Genetic Future

Update 14:35 EDT 04/08/10: I just received a phone call from a 23andMe representative indicating that these rumours are not true; more details to follow once I have a written statement.

Update 17:26 EDT 04/08/10: Here is the written statement from 23andMe’s PR firm:

We will continue to work with Illumina as our discussion with the FDA continues.

I’ve requested further clarification.

Update 17:59 EDT 04/08/10: I’ve received a (slightly) longer written statement from 23andMe clarifying that they have “no reason to believe there will be an interruption in the supply of Illumina chips at this time”.

I’ve been hearing ominous rumours that the FDA is planning to completely choke off the supply of genotyping chips to direct-to-consumer genomics companies. GenomeWeb reports that the FDA met with Illumina last week (following its letter to the company back in June) to discuss the company’s supply of genotyping products to DTC companies:

Illumina CEO Jay Flatley said last week that the company is in discussions with the US Food and Drug Administration regarding its supply of arrays to firms that conduct direct-to-consumer genetic testing.

During a conference call to discuss the company’s second-quarter financial results (see story, this issue), Flatley said that Illumina met with the FDA during the quarter to discuss the regulatory framework for DTC genotyping, and that the company intends to “fully comply with the FDA’s guidance.”

He noted that the revenue Illumina currently generates from sales of arrays to the DTC market is “immaterial.”

Illumina was the sole array vendor to receive a letter in June when the agency contacted 23andMe, Navigenics, Decode Genetics, and Knome about their services (BAN 6/15/2010).

According to the letter, which was addressed to Flatley, Illumina was contacted because 23andMe and Decode use its research-use-only Infinium HumanHap550 array in their direct-to-consumer genetic-testing service. In the letters, all dated June 10, the FDA asserts that the gene-analysis systems the companies are using are medical devices and therefore need to be approved by the agency. The FDA, which cites section 201(h) of the Food, Drug, and Cosmetic Act, 21 USC 321(h), urges the firms to contact it to obtain information that would enable them to “legally market” their services.

The rumour on the grapevine is that Illumina will soon be discontinuing its supply of chips to DTC providers, if it hasn’t already. However, I’ve also heard indications to the contrary, so I’ll provide updates as and when I find out anything official.

Either way, it’s clear that the FDA does intend to target DTC companies via chip providers. Let me spell out the implications: if the FDA has “requested” that Illumina stop supplying its chips to DTC companies, Illumina will “fully comply” – the company has little choice but to stay in the FDA’s good books if it wants to move heavily into the US diagnostic market. Illumina supplies chips used by 23andMe and deCODEme, so both these companies would lose their ability to genotype customers as soon as their current supply contracts run out.

Moreover, it’s unlikely that other chip suppliers (such as Affymetrix) would provide an alternative source: as Mya Thomae notes, “FDA could have sent that letter to dozens of companies, but they only needed to publicize one in order to put all suppliers on notice.” Like Illumina, Affymetrix won’t jeopardise its relationship with the FDA to save its tiny income stream from DTC providers.

With their chip supply strangled, DTC companies will be completely at the mercy of whatever the FDA demands. If the FDA decides it wants to kill off the DTC industry entirely, forcing the industry to move towards the physician-only route (as Navigenics and Pathway Genomics have done already), it can do it. If it wants DTC companies to abandon all disease risk prediction and move to exclusively ancestry/genealogy products, they will have no choice but to comply.

In addition, targeting chip manufacturers means the FDA can reach far beyond the US. Any DTC genomics company, whether it operates from Silicon Valley or Shenzhen, currently relies on genotyping products from these large suppliers.

I’ll update you if anything new emerges.

Comments

#1 Keith Grimaldi
August 4, 2010

This is incredible if true – even more underhand than the GAO, and this in the “land of the free”. I had always heard so much about USA being the place for entrepreneurship, business growth free of shackles and bureaucracy (but when we actually went there I was very disillusioned).

