A couple of years ago I posted an article (available here) discussing the case of Monsanto v. Cefetra. Essentially, in that case a European court held that Monsanto patents claiming the gene responsible for its Roundup Ready trait were not infringed by the importation of soy meal containing the gene, because the grinding of the soybeans to make the meal rendered the DNA incapable of expressing the encoded protein, and thus unprotectable pursuant to the 1998 European Union Directive on biotechnology.
On Tuesday, July 6, that decision was affirmed by the Court of Justice of the European Union. This is the highest judicial body of the European Union, so there will be no further appeals, and this would appear to be the final word on the subject. The decision was reported in an article posted on IPKat, a respected European IP blog, available here. I am no expert on European law, but here are a few thoughts on the significance of this case.
First, my interpretation of the decision.
In a nutshell, the Court of Justice held that although the so-called Biotech Directive, promulgated by the European Union in 1998, permits the patenting of naturally occurring DNA sequences, the scope of patent protection only extends to products incorporating the DNA if the DNA is capable of performing “the function for which it was patented." Applying this criterion to the facts of the case, the Court held that the patentable function of the gene at issue was to confer glyphosate (i.e., RoundUp) resistance upon a plant, and that since the soy meal is dead and thus incapable of expressing this function, the DNA residing in the soy meal is ineligible for patent protection.
The Court went on to hold that the Biotech Directive supersedes national law, and precludes individual European nations from enacting legislation that would permit patenting of DNA sequence per se. It also held that the prohibition against patenting DNA sequences per se applies retroactively to patents predating the EU’s adoption of Directive in 1998. In short, this unappealable ruling appears to be binding upon all 27 European Union Member States, and all DNA patents, no matter when they were issued.
The ruling could have a substantial impact on biotechnology. Obviously, it provides an opening for growers to circumvent gene patents used to protect genetically modified crops by growing the crops in a country where the gene is not patented, such as Argentina in this case, or in a jurisdiction with weak patent enforcement, and then importing the product into a European Union member wherein the patent is in force with impunity.
According to the logic of the decision, if viable seeds were imported, they would likely be found infringing because the DNA would still be capable of performing the function for which it was patented. But if the product has been processed, and as a result no longer viable, as exemplified by the soy meal at issue in this case, importation should not be found to constitute infringement.
The ruling could also have implications for the scope of protection afforded by gene patents for genetic diagnostic testing. Genetic diagnostic testing, such as by DNA sequencing, typically involves the isolation and/or amplification of the gene sequence being tested. This isolation and amplification might be characterized as the manufacture of the DNA, which could constitute infringement of a patent claiming the DNA sequence. For example, the ACLU challenge to Myriad’s BRCA gene patents is based in part on an assumption that the patent claims reciting isolated BRCA sequences would be infringed by unauthorized testing for mutations in the BRCA genes (which typically involves DNA sequencing).
However, this assumption is premised on the patents covering the isolated DNA sequences per se. But arguably, the isolated and amplified DNA fragments generated in the course of DNA sequencing and testing are not capable of performing their function, and thus after this week’s Court of Justice decision are ineligible for patent coverage in the European Union.
Of course, this would depend upon how one defines the "function of the patented gene. For example, one might argue that one important function of BRCA sequences claimed in Myriad’s patents is for diagnostic testing, and hence isolated DNA sequences generated in the course of genetic testing are performing a function for which they are patented
However, under a more restrictive view the function of the BRCA genetic sequences is to code for the BRCA protein product, i.e., the actual physiological function of the BRCA gene in the human body. Myriad’s patents specifically discuss use of the isolated BRCA gene for production of the BRCA protein in a recombinant cell, which could be a useful research tool in drug discovery. As discussed in earlier posts relating to the ACLU lawsuit, I point out that in practice BRCA testing does not involve the amplification of the full length BRCA coding sequence, but only amplicons representing fragments of the gene. These fragments would not be capable of achieving the function of expressing the full-length protein, and thus arguably lack the requisite functional potential.
