In November the Food and Drug Administration (FDA) warned the direct-to-consumer gene testing company 23andMe to stop marketing its $99 genotype screening test. The FDA's letter gave 23andMe 15 days to respond, warning that if the agency didn't like what it heard, the potential results could "include, but are not limited to, seizure, injunction, and civil money penalties."
Two weeks later, evidently to fend off further regulatory ire, 23andMe announced that it was suspending its Personal Genome Service for the duration of the FDA's review process. There will be no new DNA-testing customers until the federal government says it's OK.
The FDA worries that purchasers of 23andMe's personal genome services will do something dangerously stupid in reaction to what they learn. The letter suggested, for example, that customers could be misled by the results of a test for BRCA breast cancer genes. "If the BRCA-related risk assessment for breast or ovarian cancer reports a false positive," the agency argued, "it could lead a patient to undergo prophylactic surgery, chemoprevention, intensive screening, or other morbidity-inducing actions, while a false negative could result in a failure to recognize an actual risk that may exist."
Researchers around the world use the same biochip that 23andMe employs-the Illumina OmniExpress-and find that it provides highly accurate results. Still, all diagnostic tests produce false positives and false negatives, so the question becomes: What is unique (and uniquely dangerous) about patients obtaining this information through 23andMe?
First of all, the company clearly states on its website that it screens for only three of the more prevalent BRCA breast and ovarian cancer mutations, and emphasizes that there are hundreds of potentially carcinogenic mutations in the BRCA1 and BRCA2 genes for which it does not test.
But suppose a customer tests positive for one of the BRCA mutations. What is she likely to do next? FDA bureaucrats seem to think she will immediately search for a surgeon to remove her breasts or start gobbling tamoxifen pills to ward off cancer. But those scenarios are ridiculous. The most likely outcome is that the customer would first consult with her physician about the results and then get another, more comprehensive genetic test. That is exactly what Huffington Post blogger Jill Steinberg did when her 23andMe results came back showing that she carried one of the BRCA mutations. Her subsequent test confirmed the 23andMe finding, and she eventually decided to have a preventive double mastectomy.
What if the 23andMe test missed one of the three mutations it's designed to detect? In that case, a customer might indeed wrongly assume that she is not an especially high risk for breast cancer. But let's put that error in context. Consider, for example, the OraQuick at-home HIV test, an FDA-approved product that the agency acknowledges produces "about one false negative result out of every 12 tests performed in HIV-infected individuals." A person who wrongly relies on the negative HIV result, like the person with an incorrect negative BRCA result, may be harmed by not seeking appropriate medical care.
The FDA does warn users of the HIV test that it is "important never to use a negative test result to decide on whether to engage in behavior that puts you at risk for HIV infection." Similarly, 23andMe explains to its customers, "The absence of these mutations does not rule out the possibility that a person may have another genetic variation that increases the risk of these diseases." Why should one test be permissible and the other not?
In 2013, 23andMe researchers conducted a peer-reviewed study of how customers reacted to their BRCA results. It found that those who tested negative "did not report inappropriate actions, such as foregoing cancer screening." The study also noted that "all but one of the 32 mutation-positive participants appreciated learning their BRCA mutation status." Similarly, a 2009 Genetics in Medicine study analyzing how patients responded to the results of a suite of eight genetic vulnerability tests reported, "We found no evidence that those who considered or sought testing were inclined to overestimate the contributions of genetics to common health conditions or to underestimate behavioral risk factors."
The FDA also suggested a scenario in which 23andMe customers would learn how sensitive they are to the blood-thinning drug warfarin and then use the results to "self-manage their treatments through dose changes," producing dangerous blood clots or bleeding incidents. As it happens, my 23andMe results indicate that I have increased sensitivity to warfarin. Fortunately, I don't have to use the drug right now, but I certainly would let my physician know these results in the event that I needed such treatment.
23andMe tests for the same variants relevant to warfarin sensitivity that tests approved by the FDA do. (By the way, stand-alone warfarin sensitivity tests cost between $200 and $400, while 23andMe's warfarin sensitivity results are included in its $99 fee.)
In any event, I seriously doubt that 23andMe customers who are not using warfarin will suddenly start anti-coagulant therapy for the fun of it. Nor are those currently on warfarin likely to stop listening to their doctors and begin to self-treat by swallowing d-CON pellets. 23andMe itself offers this advice at the top of the webpage reporting my warfarin results: "Only a medical professional can determine whether warfarin is the right medication for a particular patient. The information contained in this report should not be used to independently establish a warfarin regimen, or abolish or adjust an existing course of treatment."
23andMe has not always behaved prudently in this perhaps-inevitable conflict. The FDA's letter noted that the agency "has not received any communication from 23andMe since May," although the company rolled out a big new marketing campaign for its personal genome service in the fall. It is not very smart for a company to snub the regulators who have the power to shut it down at a whim.
But when it comes to the company's relationship with its customers, it's hard to find fault. So far, nearly 500,000 people have purchased 23andMe's genotype screening tests. The FDA's warning letter does not mention any substantial customer complaints about the service.
The development of all new technologies involves a social learning process in which some early adopters try it out, explain to others how it works, and find its flaws, which newer innovators then fix. The genetic testing companies that tease out and explain useful information about disease risks and pharmaceutical interactions will be the ones that succeed. Despite the FDA's parade of horrible hypotheticals, the bigger risk is that increased regulation will slow down the progress of beneficial genetic testing and treatment.