<p>My understanding is that the SNPs they test on the X chromosome and on 20p11 account for about 13.7% of the variance of male pattern baldness, and as such, there are a number of other genes that have been implicated in baldness (and I’m assuming their talking about androgenic alopecia, here). They don’t test for other autosomal or X-linked mutations, nor do they test for the other mutations or repeat expansions in the AR gene that have been linked with male pattern baldness (which one of the papers discusses). They don’t talk about the increased risk for women (which one of their references talks about). And of course, all of this is changing all the time.</p>
<p>I’m not saying that their information is bad or wrong, but I do think that they’re testing is not necessarily as complete or thorough as it may appear. And I know many people understand that, but not everyone necessarily will. Some may think: 1.77x increase? I’ll probably be bald when I get older! Or 0.38 decrease? Eh, I don’t have to worry about that. For baldness, it doesn’t really matter, but for another condition, it might. And I know not everyone will automatically make those leaps, but for some, it could have implications. Even if it’s not in their actual health actions, it could have implications in how they think about genetics or their condition or the effect that this could have on potential offspring. Some of those misunderstandings can be harder to correct, especially when they’ve been there for years.</p>
<p>Do I think people need an MD to explain all of this? No, certainly not. In fact, many MDs probably aren’t qualified to explain or aren’t as comfortable explaining all of the implications of these tests with the one genetics course they take at the beginning of medical school. But I do think that it would be GREAT if everyone had access to a genetics healthcare professional who could explain the genetics involved in these tests, implications of the results to family members (and not just kids, but siblings and parents, as well as extended relatives), and what the test can and cannot tell you. And there are people who do misunderstand some of the genetics involved in these situations and that can result in difficulties down the road (be they be healthwise or just psychological implications).</p>
<p>I don’t know if I really understand what the FDA wants from 23andme, but they seem to have been asking for it for a while. My understanding is they want proof that the tests are doing what they say their doing? It’s always seemed like 23andme was clear that these aren’t diagnostic or clinical tests, but maybe it wasn’t clear enough for the FDA.</p>
<p>I think the link posted in #32 is pertinent. This is not really about entertainment , this is business with a very clever model. It’s naturally intriguing to find out more about oneself, but as has been pointed out there doesn’t appear to be enough data yet to draw valid conclusions about disease risks and the FDA is clearly concerned about that and the claims being made. Not that I’m cynical or anything, but it does seem plausible that their long term goal is to collect genetic data and then market it. Usually one has to pay subjects to collect data from them, so it’s really ingenious to get paid by people to give you a sample of their most personal information. And incidentally, once you’ve sent it to them, who owns your DNA?</p>
I think the idea is to market the results from the data – the site is full of surveys, and the survey results can be correlated with the genetic data to help validate or refute some assumptions about the genes.</p>
<p>For example, I apparently have a gene that’s usually associated with blue eyes. My eyes are not blue. (I don’t really know what color they are, they are kind of greenish gray and I just call them “hazel.”) So when I answer a survey telling 23andme my eye color, and thousands of other subscribers do the same – that in turn provides more data. For example, let’s say they thought that gene was associated with blue eyes 70% of the time – maybe it turns out that it is only 60% associated – or maybe it’s associated with blue eyes 90% of the time for Scandinavians but only 30% of the time for Ashkenazi. The more participants in 23andme surveys, the more they can tease out that sort of thing.</p>
<p>
Actually there are huge ethical issues about paying subjects for data, especially in the field of medical research. It’s done, but if you are structuring a research study, you are going to have to get that past an ethics board, to make sure that the amount you are paying is not so high as to tend to bias the study, or as to put undue influence to entice people to participate in a potentially dangerous study, like a drug trial. </p>
<p>Baktrax – have you seen what the 23andme reports look like? Because your paragraph about all the different factors and the genes that have yet to be identified or tested really could have been taken from just about any report on their site. That was my biggest takeaway from the process: reading about my testing made me much more aware of how complex genetic research is, and how it is a science only in its infancy. </p>
