As Gene Tests Spread, Questions Follow
By GAUTAM NAIK
December 13, 2007; Page D1
online.wsj.com
Ever since the human genome was deciphered seven years ago, companies have been rushing to sell genetic tests directly to consumers. But buyers, beware: Many of the claims that accompany these tests are not fully supported by science.
BEHIND THE GENES
Some genetic tests and questions surrounding the science:
• Consumer Genetics' CaffeineGEN test for caffeine- metabolizing genes: Results of a study linking genes to heart attacks haven't been replicated.
• deCODE genetics' test for a gene variant linked to Type 2 diabetes: Some research says the predictive value is weak.
• Salugen's GenoTrim test for weight-loss-related genes: Study found only small weight loss when using nutritional supplements based on the test.
Some 1,400 genetic tests are currently available. There are tests that aim to identify the gender of a fetus as early as seven weeks after conception. Some companies test for traits such as a "sweet tooth gene," and then sell tailored supplements to hasten weight loss. And there are disease-specific tests to predict a person's risk of cancer, diabetes and numerous other afflictions.
But often, the link between certain genes and a condition can be fuzzy. For instance, some of the most popular tests claim to identify a person's risk of getting a common ailment such as heart disease or diabetes. The roots of these conditions often lie in multiple genes, many of which scientists think haven't even been discovered yet. Lifestyle factors -- such as diet and exercise -- can have an impact on risk, too. A person may also carry several other genes whose workings counteract the ill-effects of the gene being tested for.
Some genetic counselors also note that people may be no better off when they learn the results: If a man finds out, for instance, that he is at marginally higher risk of getting prostate cancer, it's not clear what he should do with the information.
"The significance of the risks uncovered by these tests is very, very small," says Stuart Hogarth, a fellow at the Institute for Science and Society at the University of Nottingham, England, who has studied the genetic-testing industry. "Commercialization of genetics tests at this stage is premature."
Genetic testing can be a useful diagnostic tool for some diseases, such as cystic fibrosis. Would-be parents can learn if they risk passing on a gene for CF to their children, and perhaps opt for in-vitro fertilization, where embryos can be screened for the mutation.
But many experts note that most genetic tests are only loosely regulated, and aren't required to provide the type of rigorous scientific studies to back up their claims that the pharmaceutical industry must provide.
A handful of companies are starting to offer a genome-scanning service to reveal details about a customer's ancestry and disease risk. Such services are expected to run from a few hundred dollars to nearly $1,000. A newcomer called Knome of Cambridge, Mass., hopes to cash in with an even bigger project: It will identify every single one of the six billion alphabets in a customer's genome, place it on a CD, and get specialists to analyze the result and offer medical advice. The price tag: $350,000.
'Groups of Genes'
"We can look for groups of genes that have been associated with cancer predisposition, or genes that tell you whether you'll have a reaction to a drug," says Jorge Conde, chief executive officer of Knome, which has fewer than 10 employees and was started just six months ago.
Given the current state of genetic knowledge -- in which there often aren't clear-cut associations between genes and disease -- Knome's test "sounds like a genetic horoscope," says Gail Javitt, who leads a genetic-testing quality initiative at the Genetics and Public Policy Center at Johns Hopkins University.
Some tests are based on relatively obscure links: Consumer Genetics Inc. of Sunnyvale, Calif., offers a test called CaffeineGEN to tell whether a person has a "slow" or "fast" caffeine-metabolizing gene. The pitch on the firm's Web site: "Did you know that your daily caffeine consumption may put you at risk for a nonfatal heart attack?? But, if you have a certain gene, having 2-3 cups of coffee may reduce your risk by 22%!" The cost of CaffeineGEN: $139 for a standard test, with results in seven business days, or $179 for "express" results in three.
Consumer Genetics says the test is largely based on a study published last year in JAMA, the Journal of the American Medical Association. Ahmed El-Sohemy, a professor of nutrigenomics at the University of Toronto and a co-author of the JAMA research, says the study provides "the best evidence yet" that caffeine may trigger nonfatal heart attacks in some people.
However, Prof. El-Sohemy, who has no connection to Consumer Genetics, says was surprised when he learned that his research -- which hasn't been replicated by anyone else -- was the basis for a consumer test. "It's premature to sell these tests," he says. "I'd like to see further validation."
