An Analytic Approach to Patent Eligibility

By Kevin E. Noonan --

The transcendental conundrum in patent law in these times is how to overcome the misinterpretation of the Supreme Court's decisions on patent eligibility law by district courts and the Federal Circuit.  That these courts cannot overcome the precedential tangle they have created is firmly established by the Court's own decisions; Judge Moore in Athena Diagnostics, Inc. v. Mayo Collaborative Services, LLC stated the obvious when she said in her dissent:

My colleagues' refusal deflates the Amici's hopeful suggestion that our precedent leaves the eligibility of a diagnostic claim in front of the Federal Circuit "uncertain."  It is no longer uncertain.  Since Mayo, every diagnostic claim to come before this court has been held ineligible.  While we believe that such claims should be eligible for patent protection, the majority of this court has definitively concluded that the Supreme Court prevents us from so holding.  No need to waste resources with additional en banc requests.

And in response, the Supreme Court has eschewed any and every opportunity to revisit their twisted precedent denying certiorari in over 50 cases (see "U.S. Supreme Court on Eligibility: Nothing to See Here, Move Along").

Part of the problem is that it has been maddeningly difficult to define not what patent eligibility is (you cannot go wrong with "anything under the sun made by man") but rather what it is not.  In the high technology class of inventions, this has come down to deciding without defining what an abstract idea is and when its abstractness prevents patent eligibility; see, e.g., "Stupid §101 Tricks").  (The other, related type of ineligible subject matter are business method patents, the exclusion of which is almost categorical (see "Bilski v. Kappos, Alice Corp. Pty. Ltd. v. CLS Bank Int'l"; "CyberSource Corp. v. Retail Decisions, Inc."); this has the benefit is requiring little interpretation and hence maximal certainty regarding what is ineligible.)

But the ineligibility of the latest iGadget, while sometimes tragic, is not as existentially problematic as the havoc that these precedents have wreaked on life sciences patenting.  For both diagnostic methods and to a slightly lesser extent natural products, the philosophically lost proscriptions by the Court, bolstered by plain illogic in district court (see "Ariosa Diagnostics, Inc. v. Sequenom, Inc.") and Federal Circuit (see "Federal Circuit Denies Rehearing en banc in Ariosa v. Sequenom") decisions, has rendered pursuit of patent protection for these inventions to be relegated to the ranks of the foolhardy.  The effect on investment and hence progress and innovation has been as expected; perhaps the only silver lining from the SARS-Cov-2 epidemic has been that in the frantic and desperate struggle for both diagnostics and vaccines the usual market forces have been collapsed by government investment (which is not usually a recipe for economic success).

But if policymakers cannot find a path to resolve the political differences that have prevented a legislative solution to the patent eligibility issue (see "The Zombie Apocalypse of Patent Eligibility Reform and a Possible Escape Route"), perhaps it is because the various stakeholders have not been able to arrive at a consensus over what should and should not be eligible (besides interest groups like the ACLU shouting "hands off my DNA"; see "ACLU (Predictably) Opposes Patent Subject Matter Eligibility Proposal").  This situation suggests it might be better to start with examples of what should be patent eligible (or at least something that would focus the basis for opposition to eligibility) and walk backwards to the point where various stakeholder groups could raise objections that might define the metes and bounds of why certain inventions have or should be deemed unprotectable under patent law.

For diagnostic methods, it might be best to start with that most tangible of subject matter classes, a machine (see "American Axle v. Neapco Holdings" and "Ex parte Itagaki and Nishihara" perhaps to the contrary).  In this example, the claimed invention would be a machine configured to perform a diagnostic assay on a biological sample, wherein the machine provides a diagnosis based on detecting the presence of a naturally occurring biomarker in the sample.  The specifications on the machine can be limited so that each embodiment of the machine is calibrated to perform only specific diagnostic assays on specific samples to detect specific biomarkers that are specific for only a specific disease or class of diseases.  As so limited, should this machine be patent-eligible?  Why or why not?  And in making these determinations of course the patent eligibility of other diagnostic machines, from the MRI to the EKG to the stethoscope should guide the eligibility decision.

If this machine passes the eligibility test, the next-level question could be whether the processes used by the machine (e.g., the chemistry) is patent-eligible, and does eligibility depend on whether these assays are novel?  What is the significance of whether they are adapted to detecting the biological phenomenon informative for the diagnosis?  And what if they are specifically adapted (along with all the specificities considered above regarding patent eligibility for the machine) to use with the machine?

Another consideration is whether these processes need to be performed by the machine or whether they could be performed outside the machine itself?  Would eligibility of the diagnostic assay depend on claim infringement requiring use of the machine, i.e., producing the "inputs" from the biological sample cannot be interpreted unless evaluated by the machine?

And does the complexity of the biomarker inputs make a difference, i.e., should it matter if the machine evaluates the input information using a computer, for example?  Does it matter if the inputs are so complex they can only be understood using a computer?

The answers to these questions can then be used to evaluate the distinction(s) between these features/characteristics of these different embodiments of the invention that vitiate patent-eligibility.  In other words, where do we draw the line?

