Federal Circuit doubles down on ineligibility of diagnostic correlations

The latest precedential decision covering Section 101 patent eligibility deals a blow to patent owners, but there may still be hope for patenting diagnostic methods

On 6 February 2019 the US Court of Appeals for the Federal Circuit held that detecting the presence of a previously unknown disease corollary is patent ineligible as a natural law.

The technology

The claims at issue in Athena Diagnostics v Mayo Collaborative Services related to a diagnostic method involving muscle-specific tyrosine kinase (MuSK). The inventors of US Patent 7,267,820 (titled “Neurotransmission Disorders”) discovered that individuals with a particular type of neurological disorder exhibited a heightened concentration of MuSK antibodies. Athena Diagnostics – exclusive licensee of the ’820 patent – commercialised a diagnostic kit to detect the antibodies and thus identify the disorder.

After Mayo developed two competing tests, Athena Diagnostics sued it for infringing the ’820 patent. Mayo moved to dismiss under Rule 12(b)(6), arguing that the claims were invalid under 35 USC §101. The US District Court for the District of Massachusetts agreed; Athena Diagnostics appealed the district court’s decision regarding Claims 6 to 9.

Each of the claims depended on independent Claim 1, which recited:

1. A method for diagnosing neurotransmission or developmental disorders related to muscle specific tyrosine kinase (MuSK) in a mammal comprising the step of detecting in a bodily fluid of said mammal autoantibodies to an epitope of muscle specific tyrosine kinase (MuSK).

However, as is the case with many diagnostic patents, none of the claims recited an actual method of treatment. Rather, they focused on identifying the neurological disorder by detecting the MuSK antibody. The Federal Circuit agreed with the district court that the ’820 Patent claims thus did not satisfy Section 101.

Invention versus discovery

It has been said that an invention is sometimes the identification of a problem, the solution to which may then become obvious. In the case of diagnostic methods, however, the courts have repeatedly found that such an identification – however meaningful – does not constitute a patent-eligible invention.

In Mayo v Prometheus, the US Supreme Court found that measuring metabolite levels after administering thiopurine in order to detect drug efficacy was not an eligible invention, but merely an observation of a naturally occurring effect. In Ariosa Diagnostics v Sequenom, the detection of cell-free foetal DNA in the bloodstream was deemed a mere observation. In Cleveland Clinic v True Health, the Federal Circuit found that detecting cardiovascular risk by measuring a correlated enzyme in the bloodstream was not eligible.

In all these cases the inventive concept was the discovery of a correlation between a measurable element and a disease. The claims recited only conventional ways of observing that correlation (eg, Prometheus and Cleveland Clinic) or detection at such a high level of detail as to potentially pre-empt all other methods (eg, Ariosa).

In this regard, Athena Diagnostics is no different and the result is therefore unsurprising. As stated by the Federal Circuit, “[h]ere, [the relevant natural law] is the correlation between the presence of naturally-occurring MuSK autoantibodies in bodily fluid and MuSK-related neurological diseases like MG. This correlation exists in nature apart from any human action” (Slip Op, pp 9-10). Once that correlation was discovered, the techniques used to detect the autoantibodies were admittedly conventional and well known. The additional elements of appealed Claim 6 were also deemed to be conventional.

So is there hope for protecting valuable diagnostic methods? The Federal Circuit hints that it may be possible by focusing on details that differ from mere identification of a correlation followed by detection of the correlative entity. Footnote 3 of the decision states:

Our cases have not described a claim to the binding of two molecules during a sequence of chemical manipulations (here, after MuSK labeling and before immunoprecipitation) as a claim to a natural law, even if such binding occurs according to natural laws. We need not resolve that issue here, as we agree with Mayo’s identification of the natural law.

This implies that a more detailed claim reciting the sequence of chemical manipulations needed to provoke the natural binding may be patent eligible.

In CellzDirect, claims were found eligible because they recited the specific steps of freezing, separating and refreezing cells to preserve them, thus resulting in a practical application of the natural law. Adding such process details, where they themselves are inventive (or at least non-traditional), may result in an eligible claim.

However, such pathways to eligibility do not answer the fundamental question: if the diagnostic invention is actually a discovery combined with standard techniques for observing the result of such a discovery, does it still warrant patent protection? Article I(8)(8) of the Constitution provides “inventors” with the exclusive right to their “discoveries”. What level of human intervention in observation of a discovery is needed to be considered a patent-eligible inventive concept? For now, the courts are clear: once a natural correlation is known, use of a standard detection technique cannot make observation of that correlation eligible for patenting. Any other answer is in the hands of the US Congress.


Get unlimited access to all IAM content