[ExI] FDA Tells Google-Backed 23andMe to Halt DNA Test Service

Adrian Tymes atymes at gmail.com
Wed Nov 27 19:57:35 UTC 2013


On Wed, Nov 27, 2013 at 11:14 AM, Dave Sill <sparge at gmail.com> wrote:

> On Wed, Nov 27, 2013 at 1:59 PM, Adrian Tymes <atymes at gmail.com> wrote:
>
>> On Wed, Nov 27, 2013 at 7:14 AM, Mike Dougherty <msd001 at gmail.com> wrote:
>>
>>> So you can "ask your doctor" about drugs for your scaly, red, itching
>>> skin that has a side effect of "death" or some heartburn meds that give you
>>> low magnesium levels and risk of bone fractures - but you can't swab some
>>> spit into a cup because the resulting test(s) has the effect of education
>>> about information you might not be prepared to learn?
>>>
>>
>> No.  Because the information stands an unacceptable (to the FDA) risk of
>> being flat-out wrong.
>>
>
> How much risk of wrong genetic test results is acceptable to the FDA? How
> much does 23andMe have? Doesn't the use of a CLIA-certified lab address the
> quality of the results?
>

Worthy questions; these items are likely in the details of the FDA's
communications.  Although the last one can be answered "no": it's not "do
you have this marker" that's the problem, but the information they attach
to it.


> It's one thing if it advises someone of a condition they didn't know they
>> had, for which there is a risky treatment option but less risky than doing
>> nothing.
>>
>> It's another thing entirely if this advice is incorrect, and the person
>> never had that condition in the first place.  This is one of the things the
>> FDA exists to prevent, and what it says is going on in this case.
>> (Specifically that there are high enough odds that it's happening to a
>> significant number of 23andMe's customers.)
>>
>
> What evidence is there of that, and that it's resulted in actual harm due
> to people not following 23andMe's disclaimer that their reports aren't
> intended for diagnostic use?
>

Again: a good question to address to the FDA.  Although the FDA's policy is
"prove it doesn't do significant harm", not "prove it has caused
significant harm".  While this does veer toward the precautionary
principle, and the high costs of meeting their standards are a continuing
problem in the biotech industry, many cures and treatments over the years
have been able to meet those standards: it's not impossible for stuff that
works (even if it is a lot costlier than it could be).


>
> Rake the FDA over the coals for what it actually does do badly (and
>> doesn't do that it should), but let's confine the accusations to what's
>> actually going on, please.  There's more than enough without exaggerating
>> or making stuff up.
>>
>> (It's possible that you might have simply misread, but it seems at least
>> some people seem to be willfully misinterpreting what's going on here.)
>>
>
> Can you give a hypothetical scenario that demonstrates the unacceptable
> risks presented by 23andMe? Honestly, I want to understand.
>

I did (in the "one thing...another thing" quoted above), but I'll restate:

Let's say 23andMe says you have a 95% chance of having condition X.
Condition X has a 90% mortality rate over 5 years if left untreated.
However, there is a treatment that completely cures this condition,
although it itself has a mortality rate of 10% over 20 years.

Obviously you would prefer to take this treatment, right?  Well...what if
they're wrong, and you actually have less than a 1% chance of having
condition X?  (I.e., you have certain gene markers, but those markers
actually have less than a 1% correlation, not a 95% correlation.  They
haven't provided proof that it's 95% and not 1%.  Most gene markers have a
less than 1% correlation here, so without evidence to the contrary, this
gene marker is most likely the same.)

You're out the money for the treatment, and you now have a 10% chance of
dying in the next 20 years that you didn't have before, without having
gained anything.

The FDA doesn't care whether or not 23andMe has any connection to the
providers of this treatment.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20131127/9b2a4bb0/attachment.html>


More information about the extropy-chat mailing list