Are they just trying to chop off hydra heads here though – are there chip sources from abroad…China, Singapore, etc? Some company surely would step into the breach.
#2 Mark Henderson
August 4, 2010

Very interesting Daniel, as ever. It occurs to me that this could hand a big advantage to Pathway, who use their own custom chip. Not sure of the extent to which this relies on Illumina/Affy technology.

Perhaps there was a cunning plan behind the Walgreens episode all along!
#3 Daniel MacArthur
August 4, 2010

Hi Mark,

My understanding is that Pathway still relies on a custom Illumina array as a core product, although this is also supplemented with other genotyping technologies. That information might be out of date, though.

Following up on Keith’s comment, does anyone know if there are East Asian companies manufacturing high-quality genotyping arrays?
#4 beebeeo
August 4, 2010

Hi there
It seems to me that if DTC genotyping is used to predict disease related traits then the FDA has every right and obligation to regulate those. Just like test done in lab in a hospital is usually provided by other companies, it is those companies that have to seek FDA approval, just like Illumina in this case. So nothing unusual there. If the Illumina or the DTC companies cannot prove the accuracy of their results and substantiate the predictions that they make, then they shouldn’t be in the business.
The companies should refrain completely from making any medically related predictions (nowhere on their website, not even a hint of a link).
It might be patronizing to the consumers but the truth is that most do not have the knowledge required to evaluate the information given to them by the companies.

Even if other sources for chips are found outside the US, the companies would still need FDA approval if medically relevant predictions are going to be made.

The problem of course is that because the arrays could be used for disease prediction, would Illumina have to stop selling them even to companies that don’t do any disease prediction?
I too am enthusiastic about the use of genotyping to assist personalised medicine (prediction & prognosis, prevention and treatment) but I do not think the array technology has proven itself to be very valuable for that, yet. It will happen of course, at some point (maybe whole genome sequencing will have replaced arras by then) and a new regulatory framework has to be created.
If disease risk prediction is really so important for 23&me etc to succeed, then I fear they have jumped into this business too soon and will deservedly fail.
#5 Keith Grimaldi
August 4, 2010

Regulation is one thing, blocking Illumina selling them a chip is another thing altogether, has anything like this ever happened before to any industry?

Apart from sourcing overseas chips there must be many ways around this surely. e.g. setting up a stealthy overseas subsidiary to buy the chips from Illumina and sending back to parent company
#6 Albert Vernon Smith
August 4, 2010

In the context of this, one needs to remember that Illumina owns 16% of the “new” deCode (according to the bankruptcy motion of DCGN). I find it hard to believe that Illumina could be pressured to cut off supplies to deCode. However, it wasn’t so long ago that deCode doubled the prices of their services DTC well beyond the competitors; deCode needed to price DTC higher because they didn’t want to underprice their direct-to-medical-professional business, which is where they really see their future. Given all this, one could conceive that Illumina might not be completely opposed to limiting the DTC business, because it could push more business to higher revenue testing at deCode, where they would see a greater return on their investment.
#7 symball
August 4, 2010

-Medical diagnostics are regulated by the FDA.

-These companies are selling a medical diagnostic service.

-Therefore any kits or components that they use are covered by the regulations surrounding medical diagnostic devices.

The FDA are just doing their job. If you want to go back to allowing any old snake oil salesman to make whatever claims they like then remove the regulations. The reason the industry is so tightly regulated is the immense harm a bad diagnostic can cause. If you could get good clinical diagnoses using these chips, rather than vague probabilities that something might occur if a whole load of other unknown factors also exist, the companies would be busting a gut to actually get these licensed properly.

