Did Washington Post's Rob Stein exaggerate negative stories about personal genomics?

i-30121f3d1da342a03e46b8de14a42e15-50s_horror.jpg

I mentioned last week on Genetic Future that a reporter with the Washington Post, Rob Stein, had emailed the National Society of Genetic Counsellors to search specifically for people with negative experiences of personal genomics for an upcoming article on the industry. At the time I called for personal genomics customers with positive experiences to email Stein, to ensure that he had a balanced view of the industry's impact.
Stein's article came out a couple of days ago, presumably timed for maximum impact in advance of the FDA's meeting on lab-developed tests over the last two days. Overall it's less unbalanced than I'd feared - the piece concludes with a fair discussion of the pros and cons of increased regulation, including quotes from me and Dan Vorhaus (fellow Genomes Unzipped blogger).
However, Stein did manage to find some negative stories from personal genomics customers, all of which center around the mistake made by LabCorp back in June that resulted in incorrect data being received by 87 23andMe customers - and unfortunately those stories are what lead off his piece:
One woman panicked when the genetic test she had ordered over the Internet concluded that her son was carrying a life-threatening disorder and, even more disturbing, that he was not -- genetically -- her son. Another, who always thought she was white, was flabbergasted to find her genes were mostly of African origin. A third woman's result was still more stunning: She was a man, it said.

"I thought, 'Oh my God. Am I really a man?' " said Denise Weinrich, 48, of St. Peters, Mo. "I thought, 'What's the matter with me? I'm not who I thought I was. How am I going to tell my children?' DNA doesn't lie."

Firstly, it's remarkable that despite two separate emails to the NSGC, Stein only managed to uncover three negative stories from personal genomics customers, all of which refer to a single (rapidly corrected) laboratory error. That's testament to the fact that - despite the fear-mongering of critics - the majority of personal genomics customers are not tormented by the experience.
Secondly, even these anecdotes appear to have been very selectively presented by Stein. Cece Moore, who writes the blog Your Genetic Genealogist, has personal insight into two of the stories relayed by Stein, and suggests in a post that both of these are exaggerated. Firstly:

"flabbergasted" - This is a misstatement. I know this because the person to whom Stein is referring is my niece. He never spoke to my niece directly and I never said, nor implied, that my niece was "flabbergasted." Actually, I expressed the exact opposite to Stein. First of all, I told him that when she initially called me about her test, she very matter-of-factly stated, "My genes are most similar to a Nigerian" with, surprisingly, little reaction. 

Secondly, I told Stein that I immediately informed my niece that it was a mistake, so there was never any real question of whether she was actually African genetically. How he got "flabbergasted" from that, I do not know.

Secondly, Moore quotes an email from the third person mentioned in Stein's piece, Denise Weinrich:

"This is so sad....95% of what I said was how thankful I am for the testing...Told the story of my adopted children and the 30+ Guat(emalan) kids who also tested with 23andMe. How I would never want DNA testing regulated because it would be hard to do for cash strapped adoptive parents, etc. 

No wonder people complain about the interviews they give to reporters." [emphasis added by Moore]

Moore also notes that the wording of the first experience is also shaped for dramatic effect; the woman in question was informed that her son was a carrier of a single haemochromatosis variant, which doesn't actually mean that he would suffer from the disease.

 
As I wrote back in June, the mistake made by LabCorp was deeply unfortunate, but the negative impacts of it on customers were swiftly minimised by 23andMe, and as far as I can tell there's no evidence of any long-term harm occurring as a result of the incident. It's extremely telling that this is the best example of negative experiences that a Washington Post reporter can uncover, and deeply unfortunate that these incidents will be the only part of the article that most readers walk away with.
Speaking of fear-mongering, there's another quote in Stein's piece that warrants attention:

"It's come to the point where really there's a need for some oversight," said Alberto Gutierrez, who heads the FDA's Office of In Vitro Diagnostics. Results from questionable tests can be unnecessarily alarming, Gutierrez said, adding that some women have undergone surgery, for example, based on tests that purport to gauge the risk for ovarian cancer. 

