“People are getting these tests because they believe the science is all-powerful. They do not understand the nascent state of genetics.”
Empowerment through genetics is now at everyone’s fingertips—or so we’re told. A quick trip online will direct you to tests that claim to identify your genetic propensity for diabetes, breast cancer, cystic fibrosis, or iron overload disease. You can secretly check your baby’s paternity and learn its gender in as few as six weeks from conception. You can go to Target.com and buy a genetic home collection kit for the whole family, with results that tell you how your DNA makes you unique. Two new companies, Navigenics and 23andme, even go beyond individual tests and promise to bring whole-genome scanning home to everyone.
“Your genes offer a road map to optimal health,” Navigenics asserts in its marketing blurb. When consumers get their hands on information about their own genomes, according to this Silicon Valley startup, they’ll be able to find out what conditions they’re likely to get later in life and start preventing them now.
But what kind of a road map do our genes really offer?
Despite the glamour and the hype, many geneticists are turning away from DNA as the definitive answer to health and longevity. Understanding disease, they have come to believe, will require more than pinpointing the responsible polymorphisms, or genetic variations. High-throughput sequencing tools have dramatically boosted reports of gene-disease associations. Yet there are, warns Muin Khoury, director of the National Office of Public Health Genomics, in a letter with four colleagues to PloS Medicine, troubling inconsistencies in the underlying studies as well as the systematic reviews that cite them. In many cases environmental contributors to disease, like smoking in lung cancer, far outweigh variation in genetic susceptibility.
The original funders of the Human Genome Project—the National Institutes of Health and the Department of Energy—increasingly describe human health as a complex equation built from many factors that interact with an individual’s DNA. Chemicals in the environment can control gene expression. Behavior and diet can alter the chromosomes without any change in the nucleotide sequence itself—yet shift gene activity for generations. Biological networks outside the cell can influence its internal operations. No one is throwing away the genome—geneticists are leading this science, after all—but the latest research is making it clear that our DNA is neither a book of truth nor a magic guide to treatment decisions.
Even so, the message of increasing complexity hasn’t yet filtered out of the lab and into the consumer marketplace. “The only support for this kind of testing right now is genetic determinism,” says Edward R.B. McCabe, who heads UCLA’s pediatrics department and is also physician-in-chief of Mattel Children’s Hospital. The whole-genome screens may be worst of all, according to McCabe, potentially sending people off on an anxious, endless search for answers to unexplainable single-nucleotide variations. “People are getting these tests because they believe the science is all-powerful. They do not understand the nascent state of genetics.”
Why would they? The rapid-fire announcements that tie disease susceptibility to specific gene variations haven’t slowed. Science-savvy consumers want to know how the findings relate to their own lives. They may not get answers from their doctors, who often have a thin knowledge of and lack of comfort with genetics. People also may be skeptical about the ability of the healthcare system to apply new genetics tools when time is short and costs high.
Consumers have become more comfortable taking a stronger hand in their own healthcare decisions, spurred by trends such as direct-to-consumer pharmaceutical ads and pre-tax health savings accounts. Home tests for pregnancy, hepatitis C, prostate-specific antigen, and lipid levels can easily be found at the drugstore or on the Web. Even HIV screening, once thought only appropriate to deliver in a health care setting, now can be handled through a mail-in kit and follow-up call to learn the results. Going online to seek and to share health information is common, with four of five Internet users having used the Web for this purpose. Sites that sell at-home genetic tests highlight these trends and inform consumers that they can now get cutting-edge personalized medicine in the privacy and comfort of their own homes.
The flood of news and marketing and the lure of the genome have generated a sort of Wild West of online genetic testing. Demand does seem strong, but the U.S. Food and Drug Administration (FDA) has taken little responsibility for quality control. Genetic science is exploding, but the significance and usefulness of these tests remain to be seen. In some cases, it’s even questionable whether any DNA is being analyzed. The very proliferation of testing options and claims has created an opportunity for startups such as Redwood Shores, California-based Navigenics, which promises to cut through the confusion and deliver consumer-directed genetic screening services tied directly to clinical medicine. In a similar vein, 23andme in nearby Mountain View offers to help “make sense of your own genetic blueprint.”
