The Relevance of In re Bilski to the Patentability of the Metabolite Claim

    By Kevin E. Noonan --

In the Bilski case, the Federal Circuit confirmed that the judges have learned the hard lessons of ten years of Supreme Court reversals.  The decision is based solely and completely on what the judges seem to believe is the "test" mandated by Supreme Court precedent, based in large part on the Supreme Court's decision in Gottschalk v. Benson, 409 U.S. 63 (1972).  The test is a simple one:  for a process to be eligible for patenting (i.e., to be deemed patentable subject matter under 35 U.S.C. § 101), it must be claimed to include one or the other of these elements:  (1) it is tied to a particular machine or apparatus, or (2) it transforms a particular article into a different state or thing.  This "machine-or-transformation" test is supported, according to the majority, by later Supreme Court case law (Diamond v. Diehr), as well as prior U.S. patent law (all the way back to the 1793 Patent Act) and the British Statute of Monopolies.  The opinion is replete with citations to earlier Supreme Court precedent (Tilghman v. Proctor, 125 U.S. 136 (1888); Cochrane v. Deener, 95 U.S. 355 (1877)) and expressly disclaims the analytical framework of much of the Federal Circuit's own jurisprudence (perhaps surprisingly, not including the State Street Bank & Trust Co. v. Signature Financial Group case, which the majority of the en banc court distinguishes on the grounds that the State Street claims were to a machine).

Bilski was almost the perfect case for this opinion, being a purely business methods case with little to tie it to any tangible result (the claims were directed to a method for managing the risk of selling commodities like coal or wheat).  The Federal Circuit recognized the problem:  patent pre-emption of purely mental methods (or, as Judges Dyk and Linn said in their concurring opinion, "processes for organizing human activity").  While rejecting an absolute requirement for "physical transformation" to be associated with patentable processes, the Court did reconcile this decision with cases like Diamond v. Diehr by emphasizing that the algorithm in Diehr (the Arrhennius equation) was not totally pre-empted by Diehr's claims, but was limited to the application of the algorithm to the process of curing rubber.

Judge Newman (at left), in dissent, seems to have identified the very cases cited by the majority in support of its decision to contain what she terms "unequivocal" statements that the Benson test is not the Supreme Court's test.  Citing Benson itself, she notes:

It is argued that a process patent must either be tied to a particular machine or apparatus or must operate to change articles or materials to a "different state or thing."  We do not hold that no process patent could ever qualify if it did not meet the requirements of our prior precedents.  It is said that the decision precludes a patent for any program servicing a computer.  We do not so hold.

Benson, 409 U.S. at 71.  Judge Newman:  ". . . there is nothing equivocal about 'We do not so hold.'"

And Parker v. Flook, 437 U.S. 584 (1978):

The statutory definition of "process" is broad.  An argument can be made, however, that this Court has only recognized a process as within the statutory definition when it either was tied to a particular apparatus or operated to change materials to a "different state or thing."  As in Benson, we assume that a valid process patent may issue even if it does not meet one of these qualifications of our earlier precedents.

The line between a patentable "process" and an unpatentable "principle" is not always clear.  Both are "conception[s] of the mind, seen only by [their] effects when being executed or performed."

Flook, 437 U.S. at 589 (alterations in original) (quoting Tilghman v. Proctor, 102 U.S. 707, 728 (1880)).

And Diamond v. Chakrabarty, 447 U.S. 303 (1980):

The subject-matter provisions of the patent law have been cast in broad terms to fulfill the constitutional and statutory goal of promoting "the Progress of Science and the useful Arts" with all that means for the social and economic benefits envisioned by Jefferson.  Broad general language is not necessarily ambiguous when congressional objectives require broad terms.

A rule that unanticipated inventions are without protection would conflict with the core concept of the patent law that anticipation undermines patentability.  Mr. Justice Douglas reminded that the inventions most benefiting mankind are those that push back the frontiers of chemistry, physics, and the like.  Congress employed broad general language in drafting §101 precisely because such inventions are often unforeseeable.

Flook did not announce a new principle that inventions in areas not contemplated by Congress when the patent laws were enacted are unpatentable per se.

Chakrabarty, 447 U.S. at 315-16.

