Time Out: Behind the FDA’s Decision to Halt Direct to Consumer Genetic Testing

  • Share
  • Read Later
Anne Wojcicki, co-founder of 23and Me Inc. at the 2008 Milken Institute Global Conference in Beverly Hills
Fred Prouser / Reuters

Anne Wojcicki, co-founder of 23andMe Inc.

If anyone is eager to turn the page on 2013, its Anne Wojcicki.

The founder of 23andMe, Wojcicki suffered a public split  from her husband, Google co-founder Sergey Brin when his affair with an employee became public last summer. And now she’s had a setback on the professional front. Since 2009, she and her team at 23andMe have been working hard to convince the Food and Drug Administration (FDA) to give the company’s home genetic test—which uses a saliva sample to provide a quick read of selected bits of the genome—the legitimacy that it needs to make it more than a commercial novelty.

That just become harder since the FDA sent a warning letter to the company asking it to stop marketing and selling the direct to consumer service unless it can provide satisfactory evidence that the results are reliable and won’t jeopardize consumers’ health.

“We are issuing the letter today because of the company’s failure to address our repeated concerns since 2009,” says Erica Jefferson, spokeswoman and deputy director of the office of media affairs at the FDA. “We have been working with them during that entire time frame and provided extensive feedback on the type of information they need to submit to us to make this a legally marketed product. Despite numerous requests, they haven’t submitted the information. We have concerns because the product may be potentially putting consumers at risk of receiving inaccurate information that could result in delayed or inappropriate medical care.”

A spokesperson for 23andMe said the company did not have a comment at this time. In a written response to the letter posted on its website, the company said “We recognize that we have not met the FDA’s expectations regarding timeline and communication regarding our submission. Our relationship with the FDA is extremely important to us and we are committed to fully engaging with them to address their concerns.” The company has 15 days to respond to the agency, and in the meantime it must obey the order to halt sales and marketing the service.

The FDA alerted four other genetic testing companies that offer direct to consumer services about similar concerns. In its warning letter to 23andMe the agency said that the company’s Personal Genome Service, which, for a charge of $99 (down from $300-$400 when it launched) tells consumers about such potentially life-and-death questions as whether they are likely to respond to medications including the anti-blood clot drug warfarin or the anti-stroke drug clopidogrel, qualifies as a diagnostic device and is subject to the Medical Device Amendments Act. That legislation requires premarket approval of medical devices to ensure that they provide accurate results that can be used to make clinical decisions. “We do have serious concerns about the potential health consequences of false positive or false negative assessments of such high risk indications,” says Jefferson.

Those concerns have been simmering for a long time, not just at the FDA but within the scientific community as well. When Wojcicki and her team launched 23andMe in 2006, geneticists scratched their heads over what, exactly, the company would be able to tell people that was of value. Even as the company sponsored spit parties at which high powered celebrities from Barry Diller to Wendy Murdoch  and Harvey Weinstein eagerly spit into test tubes to learn more about their genetic destinies, those in the genetic-testing community knew that matters are far more complicated than just matching up a particular gene or mutation to a disease. “Every person is walking around with different variants, and each genetic variant may turn the dial [on a disease] a little bit,” says John Stamatoyannopoulos, an associate professor of genome sciences and medicine at the University of Washington. “But the problem is that it’s like a giant mixing board at a studio. We’re not Quincy Jones, and we do not know what is going on with the whole picture if each of these dials is turned a different way.”

Does a genetic variant linked to a higher risk of Alzheimer’s for example, mean that everyone with that particular string of DNA will develop memory loss? Not exactly, since that gene may need to interact with a host of others, some of which may be identified already, and some of which may still lie buried in the genome’s vault.

On top of that, says Dr. Wendy Chung, director of the clinical genetics program at Columbia University, the 23andMe panel includes a mix of variants that strongly predispose to serious diseases such as breast cancer as well as those that are smaller players in causing the condition. “There are a few zingers in there, such as the mutations for BRCA1 and 2 that predispose to breast cancer. These are absolutely really powerful genes that are medically actionable and genes that people should be acting upon,” says Chung. “But the same panel for breast cancer testing including alleles that aren’t as penetrant, and I don’t think consumers understand the difference, so they don’t know whether to take them seriously or lightly.”

For that matter, scientists themselves can’t always be sure. Even today, more than 50 years after the discovery of the structure of DNA, and over a decade after the mapping of the human genome, experts are constantly rewriting the book on human genetics. In 2012, for example, investigators learned that the huge swaths of DNA that they had previously dismissed as inert, barren stretches of genetic code that sat in between the business parts of the genome, were far more active than they had imagined. In fact, these regions of non-coding DNA regulate and govern the way the genes are read and expressed, making them an essential, rather than ignorable, part of the genome.

