Hacker News new | past | comments | ask | show | jobs | submit login
23andme has suspended health-related genetic tests (23andme.com)
50 points by Taeram on Dec 6, 2013 | hide | past | favorite | 63 comments

This seems to be the nuclear option. As a (now fortunate) previous purchaser of 23andme, I still have my health information. The information is really valuable, and considering the price ($99!) it is cheaper than a lot of standard medical exams and bloodwork. Obviously not a replacement for any of those things, but it's amazing they can offer DNA health results for so little.

The FDA seems to be mostly concerned that the results of 23andme could be misinterpreted, or wrong. As far as I am aware, 23andme was advertising the health assessment as the first step in prevention - not a one stop shop for diagnosis. I am not very impressed with the FDA.

Here's the deal: you have a report. You have no way of knowing if the information in the report is correct, or not. You may have paid $99 for something that's completely fictional. You just don't know.

Is that your "right"? Sure, to some extent. A sucker is born every minute, and if you desperately want to be one of them, the US Government is mostly content to let it happen. We'll even let companies do patently absurd things like sell you pills that are supposed to make your dingus larger. We just limit what they can say when they're selling them to you, to protect the most credulous among us.

Like it or not, our society has decided that certain forms of charlatanism are so egregious that we act proactively to prevent them. Medicine is one area where we are strict, because our country has survived a rather nasty era where unfettered "medical innovation" was allowed to kill large numbers of people. I have a vintage Merck manual sitting on my desk, and it recommends such wonderful treatments as oral arsenic for hair loss (to be fair, you probably care a lot less about being bald when you're dead...) and vinegar (via injection!) as a cure for cancer. That's "medical innovation" without regulation to enforce safety and effectiveness, and I don't particularly want to go back to it. Maybe you do.

The FDA is doing its job rationally here, it is well within its mandate, and as far as anyone who knows what they're talking about can tell (including the leaders of the company), 23andMe is just clearly in the wrong on this issue. So if you don't like the laws we've set up to protect consumers against medical malpractice, I guess that's your opinion, but it's obviously dumb to blame the FDA for doing exactly what it's supposed to be doing.

This is ridiculous, you take extreme examples to prove your point. So let's see, you can still read about all this and worse on the internet, do we still have people taking arsenic en masse to cure baldness? No. Because most people are responsible and giving them the freedom to do what they want with their body is more important than anything else. Plus, if someone is ignorant enough to drink arsenic to cure baldness, he'll probably find other ways to hurt himself anyway, FDA or no FDA.

So let's go back to the FDA, is it so effective? Here is the FDA authorizing a drug that killed lots of people http://www.undergroundhealth.com/fda-approved-prescription-d... Also, the time the FDA takes to approve a good drug can take more than 10 years, how many people die because of this unseen consequence? http://en.wikipedia.org/wiki/Criticism_of_the_Food_and_Drug_... Not to forget its reluctance to approve drugs for political reasons (canabis).

So let's see, many people die because of wrongly approved drugs and many die as it takes too long to approve good ones because of the FDA. Now let's imagine a world with no FDA, many people would still die, but at least they'd have the choice to do so on their own decision and not have to wait on a nanny state bureaucracy to tell them what's good or bad for them and many times wrongly so. Why not let people freely decide what they can do to their own body when not doing so still kills a lot of people too? Plus, allowing so would make it much faster to detect when a drug works or not and save way more lives. If you don't even own your own body, what do you own?

"do we still have people taking arsenic en masse to cure baldness? No."

Yeah, we have things called "clinical trials" that pretty much put an end to the bad old days of patent medicine. Guess who enforces that stuff?

The rest of your post is a straw-man. Does the FDA make mistakes? Sure. That doesn't invalidate the need for the agency. Is it your right to do stupid things with your body? See previous comment.

>> "do we still have people taking arsenic en masse to cure baldness? No."

> Yeah

No, we don't. I said "en masse", that is to say, do we have a massive amount of people taking arsenic everyday to cure boldness? No we don't, and yet anyone can read about it on the web.

> Does the FDA make mistakes? Sure. That doesn't invalidate the need for the agency.

