I'd like to think that I'm libertarian on points where it matters. I think this is one of those points.

23andme is a site where you get genetic sequencing done through the mail for about $250 (although the exact running price has been in flux). You then log into the site and it tells you what academic papers predict you will have or won't have. Sometimes different papers conflict, and it tells you that.

For almost its entire existence, the organization has had a mallet hanging over their heads in the form of regulation. This is the latest update.

http://spittoon.23andme.com/announcements/23andme-takes-first-step-toward-fda-clearance/

The gist is that they submitted an application to the FDA. Why?

Quote :

"The FDA responded to some of these concerns by indicating that direct-to-consumer genetic testing services require regulatory review to remain on the market. We remain strong in our belief that consumers have a fundamental right to their personal genetic data. We believe personal genetic data will power a revolution in healthcare. But we also recognize that appropriate oversight of this industry can be a stepping stone on the path to realizing that revolution."

In other words, the organization has given up the fight to avoid the FDA and recently submitted an application. Their past posts IMO sounded clearly more defiant. I think their wording is political at this point... considering they have a standing application that could be hurt.

I don't think it's terribly likely, but if it's shut down in a year or so, I would hope that we see outrage from more than its 150k user base. This is a clear case where neither the vendor nor the customers want the regulator involved. Even if it's not shut down, the silent damage is done - the prices of genetic testing were supposed to fall in an exponential curve vastly outpacing the improvement rate even transistors saw. That might not happen, and you might have government to thank for it.

This has been an awfully biased OP, who has a different opinion? Why do I need doctors involved in this website I'm using?

7/30/2012 3:04:46 PM

It should be regulated under the Constitution. If it's not violating the Constitution, then it's fair game.

If a state would like to set their own regulations, that's fine, too, as long as the state's rules don't violate the constitution.

7/30/2012 4:01:54 PM

actually one of the fights they previously had was with California.

The state had some rules someone realized could be interpreted as requiring a physician to be involved with genetic counseling.

the problem is, nothing about the process even remotely makes a user think they're getting medical advice from certified doctors. The requirement is tantamount to writing laws saying that citizens can't do research about health, because only doctors should do that.

7/30/2012 4:21:40 PM

You mean kind of like that poor guy who had a blog about his own weightloss?

7/30/2012 5:21:30 PM

Do we know what the nature of the FDA regulations are? I can see some need for some sort of oversight of this service. Since the data is going to be used by people to help make health decisions, there should be some safeguards in place to ensure that the company's processes and methodologies are valid, that their quality control checks are in place, etc.

7/30/2012 5:28:36 PM

^ IMO your argument is totally and completely wrong, and I'm going to show you why, and it all comes down to the use of one single word.

Quote :

"Since the data is going to be used by people to help make health decisions"

Is there any possible way to limit the set of tools and knowledge resources that could help someone make health decisions? Wikipedia and TWW itself, for heaven's sake, are quite liable, and not just mildly or in a superficial sense.

This is the problem with the encroaching phenomenon:
- growing "need" for government in health care
- the ever-widening set of things that can fall into health care

^^ much like it, yes. The main difference I would point out is the # of people following this, and the fact that it's not bull shit. But we're not comparing merit here, the two fall under the same legal tent.

7/30/2012 6:47:18 PM

How is genetic testing different from any other medical diagnostic service or test currently regulated by the FDA?

7/30/2012 8:50:19 PM

The problem with the FDA / DEA and all the other alphabet soup agencies involved in health care is the chilling effect that they have on the industry, especially because investors are a nervous bunch. The company I work for was doing some really awesome things to improve patient care, but a lot of them fell into that area of the law where because no one has ever done it before, there was no laws for how to do it or even if you can do it and it was a giant legal grey area. As a small company with pretty much just the owner/CEO invested, they were willing to take on some of those risks because nothing that was being done was actually against the law, and they were triple dotting i's and crossing t's to make sure that if someone ever questioned what was going on, they had all the evidence that everything was on the up and up. And the problem is, there's no one you can ask for clarification without basically suing the government, not exactly something you want to do when you're operating in a grey area.

