Last week, the US Food and Drug Administration told 23andMe to discontinue some of their personal genetic testing services. They were sent a letter requiring them to stop marketing their 'spit kits' because they were in violation of the Federal Food, Drug, and Cosmetic Act (FD&CA).
The saliva collection kits were being marketed as providing health information to help prevent illnesses including some of the most serious types of cancer. This use makes the tests medical devices under the FD&CA. To obtain the proper marketing authorisation, 23andMe had to submit evidence that the tests had been analytically and clinically validated for the intended uses. None of this evidence had been submitted to the FDA, who state in their letter that many of the necessary trials had not even started.
23andMe continued to market their tests as health reports, adding new indications that could be tested for, and embarking on an extensive television advertising campaign. All this, without having marketing authorisation in place, and without continuing to communicate with the FDA. Matthew Herper suggested that 23andMe were either trying to force a fight with the FDA - which they will almost certainly lose - or they had engaged in the 'single dumbest regulatory strategy' he had seen in his time.
23andMe has accumulated a database of 500,000 genomes since 2007. It underwent no ethical approval process before it started selling DNA testing kits. There are no ethicists on the advisory boards, or on the team as advertised on the website. There is no evidence of 23andMe using independent ethics advisors.
UK Biobank also has a database of at least 500,000 genomes. It is a national population tissue bank that collects genetic information from its participants, as well as detailed health and lifestyle information. Scientists can apply to the bank for access to the data to be used in ethically approved research projects. There is an independent Ethics and Governance Council overseeing compliance with the ethics and governance framework which was subject to an extensive public consultation process. UK Biobank also has approvals in place from the North West Multi-Centre Research Ethics Committee, the Patient Information Advisory Group and the Human Tissue Authority.
These are both large genetic databases. They are big enough to engage in serious population genetic research and to advance genetic research quite significantly. 23andMe have made no provision to open their database to researchers from any other institutions except for their limited call for academic collaborators in 2012, the success of which is not clear from their website. The only people with rights to use the information are those already employed by 23andMe, and those to whom 23andMe choose to sell the information. As I have argued, there are no credible governance arrangements in place. The only restriction on what 23andMe can do with their database is contained within their privacy policy, which they can unilaterally change at any time.
A small change to the privacy policy, and 23andMe could sell the database to the insurance industry, who could use it for underwriting health and life insurance. They could sell it to the pharmaceutical industry who could use it for marketing new products. They could sell it to a supermarket who could combine it with their loyalty card schemes to help target their offers. It was only last year that saw reports of Target using shopping habits to identify pregnant women. Think what they could do with a database linking genetic information to shopping habits. If that information were combined with Google or Facebook, they would be in possession of an extremely powerful resource.
Imagine that you have a genetic predisposition to heart disease, as identified on your entry in 23andMe's database, and you regularly write on your Facebook page about eating junk food while not at the gym, and you use Google to search for the symptoms of angina. If your insurance company could have access to that information before you bought your insurance, your premiums would skyrocket.
Big data of this type is powerful. It has a great deal of promise. That is why we have encouraged UK Biobank and why we support the research that is being done. But UK Biobank's data use is restricted. It cannot be used except in accordance with the ethics and governance framework. It cannot be sold. It cannot be used for commercial ends. The ethical implication of their work is regularly reviewed by an independent Ethics and Governance Council. 23andMe has none of this oversight. They have halted the sale of health-related tests for the moment, but in the event that 23andMe do gain the appropriate marketing authorisations, the collections will begin again, and eventually they will have the million genomes they are seeking. Who knows what could be done with a database that big.
Leave a Reply
You must be logged in to post a comment.