Australian appeals court upholds patents on isolated BRCA1 DNA
Robert Cook-Deegan, MD
On September 5, the Federal Court of Australia (the appeals court) upheld a claim on isolated DNA from the BRCA1 gene. It dismissed Yvonne D’Arcy’s appeal of a case that has attracted international attention. Australian patent 686,004 has never been enforced, so the court decision has little real-world concrete impact. As Richard Gold and Julia Carbone explained in their classic case study, “Myriad Genetics: In the Eye of the Policy Storm,” the patent rights on BRCA1 and BRCA2 were exclusively licensed for use in Australia and New Zealand to Genetic Technologies, Ltd. (GTG), which in turn made them a “gift to the people of Australia.” When the CEO of GTG proposed taking back that gift in the summer of 2008, he provoked a firestorm and the company backed down in October, restating that it would not enforce its patent rights against laboratories offering BRCA testing. The Australian Senate held a series of hearings, and a bill proscribing DNA sequence patents was proposed, but the new government opposed it, and it lapsed. Instead, Australia enacted patent reforms in 2012 that raised the bar for utility and clarified the Australian law’s exemption from infringement liability for research and regulatory approval. Most of the provisions of that law took effect on April 15, 2013, the very day Association for Molecular Pathology v Myriad Genetics (AMP v Myriad) was argued before the U.S. Supreme Court.
It is ironic that in the United States, the only jurisdiction in which BRCA patents have been enforced, the courts invalidated Myriad’s claims to DNA molecules whose sequence is found in nature as well as most of its method claims (except one method claim on use of BRCA as a chemotherapeutic assay in drug discovery, not relevant to diagnostic testing). In Australia, where BRCA patents have never been enforced, claims on isolated DNA molecules have now been upheld at the appellate level, affirming Judge Nicholas’s trial court decision of February 2013.
In the United States, Myriad began to sue competitors that offered BRCA testing in July 2013, less than a month after the Supreme Court decision of June 13. These suits are based on claims not challenged in the case that went to the Supreme Court (with the exception of one claim on 15-base pair segments of BRCA1 DNA that was challenged but whose validity is unclear under the Supreme Court ruling). Myriad has sued eight competitors. One such competitor (Gene by Gene) settled in February 2014. Pre-trial proceedings for the other cases have been consolidated in the Utah federal District Court and assigned to Judge Robert Shelby. The competitor companies currently in litigation are Ambry Genetics, Counsyl, GeneDx, Invitae, LabCorp, Pathway Genomics, and Quest Laboratories. Several of the firms have petitioned for declaratory judgment of non-infringement, and one (GeneDx) has challenged eleven of Myriad’s patents through a request for Inter Partes re-examination by the U.S. Patent and Trademark Office. Judge Shelby denied Myriad’s request for a preliminary injunction, which means that the competitors can continue to offer testing, although they will be liable for damages if they are found to infringe valid patent claims. They could also be subject to a permanent injunction at the conclusion of the case.
The Australian appeals court’s ruling was a unanimous decision of the five-judge panel, which included the chief judge. Australian court decisions are not “signed” by particular judges, unlike U.S. practice, but Judge Annabelle Bennett is widely regarded as the source of the key arguments, based in part on her background in biotech patent law and her Ph.D. in biochemistry from the University of Sydney. She formerly advised the Australian Law Reform Commission in its report on gene patenting.
The Australian ruling runs to fifty pages. Ten of those pages are a primer on DNA biochemistry. The core of the legal reasoning, however, clearly states that a distinction between invention and discovery is not a fruitful conceptual framework for patent law, and quite explicitly rejects the U.S. Supreme Court’s arguments in AMP v Myriad. The opinion lauds Judge Lourie of the U.S. Court of Appeals for the Federal Circuit, who upheld Myriad’s claims on isolated DNA molecules in two majority opinions that were unanimously reversed by the U.S. Supreme Court.
There are several reasons for the difference between the Australian and U.S. court decisions. One is Australian legal precedent. Australia’s 1990 patent statute rests squarely on Section 6 of the 1624 English Statute of Monopolies. Under Australian law, patents are available for an invention that:
(a) is a manner of manufacture within the meaning of section 6 of the Statute of Monopolies; and
(b) when compared with the prior art base as it existed before the priority date of that claim:
(i) is novel; and
(ii) involves an inventive step; and
(c) is useful…
The key Australian case for interpreting “manner of manufacture within the meaning of Section 6” is a 1959 case called National Research Development Corporation v Commissioner of Patents that requires that an invention apply to an “artificially created state of affairs.” The High Court (Australia’s equivalent of the Supreme Court) held in that case that “the distinction between discovery and invention is not precise enough to be other than misleading” and set a low bar for patent eligibility. This criterion in effect means that if “we did it in our lab” it will clear the threshold for patent eligibility. Given this precedent, the Federal Court’s ruling is sensible. This did not stop the court from expanding into dicta, and going out of its way to chide and take issue with the U.S. Supreme Court for its line of reasoning.