Eventually newer corporate investors were brought on board, and playing in a legal grey area scares the crap out of them, because all it takes is one government agent having a bad day and suddenly they can be slammed with huge fines, penalties or even have the business suspended. None of these are particularly likely scenarios given how much everything was documented, but it scares the investors none the less. So over the last year or so, they've been slowly scaling back all of those really neat services and features that the company provided and finding more traditional or safer routes, even if that means pissing off the customer base.

There's nothing wrong with the FDA having regulations and certifications and procedures, but it should all be optional, like UL listings. After all, why shouldn't a consumer be able to buy a cheap test and then use that to decide if they want to have their doctor investigate the results more?

7/30/2012 11:43:28 PM

Quote :

"How is genetic testing different from any other medical diagnostic service or test currently regulated by the FDA?"

So 23andme is equated to an MRI? Do you have any grasp on the difference between the two?

I won't argue against "regulation" because that's semantic. For crying out loud, Groupon just lost a class-action because dumb users didn't understand the concept of expiration. Even if your terms are clear, if a customer buys your service with a reasonable expectation that it does so-and-such when it doesn't, your company could be liable. That applies for commerce universally, so this equating of 23andme to medical diagnostics in general is, well... very worrisome.

Of course I believe the company should be required to adhere to a level of standard, because that is in the purview of the government in our world. But you have to understand the nature of the service, and that they are already did great due-diligence toward these means. In terms of reasonable regulation, for one, we could require they offer truthful information. Check. We could ask that they provide ample warnings about the fallibility of the predicted relative risks they publish. Got that. You need to understand - the fact that this is not enough is the part that's scary. This is about using the words of "diagnostics" and "genetic testing" as catch-alls to lead to completely unknown and unchecked action against against this type of service which just saw the light of day for the common person.

The other important point, is that many people who use this service see a greater purpose in it. We were given our genome by nature. The genome goes beyond ownership to stewardship. We have no idea the depth of secrets hidden within it, and 23andme conducts surveys with the community to advance knowledge that no one else could do because they don't have the sample size. The basic proposal of regulation implies that the diligence already done may not be enough. The fact that this discussion exists implies that the project, and the advancement that comes from it, is threatened. The core of it is that government shouldn't have the right to rule out what's being done with 23andme.

[Edited on July 31, 2012 at 9:15 AM. Reason : ]

7/31/2012 9:14:06 AM

Well, if you can get ahold of an MRI and convince someone to climb inside it, why won't the medical community let you use it?

7/31/2012 10:34:55 AM

i think this is more along the lines of an otc test like a pregnancy test. if the results are concerning you're gonna go to the doc and they're gonna run their own tests.

out of curiosity, how common/expensive are these genetic tests when done by PCPs?

[Edited on July 31, 2012 at 10:46 AM. Reason : ']

7/31/2012 10:45:43 AM

^^ the problem with implied meaning is that no one has any idea what point you're trying to make

Quote :

"out of curiosity, how common/expensive are these genetic tests when done by PCPs?"

I think your doctor wouldn't do a genetic test unless there was something very specific they were concerned about, and even if there was, they haven't learned to rely on these tools yet. I don't think there is any medical organization strongly urging precautionary genetic testing, although this is a purpose that genetic testing can serve. It is entirely possible that this changes in the future. We could screen newborns like we do vaccinations today. Lots of people are afraid of this, and I'm not advocating for it, but voluntary sequencing is an entirely different thing.

A lot of people use it to find relatives, genealogy stuff.

[Edited on July 31, 2012 at 10:57 AM. Reason : ]

7/31/2012 10:46:04 AM

Quote :

"So 23andme is equated to an MRI? Do you have any grasp on the difference between the two?"

Jesus fucking Christ, lay off the condescension for 10 seconds.