I think reasonable minds could go either way on the issue, but given the current level of antipathy towards the use of gene patents to block genetic diagnostic testing, I can envision a European court adopting a restrictive interpretation of what it means for a genetic sequence to be capable of performing its patentable function, and decide that gene patent coverage does not extend to use of the genetic sequence in diagnostic testing.
In a related context, the Court’s decision could have implications for the effect of gene patents on whole genome sequencing. There is much buzz these days regarding the imminent arrival of personal whole genome sequencing, which will permit individuals to have their entire genome sequenced for a relatively small amount of money. Some have worried that a thicket of gene patents might impede the ability of firms to provide whole genome sequencing. However, DNA “manufactured” in the course of whole genome sequencing is unlikely to be found capable of performing the function of individual patented genetic sequences residing therein, in which case genome sequencing would not result in patent infringement liability in Europe.
I personally do not think gene patents will substantially impede whole genome sequencing, in the US or elsewhere. But if I am wrong, and patents on genes pose a substantial impediment to whole genome sequencing in the US, the patents could be circumvented by exporting the process to a place like Europe. The EU court decision could substantially reduce the likelihood that whole genome sequencing in Europe will result in liability for patent infringement.
On the other hand, the ruling should not affect the ability of biotechnology companies to use gene patents to protect their biologic drugs. In this context, the patented gene is performing its primary function of expressing the protein encoded by the gene, and thus satisfies the criterion established by Tuesday's decision of the Court of Justice.
Friday, July 9, 2010
Saturday, July 3, 2010
The Impact of Bilski on Biotechnology
Prior to the Federal Circuit’s 2009 en banc decision in In re Bilski (Bilski I), the Supreme Court had established a test for patent eligibility that was quite straightforward in the abstract, albeit fraught with uncertainty as to how the test should be applied in practice. In a series of decisions, mostly dating from the 1970s and early 1980s, the Court established that essentially any man-made process or product is eligible for patent protection so long as the patent claims do not cover what the court characterized as a "fundamental principle." In these decisions, the Court identified a number of specific categories of patent ineligible fundamental principles, including principles of nature, natural phenomena, abstract ideas and mental processes.
As a practical matter, up until a few years ago patent eligibility was rarely an issue for those attempting to patent inventions relating to biotechnology. So long as there was some element of human intervention, such as isolation of a gene sequence or genetic engineering of a living organism, biotechnology inventions were generally assumed to be eligible for patent protection. Patenting of biotechnology flourished, notwithstanding the protests of many who argued that some inventions based on fundamental biological discoveries, particularly gene patents and patents broadly claiming the practical exploitation of fundamental biological discoveries (such as physiological correlations useful in the diagnosis and treatment of disease), or unsuited for patent protection and should be treated as patent ineligible.
In Bilski I, a majority of the judges on the Federal Circuit badly misinterpreted Supreme Court precedent and declared the so-called "machine or transformation" test to be the universal and exclusive criterion for determining the patent eligibility of all process claims, including by implication processes relating to biotechnology. In essence, Bilski I shifted the patent eligibility inquiry from whether a claimed biotechnology process covered a fundamental principle (i.e., a natural phenomenon) to whether the process entailed sufficient involvement of a machine or transformation
The machine or transformation test was subsequently used as the basis for invalidating a number of biotechnology process claims, including Myriad’s claims to methods of detecting mutations in BRCA genes. The PTO also began to employ the machine or transformation test to reject biotechnology claims, particularly those relating to diagnostics and personalized medicine. Some commentators, including myself, expressed concern that the machine or transformation test, as mandated by Bilski I, threatened to seriously undermine the ability of companies and researchers engaged in the discovery and commercialization of diagnostics and personalized medicine to obtain adequate patent protection for their inventions, although these concerns were to some extent allayed by the Federal Circuit’s restrained application of the test to a personalized medicine patent claim in Prometheus v. Mayo.
Fortunately, the Supreme Court agreed to hear an appeal of Bilski I, and last Monday the Court decided the case, formally titled Bilski v. Kappos (Bilski II), in a manner that effectively rejected the Federal Circuit's interpretation of the machine or transformation test. The Court held that while the involvement of a machine or transformation in a claimed process is often informative in assessing patent eligibility, it is not a general prerequisite for patent eligibility, and admonished the Federal Circuit that the test for patent eligibility remains whether or not the patent claims encompass a fundamental principle.