The Australian Federal Court dismissed the concern for blocking access to genetic information by asserting that the patent claims molecules, not information. The court reasonably noted that the information about BRCA sequences is now known and can be used without infringement. The published sequences can be copied and used freely. It failed to acknowledge, however, that by granting exclusive rights to isolated DNA molecules it thereby blocked determination of the DNA sequence of—and thus the information stored in—any as yet-unsequenced DNA molecule. No DNA will ever be sequenced without being isolated. Isolation is by definition an attribute of a DNA molecule whose sequence is being determined. Such sequencing entails the hand of man, but the sequence also comprises information that was and always will be “found” from a sample to be analyzed, not invented by the team that first discovered the gene. DNA is the storage and transmission medium of genetic information, and by granting patent rights to isolated DNA molecules, the decision means that any study or analysis of a natural gene by a human being will infringe such patent claims. The Federal Court of Australia was content to note that the information itself is not patented, but this is not the killer argument that the Court presumes. Those who view DNA as a unique informational molecule will not be persuaded that the Australian court has fully appreciated the consequences of its focus on DNA as a chemical without acknowledging its informational role. Indeed, the Australian court seems not to have dismissed the U.S. Supreme Court’s concern, but rather to have entirely missed it.
The Myriad/Utah team cloned and sequenced the BRCA1 gene and identified several cancer-associated mutations in it. This was hard work and they won a hard-fought global race to find a gene known to be residing on chromosome 17, set off by Mary-Claire King’s landmark 1990 publication establishing linkage to a cancer susceptibility locus she named BRCA1. In discovering the gene, they also invented (or discovered—here there is little need for a distinction) a way to detect sequence variations for purposes of research or diagnosis. By the logic of patent law, they were entitled to a patent right. Moreover, as the Federal Court of Australia explicitly pointed out, they also provided sufficient information (if one grants the right to correct the many sequence errors in their initial sequence) to produce the BRCA1 protein. One does not need more than the sequence specified in the patent application to do that, and if the BRCA protein had been an effective treatment, this could have been a potentially lucrative patent analogous to patents on genes for making insulin, growth hormone, erythropoietin or other therapeutic proteins.
In finding the gene, the Myriad/Utah team found a way to detect sequence variations, and that is an invention. But they claimed far more than they invented by claiming the DNA molecules—both the ones they actually characterized and those yet to be discovered. This includes either DNA extracted from samples or made into copies. The Australian court and most commentators have focused on how the patent right should extend to DNA molecules themselves—and in some contexts that is undoubtedly true. But by allowing claims on isolated DNA molecules, the court thereby caused a serious problem of pre-emption of future research, because of the informational role of DNA.
Many molecules have been patented that are found in nature: adrenaline, antibiotics, vitamins, hormones and growth factors, vaccines, and other products. In those cases, the molecules are put back into bodies to change them, and it is natural to conclude those result from human intervention and “invention” in a meaningful sense. When doing a diagnosis or conducting an experiment, however, the “isolated” DNA molecule being sequenced is not useful in itself, but is only an informational intermediary. It is useful if and only if the “isolated” DNA exactly copies the information content of the sample DNA. If the “isolated” DNA differs from the sample in its information content, it is useless as a diagnostic test or scientific method; it is a sequencing error. The sample DNA, however, is not invented by anyone; it is found in nature. According to the U.S. Supreme Court, it should not be claimed in a patent.
Here the distinction between discovery and invention does hard work. The DNA sequence of the sample to be analyzed cannot be known in advance, and the sequence itself is not the result of human work; for purposes of diagnosis, the DNA molecule is simply the embodiment of the sequence information, and the purpose of the test is to capture the information via the DNA sequence, not to obtain the DNA molecule itself. The crucial input for diagnosis or scientific inquiry is sample DNA but the crucial output is information. There is thus a two-pronged test to distinguish claiming an isolated DNA molecule to direct production of a protein, which is patentable, from claiming the molecule when the output is information, where the U.S. and Australian courts disagree about whether it is patentable. Does the DNA molecule’s claimed sequence reside in nature and is it, by definition, not known in advance? The very purpose of a genetic test is to extract information residing in the sample DNA and to resolve uncertainty about whether it harbors a deleterious mutation. That is, the input is sample DNA of unknown sequence and the output is information; the DNA molecule is merely an intermediary. The second distinction is that the information is the only product; whereas for gene patents on DNA encoding therapeutic proteins (or for patent claims on antibiotics, vitamins, vaccines, hormones, growth factors and other patented biological molecules found in nature), the end product is a molecule that will be put into someone’s body in order to change its biology.