For natural products, the issue may be less complex, based on the distinction drawn by the Court in Myriad that "mere" isolation of genomic DNA rendered that DNA ineligible for patenting, and that conversion of naturally occurring messenger RNA (mRNA) to complementary DNA (cDNA) involved sufficient human intervention to render cDNA patent-eligible.  But as a general rule, this distinction with other natural products is not satisfactory for at least two reasons.  First, it flies in the face of established practice that "mere" isolation could pass the patent eligibility test.  And second, because natural products in their natural state are not usually useful, their utility arising from isolation (and purification, and concentration) and use in applications for which they are not used in nature (the canonical example being the anticancer drug taxol which, while intrinsically having anticancer properties does not exhibit this capability naturally because trees don't get cancer).

The issue may be as simple as recognizing (as District Court Judge Sweet did in his opinion invalidating Myriad's claims) that DNA is sui generis, the "physical embodiment of genetic information."  "Mere" isolation of DNA was (perhaps) less relevant because the function of the DNA was unchanged by its isolation; before this statement enrages biotechnologists it should be recalled that all the ways that isolated DNA has been used remain patent-eligible (incorporated into vectors, introduced into heterologous cells, producing valuable biological molecules, etc.).  For almost every other natural product, however, its utility is not found in nature, so perhaps that transformation, produced by human intellect, should be enough to render most important "natural" products, converted most unnaturally, to be patent eligible.  After all, a claim to "Penicillin" per se has never been patent-eligible because it reads on penicillin as it exists in nature, and a claim to "a pharmaceutical composition comprising penicillin [at a useful dose and formulated for pharmaceutical use]" has always been (and still is) patent eligible.  What is uncertain today is whether a claim to "an isolated/purified preparation of penicillin [having a specific utility, claimed or disclosed in the specification]" is patent eligible, and the outcome of this thought experiment on this subject matter is to determine why or why not?

It may only be by "walking backwards" from exemplars of these different technologies that must be patent-eligible to the point where that eligibility can be called into question that some insight may be gained as to why some subject matter (or embodiments thereof) are patent-eligible sheep and others patent-ineligible goats.  Almost a decade of letting courts try to come to these distinctions in their case law has failed to arrive at any clarity, Blackstone be damned; maybe it's enough to make the case for a more analytical approach to recognize that the result can't be any worse.

Comments

[Profesor Joshua Sarnoff]

"Experience without theory is blind, but theory without experience is mere intellectual play." Immanuel Kant.

For a theory that is coherent (creative application of an ineligible discovery is required to constitute an "invention" in the meaning of patent law), and actually maps (incompletely - e.g., cDNA in Myriad) to the SCT and some of the incoherent CAFC case law, but which excludes significant subject matter (including most diagnostics) that others would like to see eligible, others (as you no doubt have already read them Kevin) can see my briefs in Mayo and Myriad, my articles and book chapters, and my testimony to Congress (some here: https://papers.ssrn.com/sol3/cf_dev/AbsByAuth.cfm?per_id=29684).

It's not hard to have a coherent theory of ineligibility. It is hard to get consensus for what should and should not be eligible, and the lack of consensus (experience) drives the incoherence of theory here. I hope that better consensus can result in legislation, but it will likely look like gerrymandering and not a coherent theory of eligibility that courts can apply consistently and sensibly.

Happy holiday season to all. Josh

[skeptical]

While your call to reason is admirable (as is your recognition that the score board is broken a proper starting point), I must take exception with the 'relativeness' and apparent 'special treatment" (perhaps only inadvertent) your phrasing calls forth. That 'little gadget' deserve every bit the same level of LEGAL protection as your life science advances. Substantive patent law -- on purpose -- is NOT built for silos, as that IS NOT how innovation works. See the James Burke series Connections for a fun reminder.

As second not-minor point is that in your attempt to 'walk back to' a 'reasoned' (or at least a negotiated) level, you step into a a very source of quagmire that 'got us here:' conflating prior art and eligibility.

Simple admonition on that point: DON'T DO IT.

[Scott Elmer]

Great idea Kevin! Perhaps the newly composed Supreme Court will have the courage to take up one of these cases and admit that making up exceptions to patent eligible subject matter that are found nowhere in the statute was and is wrong. These decisions have real world consequences. The most significant consequence from where I am sitting is that academics and industry can no longer effectively collaborate on diagnostic research since trade secret protection and publication do not mix.

[Karl P. Dresdner, Jr. Patent Agent 63,319]

Kevin,

Thank you for your analysis of patent eligiability and how to deal with this currently. I agree with your thinking, especially about claiming a machine approach when computing is a necessary invention element to be claimed.

[Kevin Noonan]

Thanks, Josh. How do you reconcile the express use of the term "discovery" in both the statute and the Constitution with "creative application of an ineligible discovery is required to constitute an "invention" in the meaning of patent law"?

[Martin H Snyder]

The problem is information.

