On September 16 the Federal Circuit decided a patent case called Prometheus Laboratories Inc. v. Mayo Collaborative Services (pdf). Prometheus sued Mayo for infringing two patents on a method of optimizing drug therapy for autoimmune diseases. The question in the case was whether the method met the patentable subject matter standard of section 101 of the Patent Act, as interpreted in the Federal Circuit’s 2008 In re Bilski decision (pdf) (which is now before the Supreme Court). The patentable subject matter standard is an initial threshold that must be crossed on the way to patentability. To satisfy the standard, the claimed invention must be within the broad categories of things (machines, manufactures, compositions of matter, or processes/methods) that the Patent Act permits to be patented. If not, the inquiry ends; if so, the invention must still be shown to be novel, useful, and nonobvious. This is the standard that the ACLU has focused on in its motion for summary judgment in the Myriad Genetics case, which we have been following closely.
Claim 1 in Prometheus’ U.S. Patent number 6,355,623 claims “a method of optimizing therapeutic efficiency for treatment of an immune-mediated gastrointestinal disorder.” The method comprises “administering” a specified drug to a patient and then “determining” the level of the drug in the patient. The remainder of the claim specifies threshold levels of the drug’s metabolites (the chemical products of metabolism in the body) in the patient’s blood below which the dose should be increased (because of lack of efficacy) and above which it should be decreased (because of potential toxicity).
Bilski involves a business method patent, but its logic also applies to diagnostic and therapeutic methods. Under Bilski (at least as it now stands), the method must either be tied to a particular machine or transform an article into a different state or thing—hence the shorthand “machine or transformation” test. The question in Prometheus was whether the claimed method satisfied the transformation branch of the test. The Federal Circuit, reversing a ruling by a federal district court in California, held that it did.
The Federal Circuit found that both the “administering” and “determining” steps in the patent brought about sufficient transformations. In the administering step, “The transformation is of the human body following administration of a drug and the various chemical and physical changes of the drug’s metabolites.” At the determining stage, the court found that measuring the levels of drug metabolites in the body “necessarily involves a transformation, for those levels cannot be determined by mere inspection.” Transformative methods might include high-pressure liquid chromatography or the extraction of metabolites from the body. Apparently, the transformation at either the administering or the determining stage would have been sufficient to satisfy Bilski.
The court also dismissed other objections to the patent based on language in Bilski and its predecessor cases. In response to the argument that the claims merely covered natural correlations (laws of nature, which are not patentable subject matter) and data-gathering (also not patentable), the court held—sweepingly—that “The asserted claims are in effect claims to methods of treatment, which are always transformative when a defined group of drugs is administered to the body to ameliorate the effects of an undesired condition.” It also held that the presence of “mental steps” (a red flag in a method patent) in deciding how to respond to the metabolite levels did not negate the transformation that occurred. Finally (dismissing another red flag), the court found that the patent did not “wholly preempt” the use of correlations between metabolite levels and efficacy and toxicity.
An interesting legal aspect of the case is its consistency with a couple of older Supreme Court cases dealing with whether software-based inventions were patentable subject matter. In 1978, in Parker v. Flook, the Court rejected a claim to a “method for updating the value of at least one alarm limit” in a catalytic converter that involved the repeated measurement of temperature and the repetitive calculation of a known equation with the newest temperature reading. But three years later, in Diamond v. Diehr, the Court allowed a claim to a “method of operating a rubber-molding press . . . with the aid of a digital computer” that involved the repeated measurement of temperature and the repetitive calculation of a known equation with the newest temperature reading. The only difference between the two seemed to be the claim language: Flook (invalid) claimed the measurement and calculation process itself, whereas Diehr (valid) claimed an industrial process that involved an indistinguishable measurement and calculation process.
It seems that Prometheus is putting similar emphasis on the claim language. Although the Prometheus claim involves measurement, correlation, and interpretation of significance, it is framed as a method of performing a therapeutic intervention—a treatment. As the Federal Circuit held, a method of treatment is “always transformative when a defined group of drugs is administered to the body” (my emphasis). So the various mental steps are folded into a conventionally patentable process (as in Diehr) as opposed to being the essence of the claim (as in Flook).
In one sense, this is an invitation to semantic gamesmanship: hide what are essentially claims to measurement, correlation, and interpretation—pure mental steps—in some nominal physical activity and you survive section 101. But there is more to it than that. The act of framing mental steps in transformative activity limits the scope of the patent. Prometheus does not own the mental processes, just the use of those processes in this particular therapeutic method.
Patent lawyers will argue about how significant or trivial this limitation really is, and their positions will largely depend on whether they think patents like this help or hinder the progress of medicine. The Prometheus decision is likely to make it easier to patent other kinds “personalized medicine” interventions. Any kind of treatment that changes the body or its contents in some way should be found to be transformative and thus patentable subject matter (but again, it still must be novel, useful, and nonobvious). The status of interventions that are merely diagnostic—for example, methods for finding genetic mutations—is less clear. On the one hand, it seemed important in Prometheus that the claim was to “a method of optimizing therapeutic efficacy.” But on the other, the court did hold that measurement involving more than “mere inspection” is itself transformative, which would seem to open the door to just about anything.
For the record, my own view is that the Federal Circuit (and the Supreme Court, should it take the case) will have second thoughts about just how widely it has opened the subject matter door. I predict (or at least hope) that future cases will worry about the impact of this sort of method patent on medical practice and research and look for ways to limit Prometheus, maybe by focusing on the therapeutic aspect. But don’t bet the house and the dog on this or any other outcome just yet.