FDA shuts down 23andme

[cosmicalstorm]

Shockingly stupid decision. I had my own test done 3 months ago.

For a couple of years, I have been warning all my friends and colleagues to purchase $99 personal genome testing from 23andMe before the Feds banned it. Well, now the Food and Drug Administration has banned it sending the genome testing company a warning letter:

The Food and Drug Administration (FDA) is sending you this letter because you are marketing the 23andMe Saliva Collection Kit and Personal Genome Service (PGS) without marketing clearance or approval in violation of the Federal Food, Drug and Cosmetic Act (the FD&C Act).

This product is a device within the meaning of section 201(h) of the FD&C Act, 21 U.S.C. 321(h), because it is intended for use in the diagnosis of disease or other conditions or in the cure, mitigation, treatment, or prevention of disease, or is intended to affect the structure or function of the body. For example, your company’s website at www.23andme.com/health (most recently viewed on November 6, 2013) markets the PGS for providing “health reports on 254 diseases and conditions,” including categories such as “carrier status,” “health risks,” and “drug response,” and specifically as a “first step in prevention” that enables users to “take steps toward mitigating serious diseases” such as diabetes, coronary heart disease, and breast cancer. Most of the intended uses for PGS listed on your website, a list that has grown over time, are medical device uses under section 201(h) of the FD&C Act. Most of these uses have not been classified and thus require premarket approval or de novo classification, as FDA has explained to you on numerous occasions.

The FDA says it is concerned that consumers would misunderstand genetic marker information and self treat. For example, the agency cites the company for testing for versions of the BRCA gene that confers higher risk of breast cancer worrying that women might get a false positive test leading "a patient to undergo prophylactic surgery, chemoprevention, intensive screening, or other morbidity-inducing actions...."

What the test results would actually lead patients to do is to get another test and to talk with their physicians. The FDA also cites the genotype results that indicate the sensitivity of patients to the blood-thinning medication warfarin. Again, such results would be used by patients to talk with their doctors about their treatment regimens should the time come that they need to take the drug. In fact, in 2010 the FDA actually updated its rules to recommend genetic testing to set the proper warfarin dosages for patients.

It is notable that the FDA cites not one example of a patient being harmed through the use of 23andMe's genotype screening test. Nevertheless the agency orders that...

...23andMe must immediately discontinue marketing the PGS (Personal Genome Service) until such time as it receives FDA marketing authorization for the device.

The FDA bureaucrats think that they know better than you how to handle your genetic information. This is outrageous.

For more background, see my 2011 Reason article on my own genetic testing experience here and go to SNPedia here for even more information on my genetic flaws.

H/T Mike Riggs and Andrew Mayne.

http://reason.com/blog/2013/11/25/fda-s ... ously-bann

[GrandMasterTerwynn]

No, it is not "shockingly stupid." The company has been repeatedly warned that their test fits the definition of "medical device," since they claim it can be used to diagnose disease (in fact, this is apparently prominently advertised.) At every turn, the company has stonewalled the FDA, or has been tardy in conducting/producing the study data required to have the test certified under the law. This is really more a case of the FDA finally running out of patience with the company.

[Serafina]

Same principle as selling any other diagnostic device, such as for blood sugar measurement, pregnancy test kits etc. - sure, they are safe to use even for laypeople - but that does not necessarily mean that laypeople can interpret the results properly, and it most certainly does not mean that the reliability of the test is unimportant!

[Guardsman Bass]

They're worried that people will use the tests to get unnecessary surgery/medical treatment, or to deviate from prescribed drug regimes. I don't know if there are any proven cases of that happening (it's been on the market for years), but it's something to consider. And the company's tardiness in dealing with the FDA rules on this undoubtedly didn't help - they've especially been stonewalling the FDA since May, while introducing new marketing campaigns for the product.

That said, I'm not thrilled by the idea that this might ultimately lead to a ban on genetic test interpretation set-ups that aren't done through a doctor/clinical setting.

