What’s better than a holiday card? “Possibly the worst FDA letter of all time.”

Building on Lukas's post here yesterday on the FDA's warning letter to 23andMe, I asked Dr. Margaret Curnutte, a postdoctoral fellow at Baylor School of Medicine, to write a guest post on the topic. Maggie wrote her dissertation on the business and politics of direct-to-consumer genetic testing and has continued to follow the next generation genome sequencing industry closely.

Holiday greetings from the Food and Drug Administration (FDA) came a little early this year for direct-to-consumer (DTC) genetic testing firm, 23andMe. All seemed to be quiet since 2010 when 23andMe and a handful of other DTC firms came under the scrutiny of Congress and FDA. The ebb of critique had subsided so much that it became unfashionable at academic meetings to mention anything DTC. And now FDA’s most recent letter to Ms. Wojcicki, 23andMe’s CEO and co-founder, has put the company back in the spotlight. Time to write another chapter in the DTC saga.
Seasons Greetings from the FDA!
     In its recent letter FDA informed the firm that its Personal Genome Service (PGS) is without marketing clearance and in violation of the Federal Food, Drug and Cosmetic Act. To translate, back in 2010 FDA had sent a similarly friendly letter to tell the company that its service was a medical device under said law. This meant that depending on how the device was classified (i.e. Class I, II, or III) it was subject to premarket (in practice post-market) approval. This sort of approval seems to present an existential crisis, but it looked like 23andMe had been trying to work through this “premarket” approval while continuing to be fully operational. Then on November 22, 2013 FDA asked the firm to immediately discontinue marketing of PGS. The major concerns have been and continue to be about the analytical and clinical validity of the service, meaning its accuracy and reliability.