The FDA regulates in-vitro diagnostics.

Quote :

"In vitro diagnostic products are those reagents, instruments, and systems intended for use in diagnosis of disease or other conditions, including a determination of the state of health, in order to cure, mitigate, treat, or prevent disease or its sequelae. Such products are intended for use in the collection, preparation, and examination of specimens taken from the human body."

This includes things like blood tests and home pregnancy tests, and would certainly seem to include genetic testing.

So, again, how is genetic testing different from any other medical diagnostic service or test currently regulated by the FDA?

Quote :

"The basic proposal of regulation implies that the diligence already done may not be enough. The fact that this discussion exists implies that the project, and the advancement that comes from it, is threatened. The core of it is that government shouldn't have the right to rule out what's being done with 23andme."

Regulation does not necessarily imply that diligence hasn't been done, or it's an inherently unsafe activity, or the government is attempting to ban or otherwise stymie genetic testing. In fact, most in-vitro diagnostic regulations are in the vein of quality assurance requirements--things which, if 23andme is as fabulous a company as you seem to think, they should be doing already.

7/31/2012 1:40:15 PM

Quote :

"The FDA regulates in-vitro diagnostics."

Quote :

"This includes things like blood tests and home pregnancy tests, and would certainly seem to include genetic testing."

This didn't introduce new information.

Quote :

"In fact, most in-vitro diagnostic regulations are in the vein of quality assurance requirements--things which, if 23andme is as fabulous a company as you seem to think, they should be doing already."

So are you classifying 23andme as a in-vitro diagnostic device? In the FDA's own words, these devices are regulated such that they have "sufficient valid scientific evidence that provides reasonable assurance that the device is safe and effective for its intended use or uses."

Can you identify the intended use of 23andme? I can for a home pregnancy test.

7/31/2012 2:47:46 PM

Quote :

"So are you classifying 23andme as a in-vitro diagnostic device? In the FDA's own words, these devices are regulated such that they have "sufficient valid scientific evidence that provides reasonable assurance that the device is safe and effective for its intended use or uses." Can you identify the intended use of 23andme? I can for a home pregnancy test."

The intended use of 23andme is to test for and indicate markers which may be indicative or at least predictive of certain conditions, disorders, and diseases. The analogy of the in-vitro diagnostics is a pretty good one as they are classified as "tests that can detect diseases, conditions, or infections" - though the name does more than strongly imply this is limited to fetuses. Personally I don't think they should be hamfistedly shoved into that classification and would rather see them be placed under a more appropriate categorization which would include home test kits such as the home HIV tests or blood glucose meters.

Given their function though, they can perform many of the functions of in-vitro diagnostics kits - about half-way and with a bit more vagueness. Gotta remember where the baby's genes comes from.

Also it might be important to note that the FDA has said the product must be reviewed to remain on the market, not that it must be taken off the market until it had been reviewed. That seems to be a relatively reasonable decision and furthers their mission while making a distinction in approach between a home-test kit like this and something like an aids drug where efficacy would need to be demonstrated prior to even beginning to sell the product.


To answer the title question though-

I think that efficacy testing requirements for home kits that purport to be able to identify markers predictive of a wide array of health problems (of which a few particularly terrible ones are close to 100%) then requiring they be reviewed for accuracy/effectiveness might be a good idea. Since these devices do offer medically significant diagnostic information specific to the individual consumer it is fairly easy to see how this differs both practically and conceptually from researching information of a far more general and less diagnostic nature.

Personally i would be far more interested in the security of their database and their gaurantees of privacy than just the accuracy. Classifying it as a medical device and service means that the information collected by them would also fall under some sections of HIPPA and that the sale or release of any such information would carry hefty penalties.

[Edited on July 31, 2012 at 11:17 PM. Reason : fg]

7/31/2012 10:52:22 PM

^^ Given the FDA's existing definition of an in-vitro diagnostic product, it's hard to see how 23andMe would not be considered such.