[Luke Starkiller]

That said, I'm not thrilled by the idea that this might ultimately lead to a ban on genetic test interpretation set-ups that aren't done through a doctor/clinical setting.

What's wrong with needing to go to a doctor or use an approved, licensed device for diagnostics?

[Simon_Jester]

The main problem is that one can make a reasonable case that I have a right to know what my own DNA says, and that there's no particular reason this service should be made artificially expensive. If the problem is that I might ignorantly make poorly-considered medical choices for myself based on that information... well, it's no different than if I made similar decisions based on family history or anything else.

[White Haven]

That logic would ban WebMD as well. Now, whether that's a good idea or not is another question, but 'oh my, people might draw incorrect conclusions!' is not a reason to ban something.

My personal take is that as it is neither a drug nor a food, the Food and Drug Administration can go fly a kite. Not only is their name quite specific, but even if you allow them to balloon their remit to be arguably relevant, they're still supposed to be concerned with actually harmful products, not knock-on social effects of services.

[B5B7]

My guess is that people want to use this service for privacy reasons also. If done via non-secure methods (through doctors, for example) then insurance companies, employers, and other third parties can obtain this data to disadvantage people.

[Lord MJ]

If I understand correctly, WebMD is a site where you read articles about different medical issues and/or talk to actual medical professionals.

23andme is a device or technology that takes DNA samples and makes diagnoses. If that is the case, they have to go through clinical trials. Even if it just reports, "this is your DNA makeup" it still needs to go through trials.

If WebMD had a feature like "Take a retinal scan online and we will tell you what conditions you may have" it would require trials. If WebMD or any other service had an automated diagnoses of ANY sort, it would require FDA approval.

The only way this company could conceivably get away with it, is all they were doing is collecting the DNA samples. The minute they try to compute inferences based on the samples, the FDA approval requirements kick in.

[Exonerate]

Supposedly as a warning letter, it's not legally binding... but of course, ignoring it would probably upset the Feds even more and bring the hammer down. Reading through the letter, it seems to be primarily directed towards how it is marketed, not the actual testing procedure itself, meaning if they don't market it as diagnostic, preventative, etc, they're in the clear. I suspect if they sold it as a service to check for genes known to contribute to various risk factors for diseases, it would pass the letter of the law. I actually just sent in a kit for myself last weekend - I hope it'll still get processed - so I'm obviously a believer in this service, but at the same time, I do think it's important that anybody who goes into it understand the uncertainty involved.

[PainRack]

Say what? The FDA remit on medical devices has been a relatively longstanding commitment. What changed to make you take up that position White Haven?

[Napoleon the Clown]

My guess is that people want to use this service for privacy reasons also. If done via non-secure methods (through doctors, for example) then insurance companies, employers, and other third parties can obtain this data to disadvantage people.

The employers part doesn't even enter the equation because of HIPPA laws stating that your medical records cannot be disclosed to anyone without your express written consent, with each party you want to have access requiring separate paperwork. Violation of HIPPA can result in massive financial penalties and possibly more. I don't remember exactly what penalties can be incurred.

Due to the nature of health insurance, they will basically always be a party you have to give a waiver for. Your employer? None of their damn business so they won't know unless you release the information to them or say they can get it. Same with any other third party. And the ACA means that your health insurance can't drop you for what a test finds.


As far as diagnostic tools go, people self-diagnose enough as it is. Stop and think about how often doctors will prescribe shit just to get a patient out of their hair. Antibiotics, for example. The doctor can know for a fact that the patient doesn't need them yet will still write it just so the patient will fuck off and let them see other people. Major stuff probably won't result in a doctor being willing to do that, but people will make medical decisions based off a product that hasn't actually proven itself. This isn't like a blood pressure cuff where you can see the numbers. It isn't like a scale that says how much you weigh. This claims to actually diagnose stuff. If you claim to be able to diagnose problems you had damn well better prove it.

