Data Rich, Information Poor

With hundreds of genetic tests in hand, researchers hold the promise of personalized medicine within their grasp. But important issues must be resolved before these tests can be used to predict and prevent disease.

Once upon a time, curious souls visited palm readers for a glimpse into their futures. Now that the complete human genome sequence is available, DNA tests could put these sages out of business--at least when it comes to predicting an individual's health. Scientists have linked hundreds of genetic alterations to conditions ranging from heart disease to Alzheimer's to cancer, and tests for such variations promise not only to reveal the likelihood of developing these disorders but, in many cases, to suggest a path toward prevention.

Anyway, that's the idea. And it's a concept that companies are already exploiting. "Genovations' proprietary testing helps patients and their doctors determine what specific health risks might lie ahead--long before any physical problems appear," reads a press release sent out by the Great Smokies Diagnostic Laboratory (GSDL) in Asheville, North Carolina. Results from the company's Genovations test panels for heart disease, osteoporosis, immune problems, and "detoxification defects" can help patients tailor their nutritional supplements, lifestyle choices, and pharmaceuticals, says Kay Patrick, product manager for Genovations at GSDL. Patrick, for example, has a genetic deficiency that prevents her from optimally absorbing vitamin D. "I may not have a direct clinical condition yet as a 30-something year old woman," she says, "but I might need to take in more calcium than the RDA."

GSDL is not alone in the personalized-medicine market. Other companies are offering similar genetic tests said to measure vulnerability to oxidative stress, and at least one firm is marketing skin creams and cosmetics purportedly suited to an individual's DNA.

No one can say who is purchasing these products or how they are being used; none of the companies contacted by SAGE Crossroads agreed to share information regarding sales. But the fact that such tests are currently being marketed alarms geneticists and consumer watchdog groups. "It's too early to offer this kind of predictive information to help people change diet or lifestyle," says Muin J. Khoury, a geneticist and director of the CDC's Office of Genomics and Disease Prevention in Atlanta. Scientists don't yet fully understand how variations in these genes relate to health or disease.

Patrick insists that GSDL's approach is valid, as the Genovations tests are based on links between genetic variations and disease that are described in peer-reviewed studies available in scientific databases such as MEDLINE. But pointing to journal articles does not make a test meaningful, says Helen Wallace, a mathematician and deputy director of the watchdog group GeneWatch UK in Derbyshire. "A lot of the early links turn out to be wrong," she says. An analysis published in Nature Genetics in 2001 examined 370 studies of 36 disease-gene correlations and concluded that "the first study often suggests a stronger genetic effect than is found by subsequent studies." Follow-up work sometimes fail to confirm the link at all. Wallace worries that such ambiguities aren't relayed to consumers by companies such as GSDL. "These companies certainly have a tendency to pick papers that support their claims," she says. "It's a fantastic marketing ploy, but a bit weak on the science."

Few gene screens are ready for prime time. Of the 1000 that are now available, about half are used clinically to diagnose rare genetic diseases. Diseases such as Huntington's, in which a single gene determines whether a person will get the condition, are the exception, not the rule, says Khoury. Even screens for variations that are strongly linked with a disease can't predict what will happen in a given individual. "The test can't tell you how strong your symptoms will be, or when you'll get the disease, or even if you'll get it at all," says Wallace. For complex diseases such as heart disease and diabetes, Khoury says, "you're dealing with complicated genetic pathways. If you have a gene variant, it may not mean you're going to get the disease."

"The vast majority of complex diseases," adds Wallace, "rely much more on your lifestyle and environment than on one gene," a point that is often lost on the general public.

And gene tests run the risk of falsely reassuring people, says Khoury. "If you tell someone they're at lower risk of lung cancer, do you give them the message that it's OK to smoke?" he asks. On the flip side lies the danger of unnecessarily frightening people. "If you're positive for a test, you may become more fatalistic and decide it's not worth taking care of your health," says Khoury.

GeneWatch advocates establishing assurances that genetic tests are always done under the supervision of a health care professional who can explain the results, and the group calls for tighter regulations. The Food and Drug Administration (FDA) does not regulate testing services that are performed at a lab such as GSDL, says Steven Gutman, a pathologist at the FDA's Office of In Vitro Diagnostic Device Evaluation and Safety in Rockville, Maryland. Two years ago, the Health and Human Services Secretary's Advisory Committee on Genetic Testing recommended that the FDA regulate new tests before they enter the market, but the committee was dismantled after the new Administration took office, and the recommendations were never enacted. However, the FDA is currently formulating a proposal for public comment, says Gutman, and a new Advisory Committee is likely to take up the issue in the future.

Devising such regulations won't be simple, says geneticist Joann Boughman, executive vice president of the American Society of Human Genetics. "The first thing you want to look at is the validity of the test: Does it test for what they say it's testing for?" she says. "Then, does the result mean what you think it means, and if so, does it make any difference for the patient?"

Until genetic tests can point people toward specific prevention strategies, they're likely to remain about as useful as a palm reading. Right now, says Khoury, "they're telling people to eat more broccoli and exercise more. You could have done that without the tests."

Christie Aschwanden is a Switzerland-based writer who doesn't need a test to know that she has inherited the chocolate-loving gene.