23andMe's intended use is to provide "genetic testing for over 100 traits and diseases as well as DNA ancestry" (from their website). The prominantly displayed health section on the website strongly indicates 23andMe's desire to be involved with individual health care.

7/31/2012 10:59:59 PM

Missed the 30 minute edit window...

Quote :

"The analogy of the in-vitro diagnostics is a pretty good one as they are classified as "tests that can detect diseases, conditions, or infections" - though the name does more than strongly imply this is limited to fetuses. Personally I don't think they should be hamfistedly shoved into that classification and would rather see them be placed under a more appropriate categorization which would include home test kits such as the home HIV tests or blood glucose meters."

In-vitro does not imply it's limited to fetuses.

Also, I'm pretty sure blood glucose meters are considered in-vitro diagnostics.

7/31/2012 11:49:39 PM

Hah you're right, not sure why I thought it meant that. For some reason I was confusing it with in-utero. Then yeah that completely applies and that actually is the broader tissue-testing classification I imagined would make more sense.

[Edited on August 1, 2012 at 1:06 AM. Reason : l]

8/1/2012 1:05:25 AM

Quote :

"The intended use of 23andme is to test for and indicate markers which may be indicative or at least predictive of certain conditions, disorders, and diseases."

Except that's wrong. It is a part of the purpose of 23andme, but it's not right to say it this way. Even these markers aren't what you're making them out to be. Let me list what you said:

- conditions
- disorders
- diseases

This shows a fundamental lack of understanding of the service and genetic sequencing in general. Really, none of these three are right. Look at the stories on the site.

https://www.23andme.com/stories/

Quote :

"She made the connection with her half-sister, who was battling cancer, just after getting an update from 23andMe about her own elevated risk for breast cancer. When she went in for her annual physical, her doctor noticed something on the mammogram, but thought it was likely some calcification."

Is a higher risk of breast cancer a disorder? No, we all have higher risks for different things. I have higher genetic risk for Type 2 Diabetes, and 99% of the population will interpret that statement wrong. If you correct for other factors, I am vastly less likely to get diabetes than the rest of the population, and a large fraction of the population has the same elevated risk that I do. It's not an abnormality, it's just genetics. It is about analyzing your normal health state.

This is the core of it, 23andme doesn't check for medical problems or clear events. Find me a story of someone who found out about a condition by using the site. I don't even see how this request makes any sense... it doesn't.

The closest thing that 23andme ever comes to identifying a condition is the carrier status, and it's just that, the status of carrying genes that might lead to conditions if combined with other genes. For instance, I am a carrier of a variant of the Alpha-1 Antitrypsin Deficiency, by having an S form of the SERPINA1 gene. If I had a child with someone who also had one S form, our child might then have two copies, an SS form. If they did, we don't know what that would mean, because the S form is so rare they've never been able to do studies on it, but studies have reasoned that people with the variant probably wouldn't be at increased risk of lung or liver disease. Otherwise, the M type is "normal" and the Z can be harmful. The implications of the S type aren't well studied.

So, I just got diagnosed? Is that what you're telling me? It also diagnosed me for having curly hair. Guilty as charged.

Quote :

"Personally i would be far more interested in the security of their database and their gaurantees of privacy than just the accuracy."

In this sense, it already is regulated. I think we'll certainly all agree on this point, along with the legislators who have already introduced genetic privacy laws that restrict organizations (employers, health insurance) from making decisions based on someone's DNA. We've probably all seen Gattaca.

[Edited on August 1, 2012 at 8:40 AM. Reason : ]

8/1/2012 8:38:19 AM

Quote :

"but it's not right to say it this way"

It's pretty close to how 23andMe itself describes its service:

"[To provide] genetic testing for over 100 traits and diseases as well as DNA ancestry"

Whether 23andMe's service diagnoses a specific disease or not (as opposed to future likelihood), it still seems likely to fall into the in-vitro diagnostic category, which is well within the FDA's regulating territory.

