Prometheus v. Mayo – The Wrong Rat?

A decision with the right outcome but for the wrong reasons can confound jurisprudence nearly as much as a decision that is entirely wrong.

In our field an example is provided by the opinion of  Justice Douglas in Funk Brothers Seed v Kalo Inoculant, 333 US 127 (1948) which has recently been exhumed following decades of obscurity. The invention in that case concerned an inoculant for leguminous plants containing mutually non-inhibitive strains of Rhizobium bacteria. Modern readers would find it a surprising proposition that identifying such strains and enabling a farmer to conduct a single treatment of his fields when formerly six separate treatments were needed did not amount to invention. The rat which Justice Douglas smelled existed but was mischaracterised. As set out in the concurring opinion of Justice Frankfurter it was not enough to conceive of the idea that non-interfering strains might exist: it was necessary to go further and identify particular operable strains of bacteria (this was, of course, well before deposit systems for bacteria and other living species were established). The proper ground of non-enforceability was non-enablement, not obviousness.

Similarly in Mayo v. Prometheus the proper objection, it is submitted, was obviousness, not patent-eligibility. In its desire to issue a decisive opinion and to produce a correct outcome it is submitted that the court over-stretched the principles of patent-eligibility under 35 USC §101 and inadmissibly combined the §101 objection with §§102 and 103 objections, as suggested by commentators. To understand why the court may have been tempted to do so, it is necessary to consider the relevant factual background.

The thiopurine drugs were first used by Sir Roy Calne in Cambridge, England in the 1970’s to minimise transplant rejection. Subsequently they were suggested for use in the treatment of inflammatory bowel diseases such as ulcerative colitis and Crohn’s disease in order to avoid or minimise the use of steroids.

The inventors of the patent in issue, Ernest Seidman and Yves Théorêt, worked at a hospital in Montreal and suggested a metabolite-based blood test which would facilitate safe administration of drugs in this family. Patient monitoring over a period to maintain metabolite levels within a safe range was needed, as the chart below from a brochure of the licensees Prometheus illustrates.

The advantages of the Prometheus test are explained in their merits brief to the Supreme Court:

The practical problem addressed by the patents-in suit is determining the proper thiopurine dose for each patient. Because of variations in the activity level of the enzyme that breaks down these drugs (thiopurine methyltransferase, or TPMT), everyone metabolizes thiopurines differently. If a dose turns out to be too high for a particular patient, it could cause severe, potentially fatal, side-effects. Even “minimal doses” can be fatal for a minority of the population. These risks typically led doctors who used thiopurines to start with a very low dose and work slowly upwards while continually monitoring and testing their patients’ blood and liver for toxicity. This “start low, go slow” approach multiplied office visits and costs and dramatically delayed the drug’s effectiveness. And doctors still risked toxic consequences if they overshot the appropriate dosage for a particular patient. Many physicians, accordingly, were reluctant to treat their patients with thiopurines despite the potential benefits.

Researchers appreciated, in theory, that determining a patient’s thiopurine metabolite levels could provide valuable information about the efficiency with which that patient metabolizes the drug. And laboratories had the technology to determine those metabolite levels from blood or bodily samples.  But developing and validating a metabolite-based treatment protocol proved difficult because the complex metabolic pathways at issue were not completely understood…

No evidence in this record supports Mayo’s suggestion that, when that application was filed, doctors were already using measured metabolite levels clinically to calibrate thiopurine dosages for patients suffering from autoimmune disorders. To the contrary, at the time, there was persistent skepticism about whether monitoring metabolite levels would improve patient treatment. In 1997, for example, Mayo’s own Dr. Sandborn wrote that “measurement of [red blood cell] 6-TGN and 6-MMP in patients with Crohn’s disease treated with AZA or 6-MP remains investigational and cannot be recommended for routine clinical use.” It was not until 1999, when Prometheus introduced its test, that metabolite measuring tests became commercially available to practicing gastroenterologists…

Since their introduction, Prometheus’s patented methods have improved the lives of countless individuals suffering from autoimmune diseases. With these tools, doctors are both more willing and better able to treat patients effectively with thiopurines and minimize resort to more toxic or less effective drugs.