Yes it does, I've just explained that the FDA is responsible for more death than if we had no FDA while taking away people's freedom do cure themselves the way they want.

> Is it your right to do stupid things with your body? See previous comment.

Well, according to you, only if the FDA approves so, the same FDA responsible for countless deaths. No thanks.

"The SNP chip method that 23andMe uses was never very good at providing useful genetic information. Its advantage has been its low cost. But as sequencing gets cheaper and cheaper, SNP chips have already largely become obsolete for most genetic testing."[1]

Basically, 23andMe has always oversold the results available from the crude SNP chip technology they were using as a hook to have people PAY to send them DNA samples that would make them "members" of 23andMe. Then the members are reminded to send in personal health history information to be put into the 23andMe database so that 23andMe can data-mine that and try to figure out what the DNA actually means[2]--which no medical geneticist knows to any precision at the moment. That business model of suckering in paying clients today for the promise of an unknown amount of actionable information in the future is why 23andMe was trying to stonewall the FDA while marketing what was plainly a direct-to-consumer medical device. It's not surprising (see first link here) that the FDA is not going to let 23andMe get away with that.

[1] http://venturebeat.com/2013/12/03/why-the-fdas-anvil-dropped...

[2] http://www.fastcompany.com/3018598/for-99-this-ceo-can-tell-...

I'd like to add that 23andMe also wants to be a resource for finding patients for studies. They boast being able to find a large number of people with a certain condition in a short amount of time (a real benefit in human studies) because of the health information they have.

I agree. Trying noninvasive but inaccurate tests before accurate but invasive tests is just par for the course. 23andMe is certainly noninvasive, and as far as I can tell more accurate than most such preliminary tests.

The accuracy isn't the problem. It's that the accuracy is unknown. And evidently 23adMe haven't been attempting to prove it to the FDA.

As far as I'm concerned at that stage you might as well take homeopathy.

> and as far as I can tell more accurate than most such preliminary tests.

Oh thats good? I'm sure that anecdotal well set the FDA at ease.

As far as I'm concerned at that stage you might as well take homeopathy.

How does this follow? No test is equivalent to ignorane. Using the test as a non-invasive prelim screen may be imperfect, but its hard to say the "null hypothesis" is equivalent. That would require that no information be present, which seems unlikely (if unproven). Or did you mean something else?

Is there any way to export all existing test results and not just the raw data?

Thanks, but that's only (OK, at least) an overview titled '23andMe Genetic Health Overview'.

Is there a more comprehensive and maybe configurable export feature available?

I don't see how the results could be harmful or negative. My highest risk item was a 54% chance of Type 2 Diabetes. It's interesting information, but I'm not making any huge life decisions based on it..

Why haven't you made any huge life decisions based on it? Wasn't that the point of you ordering the test?

If you do strenuous exercise several times a week and adopt a healthy diet, your risk will drop dramatically. Was it because you already do these things?

I got this test too, but it didn't show any terrifying results.

I wanted to know if I have awesome genetic material and it was true ;)

We can learn a bit more about this dispute from how 23andMe has responded here.

Given that 23andMe is still willing to release their raw genotype results, it seems clear that the FDA isn't objecting to that level of their work. That's comforting: evidently at least that is considered more or less accurate.

Instead, they've only redacted "23andMe’s interpretation" of the data from their reports to new customers. My take on this is that the FDA wants some sort of rigorous demonstration that 23andMe isn't making errors in its interpretation of medical data. I'm thinking specifically of stories like this one: "My deadly disease was just a 23andme bug" https://news.ycombinator.com/item?id=6796219 It's hard to know (as a simple consumer) how common that sort of error might be: I rely on folks like the FDA to set and enforce quality standards on such things.

Similarly, I honestly don't have a great sense of how well 23andMe's interpretations reflected the full current state of known science. For instance, my memory is that they were only able to test for a few of the known breast cancer risk factor genes because others are still covered by patents. How clearly did their interpretation pages convey those gaps in what they were presenting?

I hope that 23andMe finds a way to get fully back online before long. But honestly, I'd be happiest if they manage to satisfy the regulators that they really are getting things right. Someone needs to verify claims like that for the rest of us!

That's comforting: evidently at least that is considered more or less accurate.