And even Diamond v. Diehr, a case the majority believed was consistent with its approach:

[W]hen a claim containing a mathematical formula implements or applies that formula in a structure or process which, when considered as a whole, is performing a function which the patent laws were designed to protect (e.g., transforming or reducing an article to a different state or thing), then the claim satisfies the requirements of §101.

Diehr, 450 U.S. at 192.  Judge Newman also cited more recent Supreme Court precedent to cast doubt on the soundness of the majority's position:

In J.E.M. Ag Supply, Inc. v. Pioneer Hi-Bred International, Inc., 534 U.S. 124 (2001), the Court described Section 101 as a "dynamic provision designed to encompass new and unforeseen inventions," id. at 135, that case arising in the context of eligibility of newly developed plant varieties for patenting.  The Court stated: "As in Chakrabarty, we decline to narrow the reach of §101 where Congress has given us no indication that it intends this result."  Id. at 145-46.

The relevance of this to biotech and pharma patenting is, of course, the Laboratory Corp. v. Metabolite Labs., Inc. case, and Justice Breyer's criticism of the scope of that claim under a patentability analysis.

The claim at issue in the Metabolite case concerned a diagnostic method for identifying a vitamin deficiency based on increased levels of homocysteine in blood:

13.  A method for detecting a deficiency of cobalamin or folate in warm-blooded animals comprising the steps of:  assaying a body fluid for an elevated level of total homocysteine; and correlating an elevated level of total homocysteine in said body fluid with a deficiency of cobalamin or folate.

In answering the question, is this claim patentable under the Bilski majority's "machine or transformation" test, a trivial way of performing the analysis would be to say that the Metabolite claim involves a physical transformation (detecting homocysteine levels).  But that misses the point:  the chemical detecting step in the Metabolite claim merely provides the raw materials, the information, used in "correlating" levels of homocysteine with levels of risk of disease.  It is this informational component that bedevils the analysis, because it removes claims like the Metabolite claims from the comfort of typical chemical/pharma claims.

That is not to say that diagnostic claims should not be patentable; rather, it is to say that the Federal Circuit's usual penchant to treat patent law as technology neutral is imprecise and inefficient.  There is nothing wrong, in theory, with the Court's approach, but in practice it tends to lead to anomalous results.  Here, the question is whether a claim to a method for making a diagnosis is taking a "fundamental principle" out of the public domain, so that a physician could not perform the step of "correlating" higher homocysteine levels to make a diagnosis.  (Of course, insofar as the Federal Circuit, or the Supreme Court, is concerned about restricting a physician's right to practice medicine unfettered by patent law, 35 U.S.C. § 287(b) provides physicians with an almost plenary exemption to patent infringement.)

One way around the problem is to use kit claims, because (frankly) they are easier to enforce and use the principles of convenience and consistency to impose compliance.  Another is to tie the claim to a device, for example, an automated analyzer that would quickly produce a result (i.e., disease or no disease) from a drop of blood.  The third, albeit the weakest because it can be cast as "post-solution activity (which the Bilski majority said was expressly rejected by the Supreme Court as a rationale for rendering patentable otherwise unpatentable subject matter), would be to cast the claim as a method for determining a course of treatment, wherein the ultimate step would be administering the appropriate medicine.

What is clearly outside Supreme Court precedent is being able to claim the correlation between rising homocysteine levels and vitamin deficiency, something that is certainly a fundamental principle of medicine and physiology.  And the Metabolite claim is particularly vulnerable using the Bilski majority's logic, since the "correlation" step is broad enough to encompass a physician making a diagnosis from merely observing the results of a blood test.

On the other hand, it makes much more sense to say that providing a diagnosis is a different matter than claiming the correlation:  like in Diehr, the correlation permits an actor to take or not take a tangible, concrete treatment step.  This does not need incorporation of the treatment step to confer patentability, merely a recognition that the claim to the diagnostic method "does enough" to be patentable while not doing so much that it precludes use of the correlation algorithm (for example, which could be used to detect other diseases).