I live in the UK so I really don’t care either way, these services are already illegal here. although we have plenty of other problems with regulatory bodies.
#8 Keith Grimaldi
August 4, 2010

@symball – they are not illegal in the UK. As far as I know they are specifically outlawed in Germany by recent legislation while older laws for clinical genetics could possibly be used against DTC in France and Switzerland. There is nothing banning it in the UK – if there is, the HGC have just wasted 2 years on their code of practice for DTC
#9 Daniel MacArthur
August 4, 2010

Albert,

I agree that Illumina is unlikely to stop selling products to deCODE as a whole, but they could sell products under the agreement that they weren’t to be used for DTC medical risk prediction.

symball,

These companies aren’t illegal in the UK; and scientifically valid annotations of individual genetic data aren’t “any old snake oil”. Please do some background reading before commenting here; this isn’t a Daily Mail comments thread.
#10 asdf
August 4, 2010

symball and beebeeo appear to think that associations published in Nature and Science are “snake oil”. A better word would be “best of the available research”. No technology gets to version 10 without going through some earlier iterations.

Regarding pathway they are not opening up semiconductor fabs anytime soon. “Custom” is still built to spec by Illumina.
#11 Dan
August 4, 2010

Daniel-these are not scientifically valid annotations. That is the problem the FDA has with them right now. Except for a small handful of snp’s these DTC genetic companies use statistics to associate disease with genes and calculate risk, they are not clinically proven. Even if you have a gene statistically associated with a disease, chances are, you will never come down with the disease, in many cases the associations are dis-proven in animal studies a few years later. DTC companies only test for a small fraction of one type of mutation, they aren’t even looking at duplications, deletions, etc. In many cases they attempt to associate what is thought to be a polygenic disease to a single snp.
#12 Reluctant Tea Partier
August 4, 2010

“USA being the place for entrepreneurship, business growth free of shackles and bureaucracy”

Hate to get political because most of us voted for the man. but its kind of important to realize we’re in the age of Obama now. Out of the frying pan of Bush into the even hotter fire. New administration ordered the hit on DTC genomics (Gutierrez has direct quote to this effect). Regulatory action here is just a microcosm of what’s going on elsewhere in the economy and is exactly why businesses are hoarding cash. Why would you put 30 million into a new product if the fedguv is going to shut it down after reading a petition from genentech? Read Ivan Seidenberg (Verizon CEO, member of Presidents economic advisory board). His Business Roundtable put together a 54 page summary of all the new job killing regulations. Yes, 54 pages in length for the summary.
#13 beebeeo
August 4, 2010

@symball, @Daniel MacArthur

scientifically valid annotations of individual genetic data aren’t “any old snake oil”

Yes but without regulation, how do we make sure the information and the advertising are all scientifically valid.
Even predictions based on the latest research can turn out not to be validated in later studies, or the populations that the cases and controls were taken from might not be appropriate for a specific customer.
The field is just too new and the validated predictions are just too few to be of real benefit.
Again, without wanting to be patronising to consumers, it would be a bad idea to establish an industry like this without the proper regulation.
If you read any of the blogs commenting on alternative medicine here at scienceblogs, for example Orac’s blog, then you will know that what kind of quackery can be done without the proper regulation.
#14 symball
August 4, 2010

My apologies for the misunderstanding about the legality of these tests- Although I am very dubious about the ethics of this whole direct to consumer testing.

Although I was guilty of hyperbole when comparing these tests to snake oil, they most certainly are not anywhere near accurate or precise enough to do any more than give a rough guide as to whether there is an increased risk of developing a disease related to the gene, or genes found. While this is invaluable information to researchers and even some doctors (specialists rather than GP’s) when diagnosing problems, there is no good reason to be selling information like this to individuals who have very little chance of understanding the real meaning of these results.

I know this is an exiting field, and that there is a wealth of information to be gathered from tests like these, but they are not of a standard to accurately diagnose disease. Clinical Diagnostic tests have massive hoops to jump through before they are licensed, because false or inconclusive results have a massive impact on the people who receive them.

If these arrays produce data that is accurate enough to diagnose disease to the regulatory standard then they should apply for a license. Otherwise it might be better to stick to the research market where they have such an important role to play.
#15 Orac
August 4, 2010

It seems to me that if DTC genotyping is used to predict disease related traits then the FDA has every right and obligation to regulate those. Just like test done in lab in a hospital is usually provided by other companies, it is those companies that have to seek FDA approval, just like Illumina in this case. So nothing unusual there. If the Illumina or the DTC companies cannot prove the accuracy of their results and substantiate the predictions that they make, then they shouldn’t be in the business.