We know of reports of people who have found a test, found a doctor that is willing to order the test since they are so afraid of the disease, and even removed ovaries based on questionable results," Gutierrez said.

It's a serious, disturbing claim - and one that requires evidence, which is not provided. If Gutierrez intends to keep using this anecdote to justify increased regulation of the genetic testing industry, he needs more than just a vague "we know of reports". Gutierrez should provide details of the testing company involved, and explain exactly why he regards the results of the test as "questionable". If indeed someone has had ovaries removed due to spurious genetic testing results the surgeon should be immediately investigated for malpractice (if that hasn't happened already). If Gutierrez can't provide these details he should quit courting media attention with an unverifiable anecdote.
Mind you, Alberto Gutierrez is a man who also decided to announce that the FDA regarded DTC genetic tests as illegal via a quote in the Washington Post (also via Rob Stein) - so I suppose we should not be entirely surprised at such flippant use of media scare-mongering. It's hard to create fear when you can't actually quote convincing empirical evidence that information from genetic testing is harmful - and you can't quote such empirical evidence because it doesn't actually exist. As such, the FDA is forced to resort to "we know of reports" to muster the requisite sense of horror.
As I mentioned above, Stein quoted me briefly towards the end of his piece, noting that increased regulation is not cost-free. For context, here's the rest of the email that I sent to him two weeks ago:

While regulations are sometimes necessary to protect consumers, people also need to remember that they come at a cost: increased prices, decreased competition and reduced innovation. That means we need to weigh up the potential benefits of any proposed regulation against the stifling effect it will have on progress in the field, especially for a field as crucial to the future of medicine as the genetic testing industry. 

Right now the direct-to-consumer genetic testing industry is an engine of innovation in a range of key areas - for instance, companies like 23andMe have developed ways of presenting complex genetic risk data to consumers that are far better than anything you'll get from your doctor. While there's a place for very careful regulation that punishes companies who make false claims about genetic testing, excessively heavy-handed regulation from the FDA will end up reducing the number of new companies entering the field, thus inhibiting competition, as well as increasing the cost of creating new products. Suppressing innovation in this way could easily have seriously negative effects on the long-term development of technologies for personalised medicine. 

 There are alternatives to stringent FDA regulation - for instance, forcing genetic testing companies to disclose the markers they test and the scientific evidence backing up their claims would go a long way towards cleaning up the field. Incorporating this kind of information into the NIH's proposed genetic test registry, and making registration mandatory for testing companies, would be one way of doing this. That way consumers get access to the information they need to avoid the dodgy operators, but the overall cost to the industry is much lower.

We all agree that there is room for better regulation of the direct-to-consumer genetic testing industry - something that makes it hard for outright scammers to operate, or for companies to make false claims or use unreliable tests. However, everyone should be aware that if the FDA moves beyond its remit here, and imposes unnecessarily stringent regulatory hurdles on genetic testing companies, it won't just be the companies that suffer: excessive regulation will cost consumers, and it will directly inhibit innovation in the crucial and emerging field of genomic medicine. It's not just 23andMe's fate that's at stake here, but all of the start-up companies - some of which haven't even been formed yet - that will create the products and devices that bring medicine into the genomic era. 

This needs to be done right. If we allow bureaucrats from the FDA to bury the genetic testing industry beneath piles of arbitrary demands (sorry, "enforcement discretion"), this will hurt all of us. I'll have a post up soon on Genomes Unzipped expanding on this topic.
For now, we all need to sit back and absorb the key lessons from the last two days of presentations at the FDA's meeting. A good place to start is Dan Vorhaus' summary of day one, and the excellent presentations made by Anne Wojcicki and Ashley Gould from 23andMe and Katherine Borges from the International Society of Genetic Genealogists.

