There are 24 days in Advent this year, not 23, but the Royal Institution is celebrating 23 anyway, with an absorbing Advent calendar that takes you through the karyotypes of our 23 chromosomes. On the other side of the Atlantic, the appropriately named 23 and Me genomics company is having a much less festive season.
23 and Me sells gene tests: it sends its customers (who number more than 500,000, and include me) a tube; they send back spit, and the company takes SNPs across the genome and returns a report. They’ll tell you whether you have the allele for grey or caramel earwax, for example, or for curly hair. But they’ll also tell you which ApoE allele you have, and whether you have one of a couple of BRCA mutations. The ApoE 4 allele sharply increases your risk of Alzheimers; if you have one of two BRCA mutations, your risk of developing breast or ovarian cancer before your 70th birthday is about one in two.
On 22 November, the FDA wrote to 23 and Me CEO Anne Wojcicki, telling her that it considered 23 and Me’s spit tube a diagnostic or treatment device requiring pre-marketing approval for every trait tested for. The FDA is “particularly concerned” that people might act on false positives, undergoing “prophylactic surgery” (i.e., the removal of breasts and ovaries). Note that quite a few commentators are unconvinced that the tube is a device at all.
23 and Me has responded by continuing to provide new customers with ancestry information (they really will tell you how much of a Neanderthal you are) and genetic information, but not “health-related information” - the middle bit that interpreted your results and told you what your alleles were likely to mean. This won’t protect customers, and it certainly won’t deny them the information that might prevent them following up. It will just make it harder to get: instead of clicking through a website, you’ll spend hours - and hours, and hours: the company tests for 240+ traits - on PubMed and Wikipedia.
The FDA’s actual examples are ludicrous: it’s hard to imagine a surgeon wading in without verifying a BRCA mutation, especially in the litigious US. So why is it coming down so hard on one biotech startup?
Certainly, 23 and Me has irked a lot of people. The American Medical Association, like several related professional bodies, has issued a position statement recommending that genetic testing involve the “personal supervision” of a healthcare professional; the usual counter-argument is that preventing access is paternalistic, and also contrary to the public interest in enabling people to make lifestyle changes to their health. (Bloss and colleagues survey the arguments here.) The Sacramento Bee goes even further, suspecting sinister purposes: 23 and Me, it says, is “really a way to create a giant genetic database that 23andMe can sell to drug companies and researchers to develop cures that can be patented.” It’s not a crime to want to profit. (Where do you think drugs come from?) And 23 and Me is pretty open about its data-sharing model, for which it obtains consent.
Economist Eric Crampton takes it further: "The first temptation will be searches for matches against samples from crime scenes. Even identifying the close kin of criminals can help pin them down." But that isn't a reason for a government bureaucracy to march on 23 and Me.
In fact, the FDA coming down on 23 and Me isn’t really about protecting babes in the woods, and it’s not about an upstart David fighting Goliath on behalf of the armies of God. No: it's about Canute at the tide. The FDA can chase 23 and Me, and it can stomp slowly over its competitors. But, as chemical printing, open-source synthetic biology and, yes, genomics gather speed, the intersection of health and personal autonomy is going to effervesce. You can’t regulate a wave top-down. What are the 14,648 people at the FDA going to do about that?