Patentable subject matter at the Supreme Court: An exceptional case?

The following contribution to our gene patenting symposium comes from Arti K. Rai, Elvin R. Latty Professor of Law, Duke Law School and Duke Institute for Genome Sciences & Policy.  From 2009-2010, Rai served as the USPTO Administrator for Policy and External Affairs.

In three of the past four Terms, the Supreme Court has taken a case addressing what constitutes subject matter eligible for patenting.  Association for Molecular Pathology  v. Myriad Genetics, which will be argued in April, is the third of a trio that also includes Bilski v. Kappos (decided in 2010) and Mayo Collaborative Services v. Prometheus Laboratories (decided in 2012).  In these subject matter cases, the Court may be defying some of the trends that the Court itself has established over the last fifteen years, as it has stepped up review of the Federal Circuit.  These trends include deference to the executive branch; a preference for standards over bright-line rules; and some attentiveness to the retroactive effects of decisions that change existing law significantly.  Predicting what will happen in Association for Molecular Pathology v. Myriad Genetics therefore requires analyzing not only the Court’s patent jurisprudence as a whole but also its specific interest in subject matter eligibility.  Ideally, the Court will, in deciding the Myriad case, draw less upon its subject matter jurisprudence and more upon its general patent jurisprudence.

In patent cases before the Court, the executive branch has almost always asserted its views, either as a party or an amicus.  One pervasive theme of the Court’s patent jurisprudence has been solicitude for these views. This solicitude has been particularly evident where the executive branch has disagreed with the Federal Circuit.  Since 1996, in ten of the twelve patent cases in which the Federal Circuit and the executive branch have been at odds, the Court has agreed with the executive branch.  One inference that could be drawn is that the Court, confronted with technical and abstruse legal disputes, is generally more comfortable with the judgment of specialized executive branch agencies (including but certainly not limited to the USPTO) than with the judgment of a specialized court.

In the context of patentable subject matter, however, the Court appears to be less willing to show deference to the executive branch.  In the Bilski case, a divided Court observed that although the test advocated by the government – that a process claim include a “machine or transformation” to be patent-eligible – was a “useful clue,” it was not definitive. More dramatically, in Prometheus, a unanimous Court squarely rejected the government’s view that patent statute requirements other than subject matter, such as novelty, non-obviousness, and adequate disclosure, should do the heavy lifting in determining whether a process represents an unpatentable “law of nature.”  Instead, the Court took the position that processes that encompass laws of nature are not patent-eligible subject matter unless they contain an “inventive concept” in the application of the natural law.

The Prometheus case also defied the Court’s general preference for patent standards over patent rules.   In recent blockbuster patent cases ranging from KSR International v. Teleflex (rejecting a rigid rule for determining what constitutes obviousness) to eBay v. MercExchange (rejecting a rule granting automatic injunctive relief in cases where a patent had been proved valid and infringed), the Court has railed against the poor fit between rigid rules and the policy-laden tests at the heart of the patent system.  In Prometheus, the Court similarly noted that the subject matter test is rooted in a policy concern that patents covering broad subject matter will “preempt” too much future research.  Yet the Court explicitly chose to adopt a bright-line rule against the patenting of any law of nature, whether broad or narrow, arguing that courts don’t have the institutional competence to distinguish between broad and narrow natural laws.  While concerns over institutional competence are of course an eminently legitimate reason for preferring rules over standards, such concerns have not, in the past, been at the forefront of the Court’s reasoning. 

Finally, in many of its patent cases, the Court has paid attention to arguments about settled expectations.  Indeed, in some cases it may have paid too much attention to such arguments.  For example, in i4i v. Microsoft, the Court invoked settled expectations to uphold a “clear and convincing” evidentiary requirement for challengers that seek to invalidate granted patents, even in cases where the challengers present evidence not seen by the USPTO.  By contrast, in Prometheus, the Court did not credit arguments about the potential for invalidating large numbers of existing diagnostic method patents.  To the contrary, it framed the baseline as one where patents of this sort did not exist and thus granting such patents would represent a “new protective rule” for one field over another. 

Similarly, in Myriad, the Court may not be willing to credit the retroactivity argument made by Judge Kimberly Moore in the decision of the Federal Circuit below – that is, we should accept human gene patents because tens of thousands of them exist and have existed for decades.  Additionally, the Court may feel no need to defer to the government’s position.  To the contrary, it may deem the government’s position in the case, which draws subtle, and policy-laden, distinctions between different types of human DNA, to be insufficiently bright-line.   

Like the Court’s decision in i4i v. Microsoft, Judge Moore’s position may be too attentive to settled expectations.  Where existing patent rights have significant potential to hinder innovation, settled expectations alone aren’t a reason to keep them.  By contrast, the government’s position – also adopted by Judge William Bryson in his dissent below – is worthy of the Court’s serious attention.  The government draws a distinction between claims to DNA that are merely isolated – that is, excised from their original cellular environment and other genomic material – and claims to sequences that are laboratory-generated duplicates of the select portions of the DNA sequence that actually code for a protein.  According to the government, the former set of claims, to so-called genomic DNA (gDNA), cover a product of nature, while the latter set of claims, to full-length coding DNA (cDNA), cover a man-made patentable construct. 

From a formal scientific or legal perspective, the government’s distinction is not entirely compelling.  However, as I discuss at length in a 2012 Duke Law Journal essay, the distinction tracks reasonably well some relevant economic policy considerations.   Full-length cDNA claims are much less likely to cover future research than are gDNA claims.  Conversely, while the affirmative economic case for gDNA claims is highly contested, full-length cDNA claims are typically associated with therapeutic end-product molecules known as biologics.  For such end products, patent protection is clearly a significant spur to innovation (though arguably less important than various data exclusivity and regulatory protections).

Judge Bryson’s dissent adopting the government’s position notes this economic policy rationale.  In drawing the line at full-length cDNA patents, Bryson cites a prominent report written by an advisory committee to the Secretary of Health and Human Services that discusses the potential for patent thickets created by patents on gDNA. 

Judge Bryson and the government also adopt a reasonable approach to the issue of settled expectations.  While eliminating cDNA-related patent rights in therapies would require industry to absorb a very dramatic shock, the disruption caused by the government’s position would be less significant. 

In an ideal world, the issue of how much human manipulation is necessary to make genetic material patent-eligible would have been addressed more fully decades ago.  As the Duke Law Journal essay explains, the issue’s prolonged gestation testifies to certain institutional deficiencies in the patent system.  The Supreme Court, faced with the question late in the day, has only imperfect choices before it.  In this case no less than many others the Court has taken, the government’s imperfect position makes the greatest policy sense. We should therefore hope that Myriad falls within the mainstream of the Supreme Court’s patent jurisprudence and does not represent an “exceptional case.”