Precisely. Dan’s expressing shock that the FDA is actually doing its job for a change, as if that were a bad thing.
#16 Orac
August 4, 2010

Although I was guilty of hyperbole when comparing these tests to snake oil, they most certainly are not anywhere near accurate or precise enough to do any more than give a rough guide as to whether there is an increased risk of developing a disease related to the gene, or genes found.

What hyperbole? With few exceptions these tests as currently marketed are very much akin to snake oil. There’s minimal or no clinical validation, and we don’t really know for sure what the results mean. A lot more work will need to be done to make these results meaningful, and even then it’s very much advisable to leave such tests under the purview of physicians and genetic counselors who know how to interpret the results, as we do with cancer risk genetic testing right here at the cancer center where I work.
#17 Orac
August 4, 2010

With their chip supply strangled, DTC companies will be completely at the mercy of whatever the FDA demands.

Which is as it should be if they’re going to sell a test to diagnose human disease or the predilection to human disease. Why should DTC companies be treated any differently than other test and device manufacturers?
#18 Mike
August 4, 2010

@symball
The basis of your argument is that these are clinical diagnostic tests, which they are not, nor AFAIK are they claimed to be so.
The question is are DTCG test being used to make clinical decisions? If so, how, where, when, and by whom? If they are not, then surely they do not fall under the category of clinical diagnostics.

DTCG testing doesn’t make claims about clinical utility, validity or applicability. The problem lies with people’s misunderstanding of the probabilistic nature of risk prediction. Just because it may be used in the future to predict onset of disease, doesn’t mean it is now.
#19 John
August 4, 2010

As I understand it 23andMe contract out the basic analysis. FDA should not stiffle the legitimate business of a company doing high quality basic analysis. The target, if warranted, should be the company subsequently using that data for health, genealogy .. or earthquake prediction .. purposes.
#20 symball
August 4, 2010

@Orac- Ta, I was just being polite and dialling down the language a little. I need to practice at my RI I guess!

@ Mike #18
your second paragraph sounds rather like the quack miranda. These companies cannot tell anyone with any degree of certainty that they have a disease or are at risk of developing it, otherwise we would not be having this discussion and they would be marketing as a clinical service. By your logic we shouldn’t need an FDA or an MHRA- caveat emptor instead.
#21 Keith Grimaldi
August 4, 2010

Daniel’s post is specifically about an extraordinary possibility that the FDA may destroy a business not through regulation but through opaque pressure and threats on suppliers. Not many in the field are actually opposed to some sort of regulation – the main debate is on the nature of it.

There is snake oil out there and the reputable companies would benefit from it being identified and blocked. It’s very disappointing though to see some of the ignorance in some of the comments that brand it all as snake oil. The “sceptics” sometimes are correct about alternative medicine but often get so dogmatic as to become irrelevant.

@symball goes from wrong and hyperbolic to patronising:

“there is no good reason to be selling information like this to individuals who have very little chance of understanding the real meaning of these results…I know this is an exiting field, and that there is a wealth of information to be gathered from tests like these, but they are not of a standard to accurately diagnose disease.”

Why should the information be denied to anybody whether they can understand it or not? And it is INFORMATION by the way, NOT clinical diagnosis. You are right, it is interesting, comes from research mostly paid for by public money, there is no evidence of harm but quite a bit on benefit.

#11 Dan: “Even if you have a gene statistically associated with a disease, chances are, you will never come down with the disease, in many cases the associations are dis-proven in animal studies a few years later.”

That is the whole nature of any predictive test, ditto for any blood test measuring cholesterol, liver enzymes, glucose, etc. It’s also a feature of the very hit-and-miss family history assessment for disease risk (so is family history snake oil?). As for dis-proving: this goes back to the pre GWAS era, when genetic association studies were done on about 2 people and a budget of £50. The GWAS information on which the main services are based have been very robust and reproducible.