More like this

In my previous post I quoted from Mary Carmichael's excellent interview with two officials from the FDA, Alberto Gutierrez and Elizabeth Mansfield (part of her fantastic week-long DNA dilemma series, which you should read in full if you haven't already). There are numerous nuggets of gold tucked…
It looks as though the FDA is swooping down on the direct-to-consumer genetic testing industry in a serious way, sending formal letters to five companies informing them that their tests will be regulated as medical devices: WASHINGTON -- The Food and Drug Administration is issuing regulatory…
The brief Golden Age of direct-to-consumer genetic testing - in which people could freely gain access to their own genetic information without a doctor's permission - may be about to draw to a close. In a dramatic week, announcements of investigations into direct-to-consumer genetic testing…
(This is an edited excerpt from an op-ed piece I just wrote for Xconomy, posted here as I think it provides some nuance on my views on regulation of genetic testing that was lacking from my post last week. Some context for new readers: a Congressional investigation into the direct-to-consumer (DTC…

Well said.

Gutierrez keeps repeating that ovaries claim, as in NYT last month - âIt is not unknown for women to take out their ovaries if they are at high risk of ovarian cancer,â Dr. Gutierrez said (makes it sound like they do it on their own, no doc! glad he cleared that up in the Stein bit) www.nytimes.com/2010/06/12/health/12genome.html

As you say, give us the evidence, until then it IS unknown. What test? How high is the risk? (and if it is high is surgery so unreasonable???)

I don't know what surveys say but following blogs and especially Twitter over the last couple of days the majority don't want heavy regulation. Applause at the FDA was generally louder for the Katherine Borges point of view compared to some of mouth foaming doom spreaders who spoke). The most popular is the transparency call, rather than regulation itself (apart from Dan Vorhaus this is also endorsed by HGC and PHG in the UK who are by no means libertarians)

Agree also with the innovation in communication, based on other NIH sites I sincerely hope that they will get help from 23andme et al on how to make the GTR consumer friendly, it needs to be accessible to everyone I think.

Prophylactic removal of breasts & ovaries ,a href="http://www.ncbi.nlm.nih.gov/pubmed/20644999">has been shown to be effective (though clearly drastic) way to prevent cancer in BRCA1/2 carriers.

Obviously, any doctor who did this based on a DTC test without further analysis is skating on very thin ice. How many BRCA mutations do any of the DTC kits test for? They must be risking a lawsuit from Myriad to test for any of them

I expect he is referring to (perhaps via some chinese wispers) the case of Teri Smieja. She had the 185delAG mutation in BRCA1, which gives a 40-60% lifetime risk of developing breast cancer, with reduced onset age and survival rates for those who do develop it, as well as a 40% chance of developing ovarian cancer.

After talking with a genetic councilor, Teri Smieja decided, based on this data, to have her womb, ovaries and breasts removed (after her youngest child was weaned), reducing her chance of developing cancer to slightly below the population average. Her reasoning was that she did not intend to have any more children, and that the risks and unpleasentness of surgery outweighed the risks of developing cancer.

She also wrote about this decision pretty extensively:

http://myblip.wordpress.com/
http://www.wired.com/magazine/2010/01/ff_decisiontree/2/

Whether you agree or disagree with her decision, characterising it as "stupid woman cuts off breasts because she doesn't understand genetics!" is a horribly misguided way of looking at the case.

Keith,

23andMe tests for 3 BRCA variants: 185delAG and 5382insC in BRCA1, and 6174delT in BRCA2. No patent infringement case from Myriad yet.

Any doctor who performed surgery based solely on 23andMe test results, without performing a clinically validated follow-up test, is not just skating on thin ice - that's gross medical malpractice.

Luke,

Thanks. That could indeed be the case Gutierrez is discussing, in which case (1) as you say, it's a terrible example, since Smieja's tests weren't "questionable", and (2) it would have been trivial to provide some actual detail beyond "we know of reports".