<p>I also feel much, much more informed than I do when I get testing in a medical environment. Typically that gives me a number - for example, a glucose level - but no real understanding of what my numbers mean. Generally the doctor doesn’t provide a copy of the lab tests, and the numbers are only discussed if they are in any way worrisome (to the doctor). I go home from medical visits totally bewildered. OK, the doctor said my thyroid was kind of low, but not low enough to require medication, unless I am feeling particularly fatigued. Hunh? I’d love to receive a report even half as informative as the hundreds of reports 23andme is giving me on my genotype. (Because it’s web based, you don’t read it all at once – you’ve got an overview and then you can click the links if you want to learn more). </p>
<p>The reports encourage people to seek advice of medical professionals if appropriate. But to require the medical professional add a hefty pricetag to the information – and a lot of us don’t need genetic counseling. </p>
<p>I know that I can get breast cancer even if I don’t have the BRCA gene – one reason I know that is that 23andme tested me for other breast-cancer associated genes too, including CHEK2 mutations, and genotypes for FGFR2, 16q12, 2q35, CASP8, 3p24, STXBP4, and TAB2. I had never even heard of these before – who knew? Would formal medical testing for BRCA come back with a report on all of these?</p>
<p>I feel I’ve learned a tremendous amount from reading these reports. Would seeing a genetic counselor give me more information? Not for my purposes – I would have to spend hours and hours with the counselor to touch upon the information that is included with my 23andme report – and for what purpose? who pays? There is nothing that a genetic counselor can talk to me about because I’m past childbearing age, and I don’t actually have any symptoms of any of the diseases for which I am heightened risk. But I feel I’ve benefit from learning more about these diseases, and measures I can take in terms of my own health. Maybe they are all common sense advice anyway (things like eating a healthy diet and weight control,) but reading about them in this context might make me more likely to heed such advice. </p>
<p>It seems to me that the FDA’s concern is basically that I’m too stupid to handle this stuff. </p>
<p>I’d rather have the right to make that decision on my own. If the FDA wants 23andme to print certain disclaimers on its reports or advertising, fine with me. But their claim that the little plastic spit-container is a “medical device” seems pretty far fetched to me. </p>
<p>I don’t think this really came out of the blue to 23andme – I think they are well funded and have lawyers and probably want to be able to go to court and set a favorable precedent.</p>
<p>If the FDA is regulating the plastic spit-container as a medical device, what about all the other DNA companies like Ancestry.com which use essentially the same device? In principle, 23andme could have Ancestry.com sequence your DNA, and then those who are interested in their medical info could download their DNA code from Ancestry and upload it to 23andme. Voila - 23andme is out of the “device” business.</p>
<p>Now I don’t think this is what 23andme wants to do, but it shows how ridiculous it is for the FDA to try to regulate the spit-container as an unproven “medical device.”</p>
<p>I don’t have direct experience with the 23&Me case but I know from my professional work what the FDA expects for medical devices. (And medical tests are classified as a type of medical device). There are three levels of regulation: Classes I, II, and III. Class I is for the simplest and safest devices that have a long history of safe use - tongue depressors for example. Class II is for devices of moderate risk that are in current use, or in case of newer versions of a product, that are shown to be “substantially equivalent” to similar devices in current use (a “predicate” device). This is the most commonly used classification. </p>
<p>The submission you file with the FDA for Class II device is called a “510(k)” for the section of the regulations that provides for it. The studies you must do for a 510(k) are usually mostly analytical - showing that it accurately and reproducibly measures the target molecule and performs as well or better than the predicate device.</p>
<p>Class III is for high risk and/or new devices that have no predicate. For that the company must do a full-blown clinical trial not only to show that it measures its intended target properly but also to define the safe medical use of the test. The filing for a Class III is called a Pre-Market Approval (PMA).</p>
<p>From what I’ve seen in the papers about 23&me case, the FDA is hammering them for failing to complete their 510(k) submission and/or failing to answer FDA’s questions about their original submission. They apparently have not completed all the validation studies that the FDA routinely requires. </p>
<p>This is not a merely a spat over bureaucratic technicalities. Based on the highly public nature of the FDA’s move you can tell this is serious business. When FDA brings the hammer down so publicly it means the company is probably in serious violation of the regulations, and the FDA believes the risk to public health is real.</p>
<br>
<br>
<p>And that’s the problem. All that information they are providing is far beyond what they can legitimately claim for their test. They are quoting data from scientific papers where other scientists using other tests have found these disease associations with certain genes. But they are not showing validated data for their own test. They claim they are measuring the same gene, so by implication they are detecting same disease associations and risks. But they have not shown the FDA data proving that this is case.</p>