In analogy to the distinctions drawn by the Supreme Court (and the Federal Circuit) between the Diehr (patentable method) and Benson (unpatentable) cases, there may be many uses (diagnostic or otherwise) for the information that a patient has an increase in peripheral blood homocysteine levels.  If the claim was directed merely to "methods for detecting increased blood homocysteine levels to identify a patient in need of therapeutic treatment," then the Metabolite claim would be analogous to the digital methods of the Benson case, where the Supreme Court determined that the claim encompassed all uses for the method.  But the Metabolite claim is not so broad:  the uses of the "correlation" are limited to diagnosing but one (actually, two) pathologies (deficiencies in one or the other of two B vitamins), just as in Diehr the claimed use of Arrhenius equation to regulate rubber curing did not preclude the other uses of the equation.

The tension between a claim and the information upon which the claim is based is not new to biotechnology patents:  this is essentially the problem solved by the Federal Circuit for nucleic acids in Amgen Inc. v. Chugai Pharmaceutical Co., 927 F.2 1200 (Fed. Cir. 1991), when the Court held that "a gene is but a chemical compound, albeit a complex one."  The gene (isolated as a chemical) is patentable; the sequence is not.  Accordingly, patent infringement does not lie when the sequence of a patented gene is used, for example, to interrogate a nucleic acid sequence database.  The distinction is harder to draw for method claims, which are not tangible in the way a nucleic acid is, and are necessarily comprised of steps performed, inter alia, by a human being.  The fundamental policy question remains the same, however:  how to permit inventions that satisfy the other statutory requirements, and for which the public should receive the benefits of any patented technology (disclosure), to be patented without trespassing on the fundamental principles upon which (in medicine) those methods are based.

The Supreme Court will now be given the opportunity to decide whether the Federal Circuit has made sufficient mischief on this matter; how the Federal Circuit reacts to the application of the Bilski rationale to medical diagnostic claims will depend, most probably, on the identities of the judges that hear an appeal on such claims.

Comments

[Dan Feigelson]

Nice analysis, Kevin. Thanks.

[David]

Dear Kevin,

Regarding diagnostics: since intent is usually given no patentable weight, unless the assay methods used are novel and non-obvious, I don't see how claiming a kit or a device for diagnosing a condition/risk on the basis of a newly found correlation is going to confer patentability to the invention.

Could someone please develop the idea behind multiple uses for a mental step (i.e. correlation) being an argument against a claim to one of those uses preempting the activity?

Thanks,
David

[Kevin E. Noonan]

Dear David:

Whether a claim is patentable under 102, 103 and 112 isn't my concern; I think kits and devices are more "tangible" embodiments of the correlation and thus are less at risk under the court's Bilski analysis. Don't you think that the machine part of the Bilski test would render patentable a device that "read" a drop of blood to detect increased homocysteine = need for vitamin supplements?

I think Diehr says that an algorithm useful in a claimed method can be patentable provided that the entirety of the algorithm isn't pre-empted - that was the distinction the court made in Dayton.

[Kevin E. Noonan]

Sorry - Besson

[Norman Siebrasse]

Kevin

To quibble just a bit, I think it was Benson where the court said an algorithm is patentable provided it doesn't preempt all applications; Flook said an invention is patentable provided the algorithm isn't the point of novelty; Diehr did cite the preemption argument from Benson (it also cited Flook), but I think Diehr is based primarily on the rule that an abstract algorithm is not patentable, but it is patentable if it has a practical application. That is certainly how the CAFC interpreted Diehr at least until Bilski (see e.g. Alappat). I think these three approaches to subject matter are different and incompatible. I entirely agree that the CAFC has finally decided to try to follow SCOTUS precedent. The hash in Bilski is the result of trying to follow incoherent precedent.

[Kevin E. Noonan]

Dear Siebrasse:

That's what I get for referencing Judge Newman. I guess the point is that Benson says the invention DID preempt all applications and was unpatentable, and the invention in Diehr did not preempt the algorithm and so was patentable. I do agree that the CAFC will go astray if it blindly follows incoherent Supreme Court precendent - it seems Supreme Court justices have their own brand of gobbledygook.

Thanks for the comment.

[Kevin E. Noonan]

Sorry, Norman - I am traveling and am pretty tired. I meant to say either "Dear Norman" or "Dear Mr. Siebrasse." No disrespect intended - thanks for the contribution.

[CSTLR]

If you would like another perspective on the issue, see:
W.M. Schuster's "Predictability and Patentable Processes: The Federal Circuit's In re Bilski Decision and Its Effect on the Incentive to Invent." 11 Columbia Science & Technology Law Review 1. Available at:
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1353604