     That’s a brief synopsis of the most recent drama. My interest in 23andMe dates back to 2007 when I had just started my PhD program in Milan and I received a (real) physical newspaper clipping all the way from California. Now when I think back on it I’m amazed the letter made it through the Italian post. My mom (yes, I will give her credit for helping me find a dissertation topic) sent me an article from the Palo Alto Daily News about how a new company, 23andMe, had just begun offering a form of genetic testing online. Without the mediation of a physician, people could access genetic testing over the Internet. I was fascinated and then began many years of meticulously following the practices of 23andMe as well as other DTC providers. Since then, I have expanded my research to look at the next generation sequencing industry, but I will talk about that later.
      Beginning yesterday morning my inbox has been flooded with emails from colleagues and friends over the recent news. My favorite was an email my dad, who works in the biotechnology industry, forwarded from one of his colleagues titled, “possibly the worst FDA letter of all time.” I think that has been the best summation of FDA’s newly found firmness on DTC genetic testing. This is the sort of letter you dread if you are part of the biotechnology industry. FDA’s unyielding power can make or break a company.
      So, what does this most recent development mean? I was first asking myself why this letter arrived on 23andMe’s doorstep now. I can’t help but think the recent press around genetic testing for the BRCA gene mutations plays a part in the story. Between Angelina Jolie and the recent United States Supreme Court ruling on the legitimacy of Myriad’s patents on these genes, we have been inundated with messages about the potential medical significance of BRCA testing. Then you look to 23andMe’s website, and for a mere $99 one can directly purchase testing for the BRCA mutations amongst hundreds of other genetic tests as part of their Personal Genome Service. In the most recent letter FDA mentions the potential significance of a false positive BRCA test. One might, for example, undergo a prophylactic double mastectomy. FDA also highlights what it sees as the “potential health consequences” of pharmocogenetic testing, meaning genetic tests that indicate how one will respond to a drug.
      The introduction of pharmocogenetic testing into their Personal Genome Service deserves a little bit of a backstory. When 23andMe first started out there was one alphabetical list of conditions and traits, mixing tests for one’s earwax type and asparagus metabolite detection with one’s predisposition to lactose intolerance, Crohn’s disease, and prostate cancer. In the beginning there were only two categories of tests, “Clinical Reports” and “Research Reports.” Through the company’s own (publicly available and obviously transparent) methods of scientific curation, the “Clinical Reports” were deemed more scientifically sound than the “Research Reports” based on their assessment of the relevant literature. Over time it continued such a distinction, but began to break down their tests into four categories, “Carrier Status,” “Drug Response,” “Traits,” and “Disease Risk.” At the beginning of 2010 I was amazed that almost overnight, with little notice, the company had gone from something that many termed “recreational” to offering information about how one might process a particular drug.
      This has always seemed like a particularly bold move to me since the company, from its inception, has treaded a fine line between claiming to provide something that is “health related,” yet not medical. The reasons for doing this should be fairly obvious. If your business model relies on directly accessing the genomes of consumers, who wants the medical establishment to get in your way of convincing people to pay to put their genome in your proprietary database?
      Flashing forward now to FDA’s regulatory stance, something I’ve given a lot of thought to over the years. Does the agency’s most recently expressed concern over the analytic and clinical validity of PGS mark a departure from previous concerns? Not really. As FDA has made clear in the past through their correspondence with DTC firms, the worry with a medical device is that it be accurate and reliable. That is FDA’s rationale for post-premarket review. But back in 2010 Guitierrez, the head of FDA’s Office of In Vitro Diagnostics, did not want the agency to be viewed as paternalistic and said, “We really don’t have any issues with denying people information. We just want to make sure that the information they are given is correct.”1 So, now we are back to worrying about analytic validity (the reliability of the actual genetic test, meaning laboratory performance) and clinical validity (whether the genetic variant actually corresponds to the condition or trait). In the most recent correspondence FDA clarified that 23andMe has not provided sufficient evidence to establish the reliability of their service even after the agency’s repeated requests to do so. Although FDA has intervened on medical grounds by classifying PGS as a medical device, their regulatory stance has and continues to translate into concerns about accurate information, not its medical significance.
      With an emphasis on the accuracy of genetic information, DTC providers have made the case that genetic information is just a form of personal information similar to one’s bank statement. Both kinds of information, they argue, provide useful descriptions of personal circumstances. Elsewhere I have shown that there is alignment between the companies’ discursive strategies and the regulatory emphasis on the accuracy of genetic testing results. 2 This emphasis has furthered the notion that genetic information is simply personal information to which the consumer is entitled, and that the role of the state can and should be to protect the consumer from fraud and misrepresentation. I find it interesting that FDA has not taken a stronger position that such testing can only be provided in a medical context, like Germany, for example, but rather that if 23andMe is to continue providing its service, the onus is on the company to establish test validity for the consumer.
      If, or I suppose when, 23andMe jumps this regulatory hurdle, it has huge implications for the next generation sequencing industry. Most FDA-cleared genetic tests are for a single disease while 23andMe’s would be the first to test for multiple conditions. This is the direction in which next generation sequencing services are headed. If you sequence an entire genome or exome, the point would be to simultaneously look at multiple conditions. FDA has only just begun to regulate this space. 3 In the meantime you have companies like DNA DTC, a subdivision of ancestry company, Gene By Gene, offering DTC whole genome sequencing for about $7,000, but notably without interpretation: “Customers are provided with their raw data as a fastq file. Please note that as a Research Use Only test, the whole genome does not include any analysis or interpretation.”4 Presumably FDA would have a hard time classifying this service as a medical device since they do not offer interpretation and analysis.
      In the midst of a proposed regulatory approach that emphasizes the technical aspects of genetic testing, which in turn could allow for the expansion of this new market, where does the genetic consumer stand? While 23andMe has positioned its service as empowering consumers, against the limitation of patient’s rights, I would like to conclude by pointing to the fact that the consumer is being constrained by a new set of actors. While the patient was subject to the authority of the physician, the consumer is now subject to the interests of private industry. I will continue to be interested in how this industry-regulatory dynamic shapes our rights and entitlements around genetic information.
1 Pollack, Andrew. 2009. Google Co-Founder Backs Vast Parkinson’s Study. The New York Times, March 12, sec. Business.
2 Curnutte, Margaret, and Giuseppe Testa, “Consuming genomes: scientific and social innovation in direct-to-consumer genetic testing,” New Genetics and Society 31:2     (2012): 159-181.
3 http://blogs.fda.gov/fdavoice/index.php/2013/11/gene-sequencing-devices-are-next-generation/
4 http://www.dnadtc.com/#jumpNav6


Thank you for this analysis. Very enlightening. I'm wondering if the kinds of testing that the DTC market has been offering are currently available with a physician order from other labs? If so, how would they be affected by the new FDA mandate? If the FDA is concerned about the accuracy of both the testing and the interpretation, then simply offering the same service through a physician's office could potentially improve the interpretation, but wouldn't change the validity of the testing itself. Would you agree?

That is a great question. FDA cannot interfere with the practice of medicine, so it seems unlikely that they would intervene on a clinician's analysis and interpretation of a genetic readout. If the test is for clinical purposes it should be processed by a CLIA certified lab, which is meant to ensure accuracy of lab processing. I am pretty sure 23andme's samples go through a CLIA certified lab, so yes, the real concern would be about the validity of the interpretation.

Note: Only a member of this blog may post a comment.

back to top