Lily Nguyen, product manager for Consumer Genetics, says the company stands by the test. "We're not making a recommendation to drink two to three cups or not to," she says. "That's in the consumer's hands."
Iceland's deCODE genetics sells a $300 test that looks for the variant of a gene associated with Type 2 diabetes. While about 11 other genes have also been linked to the disease, the current version of the test incorporates only one. Even if a person tests positive for the variant, TCF7L2, the risk of getting diabetes rises to roughly 14%, from about 7% in the general population.
"The predictive value of the genetic test is pretty poor," says David Melzer, a professor of epidemiology at the University of Exeter, England. Last year, Prof. Melzer and colleagues published a study based on data collected from more than 900 elderly people in villages near Florence, Italy. They found that 80% of the people who tested positive for TCF7L2 didn't get Type 2 diabetes in old age. And nearly 40% of the people who had diabetes didn't carry the gene variant at all.
Kári Stefánsson, the CEO of deCODE, says the test is useful. He says that people who carry the diabetes-gene variant will respond better to lifestyle changes -- improved diet, more exercise -- than those who don't. "Knowing your risk relative to others provides an incentive to seek treatment," he says.
Genelex Corp. of Seattle sells a $600 test for three genes, known as Cyp variants. Studies have shown that the genes play some role in determining how a person's body metabolizes antidepressants. But last year, the U.S. government's Agency for Healthcare Research and Quality published a report indicating that past data didn't support a strong link. Says a Genelex spokeswoman: "Genelex agrees that randomized trials are important and should be done, but there is no reason to not have testing available in the interim."
Genetic tests are regulated on the basis of whether a lab uses its own chemicals and experimental formulas, or whether it uses a "kit" that is manufactured and sold to labs and clinics that perform the test. Such test kits are regulated by the Food and Drug Administration as "in-vitro diagnostic devices," including those used to check for pregnancy or diagnose HIV. Very few genetic tests are marketed as kits.
"For the consumer, there's no vetting for clinical validity for most of these tests," says Ms. Javitt of Johns Hopkins.
There are growing calls for tougher regulation. In the United Kingdom, the Human Genetics Commission wants new rules requiring companies to prove the validity of their DNA tests before selling them to the public. In the U.S., two bills were introduced in Congress this year that, if enacted, could lead to greater regulation of genetic tests by the FDA.
An area of particular scrutiny is the growth of "nutrigenetics" -- vitamins and supplements tailored to a patient's DNA profile. For example, one company, Salugen Inc. of San Diego, offers a $450 test called GenoTrim, which will analyze your "sweet tooth gene," "nervous eating gene," "fat regulator gene" and others. Then, for $99 per month, you can buy the firm's "DNA-customized nutritional supplements." Including pilot studies, more than 2,000 people have done the tests and bought the supplements since March 2006, the company says.
"Each of the genes has been associated with weight problems, and each of the supplements have been linked to the particular genes," says Brian Meshkin, CEO of Salugen. However, he concedes that there are hundreds, and possibly thousands, of other genes that play a role in weight gain but aren't included in the test.
A Look at the Research
Salugen says that an "observational study" of 1,058 people who took both test and supplements yielded "statistically significant" results. But according to the 2006 study -- which wasn't randomized and hasn't been published -- people lost an average of only 2.5% of their body weight.
Mr. Meshkin says the effect could have been greater if the company tried to control what the participants ate. In the future, the company plans to provide dietary recommendations as well.
In August 2006, the U.S. Government Accountability Office looked at several nutrigenetic tests sold on four Web sites. (GenoTrim wasn't one of them.) The agency's conclusion: "The results from all the tests GAO purchased mislead consumers by making predictions that are medically unproven and so ambiguous that they do not provide meaningful information to consumers."
The GAO found that results on one test also suggested that the ingredients for "personalized" supplements, which cost about $1,200 a year, were substantially the same as typical vitamins and antioxidants found at any grocery store for about $35 a year.
Write to Gautam Naik at gautam.naik@wsj.com1
URL for this article:
online.wsj.com
Hyperlinks in this Article:
(1) gautam.naik@wsj.com |
