free.” People fly in from around the world to get their physical every year. It takes a lot of time, but they see it as a big gift to the world.

I know this because I’m from Framingham. And as soon as the company showed up, I got a call from people in Framingham who said, “You’ve got to come back here and help us, because this company is going to rip us off.” By the time I arrived, the people in the town who were involved in the study were saying, “If this company forms, we’re not going to participate anymore”—losing the best study in the history of medicine.

Framingham Genomics fell apart. It doesn’t exist. The data is still in the boxes. What the company proposed to do never happened. The NIH will not fund the transfer of this gigantic database: It’s too expensive and NIH doesn’t do this kind of data work. So it sits there.

Should there be a Framingham Genomics? Absolutely. This is nuts. Framingham has looked at something like 10,000 people for 50 years. It has everything you’d ever want, and it fell apart because of ethics failure. The people who built the company never thought about what it would mean to ask people who had made a gift to let a commercial venture come in.

Could they have done it? Sure. The company could have said, “If we build this registry, we promise certain returns to the town. We promise certain things will be done for the general public in the spirit of what you were trying to do with this data set. We promise to protect your anonymity. We promise to tell you if we find out things that put you at risk.” There are a lot of things to negotiate. None of them were done.

It reminds me of genetic testing and screening. Here’s this wonderful thing we could all benefit from. Well, nobody’s benefiting from it, because of no attention to the ethical infrastructure.

Let me tell you a little about Viagra—the first intersection I know of between bioethics and a big drug company. About eight years ago, I got a call from Pfizer saying, “We’re doing some research on a drug to lower blood pressure. It doesn’t work, but it raises something else. And we think that might be important.” This was a complete serendipity; they didn’t set out to make a drug that would treat impotency. They said to me, “We can’t jeopardize the whole company if we get in trouble with a drug that has to do with sex. It makes us nervous, and we would like to anticipate problems that could come up if we went into this drug.” I got there just when they were first doing clinical trials. And I was there for four years.

When they got into this, I said, “Look, if you’re going to have this drug, here are the problems I can see. Conservatives or religious groups may come after you and say you’re promoting sex. One way to deflect that criticism is to treat a disease. If you can present this drug and say it fills a need, it responds to a physiological problem, then you are likely to be able to carry this product forward in the face of criticism.” So they promoted Viagra as a treatment for disease. Not as a quality-of-life drug, not as an enhancement, but as a disease drug. And that medical model served them very well.

By medicalizing it, they never got into some of the problems that other drugs have. If it had been out there as a quality-of-life drug, like baldness, they would have gotten clobbered the first time somebody died. You can’t have deaths for something that’s just supposed to let you have a good time.

I also talked to them about other issues, such as having a policy in place if someone raped somebody on Viagra. And would you sell this drug in nursing homes, since impotency increases with age? Are you going to let people find places to have sex in nursing homes? Would you sell it to people with cognitive impairment?

Pfizer has policies on all these things. Some of them they’ve used, some of them they haven’t. But the bottom line was, by carefully thinking about the ethics of that drug, they were ready to handle the third-party payer issues. When people said, “We’re not going to pay for that. It’s just some kind of sex drug,” the company could reply, “No, it treats impotency.” And within a year every payer was covering Viagra.

I will wrap up these remarks by showing you some of the relationships between ethics and the genetics industry or biotech industry. The model I used with Pfizer was what I’ll call the hacker model. I was hired to find every ethics problem in advance, think about it, and present it to them. Then management decided what they wanted to do about it.

By the way, I told them to build a corporate legacy so they’d remember what the discussions were about—treating disease. They didn’t do that, because who’s selling Viagra to you today? Not Bob Dole. It’s baseball players. Racecar drivers. We’ve shifted from medical to lifestyle. And that’s interesting, because I think it exposes them to more vulnerability. They may not care anymore because it’s an established drug.

So that’s the hacker model. Test your defenses, probe around, figure out what problems might exist. And then there’s a firefighter model. Show up when there’s a crisis, like Framingham Genomics. That was no good; the house burned down before we showed up anyway.

Coming up with preventive policies requires a team to think through the infrastructure—some industry bioethics-law alliance for deciding how to answer all of those things. I’ll remind you: That was never done for genetically modified food. The first time you found out about it was when you ate it. You’ve been eating it about eight years if you eat anything out of a supermarket that’s processed. I don’t think there’s any health problem with eating it, I really don’t. But I know that because I’ve looked at a lot of the safety studies. A lot of Americans would say, “Gee, I don’t know if that’s safe.” They should never be asking that question. Never.

The last model out there is simply a kind of ethics review. I didn’t give you an example of that type, but sometimes it’s useful to have somebody step in and look through the company. Genomics Collaborative is that. It’s a small startup that’s interested in building a DNA database. They’ve hired me and Penn to think about their informed consent forms and privacy protections. We sit there and review what the company is doing, make suggestions in a more systematic and less crisis-ridden mode for their behavior.

So those are different models to ponder. But the bottom-line lesson is, if you don’t think about the ethics in the area of biotech, you can get in big trouble. And there are plenty of failures lying around.