<p>By analogy, someone could publish a study showing that a car with a small hybrid engine gets 40 miles to the gallon. So along comes another car company and says my car has a small hybrid engine too, so therefore my car also gets 40 miles to the gallon. The EPA would say “Not so fast. Show us the results of the test of * your* car that demonstrates 40 mpg.”</p>
<p>It’s the same here. The FDA is saying “Not so fast…” to 23&Me.</p>
<p>Not really ridiculous if you think about it. FDA routinely regulates specimen collection devices (needles, syringes, collection cups, etc) for medical tests much the same way it regulates the tests themselves. This is because the results from the best test in world are meaningless if the specimen that was tested was not collected and stored in a way that preserved the specimen’s integrity and stability.</p>
<p>“It seems to me that the FDA’s concern is basically that I’m too stupid to handle this stuff.”</p>
<p>All other arguments aside, you personally might be an expert in this field, but the science education level of the general population is quite abysmal. 23andme is not marketing this to people with degrees in genetics and molecular biology, they sell their “test” to anyone willing to pay $99. While you might have the education and erudition to interpret the lengthy information, a joe the plumber will have no clue what the majority of this stuff means. </p>
<p>Thank you, coureur, for taking the time to explain the medical device approval process. Absolutely agree with what you wrote. $99 will buy me a lot of healthy blueberries ;)</p>
<p>Correct. Which is why the content of direct-to-consumer ads is also very highly regulated by the FDA - right down to the details of the wording and punctuation and also specifying the types of shiny, happy people that can be portrayed in the ad. FDA also must approve the trade names given to drugs, “Viagra” and what not. In both cases it’s to minimize the risk that consumers will get wrong information or jump to wrong conclusions about a drug.</p>
<p>Somehow, I doubt the intended targets of those ads typically read the pages of fine print in the print ads. Then there are the ads on television that don’t have as much fine print (and don’t last long enough to read it).</p>
<p>^^Correct again. People usually don’t read the fine print on those TV ads, which is why FDA requires the companies to include a voice-over reading that long list of all the possible side-effects, regardless of how extremely rarely those bad things might actually happen. </p>
<p>There is no way a company would tell you out loud that its product might cause skin rash, vomiting, diarrhea, severe bleeding, and occasional death unless the FDA made them do it.</p>
<p>I wonder how many here have actually bought and received their results. I did it out of curiosity. I found the “ancestry” part interesting, but the medical part expectedly underwhelming. Of course, having participated in the statistical evaluation of a large pool of HPV screened patients, I happen to be extremely biased against the results of the simple devices and small tubes. For the record, the same patient did routinely have very different results from the same samples depending on the test device and lab procedure. </p>
<p>There is little doubt that this company oversold the validity of their medical diagnostics, if it can even be called a medical screening test. The overall implication of relying on such test beyond the pure entertainment is strongly negative. And this is one case where the FDA is on the right side.</p>
<p>I’ve only looked at their sample reports–I haven’t done it myself because I haven’t had push to cough up the money for something that I would never really use. I’m not saying that they never mention–I’m saying that they don’t test for it. And someone who just looks at their report (I have seen where they show you an overview of your risk for different conditions on someone else’s report) and doesn’t read the literature of what it’s testing for and how relevant those tests are to you may interpret the results incorrectly or read more into it than they should. I’m not saying YOU are. I’m saying that some do, and that can cause problems down the line. There are gene mutations they’re not testing for, there’s environmental factors that we can’t quantitate, there are implications or situations that they can’t necessarily predict (for instance, germline mosaicism, incomplete penetrance, variable expressivity), etc. I’m not saying they’re not doing a good job of explaining what they’re doing, but like I was saying before, I just want everyone to read exactly what they’re testing for and what the test can tell you. I’m not saying that you’re not reading the information (and I never was), and I’m not saying that no one will. But some won’t (and I’d recommend reading the information regardless of what they say your risk is), and not everything they give you is relevant to what you need to know.</p>
<p>
</p>
<p>First of all, genetic counselors do counsel individuals past child-bearing age. One of the main specializations of genetic counselors is cancer genetic counseling, and most of their clients are past reproductive age and are interested in testing mainly because of the implications for their children–much like you were. There’s plenty that a genetic counselor could talk to you about. Genetic counseling isn’t just prenatal anymore–there’s cancer, pediatrics, cardiac, psychiatric, adult, neurogenetics, etc.</p>