That understanding, however, is in stark contrast to the more deterministic way that the lay public views the genome. For most people, genes equal disease, and companies like 23andMe capitalize on that perception, and the promise of mining the human genome to make better health decisions. And because 23andMe was backed by Google, a company known for its innovative approach to common problems, the idea of genetic testing—and its utility—perhaps gained an air of legitimacy and urgency that was premature.

“There were two things going on,” says Stamatoyannopoulos of companies like 23andMe and Navigenics, another genetic testing company that targeted consumers and was recently acquired by Life Technologies. “Number one, was the sales pitch they were giving to investors and the public. And second, I wondered about what they actually planned to do when I heard they were looking for common [genetic] variants. These are things that are completely non deterministic and likely contribute little risk so there was no way to act on or understand what they meant clinically, let alone integrate them all together.”

Chung notes that while medical geneticists who treat patients have access to the same genetic information linking certain genetic variants to diseases that 23andMe and other genetic testing companies do, they don’t conduct such tests because they still aren’t confident about interpreting the results. “The sad truth is that it is taking a lot longer than people want,” she says. “People are trying to take some of this progress into their own hands. But there is only so fast that we can accumulate the data that we need to in order to practice responsible medicine.”

Wojcicki thought she could help. In an interview in 2012, she told me that “we have a massive problem of what to do with all this information that we don’t know exactly what it means.” Her solution was to turn all the genetic datapoints from the company’s 150,000 or so consumers into a database that researchers could tap, with the customers’ permission, to test new hypotheses. Wojcicki envisioned the 23andMe customers as a constantly renewable resource, available to be contacted by scientists who might want to know about their smoking habits, say, or their family history of a specific disease, in order to learn more about how certain genetic variants linked up with risk of specific diseases.

Chung says that if the company had pushed this concept from the start, and not tried to sell a genetic testing service that promised more than it could deliver, more scientists might have supported the idea of promoting direct to consumer genetic testing.

It’s not that mining our genomes for potentially useful information isn’t an admirable, and in fact desirable thing. Even the FDA believes there is potential in it, but it’s potential 23andMe cannot yet fulfill. “We think access to these types of tests could allow consumers to take more responsibility in certain aspects of their health,” says Jefferson. “We just want to make sure they work the way they should work, and that patients are well aware of the risks and benefits and what the limitations could potentially be.”

“The way I look at it, this is a time out,” says Stamatoyannopoulos. “It’s not a game ending call, just a time out in the whole picture to sit back and understand what’s going on.”

Correction: The original story said that the FDA sent “warning” letters to 23andMe and four companies. The four other companies previously received first-time notifications from the agency known as “it has come to our attention” letters. The FDA issues warning letters only after companies fail to respond to the agency’s requests for more information or compliance with regulations.

3 comments
DGA
DGA

One point.  The statement " In 2012, for example, investigators learned that the huge swaths of DNA that they had previously dismissed as inert, barren stretches of genetic code that sat in between the business parts of the genome, were far more active than they had imagined. In fact, these regions of non-coding DNA regulate and govern the way the genes are read and expressed, making them an essential, rather than ignorable, part of the genome" is incorrect.  It likely refers to ENCODE, the results of which were purported to mean that 80% of the genome has functional importance, but this is not correct.  Geneticists have know for 50 years of the importance of regulatory sequences in the genome, but these along with the protein coding sequences total about 10% of the genome only.  Vast swathes of the genome have no known function and are indeed deletable without effect (work has been done to this effect with other creatures, notably mice).

Son-ofMan
Son-ofMan

Some of the arguments here are just silly.

I have subscribed to 23andme and found that it provides some possibly important information (like increased predisposition to say, prostate cancer), as well as information which has mostly curiosity value (dry or wet ear wax).

But ultimately, I do not think that anyone sane would see the results and immediately take out their prostate, or remove their breasts, based solely on the 23andme report. To suggest otherwise is disingenuous, at best.

In my mind, the benefits of such easily accessible and affordable tests far outweigh any concerns about people acting in unreasonable manner based on the information. Most reasonable people would use the 23andme report as a conversation starter with their doctor.

In my case, the 23andme results simply provided information which prompted me to heed my doctor's advice to go and have an exam, which I would likely have postponed much longer otherwise.

Ultimately, 23andme provides just another bit of information about our selves. Most of us understand the caveats that come with it and should have the option to access it, if we chose to do so.

My fear is that the FDA is ultimately going to increase the cost of obtaining this information by making physicians the gatekeepers, instead of the advisors.


vgupta123
vgupta123

@Son-ofMan But what if this "possibly important information (like increased predisposition to say, prostate cancer)" is incorrect. 

The company has to convince FDA (and independent scientists) that their interpretation of individual DNA regarding predisposition to certain diseases is indeed based on some scientific evidence. 

Incorrect information (or interpretation) is worse than no information!

The FDA always increases the cost of drugs and tests and so on. So we could just get rid of the FDA, I suppose.