This is the head of a major branch of the FDA - scare-mongering based on flimsy anecdotes simply to justify a regulatory power-grab is disgraceful.

Interesting - if that is indeed the case he is referring to then it is absolutely disgraceful. According to her blog the test was Myriad and her lifetime risk was 87%...

"...found a doctor that is willing to order the test since they are so afraid of the disease, and even removed ovaries based on questionable results"

He could be on thin ice himself (Gutierrez) - whether or not it's based on the case Luke cites, it's actually pretty nasty to trivialise something that many women have had to go through, he should be apologising and watching out he doesn't get sued

Also, the other speaker who brought up the testing/oophorectomy example at the FDA LDT meeting was Sue Friedman of Facing Our Risk of Cancer Empowered. She spoke of a woman who decided to have her ovaries removed based on a "mutation" that eventually turned out only to be a "variant". (In Myriad-speak, a "mutation" is a variant with known deleterious effect, while a "variant" is a variant of unknown significance.)

That story doesn't sound like Teri Smieja's story. Might be a good idea to contact her.

By Andro Hsu (not verified) on 21 Jul 2010 #permalink

It's a pity that some of these contentious interviews were not audio-recorded so that they could be posted on YouTube to shame Stein.

Reviewing the data anecdotes:

1. There's little that FDA regulation would have done to prevent the 23andMe sample swap problem. It was a laboratory quality problem, not a problem with the analytical validity, clinical validity, or clinical utility of the 23andMe test. Rather, it's a CLIA/CAP issue. Moreover, sample swaps probably happen every day in clinical labs around the US, but only with information rich tests like those done by 23andMe can you detect them so readily. There's nothing about direct patient access to testing that changes this situation. The fact that it was a genetic test only made it feasible to detect the swap -- it didn't cause the problem.

2. Myriad's testing is done based on a mix of genotyping and sequencing and is usually ordered for patients with a family history that increases the prior probability of carrying one or more deleterious mutations. It's probably safe to guess that Myriad's analytical validity is effectively 100%. Their clinical validity is a function of their ability to predict the pathogenicity of detected variants. This process involves trade-offs and can never be perfect. If FDA were the lock Myriad's reporting to an approved list of variants, the variant reclassification issue could be eliminated, but only to ill effect. Myriad's board-certified laboratory directors are in the best position to make judgments about the interpretation of their genotyping/sequencing results based on the best available evidence at the time. All FDA could do is lock down that list and keep it from changing without a new cycle of external approval. The kind of judgments that Myriad has to make about pathogenicity of a variant happens every day for a wide variety of complex laboratory tests, and it's the reason we require clinical laboratory directors to be board certified -- they are either MDs or PhDs. Like any medical practitioner they will make "mistakes", even when that mistake is to make a false report that was otherwise based on the best available evidence at the time. Controlling Myriad's in that fashion would be a grab by FDA to further control the practice of medicine, which is beyond the spirit if not the letter of their Congressional mandate.

I'm curious if anyone can offer a clear-cut explanation why personal genomics tests that predict disease risk don't need FDA regulation (e.g. 23andME), while 'conventional' diagnostic tests that predict disease risk do (e.g. Oncotype DX, BRCAnalysis). In other words, what (if anything) distinguishes tests that should have to prove clinical validity to FDA's satisfaction before marketing, vs. tests that shouldn't have to?