The Federal Court of Australia is persuasive that Myriad invented something useful, but unpersuasive that its claims should be as broad and should extend to naturally occurring sequences. By granting the rights claimed, the Federal Court of Australia has in effect blocked any way of getting access to that information by what truly does seem to be a “lawyer’s trick,” the use of the magical word “isolated.” The court has extended an exclusive right to the study of DNA; any means of determining the sequence of a DNA molecule containing one of the claimed mutations from a person’s cells will infringe the claim during the term of the patent. The Myriad/Utah team discovered the gene, and thereby invented something that is new, useful, inventive, and commercially valuable. But they did not invent everything that they claim. This is not a case of rewarding or failing to reward the act of invention, but of precision in claiming what has actually been invented—and what has not.
Sometimes discovery and invention are close synonyms, and sometimes they are not. By rejecting any distinction between invention and discovery across the board, the court has bred more confusion than it avoided. It has failed to acknowledge a situation in which the distinction between invention and discovery can do hard legal work. The Myriad/Utah team did indeed do something valuable and inventive, but they did not invent the molecules claimed as:
An isolated nucleic acid coding for a mutant or polymorphic BRCA1 polypeptide, said nucleic acid containing in comparison to the BRCA1 polypeptide encoding sequence set forth in SEQ.ID No:1 one or more mutations or polymorphisms selected from the mutations set forth in Tables 12, 12A and 14 and the polymorphisms set forth in Tables 18 and 19.
They characterized those molecules by making laboratory copies, but they did not invent them. The DNA molecules so described encode a mutant BRCA1 protein. And the Myriad/Utah team thus did indeed invent a way to make the gene in a laboratory and turn it into protein, and they can claim that invention as a process. Or they can claim specified DNA molecules that do not reach to all ways of determining DNA sequence. But it is not clear they should be able to claim DNA molecules they discovered and did not invent. Yet the scope of the claim encompasses such molecules in addition to those they actually did make and describe in their patent. The Australian court has said yes, they can claim them; the U.S. Supreme Court has said the opposite.
In its zeal to protect patents on the DNA that can produce the BRCA1 protein, the Australian court passed over in silence how DNA molecules are used for diagnosis and when doing research. When doing a diagnosis of a sample sent for analysis or when doing research, the molecules created in the laboratory’s “artificial state of affairs” come not from Myriad but depend entirely on the sample DNA being analyzed. That DNA was not invented by the Myriad/Utah scientists. Both Judge Nicholas and the Federal Court of Australia argued that the DNA molecules that Myriad made in the lab when discovering BRCA1 are different from those in the body. By the same logic, then, the “isolated” molecules Myriad claims are also necessarily different from the natural DNA in cells and bodies that are the subject of diagnostic testing or research. The molecules are by definition being created de novo for each test or experiment, based on a natural template. Should they fall under the scope of Myriad’s claims? Whether the discoverers of the gene should be able to claim the molecules copied from the natural source is not an obvious or easy decision. It is certainly not the slam dunk argument the Federal Court of Australia took it to be.
Infringing uses of the claim as granted include diagnosis and also research. In Australia, some research will infringe, but be exempt from infringement liability because of the research exemption that was clarified in Australia’s 2012 statute. The U.S. has no such brake on patent rights, however. Since the Federal Circuit’s decision in Madey v Duke University (2002), the absence of a meaningful research exemption in U.S. law has been apparent. Any research that has practical uses, even nonprofit research at academic institutions, infringes patent claims. Only research that is purely for satisfying curiosity has a chance of escaping infringement liability in the United States.
The U.S. Supreme Court was therefore quite right to be worried that exclusive rights can be enforced against research. The vast majority of the thousands of papers published on BRCA1 and BRCA2 infringe the kind of patent claims upheld by the Australian court. Myriad initiated a suit against the University of Pennsylvania when it was going to be a BRCA testing hub for National Cancer Institute grants that would involve BRCA testing in research, so there is precedent for being concerned about infringement liability in research. While the Australian Federal Court gave short shrift to pre-empting research, it was a major concern of the U.S. Supreme Court. Myriad’s pledge not to enforce patents against noncommercial research—a pledge made just as it filed suits against competitor commercial testing laboratories in July 2013—will give cold comfort, since enforcement is entirely at the discretion of the patent-holder, not a matter of law.