In essence, the Supreme Court has turned back the clock, negating Bilski I and the Federal Circuit's ill-conceived experiment with the machine or transformation test. Significantly, however, the Supreme Court provided little if any guidance with respect to what it means for a patent to claim a fundamental principle, and absolutely no guidance with respect to how to apply the test when the fundamental principle is a biological natural phenomenon, leaving it to the lower courts to sort out this important issue
I predict that the issue of patent eligibility for biotechnology inventions, both products and processes, will continue to be actively litigated well into the foreseeable future, as courts and the patent office struggle to figure out just what exactly it means for a patent claim to cover a biological natural phenomenon. In this article, I summarize important developments in the patent eligibility doctrine leading up to and including Bilski II, particularly as they relate to biotechnology, and discuss the issues relating to patent eligibility that will need to be addressed in a post-Bilski world.
Patent Eligibility Prior to LabCorp
There are four requirements of patentability which an invention must satisfy in order to be patentable : patent eligibility, practical utility, novelty and nonobviousness. The first, patent eligibility, has been characterized as the threshold criterion, used to exclude purported inventions that are useful and innovative, but nonetheless unsuited for and hence ineligible for patent protection.
Prior to the advent of modern biotechnology, some biological subject matter, particularly living organisms, were generally considered patent ineligible. This explains why Congress saw fit to enact legislation explicitly creating alternate forms of intellectual property protection for plants (plant patents and plant variety protection certificates). However, with the increasing commercial significance of biotechnology in the 1970s, the question of whether living organisms and other biotechnological inventions of were eligible for patent protection became of increasing importance, and in 1980 the Supreme Court answer to the question in the affirmative in its landmark Diamond v. Chakrabarty decision.
In Chakrabarty, the Supreme Court essentially held that any man-made product or process is patentable, so long as the patent does not purport to claim a "fundamental principle." In Chakrabarty and other Supreme Court decisions relating to the doctrine of patent eligibility, the Court has at various times identify specific categories of fundamental principles, including natural phenomena, physical phenomena, principles of nature, abstract ideas, and mental processes.
As a practical matter, after Chakrabarty patent eligibility was not a major issue for biotechnology. So long as there was some degree of human intervention, biotechnology inventions were generally deemed patent eligible by the patent office and the courts. While living organisms and other biological organisms could not be patented as they exist in nature, genetic engineering of an organism was deemed sufficient to render the non-naturally occurring modified organism patentable. Similarly, isolation of a naturally occurring DNA sequence was deemed sufficient human intervention to render the isolated molecule patentable.
It was not until the Supreme Court granted certiorari in LabCorp v. Metabolite in 2005 that patent eligibility once again took on significance in the context of biotechnology. The patent at issue in lab were was based upon the discovery of a natural correlation between the level of total homocysteine (a naturally occurring metabolite) in the human body and existence of a vitamin B deficiency. Based on this discovery, the patent broadly claimed any method of detecting a vitamin B deficiency that involves assaying a patient's blood for total homocysteine content and correlating elevated total homocysteine with such a deficiency. Although the issue of patent eligibility was never directly addressed in the lower courts, the Supreme Court agreed to hear the case specifically to address the question of whether or not the patent violated the prohibition against the patenting of fundamental principles, i.e., natural phenomena, principles of nature and abstract ideas.
Upon further consideration, however, the Supreme Court had second thoughts about taking the case, and dismissed the LabCorp appeal without deciding it. However, three of the Justices signed on to a dissenting opinion penned by Justice Breyer stating that they considered the claimed process patent ineligible for embodying the natural phenomenon of a correlation between total homocysteine and vitamin B deficiency. Although this dissenting opinion by three of the nine justices has absolutely no binding legal authority, it has proven highly significant because at least established the viability of asserting the doctrine of patent eligibility to challenge patents based on fundamental biological discoveries, particularly patents broadly encompassing the use of natural correlations in a diagnostic method, which had become increasingly common with the emergence of genetic diagnostic testing and personalized medicine.
Patent eligibility after LabCorp
No doubt encouraged by Justice Breyer's dissent, accused infringers began raising the defense of patent ineligibility in patent infringement litigation. Three of these cases, Ariad v. Lilly, Classen v. Biogen, and Prometheus v. Mayo, have been extensively discussed in previous articles on my blog. Significantly, in all three of these cases the case for patent ineligibility was based upon an argument that the challenged claims effectively preempted all practical uses of a biological natural phenomenon, essentially the same rationale propounded by the dissenting justices in LabCorp.
These patent ineligibility-based challenges met with mixed success at the district court level. In Ariad, the district court rejected the challenge, based on a rationale that frankly made absolutely no sense to me. In any event, the issue was rendered moot because the court ruled the claims invalid on the alternate grounds that the claims violated the written description requirement, decision that was ultimately affirmed by the en banc Federal Circuit.
However, in Classen and Prometheus the district court found for the accused infringers and held the asserted claims patent ineligible for broadly preempting the practical exploitation of biological natural phenomena. The patent owners both appealed their cases to the Federal Circuit, but before the Federal Circuit heard the cases issued its en banc decision in Bilski I, which effectively changed the test for patent eligibility, as discussed in the next section.
Patent eligibility in the time between Bilski I and Bilski II
In Bilski I, the Federal Circuit essentially ignored Supreme Court precedent establishing that the test for patent eligibility is based on an inquiry into whether the patent claims a fundamental principle, and announced that the test for patent eligibility of any process claim was whether or not the claimed process included sufficient involvement of a machine or transformation. Bilski I was flawed in a number of regards. For one thing, the holding that no process could be patent eligible without substantial involvement of a machine or transformation was entirely inconsistent with Supreme Court precedent. In attempting to justify its holding, the Bilski I majority pointed to Supreme Court precedent suggesting that a process involving a machine or transformation must be patent eligible, and concluded that this meant that a patent eligible process must include a machine or transformation, a fundamental error in logic that effectively transformed a safe harbor for patentees into an absolute prerequisite for patent eligibility.
As I explained in an earlier article, the machine or transformation test is fundamentally inappropriate for assessing patent eligibility of some processes, especially those involving biotechnology. In crafting the test, the court seemed focused on the perceived problem of so-called business method patents and the like, such as the specific patent claims issue in Bilski I, and erred by declaring the test applicable to all process claims, without adequately considering the implications.
Despite its flaws, Bilski I was an en banc decision of the Federal Circuit, and thus controlling law on the subject absent Supreme Court intervention. Even though Classen and Prometheus were both decided at the district court level prior to Bilski I, based on arguments that focused on alleged preemption of biological natural phenomena, by the time the appeals made it to the Federal Circuit the law had shifted, and as a consequence the parties had to change their arguments to focus upon the involvement (or lack thereof) of a machine or transformation in the claimed processes.
The Federal Circuit decided Classen first, and upheld the district court determination, concluding that the claim, which was originally held invalid or in compassing a natural phenomenon, was invalid for failing to satisfy the machine or transformation test. The decision was only a few sentences in length, and provided absolutely no explanation as to the basis for the court’s decision.
The Federal Circuit then issued a lengthy decision in Prometheus, this time reversing the district court and finding the challenged patent claims to be patent eligible. Specifically, the Federal Circuit found that the steps of administering a drug and determining the level of drug metabolites in a patient's body were both sufficiently transformative to satisfy the machine or transformation test.
The Bilski I machine or transformation test was also the basis relied upon by the district court in invalidating the method claims in the ACLU/Myriad gene patent case, as discussed in previous posts to this blog.
Bilksi II
In Bilski II, the Supreme Court identified the flawed logic in Bilski I, and rejected the Federal Circuit’s machine or transformation test, holding that while the presence of a machine or transformation in a process claim can be relevant in assessing the patent eligibility, there is no general requirement that a patent eligible process include a machine or transformation. The Court reiterated that the test for patent eligibility hinges upon whether or not the claim encompasses a fundamental principle, and held that because the claims at issue were directed towards an abstract idea, one of the fundamental principles expressly identified in earlier Supreme Court decisions, they were patent ineligible.
In essence, Bilski II has voided the Bilski I decision and turned back the clock to where we were prior to Bilski I. The patent eligibility of biotechnological inventions is still very much in play, and there will be much for the courts and patent office to sort out over the next years, but the focus will again be upon figuring out what it means for a patent claim to cover a fundamental principle, not whether it involves a machine or transformation.
The Supreme Court has vacated the Federal Circuit decisions in Classen and Prometheus, and sent those cases back to the Federal Circuit to decide based on Bilski II. Since Bilski II is essentially the law prior to Bilski I, this will require the Federal Circuit to consider the arguments that were actually made a district court level which focused upon preemption of biological natural phenomena, the correct test for patent eligibility. To a large extent, the Federal Circuit will be working on a blank slate. Although the Supreme Court has made clear that the determination of patent eligibility should be based on an inquiry into whether the patent claims encompass a biological natural phenomenon, the Court has provided very little specific guidance as to how this test is to be applied in practice.
An alleged failure to comply with the machine or transformation test was the basis for the invalidation of Myriad’s method claims in the ACLU gene patent challenge. These were the broadest and most relevant of the claims challenged in the case. In my view, Myriad’s product claims directed to polynucleotides are either highly susceptible to invalidation based on prior art (this applies to the fragment claims) or clearly not infringed by BRCA testing (the full-length coding sequence claims). I also think it is highly likely that the Federal Circuit will reverse the lower court’s ruling these product claims are patent ineligible.
However, I thought that the Federal Circuit would probably uphold the finding of patent ineligibility for some of Myriad’s method claims broadly claiming the detection of mutations in the BRCA genes. The processes recited in these claims did not seem to require the use of any machine or transformation, and the district court was probably correct in holding that the claims failed the machine or transformation test.
I think it is pretty apparent that if the district court had decided the case prior to Bilski I, or subsequent to Bilski II, it would have found the method claims patent ineligible for preempting natural phenomena. It just so happened that the district court decided the case during the brief tenure of the machine or transformation test, and thus through no fault of its own conduct the analysis based on the incorrect test. It remains to be seen how the claims will whether a patent ineligibility challenge based on the correct test as set forth in Bilski II.
Bilski II will also force the PTO to alter its examination practices. After Bilski I, the PTO began applying the machine or transformation test as a basis for rejecting patent claims, including claims relating to biotechnology methods, particularly with respect to inventions relating to diagnostics and personalized medicine. Since Bilski I was an en banc decision of the Federal Circuit, the PTO was required to implement it, but now that the Supreme Court has rejected Bilski I rejections based solely on the lack of a machine or transformation are improper. The PTO will need to either forgo rejections based upon a patent ineligibility, or if it chooses to continue on this route, to ground them in a finding of preemption of a fundamental principle, such as a biological natural phenomenon.
Patent eligibility post-Bilski II
I predict that patent eligibility will be a relevant issue for biotechnology in the foreseeable future. Now that the Supreme Court has squashed the red herring of the machine or transformation test, the focus is again upon whether a claim preempts a fundamental principle. For claims like those at issue in Bilski, including claims to so-called business methods, financial methods, and the like, the category of fundamental principle most implicated has been "abstract idea."
Biotechnology inventions do not implicate the claiming of abstract ideas, so much as the patenting of a natural phenomenon. Bilski II provided almost no guidance with respect to what it means for a patent claim an abstract idea, other than pointing out that methods of hedging risk are well known in the economics literature, and concluding that a claim directed to a method of hedging risk improperly encompasses an abstract idea. The Supreme Court has provided even less guidance with respect to what it means to claim a biological natural phenomenon.
Moving forward, the courts, and the PTO if it chooses to base rejections of biotechnology patent claims based on lack of patent eligibility, will need to grapple with two fundamental questions. First, what exactly is a biological natural phenomenon? And second, what does it mean for a patent to impermissibly claim a natural phenomenon?
The first question was addressed in Prometheus. The inventors had discovered that for certain drugs there was a relationship between the level of specific drug metabolites in a patient's body and the optimal dosage of the drug. The patent claimed a method that involved determining the level of the specific drug metabolites in a patient's body and using that information to adjust the dosage of the drug, thereby tailoring the dosage to the physiology of an individual patient, an example of a personalized medicine. The district court held that the relationship between the level of drug metabolites in a person's body and optimal dosage was a patent ineligible "natural phenomenon."
On appeal to the Federal Circuit, I filed an amicus brief arguing that the District Court had erred in this regard, and that the interaction of a man-made, non-naturally occurring molecule (such as a drug metabolite) with the human body is not a natural phenomenon. In my brief, I pointed out that characterizing the interaction of a non-naturally occurring molecule with the human body as a natural phenomenon simply because the interaction is governed by natural principles of chemistry and physiology is akin to characterizing the interaction of an airplane with the atmosphere as a natural phenomenon simply because the interaction of airplane with the air is governed by fundamental scientific principles. I also pointed out that if the interaction of a drug metabolite with the human body is treated as a natural phenomenon, then logically this implied that in general the interaction of drugs with the human body are natural phenomena, which would generally cast doubt upon the validity of many drug patents.
On appeal, the Federal Circuit avoided the issue by simply relying on the machine or transformation test, but now in the wake of Bilski II the Supreme Court has remanded the case to the Federal Circuit, and on remand I think the court needs to address head-on the question of whether or not the interaction of a drug metabolite with the human body is a natural phenomenon.
In contrast, in LabCorp the claim was based on a correlation between a naturally occurring metabolite (homocysteine) and a physiological condition (vitamin B deficiency). This looks much more like a natural phenomenon to me than the correlation between a non-naturally occurring drug metabolite and optimal drug dosage. All of the examples of natural phenomena that have been provided by the Supreme Court involve phenomena that are present absent any human intervention, such as the law of gravity or E=mC2. In Chakrabarty to Supreme Court made clear that for biological inventions human intervention is the key to determining patent eligibility. By drawing a line between biological phenomena that occurs absent human intervention and phenomena that occurs as a result of human intervention, one could have a principled basis for finding the LabCorp claim patent ineligible while upholding the eligibility of the Prometheus claims, and drug patents in general.
This approach could also be a way of dealing with gene patents. Myriad's method claims are directed towards identification of naturally occurring mutations in the BRCA gene. Like a correlation in LabCorp, the correlation between naturally occurring genetic mutation and susceptibility to cancer, or even simply the presence or absence of a national mutation, could plausibly be characterized as a natural phenomena, and therefore a claim preempting the phenomena would be patent ineligible. On the other hand, a patent claim directed towards a correlation between a genetic mutation and a person's optimal medical treatment might be fine because it is not implicate a natural phenomenon. For example, genetic testing is used to identify patients that will benefit from treatment with the cancer drug Herceptin. Since Herceptin is a non-naturally occurring drug, the correlation between genotype and response to Herceptin treatment should not be considered a natural phenomenon, and thus a patent claim directed towards this correlation would not be susceptible to the same sort of challenge that a patent claim directed more broadly simply to the mutation itself.
The benefit of this approach is that would allow some principled means for drawing a line between patent claims broadly encompassing the practical exploitation of a naturally occurring biological phenomenon, which a sizable number of people find objectionable, and patent claims directed towards drugs, personalized medicine, and specific diagnostic applications, which most would agree warrant patent protection. Many people do not like Myriad’s claims broadly directed towards identification of naturally occurring mutations in the BRCA genes, nor did they like claims of the type at issue in LabCorp, which broadly encompass the observation of naturally occurring biological correlations, but would not object to patent protection denied for more targeted protection of methods of treatment or diagnosis, or the combination of the two embodied in personalized medicine
Furthermore, the Supreme Court has failed to articulate what exactly it means to impermissibly patent such a natural phenomena. The admonition could be interpreted as a simple statement of the obvious, i.e., that it is impossible to patent a natural phenomena per se. It is well-established that only products or processes can be claimed in a patent, so inherently it would be impossible to patent a natural phenomenon per se. For example, it would be impossible to simply claim “gravity,” since gravity is neither a product or process. If the prohibition against patenting natural phenomena were interpreted this narrowly, the Supreme Court dictum regarding the patenting of natural phenomena would be entirely gratuitous, and provide absolutely no additional limitation on patenting beyond the statutory limitation of patent protection to products and processes.
However, with all the talk in Supreme Court decisions about the patent ineligibility of natural phenomena, we should assume that the Supreme Court envisions some actual limitation on the scope of patentable subject matter, implying that the prohibition must extend beyond the mere patenting of natural phenomena per se. Clearly Justice Breyer and the other dissenting justices in LabCorp saw its reach in much broader terms. The claim at issue in LabCorp was not directed towards the physiological correlation per se, but rather to a method that involves assaying the blood for total homocysteine level and using that information to diagnose the vitamin B deficiency, thus requiring substantial human and all. Nonetheless, the court concluded that this process amounted to an impermissible attempt to claim the natural phenomenon itself.
On the other hand, it cannot be the case that an invention is patent ineligible simply because it involves a practical application of a natural phenomenon. Indeed, it is difficult to imagine a technology that is not based to some extent on the practical exploitation of natural phenomena, a point made in my Prometheus amicus brief.
If there is indeed a prohibition against the patenting of natural phenomena, and the Supreme Court has repeatedly stated that there is, logically it must go further than simply barring the patent eligibility of natural phenomena per se, but it cannot go so far as to declare all inventions involving the use of a natural phenomena patent ineligible. The line must be drawn somewhere, and there is language from the Supreme Court suggesting that the line is crossed when a patent claim effectively preempts all practical uses of a natural phenomenon. It seems to me this is the most sensible place to draw the line, and this is the criterion that was applied by the district court in Prometheus.
In that case, the district court held that the process claims effectively encompass all practical uses of the natural phenomena at issue, i.e., the correlation between drug metabolite and optimal dosage. In the ACLU gene patent decision, the district court also concluded that the claims at issue effectively encompassed all practical uses of the claimed genes.
If this is correct, a determination of patent eligibility based on the patenting of a natural phenomenon will require two showings: first, a true natural phenomenon will need to be identified, and second, if a natural phenomenon is identified, that must be determination of whether the claim effectively preempts all practical uses of the phenomenon. The impact of the patent eligibility doctrine on biotechnology will depend in large part on how "preemption" is defined. For example, arguably the claim at issue in LabCorp covers any practical method of exploiting the discovery of a correlation between total homocysteine levels and vitamin B deficiency. A court might conclude that this claim is so broad it entirely preempts the practical use of the correlation, thereby preempting a natural phenomenon and thus rendering the claim patent ineligible. However, one could alternatively find that since the claim has no impact on the correlation as it exists and functions naturally in the human body, the phenomenon is not entirely preempted by the claim.
Similarly, some of the Myriad gene patent method claims broadly cover identification of mutations in the BRCA genes. One could consider this preemptive, to the extent it covers any diagnostic use of the information regarding the mutations. On the other hand, the claim does not cover the mutations as they exist in nature, nor the effect of changes in the BRCA sequence on human physiology, so arguably the claim does not preempt the natural phenomenon. Depending on how a court conducts its preemption analysis, patent in eligibility based on preemption of a natural phenomenon could have little impact on these patent claims, or if interpret more restrictively, could constitute a new in substantial obstacle to broad patent claims such as those at issue in Prometheus, Classen in the ACLU patent lawsuit.
Although the natural phenomenon category is probably the most relevant for biological inventions, mental processes has also been identified as a category of patent ineligible fundamental principle, and some biotechnology process patents might be challenged for claiming a thought process. For example, it has been argued that the LabCorp claim, the Prometheus claims, and the Myriad claims directed to methods of detecting mutations all could be infringed by a doctor simply recognizing and thinking about the claimed correlation. Some of the Myriad claims appear to be broadly directed at simply identifying or observing a mutation in a BRCA gene, which could be challenged as improperly covering the mental process of simply noting the existence of the mutation.