Besides the genetic profiling companies, the health-related direct-to-consumer testing options available today can be lumped into four general categories:
· On the periphery: This group capitalizes on wellness and complementary medicine trends. Nutrigenetic tests, for instance, offer to pinpoint your nutritional needs by checking for gene variants associated with metabolism, nutrient absorption, and antioxidant defense. Last year, the Government Accountability Office (GAO) checked out these claims by creating 14 fictitious customers and sending in DNA from a 9-month-old baby girl and an unrelated 48-year-old man to four web sites. The results that came back seemed tied more closely to lifestyle factors that investigators had made up for the fake test-takers than to the genetic material in the real samples. The GAO concluded that these companies “mislead consumers by making predictions that are medically unproven and so ambiguous that they do not provide meaningful information to consumers.” While the results stirred calls for more regulations, so far nothing has changed.
· Complex and hard to predict: Another set of tests searches for DNA variants associated with a risk for variable and complex disorders such as Type-2 diabetes. Geneticists are vigorously debating the value of gene tests for these conditions as they may be no more useful than simply heeding a doctor’s advice. The diabetes test offered by San Francisco-based DNADirect, for instance, looks for a variant in TCF7L2. In one recent analysis, researchers combined TCF7L2 with two other common variants associated with diabetes susceptibility and found that their collective predictive power was very small—a disease risk of about 10 percent with all three alleles present—but might still be useful. A negative result doesn’t mean the patient can cross the disease off his or her list of concerns; a positive offers a hint but not a prediction. If you find that you have a slightly higher risk for diabetes, asks Kaiser Permanente geneticist David Witt, “What can you do about it that you couldn’t do anyway?”
· Dosage advice: A number of tests check for gene variants associated with metabolizing drugs such as warfarin, tamoxofin, and Valium. New labels on the widely prescribed blood-thinner warfarin, for instance, warn that people with variations of CYP2C9 and VKORC1 might need lower doses. Somewhere between 5 to 10 percent of people have versions that may put them at risk for life-threatening bleeding. Even with information about higher drug sensitivity, though, admonishes the FDA, doctors will still need to do regular blood tests to check proper clotting. And they must take into account age, diet, and other medications. These tests’ significance when writing prescriptions is still unclear, and no clinical trials have yet demonstrated that they offer any better protection than an alert physician.
· Powerful test, small population: Some of the first genetic tests that became broadly available to consumers through doctors’ offices also quickly went online. Salt Lake City-based Myriad Genetics sells a screen for mutations in BRCA1 and BRCA2, which are thought to underlie between 5 to 10 percent of breast and ovarian cancers. Five years ago, Myriad became the first company to promote a genetic test directly to the general population through a radio, print, and television advertising campaign. In recent months it began television ads in the Northeast that suggest that a single relative with breast cancer might be reason enough to get the test. Myriad requires a doctor to be involved in ordering online. DNADirect.com, however, legally offers the same test in 25 states and Washington, D.C., without any practitioner okay. While this screen clearly has predictive power, only about 1 in 400 women are likely to have the mutation.
The at-home tests differ from carrier testing or prenatal screening for single gene disorders such as Tay Sachs or sickle cell anemia, which do offer clear predictive value—but aren’t generally offered online. Those tests also engendered debate about potential discrimination, proper counseling and, in cases such as Huntington’s Disease, the practical usefulness of genetic knowledge.
Except in such clearly defined cases, physicians and the medical systems in which they work have been hesitant to adopt any of these genetic tests for the wider patient population because of doubts about their usefulness. Kaiser Permanente, known for its cost-consciousness as well as its attention to evidence-based medicine, was one of the first to embrace broad-based carrier screening for cystic fibrosis. But geneticist Witt, who led that effort, says the system is very cautious about predictive genetic tests. For him, the added expense takes a back seat to concerns about inconclusive results that have no true clinical interpretation and can even be misleading. “Our guiding principle is doing what’s best in terms of medical care,” Witt says.
Even though some gene variants have been strongly associated with disease, that genotype alone may not be sensitive or specific enough to warrant diagnostic or predictive use. One can go online and order a $280 test for APOE4 to check for susceptibility to Alzheimer’s disease, for instance, but the American College of Medical Genetics has recommended against using it for that purpose. Not having the gene doesn’t mean a person is safe from the disease, and so far, there is no means of protection or treatment for someone at risk. “Among genetics folks that test died out quickly,” Witt recalls.
About half of U.S. states permit direct-to-consumer genetic testing. In an alert last summer about at-home genotyping, the Federal Trade Commission (FTC) told consumers: “A healthy dose of skepticism may be the best prescription.” Some of the tests lack scientific validity, the agency said, and others provide results that have meaning only in the context of a medical evaluation. In urging consumers to be wary, the FTC called upon the authority of the FDA and the Centers for Disease Control. Neither agency, it said, is aware of valid studies that prove tests like the one for Alzheimer’s give accurate results, or that genetic studies can be used effectively to recommend nutritional choices or predict someone’s ability to withstand environmental toxins.