The patent system as currently understood cannot deal with the fact the sometimes information is material, as in real and useful, and other times abstract, as perception that may or may not be useful.

I offer a simple answer to this problem, based on some simple propositions:

Claims and Inventions are distinguishable -

101 is for inventions. 102/103 112 are for claims.

A Process must have a result as a matter of law-

Not construing a result means a process invention has no reliable legal meaning.

Newly processed information is a kind of result-

Results of processes that are not information are not pressing problems in the patent system.

All information is abstract-

Mere abstraction should not be enough to achieve or deny eligibility, per Bilski.

The latin root of Abstract means to consume-
No human mind, no abstraction. Can anyone argue with that?

Abstract inventions are intrinsically abstract- when they result in information consumed by people. It's an inquiry that turns on a single fact.

Information consumed by non-humans cannot be abstract.

Abstract claims are extrinsically abstract, and so not a 101 problem, but a 102/103 112 inquiry.

Because PHOSITA varies in time and art. "Do it on a computer" patents are the classic choice: are the computer arts implicated, or the art being automated, or both, and to what extent are completely extrinsic, fact-based inquiries.

New diagnostic machines, compositions, and manufactures should be eligible.

Newly discovered diagnostic correlections should not be eligible.

Genetic information is not consumed by a human mind, and should be eligible, subject of course to the other statutory requirements.

Information inventions are the problem: the consumer of the information is the central fact that should determine eligibility.

[skeptical]

Mr. Snyder, an analytic approach rules out your "in my world" approach.

You remain unmoored to the terrain in which you would want to wage war.

Your 'traffic' is directly met with a 'red' light (as opposed to a blue squash ball).

[skeptical]

Dr. Noonan, as to your statutory construction question to Prof. Sarnoff, you may want to reference the writings of Sherry Knowles (in addition to the words of Congress at 35 USC 100(a): The term “invention” means invention or discovery).

This writing is not 'circular' (as others have remarked to another section of 35 USC 100, namely 35 USC 100(b) with its "The term “process” means process, art or method, and includes a new use of a known process, machine, manufacture, composition of matter, or material"), but instead is reflective of Congress recognizing legal terms of art and purposefully writing to expand what may be covered under the intellectual property landscape of patent protection.

"Start with the words of the Statute" is always a plus.

[skeptical]

Also, just stumbled across this older item:

https://www.ipwatchdog.com/wp-content/uploads/2017/05/IIPCC-WDC-May-8th-2017-Professor-Lefstin-presentation.pdf

[Martin Snyder]

The Bilski court famously declined to define "process". The term is unconstitutionally vague, so you can start with canons of statutory interpretation, 'cause the word is useless.

No result as a matter of law is the same as the situation pre-Markman: not reliable. Failure of due process.

The misapplication of a possibly useful claim abstraction test in Alice is icing on a pretty poor cake.

[skeptical]

Mr. Snyder,

You are quite clueless regarding the history of 101 here.

Please stop embarrassing yourself.

[Paul F Morgan]

This is simply note that it is important not to further confuse two very different 101 defenses, with very different subject-matter applications and Sup. Ct. decision bases. Namely, "abstract" claims versus claims to things pre-existing in nature. That may seem obvious, but I frequently see the two lumped together as a single problem.

[skeptical]

Mr. Morgan,

I hear what you are saying here.

However, what you are saying appears to be directly (and explicitly) contradicted by expert** opinion.

See: https://www.fr.com/defeating-alice/

** mind you, this expert opinion is more bombast than 'expert,' even as it is linked to a prominent law firm.

[Martin H Snyder]

It's abstract ideas and natural laws that should be lumped together, because they are both the same thing: information. All the judicial exceptions can be thought of as information much of the time, because the problem IS information.

Can't remember the last case where a "natural phenomenon"- other than some form of information- was in major controversy.

[Greg DeLassus]

"The latin root of Abstract means to consume"

Without offering comment on any other aspect of Martin's 17 Dec post, this etymological assertion simply is not true. "Abstract" is a compound of "ab" (from, away) and "trahere" (past participle = "tractus," to pull or drag). The "abstract," in other words, is that which one observes when one "pulls away" from the various individual particulars.

"Consuming" has nothing to do with this. It is very much possible to "pull away" from individual specifics without consuming them or using them up.

[MaxDrei]

Good comment, Greg. Is "abstract" then the same as "extract"? Same Latin root, more or less, and no consuming there either.

A dentist will happily abstract for you the root cause of your tooth pain and then extract for you your offending tooth. Extracting the cause and then abstracting the tooth seems wrong, somehow. Do the two words mean the same, or not?

So I Googled it and here's what I found:

"ex is really a more violent concept, expressing that the motion begins within one thing and goes elsewhere. ab expresses motion originating from a source, in a more vague sense (could be within the source or from its periphery), and going elsewhere"

But wait. Would the dentist agree that tooth extraction is "a more violent concept". I hope not. Then again, would patent attorneys agree that the sense of "abstract" is "vague"? I rather think they would.

Happy Holidays, everybody.