How could Theranos have proved tech without revealing trade secrets?
March 19, 2019 7:57 PM Subscribe
Suppose Theranos actually had been able to match traditional venous blood tests using capillary finger-stick blood. How could they have convinced the medical/bio-tech establishment that it was real without letting Siemens/Lab Corp/Quest Diag. use their method?
They didn't need to because investors coughed up hundreds of $millions without anything more than faked demos, but suppose they really had figured out the finger-stick test, even for a subset of the "hundreds" of the tests they claimed? They could patent some of the methods, but if they didn't want the expiration clock running and wanted to keep some of it trade secrets, how could they do the proper peer-reviewed-journal comparison tests? How is this done in biotech in general?
They didn't need to because investors coughed up hundreds of $millions without anything more than faked demos, but suppose they really had figured out the finger-stick test, even for a subset of the "hundreds" of the tests they claimed? They could patent some of the methods, but if they didn't want the expiration clock running and wanted to keep some of it trade secrets, how could they do the proper peer-reviewed-journal comparison tests? How is this done in biotech in general?
BTW, not patenting something solely in order to not "start the expiration clock running" would require you to be missing large chunks of brain tissue. You don't get any rights in the subject of a patent until you complete the patent process. Don't patent, risk someone else coming along during the twenty years you'd otherwise have, reproducing your work, and patenting it themselves. Then not only would any claim of trade secret protection be valueless, but you yourself could be excluded from practicing the patent!
posted by praemunire at 8:29 PM on March 19, 2019 [4 favorites]
posted by praemunire at 8:29 PM on March 19, 2019 [4 favorites]
Best answer: Unfortunately "an independent investigator draws vials of blood in the traditional manner, divides out a finger-stick portion from each" would not work, because essentially that is how current blood test work. The issue is that Theranos was claiming to be able to do the tests on a finger-stick sample. Blood from that location is not the same as blood drawn from a vein.
To do a proper test a subject would need to have blood drawn from both vein and fingerstick, each sample assigned a random id number, and the vein blood sent to two different labs. All 3 samples would have to agree to prove Theranos worked.
posted by Sophont at 11:06 PM on March 19, 2019 [6 favorites]
To do a proper test a subject would need to have blood drawn from both vein and fingerstick, each sample assigned a random id number, and the vein blood sent to two different labs. All 3 samples would have to agree to prove Theranos worked.
posted by Sophont at 11:06 PM on March 19, 2019 [6 favorites]
I've just finished reading Bad Blood which goes into loads of details on the attempted trials, flawed research and opportunities for testing they had. It's a great read, well recommended though did make cross at the silicon valley culture which caused such harm.
Because Theranos didn't want to get FDA approval for their devices they hosted them in their own labs which got CLIA signoff instead (though they didn't show inspectors one lab and faked testing results). Therefore anyone using Theranos didn't come into contact with the devices and any trade secrets were secure (although they had also patented a lot of it). The journalist does his own testing by getting his own and other peoples tests done with Theranos (through their Walgreens tie-up) and with other labs. This showed how bad the Theranos results were even though most of the tests were using existing commercial analysers.
The Theranos machines were also tested for calibration / certification by using special samples containing known levels which they had to match. In the book it talks about how they cheated these (excluding outliers, rerunning tests) to get the results they wanted.
So I don't think not being able to test / prove the devices while maintaining secrecy was an issue - a paper could present the methodology of the trials, comparators, population and all the data without explaining how the device itself worked.
Since the scandal broke, Theranos did publish a paper on their second device, the mini lab (and explains how it works), but it's now being treated with some skepticism.
posted by JonB at 1:52 AM on March 20, 2019 [4 favorites]
Because Theranos didn't want to get FDA approval for their devices they hosted them in their own labs which got CLIA signoff instead (though they didn't show inspectors one lab and faked testing results). Therefore anyone using Theranos didn't come into contact with the devices and any trade secrets were secure (although they had also patented a lot of it). The journalist does his own testing by getting his own and other peoples tests done with Theranos (through their Walgreens tie-up) and with other labs. This showed how bad the Theranos results were even though most of the tests were using existing commercial analysers.
The Theranos machines were also tested for calibration / certification by using special samples containing known levels which they had to match. In the book it talks about how they cheated these (excluding outliers, rerunning tests) to get the results they wanted.
So I don't think not being able to test / prove the devices while maintaining secrecy was an issue - a paper could present the methodology of the trials, comparators, population and all the data without explaining how the device itself worked.
Since the scandal broke, Theranos did publish a paper on their second device, the mini lab (and explains how it works), but it's now being treated with some skepticism.
posted by JonB at 1:52 AM on March 20, 2019 [4 favorites]
Theranos is probably a bad example, because it seems pretty clear that ti was a scam the whole time, but as others have noted just doing some controlled bake-off tests would have gone a long way.
That Theranos went out of its way to avoid this kind of comparison & testing says volumes. I mean, if the tech actually worked, setting up IP-protecting bake-offs with traditional testing systems would have been MARKETING TASK #1, right? There was (and remains) broad skepticism that the drop-of-blood approach is even viable, so demonstrating it early and often should've been a no-brainer.
Unless, of course, it never worked, and could not work. Which turns out to be the case.
posted by uberchet at 4:29 AM on March 20, 2019
That Theranos went out of its way to avoid this kind of comparison & testing says volumes. I mean, if the tech actually worked, setting up IP-protecting bake-offs with traditional testing systems would have been MARKETING TASK #1, right? There was (and remains) broad skepticism that the drop-of-blood approach is even viable, so demonstrating it early and often should've been a no-brainer.
Unless, of course, it never worked, and could not work. Which turns out to be the case.
posted by uberchet at 4:29 AM on March 20, 2019
Also FWIW, there are special patent provisions for medical devices that can give you extra time on your patent if you lose marketing time waiting for FDA approval. Theranos’s supposed IP problems aren’t, like, unique, they’re a normal part of the medical device business.
posted by mskyle at 4:53 AM on March 20, 2019 [4 favorites]
posted by mskyle at 4:53 AM on March 20, 2019 [4 favorites]
Mod note: A few comments deleted; question isn't really about Theranos, it's about the testing /demonstration process in biotech.
posted by LobsterMitten (staff) at 12:47 PM on March 20, 2019
posted by LobsterMitten (staff) at 12:47 PM on March 20, 2019
This thread is closed to new comments.
posted by books for weapons at 8:17 PM on March 19, 2019 [13 favorites]