It was suggested on the ISOGG email list that I might like to post here the email I wrote to Rob Stein, to which I received no answer, most likely because I had no complaints about my testing experiences. The text follows:

-----------------------------------------

Mr. Stein,

I was recently alerted to your efforts to gather negative experiences of direct-to-consumer genetic testing from genetic counselors for your planned article, as we see here:

> Dear NSGC Members,
>
> Washington Post reporter Rob Stein has interviewed NSGC President Liz Kearney on the subject of regulation of genetic testing. Rob would also like to interview patients who have experienced DTC testing. Rob is particularly interested in speaking with patients in the D.C. area who have had a bad experience with DTC testing or who sought the help of a genetic counselor to order the correct test and/or interpret the results. However, he is interested in speaking with any patient who has undergone DTC testing.
>
> If you have patients that would be willing to talk with Rob, please have them contact him directly at the phone number or email below. Rob needs to complete his interviews by Tuesday, July 13.
>
> Rob Stein
> The Washington Post
> 202-334-7338 phone
> steinr@washpost.com

This has greatly concerned us as members of the genetic genealogy community. My experience with genetic testing, and that of my fellow members of the International Society of Genetic Genealogy, has been highly positive and life-enhancing. I have purchased genetic tests of many different kinds, over seven years, from companies ranging from Family Tree DNA, to 23andMe, and DNA Direct. I have never felt exploited, deceived, confused, distraught, or traumatized, or whatever the negative experiences are that you seek to present.

What troubles me is your cooperation in the major campaign being conducted by the AMA, the NSGC, etc., against direct-to-consumer testing. These organizations are seeking to protect the self-interest of their members, who are directly threatened by the possibility that non-professionals will take an active and direct role in informing themselves about their health and genetic information, without the intervention of a paid professional.

As a life-long advocate of civil liberties, a member of the ACLU, etc., I see these efforts to restrict people's access to information about their own DNA as unacceptable, aggressive attempts to rob me of the freedom to know myself and learn all that I can about my own body and heritage. No professional organization has the right to proclaim themselves the only authorized gatekeepers of such personal knowledge, seeking to deny free access [to it] to the persons themselves, who are constructed on the basis of this DNA, the gift of their ancestors.

I hope you will re-think any concept of such a biased article seeking to present direct-to-consumer genetic testing in a negative light. Are you aware that the Genetic Alliance, the umbrella group representing the patients and families of those afflicted by genetic medical conditions, supports the right of consumers to access these tests? And then there are the thousands of us who find delight in pursuing our research to catch a glimpse of our deep ancestry, not only gaining new insight into our own families, but seeing how they are connected to and rooted in the family tree of all humanity.

I will be happy to speak with you if we can find some time that fits into both of our schedules. . .

By Bonnie Schrack (not verified) on 21 Jul 2010 #permalink

@qetzal: none of the tests you mention currently have FDA approval. They are all laboratory developed tests (LDTs). That's true of every genetic test except for a handful.

qetzal, both BRCAnalysis and Oncotype DX are CLIA LDTs, not FDA approved tests!

That's the whole point: the last decade of innovation in molecular diagnostics has taken place because labs had regulatory choice: go with nimble CLIA or burdensome and slow FDA.

Not to be mean, but your comments here and on other threads illustrate that you really don't know what you're talking about. You should seriously consider asking someone who actually operates a clinical lab to give you an education.

To give you some idea of what the FDA involves...consider XDX. Their approval notice only cleared three specific Biomek FXp devices.

http://www.accessdata.fda.gov/cdrh_docs/reviews/K073482.pdf

Yes, you see that right: the actual serial numbers of the instruments had to be noted in the filing.

Why? Because the FDA won't clear the instrument. Only these three instances of the instrument.

What does this mean? Those instruments are like precious gold to XDX. Despite the fact that these are mass produced devices, they cannot replace their older machines with newer ones of the same make. God help them if they need to add capacity. And if they break for any reason, XDX is just out of luck and has to cease operations to file another 510(k)!

This is of course insane and just a small piece of the bureaucratic hell that is the FDA. And if you question the policies, you have guard dogs and guns pointed at you.

It's just like the TSA: a set of regulations invented by politicians, full of sound and fury, accomplishing nothing.

@Debbie

That case doesn't really seem very relevant to DTC testing. No genetic test was performed, or offered, because Medicaid didn't cover it. Instead, the doctor recommended a double masectomy based on a (mistaken) belief that the plaintif had a family history of breast cancer.