[Simon_Jester]

To be clearer, I have no objection to requiring the service to prove that they are testing the subject's DNA accurately. But subject to that requirement, and assuming they don't dishonestly present the implications of their findings (i.e. presenting BRCA1 as "YOU PROBABLY ALREADY HAVE TUMORS!" or some such scaremongering)...

I don't think the FDA should be shutting them down, especially if competitive services offered clinically are significantly more expensive, which I suspect they are.

[Alyrium Denryle]

To be clearer, I have no objection to requiring the service to prove that they are testing the subject's DNA accurately. But subject to that requirement, and assuming they don't dishonestly present the implications of their findings (i.e. presenting BRCA1 as "YOU PROBABLY ALREADY HAVE TUMORS!" or some such scaremongering)...

I don't think the FDA should be shutting them down, especially if competitive services offered clinically are significantly more expensive, which I suspect they are.

Given how corrupt the FDA is, that might also be part of the motivation. Now, the marketing IS problematic. If they marketed their service as a non-diagnostic tool to be used ONLY to assist doctors in decision making and looking for risk factors then it would be one thing. But they are marketing it as a diagnostic, which it isn't.

[Borgholio]

I'm willing to bet it has to do with the company's quality control system, or lack thereof. I work for a company that sells medical imaging software. The FDA considers software to be the same as physical equipment, as far as regulation goes. We are registered as an FDA medical device due to the fact that if our software fucks up, it can potentially cause patient injury due to missing / corrupted / incorrect data. And since it is considered a medical device, we have to follow strict guidelines when it comes to coding, debugging, testing, patching, and logging any customer reports of glitches or other problems...no matter how minor they may seem. If we don't follow the FDA regulations, we lose the medical device classification and we're dead in the water.

With http://www.23andme.com, I would suspect that since they sell medical equipment (as opposed to WebMD which is just reference), the FDA would want them to register as a medical device and they're stonewalling.

[White Haven]

The point I was trying to make with regards to WebMD, however, is that the result of someone plugging a bunch of symptoms into it, reading the results, and declaring that they have colon cancer, Ebola, and five broken limbs is the same as someone drawing a poor conclusion from a DNA test like 23andme provides. There's no drug harming the patient. This isn't a piece of hospital equipment where things come down to 'use this piece of equipment or that piece of equipment.' This is a home test. It literally is a choice between 'have some information about potential genetic diseases' and 'have no information about potential genetic diseases.'

[Borgholio]

Its because they sell a device. If it was just reference, they would probably be OK.

[Alyrium Denryle]

Its because they sell a device. If it was just reference, they would probably be OK.

This. This Right Here.

With WebMD, you are basically automating a search through a bunch of books and it explicitly tells you "Hey guys, we are not giving you a diagnosis, just giving you possibilities. Go See Your Doctor."

This... they are doing the actual testing, they have not been properly certified, and they offer to actually help with a diagnosis and disease prevention in their marketing. That is a problem.

[Simon_Jester]

As I understand it, WebMD is just an automated version of a medical dictionary. Any fool can pick up such a book, flip through it, and conclude that his joint aches means he has rheumatoid turbocritical headexploditis or whatever. The book (or online database) isn't telling you anything that isn't commonplace medical knowledge.

A service that tests your DNA is trying to tell you special information, and should be held to a high standard of precision... but not so high that the alternative is to force everyone to be ignorant of their own DNA unless a doctor thinks to prescribe a test.

[Alyrium Denryle]

As I understand it, WebMD is just an automated version of a medical dictionary. Any fool can pick up such a book, flip through it, and conclude that his joint aches means he has rheumatoid turbocritical headexploditis or whatever. The book (or online database) isn't telling you anything that isn't commonplace medical knowledge.

A service that tests your DNA is trying to tell you special information, and should be held to a high standard of precision... but not so high that the alternative is to force everyone to be ignorant of their own DNA unless a doctor thinks to prescribe a test.