Such clear disapproval might seem a pathway to stronger regulation. But the FDA doesn’t review laboratory-developed tests, only commercial “test kits” that are sold to other labs or clinics. Aside from this, the agency has proposed to review only a sub-category of complex molecular prognostic tests for clinical accuracy. In February it announced its first approval, for a test called Mammaprint by Dutch company Agendia that predicts breast cancer recurrence risk.
The Centers for Medicare and Medicaid Services (CMS), which certifies labs to perform clinical tests, declined in August to require regular checks on lab proficiency, citing cost. In its petition asking for such federal review, the Genetics & Public Policy Center in Washington, D.C., cited several examples of mistakes in genetic tests. In one case, a lab gave a clean bill of health to close relatives of a man with a fatal, late-onset neurological disorder. After the lab realized that it had tested the wrong mutation, it conducted a retest and found that one relative did have a mutation that led to disease.
Most people mistakenly believe that the federal government already regulates the validity and quality of all types of genetic tests. In fact, quality testing is voluntary for most of them. Last year, two U.S. Senate bills were introduced to increase oversight of laboratories in this area, but neither made it out of committee. One, by Senator Edward Kennedy (D-Massachusetts), would categorize all laboratory-developed tests as medical devices and require evidence of clinical validity as well as proficiency testing —with special scrutiny given to DTC genetic tests. “Payers certainly want to know,” says Gail Javitt, law and policy director for the Genetics & Public Policy Center. “Does the gene have something to do with disease or not?”
Online companies often advertise the ability of consumers to manage their own genetic information and keep it private. But once a patient discloses the news to a family physician, it will likely become a notation on his or her medical record. DNADirect explains that it identifies test results only by number, also advising clients to secure insurance before taking a test. Under the Health Insurance Portability and Accountability Act, predictive genetic information cannot be considered a preexisting condition. At the same time, if they hide genetic test results from an insurer at any time, consumers can lose coverage entirely. There are no clear safeguards, either, that limit ways in which the companies themselves might store or use the DNA collected or related information. The Genetic Information Nondiscrimination Act passed the U.S. House of Representatives by an overwhelming majority last spring but must be approved by the Senate and signed into law.
Many bioethicists have raised concerns about consumer vulnerability to overblown promises of test value and understatement of risk. “Companies [are] exploiting a climate of genetic determinism and public anxiety to sell speculative technologies,” writes Byrn Williams-Jones, a medical bioethicist at the University of Montreal, in a critique published in New Genetics and Society. Others have more confidence in the public’s ability to spot an imposter. Benjamin Bates, communications studies professor at Ohio University, concludes from focus groups and interviews that people understand the general principles of genetics and, on their own, think about these tests in light of environmental, behavioral, and family history influences. “We’ve become very good at processing incomplete information and making the best evaluation for this situation,” he says.
Even if consumers pick and choose wisely, can technically difficult, emotionally charged medical care be delivered electronically? Witt and other specialists question whether the companies that sell tests directly to consumers offer adequate pre-test counseling, to check whether a test is appropriate, or follow up with useful interpretation afterward. With genetics becoming ever more complicated, people need help placing a result in the context of their family history, deciding on next steps, and accounting for family-wide consequences.
Barbara Loebel, a genetic counselor at Kaiser, says it’s certainly possible to counsel someone over the phone and do it well. But counseling is not always offered, and Loebel raises concerns about the time and care a small online genetics company can devote to a patient’s psychological needs. “Holding and handling the anxiety, dealing with the equivocal, waiting. There are definitely human connective counseling pieces that are going to get short shrift,” Loebel says. When patients end the call and are left with a printout, they’ll be on their own to try to integrate whatever information they can glean into their own medical care.
Today’s patients do expect to have more control over medical decision-making and often do the research to support that. While Loebel used to be able to guide a patient step-by-step through a clinical plan, she says, now “sometimes they jump you three steps and raise you one.” Recently a geneticist she works with logged onto a computer in the examination room and began showing parents Google images and journal citations about their child’s suspected genetic condition. Like him, Loebel suggests, other physicians must embrace patients’ medical use of the Internet and partner with them in that realm. “The paradigm is changing,” she says. “You have to accept the fact that this is here.”
Sally Lehrman is an award-winning reporter for many publications and broadcast media including Scientific American, Nature, Health, the Boston Globe, Salon.com and the DNA Files, distributed by NPR. Based in Northern California, her honors have included the 1995-96 John S. Knight Fellowship and a shared 2002 Peabody award and Peabody/Robert Wood Johnson Award for excellence in health and medical programming.