This isn't a case of someone "finding a doctor to order the tests", and doesn't involve ovaries being removed, so I'm not sure whether it'll be the case Alberto Gutierrez is referring to.

Has anyone contacted him directory and ask which cases he's referring to?

asdf,

Thanks for the response. For context, I do have experience working under FDA regulation, but it's mostly on the drug/biologic side, with a bit of conventional medical device exposure. I don't really have any direct experience with diagnostic tests.

So anyway, I thought tests like Oncotype DX and BRCAnalysis were FDA reviewed/approved. Thanks for correcting my error. However, a bit of searching shows that similar tests have been reviewed and approved by FDA, including a Nanosphere test for warfarin sensitivity (which is one of the issues raised about 23andMe) and the MammaPrint test for breast cancer prognosis (comparable to Oncotype). See the third and last items here. I also see that FDA asserts that it does have authority over LDTs, it just hasn't usually exercised it. (See Background here.)

So I'm still curious. What distinguishes which tests of this type get FDA approval, and which do not? More importantly, what should drive that distinction?

I'm not necessarily against the idea that personal genomics companies should be allowed to perform direct consumer testing for warfarin sensitivity or breast cancer risk. If it seemed otherwise from my comments, I apologize. However, I do think that we need consistent standards.

By analogy to pharmaceuticals (where my knowledge is a bit better), one could argue that medical diagnostic tests should be validated to prove that they meet appropriate standards for accuracy, robustness, sensitivity, specificity, etc. After all, a bad Dx test could have consequences every bit as serious as a bad drug. One could further argue that such validation should be reviewed by an outside agency, for all the same reasons that validation of drug safety and efficacy are reviewed by FDA.

So my basic question remains. On what rational basis should we judge whether a Dx test should be subject to FDA (or other outside) review? If you have links to any proposed frameworks that you think are good, I'd appreciate them.

Hey qetzal,

Thanks for engaging in a reasonable discussion. Regarding FDA and diagnostics, I'm sorry to say but many people involved in molecular diagnostics think the FDA is just not competent. To give you some idea, I'd point you to the statement of the Association for Molecular Pathology on LDT regulation:

http://www.amp.org/Position%20Statements/AMPComments_FDAMedicalDeviceWo…

1. Requirements for tests for the same analyte have increased precipitously for subsequent submissions. As an example, over the last six years the FDA has reviewed and cleared several tests for the detection of mutations associated with cystic fibrosis and blood coagulation disorders. In the review of the first molecular genetic tests (e.g. for factor V Leiden, prothrombin, and cystic fibrosis), the FDA allowed manufacturers to utilize DNA constructs and cell line DNA tested in replicates to establish the testsâ performance of rare mutations for which clinical samples were difficult to acquire. However, manufacturers of subsequent tests are required to perform testing on a greater number of clinical samples for each mutation. Instead of being permitted to utilize DNA constructs and cell lines, greater numbers of clinical whole blood samples are required, even for exceedingly rare mutations. In addition, FDA added a requirement for 3 peer-reviewed papers for each new mutation for which clearance is sought, describing 3 unrelated patients' genotype and CF phenotype. The requirement for 3 papers and 5 samples has made it very difficult to validate rare mutations and is inconsistent with the original clearance criteria. Most importantly, the scientific evidence supporting these escalating requirements has not been communicated. AMP is concerned that such inconsistencies greatly deter submissions of tests for review. We recommend that requirements remain consistent and, should these requirements need to be altered, the basis for any such changes be clearly communicated.

You can easily work out the math on that. There are 3 billion base pairs in the genome and 3 million SNPs alone. Requiring 3 publications and 5 clinical whole blood samples per mutation tested is a requirement that is completely disconnected from clinical and R&D reality. It is guaranteed to keep genetics in the 1950s mindset of one SNP, one disease when we are learning it is about multivariate predictors.