I'm not sure you can safely draw that conclusion from the way this has played out. I think it's equally, if not more likely, that 23andMe is able to continue offering raw and ancestry data simply because those things are outside the FDA's purview.

Consider, by analogy, the dietary supplement industry: the FDA has incredibly more power over a product that is advertised or marketed as intended to "diagnose, treat, or cure any disease." Notice how the manufacturer explicitly disclaims that exact verbiage on all those sorts of products? That's because anything that is marketed that way is immediately subject to a much higher degree of FDA jurisdiction, under which they can literally shut you down overnight.

When 23andMe runs your genome and sees a genotypic match for an arbitrary disease, and then tells you that you have that disease, they're diagnosing you. When they tell you that you're 72.3% Northern European, or give you a ginormous CSV (or however they encode their data dumps), they aren't.

That's the only thing I think you can meaningfully read from this situation: they're changing the way their product is marketed, and the services they provide, so as not to be subject to the FDA's nuclear option.

That's a really good point. Rats. So only if they go back to advertising health information after FDA approval will we be able to be confident that their actual genotype results meet FDA standards (as well as their interpretive text).

I agree - Here's a company that built a complex software system, loaded it with "if" statements pulled from research papers, and started marketing itself as a predictor of cancer and disease. How did they do their research? What is their threshold for determining if scientific evidence is conclusive enough to merit telling the world they could have a major disease? The FDA wants to know, and so do I.

So now, people will just export their data and third party companies will provide such interpretation and the FDA won't be able to touch them. How soon until github has an open source project to read and interpret 23andme data-dumps?

You can already do that today with downloaded SNPs [1] with Promethease [2].

1. https://www.23andme.com/you/download/

2. http://www.snpedia.com/index.php/Promethease (non-pro version)

Most likely anyone providing such information would also be in a similar boat with regards to the FDA.

Sure you could be overseas providing the data but it's not like the FDA is preventing you from getting overseas treatment.

I doubt it, since those people aren't making any medical devices. Moreover, the SNPedia can be used by anyone, so an open source tool can be written and deployed as an app. I highly doubt the FDA will be able to stop that.

Trying to stop someone from analyzing their health data dump is about as useless as trying to stop bitcoin transactions. About the only thing the government can try to do is to stop the public from being able to use DNA microarrays, but how long is it going to be before a cheap, consumer grade version of such chips can be made and sold in a way that the FDA can't regulate them?

a) they never marketed themselves as predictors of cancer or told anyone they have major disease b) the criteria for scientific evidence are here https://23andme.https.internapcdn.net/res/pdf/HIC-SXIYiYqXre... (from https://www.23andme.com/for/scientists/)

Hmm - thanks for the link! Good stuff. I think the complaint I have is that in their advertising (https://www.youtube.com/watch?v=JTIY310FGBU&list=TLdNSueMINq...), they feature:

- An individual for whom 23andMe correctly identified a heart issue that he later had.

- An individual who caught his prostate cancer early because 23andMe told him he was likely to have it.

In both cases, 23andMe successfully identified the condition before it occurred. That's about as close to marketing yourself as a predictor of disease as you can get. Granted, the word for word text in the video references "elevated risk", but the overall story they're selling to the public is that 23andMe predicts diseases you'll have. At least, that's my impression. Really interested to see how all this pans out.

Since they are discontinuing their interpretation of the data, they should just open source the interpretation engine and dataset. People still have to pay the $99 for the tests to get the data to interpret, so it wouldn't harm their business. It would keep their service out of the FDA's crosshairs, while keeping the service as useful as ever. In fact, it would likely improve it over what it will become with this latest development.

They are not discontinuing it. It is simply not available for new purchasers. I imagine once the FDA is appeased it will return.

New purchasers, or people who purchased in the past two weeks...

I think those petitions are highly dubious. Neither of them accurately references the FDA's issue with 23AM.

> highly dubious

See: most things on Change.org.

I wonder if another nuclear option is to just move the whole shebang to say, the Cayman Islands, China, or some jurisdiction outside the reach of the FDA and European regulators. Re-incorporate as a non-US entity.

Unlike buying medicine from foreign soil that can be interdicted, this is all electronic: shipping spit, and getting back HTML.