Mayo’s case was that all that might have been true, but from the standpoint of patent law it is beside the point because it does not start from the most relevant prior art. In their brief they objected that the inventors had in 1996 and well before the patents were filed freely published in a paper in an academic journal the general association between thiopurine metabolites and the efficacy or toxicity of thiopurine therapy. In an act of obfuscation worthy of the fictional Sir Humphrey Appleby (BBC Television: Yes Minister) that paper is referred to not in the Background section but in the middle of a lengthy specific description with the result that although its existence had been disclosed to both examining authorities (thus inter alia satisfying US disclosure requirements) neither the US examiner nor his European counterpart was prompted to study it and appreciate its significance. The final paragraph of the 1996 paper explains:

Our recent preliminary investigative efforts to measure 6-TG in leucocytes has shown a correlation between neutrophil 6-TG levels and responsiveness to treatment as well as drug induced leucopenia. Further research is needed to identify a therapeutic regimen for 6-MP treatment allowing clinicians to establish a balance between drug responsiveness and toxicity.

It is difficult to avoid the conclusion that all that found its way into the Siedman patents was the results of the very research that had been recommended in the 1996 paper and which Prometheus had been prompted to under-write. The more natural objection which, unfortunately, was not pursued was therefore lack of inventive step under 35 USC §103. It is submitted that this should have been enough to dispose of the issue between the parties, arguably even in a motion for summary judgment, but unfortunately it was not how the case was pleaded and argued.

The temptation to pursue the §101 objection may have been strengthened by the idiosyncratic form of the allowed claim which read:

A method of optimizing therapeutic efficacy for treatment of an immune-mediated gastrointestinal disorder, comprising:

(a) administering a drug providing 6-thioguanine to a subject having said immune-mediated gastrointestinal disorder; and

(b) determining the level of 6-thioguanine in said subject having said immune-mediated gastrointestinal disorder,

wherein the level of 6-thioguanine less than about 230 pmol per 8×10⁸ red blood cells indicates a need to increase the amount of said drug subsequently administered to said subject and

wherein the level of 6-thioguanine greater than about 400 pmol per 8×10⁸ red blood cells indicates a need to decrease the amount of said drug.

As has been indicated in essence and in its practical embodiment the invention concerned a blood test. When blood is centrifuged, there is a layer of plasma, a thin layer of leucocytes (white blood cells) and platelets and a deep lower layer of erythrocytes (red blood cells). The practical embodiments of the test were carried out by analysing the red blood cells, and although cheek cells and white blood cells are mentioned in the specification, owing to their lower abundance and the desirability of simple procedure it is doubtful whether these alternatives are realistic. What was the most advantageous, therefore, was a straightforward claim to a blood test for determining metabolite level in red blood cells.

The word “optimising” gives rise to clarity issues given the lack of any specified post-determination human activity, although no issue appears to have been raised in that regard. The administration step presumably involves the doctor or pharmacist whereas the determination step will be carried out in a remote lab, so that infringement of the claim by a single entity is difficult to establish. The determination step invokes no corresponding structure and therefore is clearly a step + function limitation within 35 USC 112(6). Its scope is therefore narrower than appears on the face of the claim, although this point does not appear to have been a subject of discussion. Implicitly it appears to require a sample from the human body for analysis. The “wherein” clauses are foreseeably problematic under 35 USC 101 since they do not define process features, nor do they define a machine, manufacture or composition of matter.

The granted claim in the corresponding EP-B-1114403 (Seidman) which corresponds to US 6355623 is more straightforward and might have created less difficulty if the same wording had been employed in the US:

An in vitro method for determining efficacy of treatment of a subject having an immune-mediated gastrointestinal disorder or a non-inflammatory bowel disease (non-IBD) autoimmune disease by administration of a 6-mercaptopurine drug, comprising

determining in vitro a level of 6-thioguanine in a sample from said subject having said immune-mediated gastrointestinal disorder or said non-inflammatory bowel disease (non-IBD) autoimmune disease,

wherein said treatment is considered efficient if the level of 6-thioguanine is in the range of about 230 pmol per 8×10⁸ red blood cells to about 400 pmol per 8×10⁸ red blood cells.

An even simpler formulation might have been:

A method of testing whether a patient suffering from an immune-mediated gastrointestinal disorder is receiving a therapeutically effective dose of a drug providing 6-thioguanine whilst being at low risk of leucopenia, which comprises separating red blood cells from a sample from the patient and determining whether the level of 6-thioguanine is within the range 230 – 400 pmol per 8×10⁸ red blood cells.

Such considerations are, however, moot bearing in mind the strength of the potential obviousness objection.  Why it was rejected by Mayo in preference to a patent eligibility objection must remain a matter of conjecture.