Dr. David Altshuler is a leader in the recent explosion of research linking genes to the risk of common diseases. But if, today, someone came into his airy office at the Broad Institute in Cambridge and dropped a $1,000 read-out of Altshuler's own disease-related genes onto his desk, he says he wouldn't even look at it.
"From a clinical point of view, it's just noise," he said. "No one knows how to use such information to improve health."
The business of analyzing such personal genetic information is way ahead of the ability of medical science to make good use of the data, and that has researchers like Altshuler, and other prominent figures in academia and government, raising a warning flag.
For fees ranging from under $1,000 to $350,000, a handful of companies in the burgeoning "personal genome" industry are beginning to offer people a chance, through a simple saliva test, to learn whether they carry genes linked to higher or lower risks for diseases ranging from arthritis to Alzheimer's.
This month, the latest company to enter the field, Navigenics, held its official launch in New York City and opened its website for business. "Your genes offer a road map to optimal health," the site states. "Know thyself," says the homepage of Knome, a Cambridge company that was cofounded by Harvard genomics pioneer George Church and offers the most exhaustive scans. Another genomics company, 23andMe, proclaims on its website: "Unlock the secrets of your own DNA. Today."
The trouble is, Altshuler and others caution, today may be too early.
In articles in The New England Journal of Medicine, the Journal of the American Medical Association, and elsewhere, skeptical scientists express many concerns. The personal genome industry is largely unregulated, they point out, and tests may be performed poorly. And the known genetic variations linked to disease risks probably make up only a fraction of the factors at work. So "good news" for a person with one variant that appears to lower the risk of cancer by a little bit could be more than offset by "bad news" variants that have yet to be discovered.
"The field is in such horrendous flux that you can't really say for sure that this gene variant is related to this disease and how much, and that will be a moving target for the next few years," said Dr. Muin Khoury, director of public health genomics at the US Centers for Disease Control and Prevention.
A gusher of data on links between genes and disease has led to more than 30,000 scientific articles over the last six years, he said. In the last year alone, 150 studies scanned the whole genome - the entire set of DNA - in many people to find links to disease. They have found promising gene-disease links to heart disorders, obesity, prostate cancer, and more. In coming years, Altshuler says, he believes genomics will be as transformative as the Internet.
But it is a long way from exciting research to reliable data on an individual's health, he and others emphasize. Khoury said he recently turned down a company's offer of a complimentary personal genome scan: "I don't see any utility in doing that," he said.
Genome researchers say that at this point their main focus is on the level of basic science: finding a gene linked to a disease is a step toward determining what has gone biologically wrong.
But their early findings are publicly available, and personal genomics companies argue that the data can already be used to help people, and will become only more helpful in the future.
One early client of the Navigenics service, a 39-year-old woman, found out that she was at heightened risk for colon cancer, and got a colonoscopy that turned up a potentially cancerous polyp, said Dr. David Agus, a company cofounder. The usual age for a first colonoscopy is 50.
Those who are part of the "backlash" against personal genomics, he said, "are arguing over the utility, and clearly some of these markers have utility . . . That woman did well in that regard."
Agus and Navigenics president Mari Baker said the company also has addressed many of the concerns raised by the research scientists. Navigenics has a team that vets scientific articles for validity, focusing heavily on the quality of the testing - the company says its tests are accurate more than 99.5 percent of the time. And the company offers round-the-clock access to genetic counselors who can explain what test results mean.
Neither the testers nor the counselors provide any actual diagnoses. But "if you want to be aware of where to focus your healthcare energy and prevention efforts, this is a way to do that," Agus said.
Baker said it is not known how many people have ordered their personal data from the new genome companies, but in five to 10 years, she predicts, it will be as routine to know your DNA as to know your blood pressure or cholesterol.
The personal genomics companies are using a newly available DNA-testing method. Older methods, now well established, test for single genes that confer great risk. Testing for the BRCA1 gene, for example, can tell a woman whether her risk for breast or ovarian cancer is dramatically elevated.
The new technique looks at many thousands of tiny genetic variations. Human DNA is mostly the same in everyone, but there are an estimated 10 million locations where the sequence of four biochemical building blocks making up DNA varies. These places are known as SNPs - single nucleotide polymorphisms - and some can indicate a modestly heightened or reduced risk for a disease.
Altshuler, of the Broad Institute, worries, however, that the new tests could give consumers a false sense of security. A woman with a family history of breast cancer might be reassured if her test turns up negative for known breast cancer SNPs. She might not realize that she could still be carrying a high-risk mutation in BRCA1, which SNP scans do not test for.
Or people may be told they have no increased risk for diabetes, and so "have an extra piece of cake" - and perhaps end up in worse health than if they had remained in genetic ignorance, he said.
Altshuler, who is also a Massachusetts General Hospital physician, pointed to other issues, as well: What about the patient who, based on SNPs, asks for an expensive diagnostic test? Should insurance pay for it? Also, he and others warn, Congress has yet to pass a bill protecting genetic privacy and prohibiting insurers or employers from misusing the data.
Among the first Navigenics test subjects is Dr. Robert C. Green, an Alzheimer's disease researcher at Boston University who leads a study on how people react to the results of genetic testing for Alzheimer's risk. After 20 years of studying Alzheimer's, he was pleased to find out that his risk for the disease appeared low, at least for the main gene known to be linked to it, he said.
Many of the concerns about the new personal genomics companies are valid, said Green, an unpaid advisor to Navigenics and other personal genome companies. But "I think we need to take a deep breath and see this as the very first step in the actualization of a very exciting trend toward personalized medicine," he said.
No matter what the initial problems, he added, "I don't think we're going to be able to put these genies back in the box."
Carey Goldberg can be reached at email@example.com.