Surprisingly given the early stage of the field most publications looking at clinical relevance or predictability of SNP panels have found them to be “no better” than classical markers, or “add little more”. This has been taken to be a criticism, I see it as a tremendous positive – they are not worse, they are not random, they actually work, and will get better. I would rather know about a potential risk via genetics than find out my blood biomarkers are all over the place. I.e. I would rather know when I am well, not when I am pre-disease or actually diseased and when I am well I only have 2 possibilities: genetics and family history, why not use both, they are complementary?
#22 Dave
August 4, 2010

beebeeo, your assessment about DTC customers is way off base and unsubstantiated. I would argue the counter point that most DTC customers are very well equipped to handle any information they receive from a company like 23andME.
#23 Daniel MacArthur
August 4, 2010

Hi Orac,

What hyperbole? With few exceptions these tests as currently marketed are very much akin to snake oil.

23andMe uses even more stringent standards for the inclusion of genetic associations than academic researchers (e.g. the requirement that they be found in two independent studies, whereas most researchers would be content with internal replication).

deCODEme is a spin-off of a company that has uncovered more replicated associations with human complex traits than any other organisation on the planet. 23andMe has accumulated a database of over 50,000 customers that it has used as the basis for peer-reviewed publications.

These companies use testing labs subject to the same quality control measures as those used by clinical labs. (23andMe, for instance, uses LabCorp, who I’m guessing also processes a fair proportion of the tests you order as a clinician.)

Are there snake oil salesmen in the broader DTC genetics industry? Sure, and they need to be stamped out. It’s not like this would be difficult: existing laws, enforced by existing consumer protection agencies, could do this tomorrow with a minimum of fuss.

But instead of sensible enforcement to weed out the bottom-dwellers, we get political theatre from the FDA and Congress. And instead of proposing simple, transparent guidelines that the industry needs to hew to, we get the FDA manoeuvering to disconnect DTC companies’ supply lines without ever actually telling them what is required for them to operate legitimately.

23andMe has met with the FDA numerous times since it launched in 2007. In 2008, it was explicitly given the green light to operate by the then FDA commissioner. Now, without any public change in its rules, the FDA is preparing to crack down on an industry without ever actually explicitly telling them what they’re doing wrong, and without any evidence of actual harm to consumers.

There’s minimal or no clinical validation, and we don’t really know for sure what the results mean.

These tests aren’t marketed as clinical tests. They’re explicitly labelled as non-diagnostic, not to be used for medical purposes.

It’s true we don’t know for sure what the results mean; but then, that’s true for a lot of medical tests as well. Should we wait until we completely understand the biology before we allow people to look at their own genome? Maybe you do genuinely believe that, but I think it’s an appallingly paternalistic view.

A lot more work will need to be done to make these results meaningful, and even then it’s very much advisable to leave such tests under the purview of physicians and genetic counselors who know how to interpret the results, as we do with cancer risk genetic testing right here at the cancer center where I work.

I’m all for encouraging people to take their results to a doctor for interpretation, if they choose to do so. Every company should explicitly encourage their customers to do this (and indeed all of the legitimate ones do; several also offer access to qualified in-house genetic counsellors, although I’d argue that consulting independent experts is a better idea in general).

But should I need to get a doctor’s permission before looking at my own genetic data? Seriously? Should we ban the sale of electronic blood pressure monitors and home pregnancy tests as well, in case the poor confused consumer misinterprets them? After all, right now a pregnancy test is more likely to have a life-changing impact than a genome scan.

This paternalism is particularly terrifying given that the vast majority of physicians and genetic counsellors actually have absolutely no idea how to interpret a genome scan; 23andMe does a far better job of contextualising this information than you would get from your doctor.