How are other DNA sequencing labs regulated?

8/1/2012 12:03:52 PM

Quote :

"Whether 23andMe's service diagnoses a specific disease or not (as opposed to future likelihood), it still seems likely to fall into the in-vitro diagnostic category"

Fine, then it just falls down the slippery slope of "all knowledge is diagnostic". Anything you read could potentially play a role in how you take action in dealing with a cancer or the possibility of a cancer, or any other disease. Sure, the distinction isn't that important for whether or not it is in the purview of FDA regulation, but I'm arguing on this point of submitting an FDA application. The distinction of whether or not the service addresses a specific clinical purpose does matter for that question, and it's obviously absurd to say that it doesn't.

Should the company who does DNA matching of dog poop have to file an FDA application? Well, they deal with animals, so we can set that aside. What about the companies (yes, plural) that offer matchmaking services based on similarity of HLA chromosomes?

8/1/2012 1:00:16 PM

Typical of our over reaching, over regulating, overly pervasive government.

8/1/2012 2:17:32 PM

Quote :

"This shows a fundamental lack of understanding of the service and genetic sequencing in general. Really, none of these three are right. Look at the stories on the site."

I think you may be misunderstanding what is meant by "predictive" in reference to gene groups. A gene group which indicates an elevated probability of a disease, condition, disorder etc. is referred to as predictive. This is as opposed diagnostic gene groups which indicate the current presence of a specific condition. There are genes whose detection would be predictive of future health problems; I did mention this in the very line you quote though you seem to feel I did missed that fact somehow. There are also genes whose detection would be direct diagnostic information though there are far fewer of them.

An example of a result which would be predictive would be genes that indicate significant increase risk of various forms of cancer. You can argue that detection of these genes is not to be considered diagnostic at all because the presence of those sets of genes provides information that is predictive in nature but tells nothing of the current state of the customer's health. That is true for some genes whose detection prior to the onset of a condition would only suggest the likelihood of disease or disorder later in life. That is not all that is being tested for though and it would be incorrect to think that the only predictive information is yielded.

Examples of diseases which are diagnosed in part or in whole by detection of specific genes include, hemophilia A, Duchenne muscular dystrophy, X-linked hypophosphatemia, fatal familial insomnia, Niemann–Pick and several dozen types of ""inborn errors of metabolism." In many of these cases a genetic test is considered diagnostic in nature because the presence of the corresponding genes is indicative of the disorder. In general these tests are conducted as a confirmational diagnostic for a patient who has come in for symptoms that might lead their medical provider to consider one of these genetic disorders or as part of what is called parental diagnostics for determining if either parents are carriers.

Many of the diseases and disorders for which genetic testing can provide diagnostic information are listed under their "carrier status" tab on 23andme's website. Setting aside that this service is referred to as part of "parental diagnostic" genetic testing in and of itself, the tests provide information that can be diagnostic of the patent themselves beyond merely indicating carrier status. Being indicated as a "carrier" by such a test for an x-linked genetic disorder would provide a male patient information beyond merely the chance of their kids having the disorders.


Its not a slippery slope to say that some of this information is clearly diagnostic in nature. While genetic tests are generally not used by doctors as the sole diagnostic tool even in those cases, that is not because the information they provide is always solely predictive. It is because it does not usually provide sufficient prognostic information nor does it indicate anything about the progression of the genetic disease or disorder merely its presence. Its also not as though they just tell you that - whether you are a carrier or not - they also provide the full data which indicates full geneotypes for each tested gene. That means that means that even for genetic disorders which they are testing for "carrier status" the results would still indicate if you had both recessive genes for the disorder and thus provides information that is very much diagnostic in nature.

[Edited on August 1, 2012 at 4:41 PM. Reason : as]

8/1/2012 4:22:02 PM

^ out of this page:

https://www.23andme.com/health/all/

I can find:

Limb-girdle Muscular Dystrophy *
Niemann-Pick Disease Type A *

But to make the argument myself, it seems like Tay-Sachs might be the best available example of what you're talking about.

