Exclusive: What Really Drove 23andMe And The FDA Apart

Is getting a $99 DNA test really hazardous to your health? Our editor-in-chief decodes an innovation dustup.

When we put 23andMe CEO Anne Wojcicki on the cover of Fast Company’s November issue, we didn’t expect the FDA to yank the company’s $99 DNA test off the market a few weeks later. Yet there’s a reason we called Wojcicki “the most daring CEO in America”: Her company’s product challenges established conventions about medical practices, research, insurance, and more. So is that challenge at the heart of the FDA’s rebuke? Or did something more basic derail Wojcicki’s mission?

Anne WojcickiPhoto by Jeff Brown

Since the FDA shutdown, I've spoken with insiders at both the agency and 23andMe. I learned that there are significant philosophical differences between the two organizations. I also concluded that the break between them had as much to do with old-fashioned miscommunication as anything else.

Three Points Of Conflict

There are three core points of dissonance between the FDA and 23andMe. The first is their level of faith in consumers. The FDA is tasked with protecting consumers, even from themselves. Among the objections the agency has to 23andMe is that DNA test results may provoke consumers to seek unnecessary treatments, self-medicate in dangerous ways, and take on needless anxiety about conditions for which there is no treatment. In contrast, 23andMe stresses that consumers are capable of managing their own medical care—and should have more information about it. The company highlights specific instances where customers’ DNA tests led to helpful, positive medical interventions. The FDA’s stated concerns at this stage revolve around hypotheticals, not observed cases of consumer-created medical mistakes.

The second core point of dissonance revolves around data. The FDA has been asking 23andMe for specific studies on the reliability and efficacy of its DNA test results as well as specific clinical analysis regarding health claims (for example, the risk of developing a condition that is associated with a genetic marker). 23andMe says it is working on satisfying those mandated requirements and admits it wishes its efforts were farther along. At the same time, 23andMe contends that the reliability of its test kits and data adheres to the highest standards, that its process matches that of a DNA test one would get through a physician. The fact that not all of its studies yet match the medical industry’s accepted clinical study procedures—and thus fail to meet FDA protocols—is, to 23andMe, not something that should be held against it.

The third core point of dissonance concerns education: What is the best way to prepare consumers for a future where DNA testing and information is more prevalent? 23andMe casts itself as a consumer-friendly way to provide useful information. If we are moving to a world where DNA sequencing is cheaper and used more frequently, then 23andMe contends it is preparing the way, introducing this new idea in a nonthreatening package. The FDA also sees this future coming: Around the same time 23andMe was curtailed, the agency approved a new gene sequencing technology from Illumina, the company that provides the back-end testing for 23andMe’s kits. What the FDA worries about is setting poor precedents for this coming wave. If any entity needs to be vigilant about creating appropriate standards for marketing health claims, it is the FDA.

What Really Drove Them Apart

Despite these three areas of tension, 23andMe and the FDA had been cooperating relatively smoothly. The FDA’s shutdown letter notes 14 meetings between the two entities—evidence that both sides have been trying to reach agreement. 23andMe launched more than five years ago, and the FDA allowed initial marketing of the DNA test kits, despite what it felt was incomplete data. Many promising medical devices and products are given leeway as real-world data builds.

What happened this fall was sparked less by philosophical or scientific differences than poor communication. Last spring 23andMe began to ramp up the marketing efforts behind its DNA kits. This included TV and video ads and a direct-marketing campaign. Given the long history of dialogue with the FDA, you might imagine that 23andMe would have sought feedback about these initiatives—to see if any concerns surfaced or tweaks were requested. Instead, 23andMe announced its plans to the FDA and proceeded independently. Perhaps the company did not want to set a precedent itself of having to get FDA approval for their marketing. In any case, the company charged ahead, heedless or unaware of the discomfort this created at the FDA.

The FDA, in turn, took no steps in the wake of this marketing wave to reach out to 23andMe and suggest caution or areas of concern. The agency had previously warned 23andMe about a deadline for submitting certain studies; when that deadline passed without the studies arriving, the agency began its internal systematic process toward suspending 23andMe’s tests. The agency felt it had given 23andMe the benefit of the doubt on its science for years (which is true). To be ignored after all that, the agency began to question 23andMe’s motives. Were sales more important than science?

In the end, there were several simple ways the FDA suspension might have been avoided: a call or visit from Anne Wojcicki; or a recognition from the FDA that, given turnover at 23andMe (and Wojcicki’s distraction due to her public separation from husband Sergey Brin), the task of initiating outreach fell to them.

But neither happened. The two sides fell into adversarial roles—like teenagers that hear about each other only through the eyes of others. It’s a common plot line in movies and TV shows, one that is invariably solved when the two sides finally meet and clear the air. The parties are now in active discussions, and 23andMe is working round the clock to get its missing submissions in order in advance of a December 16 deadline. With any luck, the parties in this case will rediscover their common ground and resolve their differences.