Do we need sensible, light-touch regulation to weed out the scammers? Yes. Do we need greater transparency from the industry? Yes. Do we need to stop consumers from accessing their own genetic data in case their heads explode, and instead make sure they can only look at their own DNA with a clinician as gate-keeper? No, no, a thousand times no.
#24 Keith Grimaldi
August 4, 2010

I agree with Dave, #22

Further, most clinicians have had very little scientific training. Some do pursue research and are excellent, quite a few spend a couple of years in the lab to get a career enhancing MD qualification (talking about UK here) and with a few exceptions most of those that I worked with were just interested in the qualy. Most though go from med school to first job. The result is that the majority are really not equipped to interpret the results and it’s just as well that the 23andEtc companies do such a good explanatory job.
#25 Jonathan
August 4, 2010

Keith I think you’re overreacting. FDA are not forcing Illumina to stop selling their chips to customers, but they are asking that, if those customers are going to be given medical information, then those chips need to have gone through FDA approval.

I fail to see how telling someone their Apoe4 or BRCA1 status is anything other than medical. Further, it’s incomplete medical advice since they’re just looking for a handful of common SNPs, telling you nothing about CNVs or new mutations.

If you want to avoid FDA regulation, don’t make medical claims. But if any of the companies came out and did that, their VC funding would disappear faster than it already is.

You cannot have your cake and eat it. If you want to play in the medical realm, then you have to play by the rules. If you just want to offer earwax and asparagus genes, and ancestry information, FDA won’t get involved.

@asdf:

If this information was clinically useful and not snakeoil, then sending your DNA off to three different companies wouldn’t get you three different answers. The fact that there are such wild discrepancies in reporting ought to be proof enough of that. You can point to individual Nature or Science papers all you like, but when 23andMe, Navi, Pathway, and deCODE all draw different conclusions from those papers, I think we can deduce that they’re hardly as concrete as you think.
#26 Keith Grimaldi
August 4, 2010

@Jonathan

Not many (myself included) are saying NO to regulation full stop, we are discussing what level is necessary. The “what is medicine” question has been done to death and will not help. I can buy quite strong medication in Tesco or Walmart, stuff that used to be prescription only. I can read on Medscape or WebMD some “frightening” fact about having a 10x increased risk of diabetes or whatever if my BMI is X, and waist circumference Y

Many interpretations of the same type of test will yield different risk results, this includes non-genetic algorithms to measure the risk value. It is not confined to genetic analysis. The differences between the 3 companies were not particulary wild and the similarities were greater. That said it is pretty much accepted that they need to improve their performance in that area. Fine. I don’t point to individual Nature papers, nor do the companies, they take all the research and put it together. Not snake oil.

Repeat – the real problem here is not regulation but the rumoured methods by the FDA – how can a few individuals have the power to bring down a business is such a way. I’m not in any of the businesses affected, I have been though and have been through various threats and I am not over-reacting.

Contrast between FDA/GAO (might as well get the FBI in there too) and the HGC is interesting – which approach is really trying to find the best solution to undoubted problems? Which is really trying to find the best way for ALL involved?
#27 symball
August 4, 2010

Daniel- I don’t think there is actually a great deal of difference between us regarding the usefulness of these tests to scientists and specialists. It is the marketing of this to the general public that is the problem here.

You claim:

These tests aren’t marketed as clinical tests. They’re explicitly labelled as non-diagnostic, not to be used for medical purposes……….23andMe does a far better job of contextualising this information than you would get from your doctor.

you and Keith have yet to state what purpose this data actually can be used for by the general public if not as diagnostics of future disease. Especially when you are stating that they are actually contextualising this information into medical diagnoses. Using your logic I could sell any drug I liked as long as it isn’t actually ingested. what other use do these tests have if not medical diagnosis of disease probability.

These companies use testing labs subject to the same quality control measures as those used by clinical labs. (23andMe, for instance, uses LabCorp, who I’m guessing also processes a fair proportion of the tests you order as a clinician.)