8/1/2012 5:00:20 PM

^ Yeah that's an "inborn error of metabolism," specifically a lysosomal storage disorder. Its probably the most familiar example of something that one might perform parental diagnostic screening for, but I had forgotten that there also is a late onset form of Tay-sachs that is indicated by combinations of two different mutant alleles (compound heterozygous.) Since the early symptoms of the progression of the disorder to disease can often mimic other far more common health problems, genetic testing would be a fairly good diagnostic tool in that case. But yeah good to point that one out, I had forgotten there was a form which was mostly asymptomatic until later in life for which this test could serve as a diagnostic tool in addition to screening.

But yeah if you are homozygous or compound heterozygous for some of these sorts of conditions then you have the disorder. This is particularly true with diseases/disorders caused by inborn metabolic errors because if you two genes coding for abnormal proteins then will have that abnormal protein and only that abnormal protein. For those diseases, genetic testing is not just predictive but an extremely powerful diagnostic tool. Perhaps this is why they list all of those under screening - to attempt to avoid the reality that the information they are providing on those gene groups is diagnostic in nature even if we presume parental diagnostics don't count.

The thing is though they provide information which can, if accurate, be used to diagnose some serious degenerative diseases while also providing diagnostic screening. The information they do present up-front and in pretty packaging can determine whether couples want to risk having children with each other. The complete test information they make available includes results which can be indicative of some (a minority of the things tested for) degenerative or chronic disorders despite what they claim in their disclaimers; some of the disease descriptions the provide admit this and state that if one is homozygous for this disease then one has this disease.

[Edited on August 1, 2012 at 8:15 PM. Reason : asa]

8/1/2012 7:51:06 PM

I'm just a little confused by the use of the word "parental diagnostics". the idea is that potential parents would use the information to decide whether or not to have a child? They certainly wouldn't use their own genetic information to diagnose a potential child, when the combining of genes isn't deterministic in the first place...

I still have trouble with yet another thing - what the role of FDA is when something is working, and really, I don't think that is the case. If it can be used to diagnose something serious, I don't think that is a reason to have agency intervention. I think the concern is different. from one of the stories:

Quote :

"In the piece that hit close to home for Carole, the professor offered up a scenario in which a consumer makes a "fatal mistake" by assuming that just because a risk was not found in their genetic test, they would skip regular screenings for breast cancer."

I see the point that products have to be regulated relative to consumer expectations, but to punish a product because it is found to be useful by its users doesn't make a lot of sense. A product can't make you forgo the correct medical treatment or consultation unless you developed an expectation that it can replace proper medical care. Like I said, you already have to sign off on a dozen wavers when using it, I'm pretty sure they covered that. Is the FDA going to write better wavers for them? I mean, how many non-solutions to problems that don't exist are we talking about here?

8/2/2012 8:51:13 AM

What do you think FDA regulation of 23andMe will look like?

8/2/2012 11:59:03 AM

^^ Well that's why I think the decision to allow the product to stay on the market while requiring the submission of accuracy/efficacy data was a relatively reasonable regulatory action. Given that the 510(k) was submitted relatively recently and the process of working with the FDA had begun 3 years ago, it seems to me that the time-frame they were allowed has also been somewhat less than overwhelmingly burdensome.

Compared to ordering a halt to sales until they were in compliance, this decision balances the mission of the FDA with the interest of the business and consumers involved. Whether this is a decision that is indicative of a reasonable and thoughtful regulator or clever negotiating on the part of 23andme is hard to say.

Whether one agrees with the intent and mission of the FDA in general or not, the story of 23andme does not seem to be an example of overly intrusive or burdensome government regulation. Call it either an exception to the rule or an example of the new direction some federal regulatory agencies are heading if you want I suppose.



[Edited on August 2, 2012 at 8:24 PM. Reason : 510k]

8/2/2012 8:12:48 PM