If 23andMe Falls in the Forest, and There’s No One There . . .

FDA v LDTGenomics Law Report has paid close attention to the FDA’s potential regulation of laboratory developed tests (or LDTs) over the years. We have decided to address the most recent development – a cease and desist letter sent by the FDA to 23andMe – in two posts — by Jennifer Wagner and by John Conley.

On November 22, 2103, the Food and Drug Administration issued a “warning letter” to Google-backed 23andMe demanding that the company “immediately discontinue marketing” its Saliva Collection Kit and Personal Genome Service (PGS), a direct-to-consumer (DTC) genomic testing service. The letter states that 23andMe markets the PGS as a source of “health reports on 254 diseases and conditions,” and “specifically as a ‘first step in prevention’ that enables users to take ‘steps toward mitigating serious diseases’ such as diabetes, coronary heart disease, and breast cancer.” Consequently, the FDA concludes, the PGS is a medical device subject to FDA regulation, as opposed to a mere source of information, as 23andMe has long contended. Because 23andMe has not obtained required approvals, the FDA claims the authority to order its removal from the market.

The letter has already generated tremendous interest in the media and among investors. Jen Wagner and I have been deluged with questions from both groups. Here are my thoughts on some of the most frequently asked questions. As you’ll see, my overall reaction is: calm down, take a deep breath, it’s not that big a deal. Jen has posted her views here.

1. What’s a warning letter? The short answer is: just what it sounds like. The FDA expresses its current views on the relevant law and facts, then warns the recipient against engaging in practices that it believes are illegal. The letter itself has no final legal effect. In this case, if 23andMe doesn’t stop marketing the PGS kit without first getting FDA approval, the FDA would have a couple of legal remedies to pursue. It could impose administrative sanctions, including “device detention”—impounding 23andMe’s stock of PGS kits. That decision could be appealed in court by 23andMe, which would result in a court ruling of the underlying questions of the FDA’s right to regulate, the standards to be applied to this product, and whether 23andMe’s marketing violates those standards. The FDA could also go directly to court to seek an injunction, an order prohibiting further marketing; this strategy would also yield judicial rulings on the underlying questions.

2. What are medical devices, and how does the FDA regulate them? A medical device is any product that is intended for use in the diagnosis or treatment of disease or other health conditions. All of these terms are very broadly construed, and the category can include anything from tongue depressors to artificial hearts. The FDA categorizes devices in three risk-based classes, with Class I the least risky and Class III the riskiest. Most Class I devices are exempt from most regulation. Class II devices require that the FDA be given Premarket Notification (often called 510(k) notification). The agency will give marketing clearance if it finds that the device is “substantially equivalent to one that already being legally marketed. Class III devices, which carry the most risk, must get Premarket Approval, which is much like the drug approval process and requires proof that the device is safe and effective. In the warning letter, the FDA says that 23andMe’s PGS is a Class III device.

3. What does the FDA claim that 23andMe has done wrong? The FDA’s principal grievance is that 23andMe has utterly failed to cooperate in the regulatory process. The letter notes that 23andMe submitted 510(k) notifications for multiple PGS uses more than a year ago, but has subsequently failed to provide repeatedly requested data and other responses. As a result, the FDA says, it cannot even tell how to classify many PGS uses. Moreover, some uses, “such as assessments for BRCA-related genetic risk and drug responses,” are “very concerning,” but the company has failed to furnish sufficient validation data. The claims of non-cooperation are exceptionally detailed and forceful.

4. What will be the effect on 23andMe? Even though this is just a warning letter, the company seems headed for serious regulatory trouble. It has three options: stop marketing the PGS kit, cooperate with the FDA in the hope of getting approval, or forge ahead without approval and challenge the FDA’s authority in court. The first amounts to surrender. The second approach could have a happy ending, but it’s clear from the letter that the agency doesn’t trust 23andMe. The approval process is long and complex under any circumstances, and likely to be even more so in this case. The third is the nuclear option. At best, it will take a long time, and I think at this point that the courts would be likely to come down on the FDA’s side.

5. What will be the effect of this letter on the broader DTC market? My initial response to this question is: What market? With principal competitors deCode (acquired by Amgen for its database) and Navigenics out of the business, 23andMe appears to be the only surviving major player in the health-related DTC market. Although 23andMe claims almost half a million users, its results have fallen far short of initial optimistic expectations. Google money can keep 23andMe alive indefinitely, but I’m extremely skeptical that this market has a future. Who are the customers? There is undoubtedly a hard core of genetic hobbyists, but how many people can there be who (1) know enough to seek out a genetic testing product, but (2) know so little that they’d rely on a cheap and vastly underpredictive consumer product to make heath or reproductive decisions? Maybe I’ll be proven wrong, but I just don’t see who these people are.

The FDA’s position that DTC tests are medical devices that it can regulate is not new—it notified the then-competitors to that effect in a 2010 letter (analyzed by Dan Vorhaus for GLR here). The current letter to 23andMe does not make a blanket claim that DTC tests cannot or will not be cleared for marketing. Instead, it focuses primarily on 23andMe’s uncooperative conduct (essentially admitted by the company in its public response), as well as some of the stronger claims that the company makes on its website about the health value of PGS. So the warning is very specific to the facts of the case, and stops short of broad new legal pronouncements.

Some of the FDA’s claims about risk also seem a bit overwrought, perhaps triggered by institutional irritation toward 23andMe. It is certainly true that a customer/patient might put unwarranted reliance on a false negative, or (less probably, I’d think) might adjust a drug dose in response to an erroneous finding about drug response propensities. But the letter suggests that a patient could “undergo prophylactic surgery” simply in reliance on a PGS false positive, which seems absurd—wouldn’t a doctor intervene? Overall, a more cooperative company making more modest claims might well expect more sympathetic treatment.

6. Who’s right? I’m a health care libertarian. I rarely buy paternalistic arguments for keeping health information away from patients. The DTC genetic testing market is a limited one, and I don’t think the customers are as dumb as the FDA seems to assume. In addition, most of 23andMe’s customers reportedly contribute their samples to a research biobank, which is a social good. So personally, I’m rooting for 23andMe. But legally, if things go that route, I think the FDA would prevail. I think the courts would ultimately accept the FDA’s determination that DTC test kits are medical devices that it has the power to regulate.

To sum up, three thoughts: (1) 23andMe’s conduct in its dealings with the FDA is inexplicable, unless it is trying to get to court to litigate the FDA’s basic claim that it can regulate DTC products like PGS as medical devices. (2) The FDA has not yet shown an intent to kill off the entire DTC genomic testing industry. This letter is only a nonbinding warning, and is very specific to the unusual facts of this case. (3) The market may do the job long before the FDA has a chance to.

Filed under: Badges, FDA LDT Regulation, General Interest, Genomic Policymaking, Genomics & Medicine, Genomics & Society, Legal & Regulatory