<p>Second of all, we all have the BRCA gene. It’s a tumor suppressor gene that when mutated can be associated with breast and ovarian cancers. Would a genetic counselor have tested you for mutations in BRCA1 or 2 with no indication and no family history of breast or ovarian cancer? Probably not, but like you said, you never would have seen a genetic counselor. If you did see a genetic counselor, could they talk to you about all of this? Absolutely. But they probably wouldn’t have tested you because it would have been a waste of your money (though genetic counselors are very non-directive, so if you had wanted to test regardless, they still would have helped you get it). If you did have a family history of breast and ovarian cancer, it would have been best to test an affected individual first (generally, someone who was affected at a younger age) because they are the most likely to have a mutation in BRCA1 or 2 (or a number of other genes associated with breast cancer risk), and if that individual had a positive result, they could have tested you (and your daughter, if you like) for that specific mutation in that specific gene, which is a much cheaper test than that whole panel. And depending on what is most likely given the history of cancers in a particular family, they may start with a particular test. It’s not just about identifying the genes (and there are even more genes associated with breast cancer risk than the handful that 23andme has chosen); it’s about testing for for mutations in those genes, and 23andme is only testing for a couple (and can’t test for things like deletion/duplications, repeat expansions, etc). So you COULD have a mutation in BRCA1 or 2 that 23andme didn’t test for. And I know you realize this, and I’m sure you read what 23andme said–but not everyone will, which is why my only advice to anyone considering this is to read the actual report. Know what this test can tell you. Because while you may be Ashkenazi Jewish and the mutations they are testing for is very relevant to you, not everyone who does 23andme is Ashkenazi Jewish so this test may tell them very little.</p>
<p>Who knew about the other genes associated with breast cancer risk? Well… genetic counselors and medical geneticists. There is a lot we don’t know, sure, but there is also a lot that we DO know. And what may seem like a huge, new concept to some (I read an article about mosaicism the other day that made it seem like this scary new discovery), may already be taught in genetics classes.</p>
<p>I’m not saying that 23andme sucks and is useless (because it’s not), but I AM saying that there are limitations. I think what they’re doing is fun and great and probably ultimately harmless, as long as everyone is aware of what they tell you and what they don’t tell you (and that you can send the same sample to multiple different direct-to-consumer testing companies and get a different risk estimate–which I believe has been done before). I’m not saying you’re unaware or that every one is (and no one’s calling you stupid), but I caution ANYONE not to just read their risk and be done with it. I want everyone to do exactly what you’re doing–read the actual report and think critically about what it can and can’t tell you in the context of your own life. If you (and not calmom specifically, but I’m talking to anyone) have a history of breast and ovarian cancer in your family and you test negative for a mutation on BRCA1 or 2, you are still at elevated risk to get breast and ovarian cancer, and you should still take whatever screening measures you would have even if you didn’t know the BRCA1 or 2 test result. Someone saying that they don’t “have the BRCA gene,” so they don’t have to be as worried about that and then skipping screening for breast cancer is what I’m most concerned of. That could have real implications for someone’s help, and it’s not 23andme’s fault. I’m just saying that everyone should read these reports, whether or not they say you have an elevated risk.</p>
<p>Would formal BRCA testing come back with results for all of these? Well, no, because if you’re only testing for BRCA mutations then that’s all they would report. If you saw a genetic counselor, could you discuss testing for all of these genes? Sure, and more, and there are comprehensive panels that you could do. Ideally, they would identify not only the genes that are most likely implicated in your situation (given you’re particular medical and family history), and they could order the tests for the right mutations in your particular situation. So for me, for instance, since I’m of East Asian descent, they wouldn’t just order the three mutations that 23andme does. They’d likely test for mutations all along the gene, as well as deletions and duplications. And if the test came back positive, they could explain what it means, could give me specific risk estimates based on empiric data given my particular family history and mutation (and could give you their sources, if you asked for them), and could give me something to send to other at-risk relatives so they could be tested for that specific mutation. And they would likely give me a letter (to me and to my physician) giving a summary of everything that occurred in the session and their recommendations.</p>