Agreed, but in this case, all they have to do is switch their marketing strategy.

[Simon_Jester]

If their accuracy is an issue, the FDA might reasonably take a dim view of that.

[Flagg]

Without getting into the details of why they are shut down, this sucks. I was going to buy a kit for my mom this year for Christmas to see her genealogy.

[Broomstick]

The FDA says it is concerned that consumers would misunderstand genetic marker information and self treat. For example, the agency cites the company for testing for versions of the BRCA gene that confers higher risk of breast cancer worrying that women might get a false positive test leading "a patient to undergo prophylactic surgery, chemoprevention, intensive screening, or other morbidity-inducing actions...."

Somehow, I don't think anyone with a cancer inclined BRCA variation will be doing home surgery, chemo, or radiation. In order to get all that you have to go through actual doctors who will do their own testing, so this doesn't hold water in my opinion.

What the test results would actually lead patients to do is to get another test and to talk with their physicians. The FDA also cites the genotype results that indicate the sensitivity of patients to the blood-thinning medication warfarin. Again, such results would be used by patients to talk with their doctors about their treatment regimens should the time come that they need to take the drug.

This, however, I think is a bad assumption. Some people just don't want to go to doctors and insist on self treatment. This is a rather long thread from another board but it illustrates this very well - a woman has skin cancer and treats it herself rather than opt for the standard medical treatment and winds up needing reconstructive surgery (so much for avoiding doctors...) This clearly meets the definition of "self harm".

I've also heard that people who receive confirmation of having something like the Huntington's gene will sometimes commit suicide. I don't have internet-based cites for that, but it's another thing to consider.

While I'm all for being informed there is some information that in isolation isn't much use or potentially harmful. Part of the rationale for testing going through the medical system is for someone knowledgeable to be there to provide information, dispell misinformation, and offer options to people receiving this information.

Some genetic information is likely harmless, like what continent your distant maternal or paternal ancestors came from. Other things can point to serious medical information that requires additional information to make good use of.

I guess, bottom line, I have mixed feelings about this one. I do agree with the folks who say part of what got them into trouble is their marketing tactics.

[White Haven]

Yes, it's a device. Whoo. Yay. Now, what is the consequence of misusing the device? It's exactly the same as the consequence of misusing WebMD, or just rolling dice and consulting the Critical Illness table, or whatever. The fact that It Is A Device doesn't somehow make it better or worse.

[Borgholio]

No, it's not the same. There is a very real difference between selling a physical object / diagnosis and providing an online encyclopedia.

Take the software company I work for, as an example. Our software primarily deals with orthodontics and oral surgery. It can be (kinda) described as photoshop on steroids. Big fucking deal, right? Well yes, it is. Our software can be used to make measurements of a patient's facial features and be used to plan upcoming surgeries or orthodontic procedures. While it's easy to say "Oh the software doesn't actually DO anything, it's just a photo editor..." the fact is that if the diagnostic results are off, even by a little bit, a fatal mistake could potentially be made.

The FDA is hyper-anal about that fraction of a % of cases where it could get ugly if something went wrong. In the case of our software, even though there has never been a recorded case of patient injury resulting from use of our software, we make damn sure that the output is accurate. We are required to follow FDA guidelines to the letter. In the case of 23andme, if the FDA has not certified that the physical testing unit is safe, the testing procedure is accurate, and the output results are accurate, then they will be very nervous about letting that company continue to operate.

Yeah, I know that WebMD can be used by any idiot on the internet to self-diagnose. They can also do that with Wikipedia, Encyclopedia Britannica, or any other reference site on the web. FDA doesn't (and shouldn't) care because those sites are just putting information out there. What the end user does with that info is irrelevant to the FDA. What IS relevant is if a company sells a physical object to collect a medical sample, tests it, obtains results, and then INTERPRETS THOSE RESULTS for the patient. WebMD isn't doing that. 23andme is.