The next step of appeal, to the Australian High Court, is subject to the discretion of that Court. The case that was heard by the U.S. Supreme Court came with a very different history. U.S. BRCA patents had been entirely invalidated by Judge Robert Sweet at the district court level, and the Federal Circuit had reversed him in a 2-1 split decision with a dissenting opinion. Moreover, the U.S. Solicitor General weighed in, and urged both the Court of Appeals and the Supreme Court to uphold patents on engineered DNA, but to consider DNA molecules corresponding to sequences that can be found in nature to be unpatentable subject matter. The U.S. Supreme Court thus faced a CAFC reversal of a district court decision, a split within the CAFC, and a Solicitor General amicus brief that tracked the CAFC dissent, not the majority. The U.S. Supreme Court’s intervention thus addressed mixed verdicts in lower courts and disagreement within the executive branch (between the Department of Justice, as represented by the Solicitor General, and the Patent Office, which had allowed the patent in the first place). In Australia, in contrast, both the trial court and the appeals court have upheld patents on isolated DNA molecules, so there is less legal uncertainty for the High Court to resolve. And there is little real-world impact because the patents have never been enforced, will soon expire, and the Australian licensee has just sold its rights to molecular diagnostic testing in Australia.
That said, the High Court could revisit its 1959 summary rejection of a distinction between invention and discovery. As Chris Dent notes in his history of the 1624 Statute of Monopolies, the law is anything but precise. It was a result of compromise and probably also deliberate ambiguity. The lexical precision (some would say pseudo-precision) of the Federal Court of Australia’s ruling confronts a long history of balancing and ambiguity in patent law. The Australian precedent relies on a 55-year old case that gave an expansive definition of patent-eligibility and obliterated the invention-discovery distinction. Many patent lawyers and scholars have since argued that the invention-discovery distinction is outdated and introduces needless uncertainty into the legal determinations of patent-eligibility. Many argue that novelty, utility, non-obviousness (or inventive step), enablement, and adequate written description can do the work of weeding out bad patent claims with greater clarity and precision than standards for deciding what kinds of things can be patented at all—the doctrine of patentable subject matter. The 1952 U.S. patent statute and dicta in the landmark Diamond v Chakrabarty case of 1980 both affirmed that patents could be sought on “everything under the Sun that is made by man,” with the emphasis on “anything” more than on “made by man” (ignoring both the archaic sexism and the clear invocation of invention versus discovery). The U.S. Supreme Court has repeatedly rejected this expansive interpretation of the U.S. Patent Act, most recently in at least one case per year for the past four years (Bilski, Mayo, Myriad, and Alice). The Australian courts to date have not walked back, and so distinguishing invention from discovery is not the basis of case law there. There are, however, several “escape clauses” in the 1624 Act that have been little exercised in modern jurisprudence, including a rejection of patents when they are “generally inconvenient.” The invention-discovery distinction could still lie dormant and the High Court could bring it back to life should it choose to do so.
The High Court of Australia may or may not accept the promised appeal. If it does not, then U.S. and Australian law will have quite different jurisprudence for DNA-sequence-based patents for the foreseeable future. The politics and consequences of this discrepancy between patent jurisdictions are only partially predictable.
The drumbeat for lobbying the U.S. Congress to “fix” the results from the Supreme Court’s decisions in both Myriad and Mayo v. Prometheus (2012) is getting louder. Some are arguing that the international Trade-Related Aspects of Intellectual Property agreement (TRIPs) might compel the United States to adopt laws that interpret patent criteria in the same way that the Federal Court of Australian and the European Biotechnology Directive of 1998 do, explicitly permitting patents on DNA molecules whose sequences correspond to those found in nature.
Those contemplating a statutory fix might, however, be chastened by the changing politics of intellectual property. It was mainly patent lawyers and industry leaders who tweaked statutory definitions in the Biotechnology Process Patent Protection Act of 1995. A few years later, in 1999-2001, the Patent Office tweaked its examination guidelines for utility and written description in a process that CAFC Judge Bryson characterized as “cursory” in his Myriad dissent. Those changes did not elicit strong responses from cancer advocacy organizations, the American Civil Liberties Union, or high-tech open-source advocates. Today, disease advocacy groups, cancer patients, scientists, civil libertarians, and high-tech open-source advocates are potent and engaged. They will surely oppose backroom deals to expand gene patent rights; it will not be an inside game. The politics of patent law are getting more complicated, noisier, more contentious, and involve broader constituencies than the patent bar is accustomed to encountering.
In Australia, rights on gene patents have not been enforced against diagnostic or research use, and yet the courts have upheld patents on isolated DNA molecules whose sequences would be found in nature; in the United States, rights have been enforced, but patent rights have been weakened and constituencies will be fully engaged and attentive if efforts are made to restore them. Battles in Congress will not be quiet or quick, so picking battles may require discretion. The big winners in this cauldron of confusion are patent scholars who have plenty to write about. And patent lawyers who can make hay 24 hours a day, because the sun is always shining on a jurisdiction with uncertain patent rights.
Robert Cook-Deegan, MD, is Research Professor at the Sanford School of Public Policy, Duke University, with appointments also in Medicine and Biology.