I can only repeat the perspective from my original editor’s letter: “There are deep and varied questions about how the information being unleashed by 23andMe will disrupt—for good and bad…We cannot definitively declare that 23andMe’s specific business will triumph. But if not 23andMe, then a company like it will eventually succeed in reconfiguring many of our most basic assumptions about privacy, preventive care, and insurance.” We stand by that assessment.

[Cell Splitting: Giovanni Cancemi via Shutterstock]

Add New Comment

11 Comments

  • gene

    Mistake in the article. FDA approved illumina Miseq sequencer. 23 is using gene chips from the same company. The two technologies are unrelated.

  • Dan Munro

    The heart of the conflict is really your second point. The function of the FDA in these cases is data accuracy - not being the "nanny" police (points 1 and 3).

    The evidence around errors from the 23andMe test are far more than just ancedotal - and something I wrote about here:

    Class Action Law Suit Filed Against 23andMe http://onforb.es/1bcr0BI

    These aren't just minor errors - these are fundamentally wrong results - which is the entire argument behind having scientific rigor - and FDA approval.

    The "dissonance" also goes well beyond simply a failure to communicate. There's a bit of a revolving door at the company with key personnel leaving - with "red flag" timing.

    Case in point is the departure of Chief Legal Counsel (overseeing the FDA negotiations) Ashley Gould. At the time that the company filed for FDA approval (summer of 2012 - for only 7 specific tests), she was quoted as saying this:

    “The result that you get when you review a report, the FDA wants to ensure that’s a valid result." Ashley Gould - Chief Legal Counsel - 23andMe (July, 2012)

    After more than 6 years with the company - Ms. Gould's departure was this last July (2013) which seems to coincide pretty closely (3-4 mos?) with the abrupt end of FDA communications.

    At least visibly - there has been no replacement for Ms. Gould's position as Chief Legal Counsel - none of which is accidental or coincidental. It's not as if the company is running out of funds.

    I do honestly hope that company is able to navigate to FDA approval. Until then - my personal vote is 23andNotMe.

  • Patrick

    This is yet another demonstration of why the FDA should be privatized and operate more like Underwriters Labs. If a company doesn't choose to get FDA certification and people are still willing to buy the product, a bunch of bureaucrats shouldn't be interfering.

  • Q

    Two thing are ludicrous here, and one thing is unforgivably left out. Terrible article. Robert, you should be ashamed of yourself for writing this drivel.

    The first ludicrous thing, as already mentioned, is the suggestion that the FDA should have initiated contact with 23andMe because of Anne Wojcicki's separation with Sergey.

    Second, Anne W. fired her chief counsel, the person responsible for liaising with the FDA, and did not hire anyone to replace her. It's pretty much since that point that relations with the FDA have deteriorated. If you did not know this, you did not do any research. The values expressed by these actions were pretty much that Anne and her team did not care about the FDA. This puts the onus on 23andMe entirely for reaching out to the FDA.

    Finally, the agreement with the FDA was always that they could sell their testing kits pre-approval so long as they did not mass-market. Without consulting the FDA, they initiated a mass-marketing campaign. Unsurprisingly, by unilaterally abrogating their agreement, they got shut down by the FDA.

    This is not a two-sided thing. 23andMe is at fault.

  • Aubree

    If Anne W. were a man would her martial status even be part of this article?

  • L

    To suggest that the FDA should cut 23andMe slack because its CEO is going through a separation, is the most hilarious thing I've read in years. I've spent my career in the medical diagnostics industry. 23andMe hasn't followed FDA's processes and respected its deadlines, and is being treated as any violator should expect. I'm sure those of us who've worked around FDA aren't at all surprised by its actions.

  • F

    As someone who used to work in the Bio-tech industry, at one of the world's largest at that, I would call this a white-wash of 23andme.

    Its clear from the documentation available that FDA bent over backwards in allowing 23andme to continue to sell its unapproved test/devices.

    The problem is actually simple, and alluded to be the author.

    The 'philisophical differences' that were at play were simply that Wojcicki does not feel bound by the law. Hmm, rather quite like her ex and his pals. Birds of a feather...

    While not a big proponent of the FDA in some areas, their concerns here are actually lucid and logical.

    Multiple refusals to reply to the FDA were not done on some middle-manager's whim, it was encouraged or enforced at the highest levels.

    Why, is the question.

    Interestingly, they are already facing potential class-action for "the complaint notes that 23andMe compiles customers' genetic information which it then markets to scientists for research."

  • M

    Interesting article but I'm surprised how at the end it shifts accountability to the FDA from 23andMe. The idea that the FDA should learn about and consider a company's turnover rate and CEO's personal life distractions in its evaluation of enforcement actions seems naive. How could the FDA be effective if they treated every company that way? "Oh, your company is in internal turmoil? Don't worry about those deadlines and requirements we set. We'll work something out." This isn't an extension on a term paper, it's business compliance with gov't regulation.

  • BobSafian

    There's no question 23andMe should have been more responsive and bears the lions shame of blame for the predicament it now finds itself in. My point is simply that the FDA had an option to share the extent of its discontent, prior to official action, and it did not.