Moreover it is a completely arbitrary and poorly thought out requirement. Consider the "3 publications" requirement. What if a publication lists a table with 10 variants in supp info? What about 100 in the same pathway? What about 1000 or 10000 as a literature review? And what if a close-to-infinite number of different deletions of the same region are all shown to have similar effects?

Try talking to a federal government bureaucrat about this, though, and you'll get pepper sprayed for your trouble.

Think again about the TSA. They mindlessly follow rules put in place because of bad press in the past. We are still taking our shoes off years later because of Richard Reid. We are still having hundreds of millions of dollars of toiletries seized on an annual basis because some guy tried to put liquids on a plane. Yet any reasonable person like Bruce Schneier can tell you that these methods do nothing to deter terrorists and everything to harass and delay hundreds of millions of innocent travellers.

And that's the TSA, which is public and seen/scorned by millions, and which only has power over you during that brief period when you're in the airport.

The FDA is not only far less public, it holds authority over regulated companies for all of time. It is now expanding that reach to hospitals and academic labs, formerly bastions of free speech. Public criticism of the agency is met with brutal retaliation. As a rare public example, the federal court case that Utah Medical won against the FDA shows how they deal with any defiance:

The 2003 FDA-483 was a sham, comprised of 14 pages with 19 major observation item numbers with 53 total discrete sub-items. The inspectors did not report what they saw. They reported what they were told to see, or what Barrell/ Trautman worded for them to be included in the FDA-483. The following Trautman e-mail to Weixel, Latish, Tilton, with copy to Barrell, on Friday March 7, 2003 illustrates the point. (Chase/Coleman had told the Company on March 6 that the FDA-483 would be issued on March 7.)

âI thought we had already arranged to review it (the FDA-483) before it was issued! Why canât Ricki (Chase) issue it Monday? If it is issued today without my review, I do not want any complaints if I do not support something because of the way it was written. This 483 is going to have to be dead on, for me to support an observation with all of the issues surrounding the inspections. I will not be able to massage it for the complaint like some cases because [lawyer Larry] Pilot will kill us in court.

In other words, FDA officials "massaged" an FDA-483 (citation of deficiencies) in order to "get" a small medical device manufacturer which had dared to criticize it. This was a clearcut example of a multibillion dollar agency vindictively trying to crush a small business. And the AMP and XDX pieces should give you a flavor for just how irrational the FDA is.

So when you say this:

> More importantly, what should drive that distinction?

The solution is very simple and has been proposed by many economists. Rather than having a single monopoly provider of certification, the FDA should be opt-in for all drugs and devices. The government can grant licenses to other certification agencies such that the FDA seal of approval is just one kind of seal.

For example, they could also issue certification licenses to the Mayo Clinic or Harvard Medical School, which are (after all) the places which write the papers that FDA consults and make the discoveries in the first place.

Think about it: this is already done in medicine (you can get your MD from a bunch of different places of different quality), law (ditto), and plenty of other areas. The variation in MD certification is particularly relevant as someone is taking a calculated risk with their health when they go with a Dr. from XYZ college rather than Harvard University.

Initially, the FDA would start in the public's mind as the "gold standard"...but once companies could speak about FDA malfeasance without threat of bankruptcy, a lot more stories like the XDX and AMP sagas above would come to light.

Moreover, an "opt-in" path for the FDA has already been the reality in molecular diagnostics for several decades, in that labs could choose FDA or CLIA. Everyone who can choose CLIA does choose CLIA...and the FDA gets less user fees and less power.

In short, the FDA does not offer a high quality certification service. Its customers, the labs, are leaving for another provider (CLIA) when given a choice. Thus it does what all former incumbents do when faced with a threat: it launches a legal attack rather than asking itself "why have we failed in making the FDA diagnostic path an attractive one".

It's all about preserving influence and growing their budget. That's why the FDA has launched this attack. Look at Jeffrey Shuren of CDRH being offered more money in today's hearing. What, you think he's going to turn it down?