<p>Would that take longer than a 10 minute appointment with your doctor? Yes, definitely. I’ve seen cancer genetic counseling appointments scheduled for 1.5 hours or more. Genetic counselors aren’t doctors. The point is for them to do what 23andme is giving you, but with more current and accurate scientific and testing information, with the weight of a medical team trying to navigate the best course for someone (especially if there’s no straightforward path), with the incorporation of your family history, and with empiric risk assessments for mutations that 23andme may not test for. I’m not saying you should HAVE to go through a genetic counselor or a doctor, but do they add value to the situation? Definitely. In fact, there are people who had 23andme testing and then go in to see a genetic counselor, which is causing a whole shift in how genetic counselors have traditionally operated (but that’s a whole different story).</p>
<p>And who pays? As with all things in America, it depends on your insurance =D But genetic counseling isn’t just for the wealthy. There are many people who come in from all sorts of socioeconomic backgrounds, so it’s not an impossible barrier. Genetic counselors deal with insurance companies all the time, and there are whole debates over the most financially efficient way to do things. It’s very similar to the 23andme conversation, but on a much grander scale. Should we just sequence the whole genome to see what we get or should we do a sequence of tests testing for specific things? There are huge ethical and financial implications on either side.</p>
<p>Baktrax – the point is, I wouldn’t get genetic screening (or see a genetic counselor) for breast cancer without 23andme. As far as I know, no one in my family has ever been diagnosed with breast cancer – there are NO known risk factors that would justify that sort of testing. But as I am sure you know, many people like me (no known risk factors) get cancer. So for $99 I get a screen that has a whole lot of information about a whole lot of things – and even gets updated over time as more information comes in. It’s not as if I am doing it as an alternative to genetic counseling --I’d be insane to waste money on genetic counseling based on what I knew about my health before the 23andme test – and no insurance company would pay for screening or counseling.</p>
<p>As it happens, nothing showed up in the testing that would lead me to seek further medical counseling. Again, it would be a waste of money, a waste of my time, a waste of the counselor’s time.</p>
<p>On the other hand – lets say I develop other issues and my doctor wants to put me on estrogen (let’s say, for example, because of issues related to bone density). Do doctors routinely do genetic testing these days before prescribing HRT? I don’t really know, but I don’t think so – I have a piece of information in my profile that’s a little red flag, that says: maybe not. I’d raise it with my doctor – I might then ask for more formalized testing, or seek genetic counseling when the need arose. But it probably wouldn’t occur to me to even ask that question but for the report I have in hand. </p>
<p>I don’t have a choice between a $99 scan and the “real thing” (medical labs, geneticists, etc.) I have a choice between a $99 scan and zilch.</p>
<p>calmom – And my point is that just because it worked out for you (as far as you are concerned) doesn’t mean it will work out for everyone. It’s important to know what could happen before it happens and you’re unprepared, which is why I encourage everyone to get educated about these topics (and about your options) and not just take it at face value. There are no guarantees, and there are a lot of things that 23andme can’t tell you (like, it can’t tell you that you don’t have a mutation in BRCA 1 or 2). My point is that no one should take it as law because there are all sorts of grey area in genetics, and the best way to become educated is to read their literature on what they’re looking for and what it can tell you.</p>
<p>I know the vast majority of people will never see a genetic counselor in their life–the only reason I brought it up at all was because you were saying that you would never receive the type of information that 23andme gives you from your doctor. And you probably wouldn’t, but I just wanted to bring up the point that there are healthcare professionals where part of their role is education.</p>
<p>Here’s an article that pretty much sums up my point of view, but says it much better than I can – and also comes from someone far more knowledgeable than I am (a law professor who teaches a course in genetics and the law, and whose work is specifically focused on legal aspects of personalized medicine, and regulation of emerging technologies)</p>
<p>This would be my biggest concern. Thank goodness we’re no longer in danger of being excluded from health insurance based on our DNA. But it could certainly be used to exclude us from, say, life insurance, or from life insurance payout if we die from a specified cause based on our DNA profile. Some employers might also be interested in this information. </p>
<p>I googled “23andme, who owns my dna.” Apparently at one time, they were willing to answer this question, but now both the current and cached versions say:</p>