So you'd like your genome to be in the hands of a Cayman Island company?

And just what are they going to do with it? Target me with a biologically engineered virus? I'm shedding DNA with every breath. Your DNA is currently private thanks to security through obscurity, but if you've ever watched GATTACA, that security will disappear as DNA testing becomes faster, cheaper, accurate, and ubiquitous.

People really need to get over paranoia over this and focus on what matters: laws against discrimination based on genetics.

Cayman Island company is no more harmful than in the hands of a (likely CIA/NSA infested) USA company.

In fact, they've not only suspended health-related genetic tests, but they've done so retroactively. If you signed up for 23andme in the past two weeks, you're not getting what you paid for.

Nearly ordered myself a test before discovering this mess on the day the FDA sent their letter. I would really hope they'd provide some sort of compensation to those who're being screwed...doubtful that'll happen though.

Everyone who has ordered since November 22nd is offered a refund. The details are on the website and will come in the email. Have some faith, dude. Of course they have to do that.

Ah, well that's good then.

Well this sucks. I was going to pay for the test this weekend.

Anyone have any idea how useful the health data is without the 23andme interpretation? I'm guessing it will be basically useless.

You can get the basically same report running Promethease on 23andMe data dump. There is a website even providing automatic 23andMe data import using 23andMe's API.

You can run the raw data through a 3rd party tool like Promethease (http://snpedia.com/index.php/Promethease) while you wait for the FDA kerfuffle to work itself out. YMMV though.

Since they're still going to provide the data to new users, I wonder if some of their interpretations could be recreated and published by existing users who still have access to the interpretations.

Yes, for the mendelian ones that can be inferred in a relatively straightforward way from the raw data. Not really, for the more complicated risk assessments.

Their interpretations were based on poorly proven associations between genes and disease-outcomes. I bet with the info you could look up on wikipedia/google scholar much of what you needed to know. But it depends on the format of this data dump.

No, that is pure FUD. 23andme provides a confidence assessment for the associations and risks. Here is a screenshot from their site: https://dl.dropboxusercontent.com/u/49805/confidence.png

All well and good that intervals of confidence exist in medical research. But good understandings of these associations have been reached with very few genes. The categories alone do not prove that they apply broadly enough to justify the product.

And no one has problems with their privacy policy?

Your DNA is pretty much public domain anyway. You drop tons of it all over the place.

Thankfully I do not label it with my name ;)

I am wondering, what about foreigners? Is it possible to get the full test and analysis if the results are shipped to, say, Europe, or is it locked for us too?

European regulation is supposed to be harsher than US one, so it is highly unlikely, but I didn't found the information on the website.

Time to open source 23andme health-risk analysis code so we can run our raw data through it?

Government, once again, deciding they know what is best for you. I really hope people are paying attention and vote with their intelligence during the next several elections. We need less government. Everywhere. Not more.

Or they could have worked with the FDA which it appears they were more than willing to do (FDA that is).

But fuck it. Libertarian nerd rage. Disrupt!! There is a reason people outside your bubble think you are assholes.

I bought 23andMe ages ago and haven't sent in my test yet. Because I purchased it quite a while ago, will I still be able to get all the info? If not, I'm gonna be pretty disappointed.

I don't think you'll get your results - at least not soon. The FDA said they needed to back up their claims that their assessment methodologies are accurate, and since they failed to respond to those requests they've been prohibited from performing the genetic tests and giving you the insights that you paid for. Don't think it matters when you bought the kit :-(

"Customers who purchased kits before November 22, 2013 will continue to have access to all the reports they’ve always had."

Ahh interesting. At first I'd read that to mean that they weren't taking data down, and existing customers would still be able to access their reports. I figured that only applied to people who had already had a report generated.

Yeah, you should be OK.

Great. How about the FDA deal with the claims made by homeopaths next?

They would love to, but they are limited in the scope of their authority. Homeopathy companies tend to label their products as "dietary supplements" so they don't fall under FDA drug regulations. When they slip up with their marketing, the FDA does get involved http://news.discovery.com/human/health/homeopathy-takes-hits... (see "Skirting Drug Regulations").

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact