Inoperable Inventions: How Theranos Was Able to Patent Technology Built on a Lie

Almost 10 years ago, Elizabeth Holmes, founder, chairwoman, and CEO of Theranos, Inc., now-infamously graced the cover of the October 2014 print of The Forbes 400, a list of the 400 Richest Americans according to Forbes magazine.  [1].  Dubbed “The Freshman” in the issue, Holmes was declared the youngest self-made woman billionaire, worth approximately $4.5 billion, at just 30 years old.  [2].  But just a little over a year later on October 16, 2015, journalist John Carreyrou published a bombshell in The Wall Street Journal: not only did Theranos’ landmark, ground-breaking, and patented, technology not work, Theranos had actually been testing patient blood using their competitor’s equipment.  [3].  For reasons that are superbly explained by Carreyrou, Theranos’ technology simply did not work.  [4].

The value of a patent portfolio cannot be overstated, and Theranos is a *chefs kiss* exemplary paradigm of this fact.  In December 2017, more than two years after the world learned that Theranos did not even use their own technology because it did not work, Holmes still managed to secure a $100 million loan from Fortress Investment Group LLC… with Theranos’ patent portfolio serving as collateral.  [5], [6].  At the time of this writing, the portfolio consists of 122 granted or issued patents, now owned by Fortress or by Labrador Diagnostics, a company owned by Fortress.  [7], [8]. 

The fact that Theranos’s patent portfolio is alive, well, and valid begs the question… how can you get a patent on something that does not work?  And potentially do so dozens of time? The answer is actually pretty straight forward: the utility requirement for patentability is quite low for public policy reasons, and we generally assume that patented inventions work. 

The Utility Requirement & Operability

A patent application for an invention must meet several patentability requirements before a patent is granted for the invention.  And while all the patentability requirements have to be satisfied to get a patent, they are not created equally in terms of difficulty to attain.  The utility requirement, codified at 35 U.S.C. § 101, requires that an invention be useful in order to get a patent.  [9]. Some legal experts have dubbed the utility requirement the “low bar” to patentability, others have even gone so far to say it is a “nonexistent” requirement.  [10].  In order for an invention to be useful, what would one need to prove?  Well, logically, one way to demonstrate that something is useful is to show that it actually works.  It would make sense then, that in order to satisfy the utility requirement, the useful invention must be operable. [11], [12].  In layman’s terms, the utility requirement requires (in part) that an invention must be capable of doing what it purports to be able to do. 

Seems straightforward enough, right?  But where things get a little tricky is how much (or how little) an inventor is obliged to demonstrate to the U.S. Patent & Trademark Office (USPTO) that their invention is operable.  In an ideal world, every inventor could easily demonstrate their invention working in its intended way to the USPTO when they apply for a patent.  However, in the real world, it would be extraordinarily cost-prohibitive to make inventors create working examples or models of their inventions.  Take, as an example, a theoretical physicist, who develops a new type of reactor that generates positron beams.  Should that physicist be required to build a full-scale reactor that would probably cost hundreds of millions of dollars in order to get a patent?  Or say, build a working model a fourth of the size (which would still probably cost hundreds of millions of dollars)?  I think most would agree we probably shouldn’t, because an inventor’s inability to afford to build their invention isn’t a good enough reason to deny them patent protection.

Not only would such a requirement essentially discriminate against someone’s socioeconomic background, in some circumstances, an inventor may not even be capable of demonstrate their invention works, even if they could afford to.  Take for example an invention directed to a chemical reaction to produce a new catalyst.  If such a reaction required a reactant that is heavily regulated by the government (e.g., because it is toxic, explosive, etc.), it may be impossible for the inventor to demonstrate the chemical reaction because they are unable to purchase said reactant. Therefore, as a matter of public policy, we don’t require inventors demonstrate that their inventions work, because it would ultimately bottleneck hundreds of thousands of patent applications into a measly few dozen.

So, the question becomes, what public policy do we put in place that correctly balances the burden on the inventor to prove operability with the necessity that inventions should work to get a patent?

Ultimately, when balancing the scales, we tip them towards reducing the burden on the inventor.  The USPTO has guidelines for their Examiners on how to assess whether a patent application for an invention meets the utility requirement. [13].  However, within those guidelines, the section that covers operability acknowledges that “[s]ituations where an invention is found to be ‘inoperative’ and therefore lacking in utility are rare, and rejections maintained solely on this ground by a federal court even rarer.” [14]. This is likely because the USPTO has a long-standing rule, still in effect today, that there is a presumption an invention is operable as described in the application.  [15].  As a result of this presumption, and that the utility requirement is a fairly easy hurdle to clear, quite a few inoperable, if not impossible, patent applications have slipped through the cracks to become issued U.S. patents.  Some fun examples include a magnetic ring that increases the health of the human body [16], a method of increasing penile blood flow by smelling perfumes [17], a method of curing hiccups by drinking from a cup that electrically shocks you [18], and a watch that displays the user’s estimated remaining life [19].  And of course, you can add Theranos’ patent portfolio to that list as well.

Therano’s Far-Fetched, Fantastical Inventions

With all this in mind, we return back to Holmes, Theranos, and the still alive-and-well patent portfolio. It is true that hindsight really is 20-20, but signs were there from the beginning that Theranos’ technology was simply not capable of operating the way it was claimed to.  In the fall of 2003, when Holmes’ was just 19 years old, she drafted her first patent application, which would later issue into a U.S. patent in 2007.  [20].  Holmes’ had spent the previous summer as an intern at the Genome Institute of Singapore drawing blood from patients and testing specimens using old-fashioned technology, syringes and swabs.  [21].  Believing there must be a better way to diagnose and treat patients, her first patent application was for a skin patch that would not only diagnose medical conditions, but treat them as well.  [22], [23].  One of the first people Holmes showed the patent application to was Shaunak Roy, a PhD student Holmes worked with at Stanford.  [24].  In court testimony years later, Roy would say that when he read her patent application back in 2003, he believed that it was too far-fetched.  [25]. 

Ultimately, this became a recurring narrative with Theranos’s technology: it was just too far-fetched to work.  Holmes wanted to miniaturize devices to a scale that could not produce accurate test results [26].  She wanted the device to test for too many conditions and diseases on too small a blood sample.  [27].  Engineers, research scientists, and employees of Theranos worked tirelessly for years to try and create a device capable of what Holmes had imagined: a single, stylish, miniature device that could test for any condition or disease one could imagine using just a tiny drop of blood, and deliver the test results near instantaneously.  [28]. But, physical reality rendered it impossible to produce an operable device that could do just that.  Despite this, Holmes was still able to amass a notable patent portfolio for dozens of inoperable devices that ultimately retains some amount of value.  And the reason she was able to is because, as a matter of public policy, we have decided it is generally better to reduce the burden on inventors to prove operability to promote the progress of science and useful arts, which after all, is why we have a patent system to begin with. 

Sources

[1] Robehmed. ‘The Youngest Billionaires On The Forbes 400: 11 Under 40.” Forbes. 01 October 2014. https://www.forbes.com/sites/natalierobehmed/2014/09/29/the-youngest-billionaires-on-the-forbes-400-11-under-40/.

[2] “Elizabeth Holmes.” Forbes. 27 June 2024. https://www.forbes.com/profile/elizabeth-holmes/?sh=49a32cb147a7.

[3] Carreyrou. “Hot Startup Theranos Has Struggled With Its Blood-Test Technology.” The Wall Street Journal. 16 October 2015. https://www.wsj.com/articles/theranos-has-struggled-with-blood-tests-1444881901.

[4] John Carreyrou, Bad Bood: Secrets and Lies in a Silicon Valley Startup, 19-20, 169-171 (2018).

[5] Id. at 296.

[6] McKenna. “Theranos closes deal with Fortress to shut down embattled firm.” MarketWatch. 17 September 2018. https://www.marketwatch.com/story/theranos-closes-deal-with-fortress-to-shut-down-embattled-firm-2018-09-13#:~:text=Theranos%20Inc.%2C%20the%20blood%2D,the%20company's%20patents%20but%20leaving.

[7] Google Patents search - Assignee = “Theranos, Inc.”; Status = “GRANT” https://patents.google.com/?assignee=theranos&status=GRANT.

[8] Hu. “Fortress Drops Ex-Theranos IP Suit To Let Test Co. Fight Virus.” Law360. 29 June 2020. https://www.law360.com/articles/1287592/fortress-drops-ex-theranos-ip-suit-to-let-test-co-fight-virus.

[9] 35 U.S. Code § 101.

[10] Lee Petherbridge, “Road Map to Revolution? Patent-Based Open Science”, 59 Me. L. Rev. 339, 356 n.90 (2007) (“The utility requirement is still properly understood as very low and generally presents a low bar to patentability.”); see Michael Risch, A Surprisingly Useful Requirement, 19 Geo. Mason L. Rev. 57, 58 (2011) (noting that inventions which fail to meet the current standard are rare).

[11] Process Control Corp. v. Hydeclaim Corp., 190 F.3d 1350, 1359, 52 USPQ2d 1029, 1035 (Fed. Cir. 1999).

[12] Donald S. Chisum, Chisum on Patents § 4.04 (2024).

[13] MPEP § 2107.

[14] Id.

[15] Chisum, supra note 12 at § 4.04.

[16] US 5,989,178 A (issued November 23, 1990).

[17] US 5,885,614 A (issued March 23, 1999).

[18] US 7,062,320 B2 (issued June 13, 2006).

[19] US 5,031,161 A (issued July 9, 1991).

[20] Carreyrou, supra note 4, at 14.

[21] Id.

[22] Id.

[23] US 7,291,497 B2.

[24] Carreyrou, supra note 4, at 14.

[25] Id.

[26] Id. at 19-29, 95-97, 113-115, 166-171

[27] Id.

[28] Carreyrou, supra note 4, at 19-29, 95-97, 113-115, 166-171.