QC does not guarantee anything other than the process has been carried out correctly. If that process is flawed and does not give significant results (I mean in terms of diagnosis- I’m sure that their very successful at accurately identifying the genes) then no amount of QC can improve that.

Should we ban the sale of electronic blood pressure monitors and home pregnancy tests as well, in case the poor confused consumer misinterprets them? After all, right now a pregnancy test is more likely to have a life-changing impact than a genome scan.

Pregnancy tests are regulated by the same rules as these arrays. When these arrays can tell you with the same degree of certainty that you are going to get a disease then I will be happy to see them marketed to the General public.

If these are not diagnostic tests then what purpose do they serve to the public?

Keith #24

you have already demolished your own argument- if the majority of doctors are not able to properly interpret these tests how are the rest of the public supposed to!

I look forwards to seeing your replies tomorrow- here in the UK it is knocking off time and I want my tea.
#28 Keith Grimaldi
August 4, 2010

@symball “you and Keith have yet to state what purpose this data actually can be used for by the general public if not as diagnostics of future disease”

Just like any other information that could have a bearing on my long term health, like going out in the sun, drinking, smoking and eating. Like most of what I can read on the health related web sites. I don’t get this clinical diagnosis stuff though, it just is not diagnosis, please stop repeating that it is. Also read this: bit.ly/4Chszs and posts by others who have written about their tests. Utility goes beyond just clinical utility, this is often forgotten, or never even considered in the 1st place.

“Keith #24 you have already demolished your own argument- if the majority of doctors are not able to properly interpret these tests how are the rest of the public supposed to!”

Well, they have excellent tools to do so provided by 23andEtc, as evidenced by the majority, if not ALL, who have written about their experiences (look up Mark Henderson in the Times for example). I have been able to compare genetic test interpretations directed at the general public with those aimed at physicians – the former are much much better.
#29 Jonathan
August 4, 2010

@Keith, you need to recognize that each of those organisations plays a different role and has a different mandate.

FDA is a regulatory agency. They have enforcement powers and are required to oversee things like medical devices, drugs, etc. It’s part of the Executive branch of government, and reports to the President (well, it reports to SecHHS but that’s where it all ends up). The US has a full history of shady snakeoil salesmen and that’s why, over 100 years ago, FDA was created.

GAO is an investigative agency that reports to Congress, not the President. It isn’t a regulatory agency and AFAIK has no oversight or enforcement powers. GAO won’t be responsible for regulating the industry, their role in this story is basically over. Think of them like the UK tv programme Watchdog, or a less toothy Esther Rantzen.

The Human Genetics Commission is a QUANGO that advises HMG, in much the same way that SACGHS advises the Secretary of HHS here in the US. Again, it’s not a regulatory agency, and I don’t see that they a regulatory role over the industry or any powers of enforcement should their principles not be followed.
#30 Keith Grimaldi
August 4, 2010

@Jonathan I do understand and I’ve actually had direct experience of all of them, both in US and UK. The role of the FDA is well understood, the complaint here, and I’m trying to stick to the topic of the post, is the methods. First they are talking to the companies, have been since 2007, then they send out the letters, then they hold their hearing (talking about the FDA here, not congress & GAO). The hearing was very good – open, live webcast and all opinions were invited and heard. So far so good, seems like an excellent collaborative information gathering process, similar to the UK. But then, on the day of the hearing they send out a load more letters followed by this apparent attempt to interfere with a critical supplier. Just because they are regulators and need to regulate does not mean that these methods are acceptable – on the contrary, exactly because they are regulators they should be very open and transparent about the process and really collaborate with the companies they are trying to regulate.

Yes there are scammers out their selling stuff that they shouldn’t be, some of the pseudo-genetic companies have been around for many years. If it is true and the FDA suceed in blocking supplies from Illumina then they will shut out the reputable companies leaving the others, who don’t use Illumia, to carry on as before. That doesn’t seem to be a good outcome, not really protecting the public. Have a look at the Genewize website, they declare that they have $millions in sales, many more customers than 23andEtc but won’t be touched by blocking Illumina (although perhaps the FTC should be dealing with them).

