Is the Patent Eligibility Act a Panacea for Claims to Diagnostics?

On January 23, the Subcommittee on Intellectual Property held a hearing in which the members of the committee heard testimony about the Patent Eligibility Restoration Act (PERA) introduced by Senators Coons and Tillis. I have considered case law directed to s. 101 of 25 U.S.C. for more than a decade, and was hopeful that PERA could achieve its subtitled goal of “Restoring Clarity, Certainty, and Predictability to the U.S. Patent System”. The eight panelists had submitted comments that preceded the hearing, and the hearing was live streamed. The panelists were a blue ribbon group of patent attorneys who testified before the Senators, and took questions.

Of the eight panelists, six supported PERA and only two opposed it. The sacrificial lambs were Richard L. Blaylock of Pillsbury, Winthrop, representing Invitae Corp., and David Jones of the High-Tech Inventors Alliance. The rest of the panel was made up of heavy hitters in the pro-patent eligibility positions, including Andre Iancu, David Kappos, Adam Mossoff, Courtenay Brinckerhoff and Philip C Johnson (I am not familiar with Mark Deem).

Listening to the testimony of Blaylock and David Jones was more interesting, since I am, of course, pro-patent eligibility, particularly in the area of diagnostics. As Brinckerhoff correctly noted, the Fed. Cir. endorsed the patent ineligibility of diagnostic tests in Ariosa v. Sequenom in 2015. What I call “If A, then B” diagnostic claims based on the discovery of a naturally occurring correlation have been uniformly invalidated since. The rare exceptions are claims that attach non-routine procedures to effect the discovery.

David Blaylock was unambiguous in his testimony that “PERA would permit the privatization of natural phenomena in the form of knowledge of new biomarkers [‘A’ above] and their clinical relevance [‘B’ above].” David Jones testified that the patent system does not need the expansion of patent eligibility and that it yields predictable outcomes in the courts. He argued that PERA would “expand patenting … to any area of human activity that uses or relies on any type of machine or manufacture…” He states that proposed section (b)(1)B(ii) would allow any human activity to be patent-eligible if “the process [of organizing human activity – ed.] cannot practically be performed without the use of a machine or manufacture.”

While Blaylock opposed patenting newly discovered biomarkers, very little of the testimony was specifically directed to diagnostic claims generally. Perhaps the testimony of the panel members in favor of the patent-eligibility of diagnostic tests under PERA was assumed, but I am not so sure, particularly since PERA does not use that word. So I took a deeper dive into the draft bill. PERA proposes to abrogate judicial exceptions to patent eligibility. Section (A) of 101 allows patents inter alia for the use of a known or naturally-occurring processes. Section B(k) is augmented with requirements that the invention or discovery has a specific and practical utility. So far, so good.

But what s. 100 giveth, s. 101 may take away. Proposed section (b)(1)(C) excludes “A process that—(i) is a mental process performed solely in the human mind, or (ii) occurs in nature wholly independent of, and prior to any human activity.” Just what kind of process is C(ii) intended to exclude? I think it could be read to exclude “If A, then B” claims.

In Ariosa v Sequenom., the inventors discovered that the presence of cffDNA in the non-cellular maternal blood can be used to diagnose certain conditions in the fetus, such as Down’s syndrome. But the basis of the claim is the discovery of a correlation between the presence of cffDNA in a plasma sample taken from a pregnant female and, e.g. an abnormality in the fetus. I can see why the cffDNA should not be patentable, since it is “An unmodified natural product, as the material exists in nature” but does PERA intend to deny patent eligibility to discoveries of naturally occurring correlations that provide useful diagnoses?

In 2006 the Supreme Court considered this exclusion in the Metabolite Laboratories, Inc. v. Lab. Corp. decision in which the correlation in question was between the levels of homocysteine and cobalamin. If homocysteine levels were high, cobalamin levels were low, which might indicate an increased risk of stroke. The Fed. Cir. agreed that the correlation claim was patent-eligible but the Supreme Court dismissed the petition as improvidently granted.

Three justices dissented, none of whom are still on the Court. They took the position that the correlation is a patent-ineligible “natural phenomenon”. Of course, the full Court did not agree with the dissenters until the controversial Mayo decision. The Fed. Circuit felt obliged to follow the Mayo rule banning such correlations, even when the results were of great value to patients. Judge Linn wrote a concurrence in Ariosa that reads more like a dissent.

In this writer’s opinion it is high time (or just the right time) to legislatively void judicial exceptions to patent-eligibility—which PERA would accomplish, and then to clearly state that patent claims based on the discovery of the utility of a naturally occurring phenomena are patent eligible as a process claim. The language in the bill that could be read to deny patents on the discovery and utility of such correlations should be clarified or eliminated.

Section (b)(i)(C)(i) should be deleted. (The defendants in Metabolite Labs. argued that the results of the correlation were solely performed in the mind of the doctor who read the results of the assay.) C(ii) should be deleted since it could be interpreted to bar any claim based on the discovery of a process that occurs in nature  “wholly independent of, and prior to, any human activity.”

That is true of every “process”, or biochemical pathway, that permits the human or non-human to function or misfunction. Perhaps the clause “prior to, any human activity” in C(ii) was intended to indicate that a “process that occurs in nature” would become patent-eligible if the process involves the discovery of a novel correlation using “a molecule never previously linked to the condition.” This is a gloss on the question presented in Athena v. Mayo’s 2019 petition for cert. in which Athena argued that “application of a law of nature even to a known structure or process” should be patent eligible. In Athena, the process was based on the discovery that MG patients generate autoantibodies to a naturally occurring protein, MuSK.

This was the “process that occurs in nature wholly independent of, and prior to any human activity.” The first step of the claimed Athena process involved the binding of radiolabeled MuSK to autoantibodies, to detect the presence of the autoantibodies to MuSK. Again, did the drafters of PERA intend that human activity based on the discovery of the utility of the correlation would render the process patent-eligible? In other words, one could not patent a process of binding autoantibodies to MuSk in vivo, while a diagnostic procedure based on measuring the existence and extent of the binding in vitro would be patent-eligible. I  hope so, but PERA does not set the bar for how much “human activity must be” expended before the natural “process” becomes patent eligible. The Fed. Cir. has denied the patentability of a claim directed to the isolation of a natural product cffDNA.

Perhaps C(ii) could be amended to read “a process—(ii) that is directed to a natural occurring phenomenon.”  This term has often been used by the courts as the starting point of their analyses of patentability. PERA would leave the fate of diagnostic claims in the hands of section 100(b), that PERA would amend to read “includes a use, application, or method of manufacture of a known or naturally occurring process.” This at least fits with the exception of B(ii).

Now I have come full circle in a post in which I tried to analyze PERA. If this bill gets out of committee, you will hear from me again.

  Back to All Resources