Ultimately if regulatory choice is not offered within the United States, it will be offered outside it. Many entrepreneurs are going to Europe (for the CE Mark, which is less onerous than the FDA) or even better to East Asia.

asdf,

Thanks for the detailed reply.

The solution is very simple and has been proposed by many economists. Rather than having a single monopoly provider of certification, the FDA should be opt-in for all drugs and devices.

I agree this would be a consistent framework. However, I think it's nearly impossible for this to happen, especially for drugs. As I'm sure you know, FDA's creation and the continuous expansion of its powers over the decades has been driven by congressional legislation, frequently spurred by gross failures by the pharma industry that led to people dying. (See here for a summary.) I seriously doubt that congress would ever backtrack substantially on FDA's powers, and I seriously doubt the average US citizen would be in favor of that. It's also worth noting that as far as I know, all developed nations have some kind of FDA equivalent. I'm not aware of any that allows the sale of drugs without government approval. (Please correct me if you know otherwise.)

That's not meant to argue that mandatory gov't regulation of drugs is the right way to do things. It's only to suggest that any other way is going to be very hard to institute.

As far as I know, there is no non-governmental equivalent to FDA for diagnostics at the moment. At least not in the US. My understanding is that CLIA regulates clinical labs, but doesn't review or certify specific tests. Is that right? Also, do you know of any other developed country that allows non-governmental groups to 'approve' Dx?

If we accept, for the sake of argument, that a fully opt-in FDA isn't an option, what's a reasonable alternative? Is there an acceptable way to classify tests such that only certain ones need FDA approval? (Obviously, FDA already has such classifications, but I won't assume that you or others here consider them acceptable.)

I'd like to interject my thoughts here too for a moment. The reason I asked my doctor to have the BRCA Analysis test done on me was because of the high occurrence of ovarian cancer in my family, coupled with the fact that I am from Ashkenazi Jewish descent, from my mother's side. The test was my request, was done through Myriad Labs (this was before the ACLU won their lawsuit against Myriad for laying claim to our genes themselves) who have done hundreds upon hundreds of BRCA testing. When my results came back that I was positive for the BRCA1 mutation, I scheduled an appointment with genetic experts at Cedars Sinai in Los Angeles. The decision I made to have my ovaries, and later my breasts, removed was not made in haste. Quite the contrary, a lot of thought and research went into my decision to go through with these preventative surgeries. I'm far from the only person with the BRCA mutation who has chosen this route, and unfortunately, though medical science has come a long way, the best way for someone like me to beat breast and ovarian cancer is to simply not get it at all. Eating vast amounts of broccoli or living a healthy lifestyle will only go so far with those who have the BRCA mutation. My code is 187delAG. The 'del' stands for 'deleted' - Everyone has the BRCA1 and BRCA2 genes, men, women, you, everyone. In the case of people like me, the 'del' in my DNA refers to the deletion of tumor suppressors that a body needs to fight breast and ovarian cancer (other cancers too, but those are the highest risk). My lifetime risk of breast cancer was 87%, and my lifetime risk of ovarian cancer was 44%. As if that weren't bad enough, with no tumor suppressors, that makes the recurrence rate up to 60%. So I could wait for the cancer to come (which I felt inevitably it would), battle and maybe even beat it the first time, only to have it come back again, and again.

So again, I'd just like to clear this up for the person who wrote this article. These decisions are not made lightly by any means. Some of us anguish for years, while watching family members around us die from breast or ovarian cancer. Our doctors are experts in their fields, well versed in the study of genetics and hereditary cancers. None of us WANT to have our ovaries removed or our breasts cut off, but we want to die from ovarian or breast cancer even less. We choose to take charge of our lives, we make those terrifying decisions because quite frankly, we don't have a whole lot of choice. Anyone is free to contact me if you have any questions at all. If I don't know the answer, I can easily put you in touch with someone that does.

Thanks for your time.

Teri Smieja