The GAO/Congress “trial for the prosecution”, with no real opportunity for defence, was very conveniently timed and will have a huge influence on FDA actions, even if it has no powers itself. The GAO have done their job and their conclusions are being uncritically repeated 1st, 2nd and 3rd hand by the mainstream press, they have got what they wanted, just as they did in 2006.

The HGC is commissioned by UK govt to provide information & recommendations and hopefully will have an equally huge influence on the UK regulators – who, as you say, do not have to accept their advice.

To repeat, again, the issue here is not whether or not there needs to be regulation or oversight, most companies and commentators agree on that. The issue of Daniel’s post is the methods and the impression that all the recent events give that they have decided at all costs to block DTC irrespective of opinions and debate. In the end the companies want at least to have clarity, which is why 23andMe have been speaking to the FDA since 2007, before they launched.
#31 Keith Grimaldi
August 4, 2010

OK, just seen this on twitter form Daniel: @dgmacarthur: Update on the FDA/23andMe/ILMN story: 23andMe rep just told me that the rumours aren’t true. More details to follow.

Good! At least it has stimulated a revealing discussion!
#32 Daniel MacArthur
August 4, 2010

Here is the written statement from 23andMe’s PR firm:

We will continue to work with Illumina as our discussion with the FDA continues.

I’ve requested further clarification.
#33 Keith Grimaldi
August 5, 2010

Hmm…interesting thread in the Mary Carmichael / FDA interview, full transcript here: bit.ly/aogZuY

MC: I saw this morning indicating that the FDA has “requested” that Illumina stop selling chips to DTC genetic test providers. Is that true in whole or in part?
EM: We have asked Illumina to work with us to bring their devices into compliance with our regulations. We have not specifically requested that they stop selling them to anybody.

So they can continue supplying 23andMe and deCODEme with chips right now?
EM: We’re working with them on that. …

[The chips are labeled “research use only.”] The DTC providers, when they sell these tests, aren’t [necessarily] doing research. Doesn’t that imply that Illumina does have to stop selling the chips to them?
EM: If they continue to label them that way, yes, that would be something that we would most likely take a little further action on. At the moment they’re working with us and we will see what they can bring to us in a reasonable amount of time without completely blowing up their business or the market or anything. …

Until that happens, though, is it correct that the Illumina chips will remain on the market for direct-to-consumer companies to buy?
AG: [pause] I think Illumina needs to figure out how they’re going to move forward.
#34 Dogface
August 10, 2010

This violates the 1st Amendment. We have a right to read our own genetic code. “Big Brother” intervention is censorship. No one has provided strong evidence that provider to consumer genotyping is dangerous.
#35 Andras Pellionisz
August 11, 2010

Maybe Juan Enriquez is right with his book (The UnTIed States of America…)

However, the USA may shoot herself in the foot in Postmodern Genomics, some lowly bureaucrats already having their finger on the trigger – but the rest of the World will march on. For instance, beyond DeCodeMe in Iceland there is already GenePlanet in Ireland and Slovenia – and a Seoul-based Korean genome testing facility is determined to capture the market of also China and India; backed by SAMSUNG.

As for miroarrays, Affymetrix already has a production plant in Singapore. Should Big Brother act up, they could easily figure out how to maintain supply for the transitional period till Full Human DNA Sequences replace the present “micro-interrogation”, way beyond SNP-s to even fractal defects mined by HolGenTech.Inc. The largest full DNA sequencing capacity presently in China – and HolGenTech’ founder (yours truly) is a dual USA-EU citizen. Can set up shop anywhere from Slovenia, Estonia, Lithuania, etc (EU since 2004; yes, in Hungary as well, but no one is prophet in his country – the 2006 PostGenetics initiative in Budapest was butchered by some communists who ruled there till this Spring…). Lithuania, however, has just outstripped even China in HoloGenomics’ downloads!

Let’s face it – the USA is no longer indispensable in our Global Brave New World.