As a practical matter, up until a few years ago patent eligibility was rarely an issue for those attempting to patent inventions relating to biotechnology. So long as there was some element of human intervention, such as isolation of a gene sequence or genetic engineering of a living organism, biotechnology inventions were generally assumed to be eligible for patent protection. Patenting of biotechnology flourished, notwithstanding the protests of many who argued that some inventions based on fundamental biological discoveries, particularly gene patents and patents broadly claiming the practical exploitation of fundamental biological discoveries (such as physiological correlations useful in the diagnosis and treatment of disease), or unsuited for patent protection and should be treated as patent ineligible.
In Bilski I, a majority of the judges on the Federal Circuit badly misinterpreted Supreme Court precedent and declared the so-called "machine or transformation" test to be the universal and exclusive criterion for determining the patent eligibility of all process claims, including by implication processes relating to biotechnology. In essence, Bilski I shifted the patent eligibility inquiry from whether a claimed biotechnology process covered a fundamental principle (i.e., a natural phenomenon) to whether the process entailed sufficient involvement of a machine or transformation
The machine or transformation test was subsequently used as the basis for invalidating a number of biotechnology process claims, including Myriad’s claims to methods of detecting mutations in BRCA genes. The PTO also began to employ the machine or transformation test to reject biotechnology claims, particularly those relating to diagnostics and personalized medicine. Some commentators, including myself, expressed concern that the machine or transformation test, as mandated by Bilski I, threatened to seriously undermine the ability of companies and researchers engaged in the discovery and commercialization of diagnostics and personalized medicine to obtain adequate patent protection for their inventions, although these concerns were to some extent allayed by the Federal Circuit’s restrained application of the test to a personalized medicine patent claim in Prometheus v. Mayo.
Fortunately, the Supreme Court agreed to hear an appeal of Bilski I, and last Monday the Court decided the case, formally titled Bilski v. Kappos (Bilski II), in a manner that effectively rejected the Federal Circuit's interpretation of the machine or transformation test. The Court held that while the involvement of a machine or transformation in a claimed process is often informative in assessing patent eligibility, it is not a general prerequisite for patent eligibility, and admonished the Federal Circuit that the test for patent eligibility remains whether or not the patent claims encompass a fundamental principle.
In essence, the Supreme Court has turned back the clock, negating Bilski I and the Federal Circuit's ill-conceived experiment with the machine or transformation test. Significantly, however, the Supreme Court provided little if any guidance with respect to what it means for a patent to claim a fundamental principle, and absolutely no guidance with respect to how to apply the test when the fundamental principle is a biological natural phenomenon, leaving it to the lower courts to sort out this important issue
I predict that the issue of patent eligibility for biotechnology inventions, both products and processes, will continue to be actively litigated well into the foreseeable future, as courts and the patent office struggle to figure out just what exactly it means for a patent claim to cover a biological natural phenomenon. In this article, I summarize important developments in the patent eligibility doctrine leading up to and including Bilski II, particularly as they relate to biotechnology, and discuss the issues relating to patent eligibility that will need to be addressed in a post-Bilski world.
Patent Eligibility Prior to LabCorp
There are four requirements of patentability which an invention must satisfy in order to be patentable : patent eligibility, practical utility, novelty and nonobviousness. The first, patent eligibility, has been characterized as the threshold criterion, used to exclude purported inventions that are useful and innovative, but nonetheless unsuited for and hence ineligible for patent protection.
Prior to the advent of modern biotechnology, some biological subject matter, particularly living organisms, were generally considered patent ineligible. This explains why Congress saw fit to enact legislation explicitly creating alternate forms of intellectual property protection for plants (plant patents and plant variety protection certificates). However, with the increasing commercial significance of biotechnology in the 1970s, the question of whether living organisms and other biotechnological inventions of were eligible for patent protection became of increasing importance, and in 1980 the Supreme Court answer to the question in the affirmative in its landmark Diamond v. Chakrabarty decision.
In Chakrabarty, the Supreme Court essentially held that any man-made product or process is patentable, so long as the patent does not purport to claim a "fundamental principle." In Chakrabarty and other Supreme Court decisions relating to the doctrine of patent eligibility, the Court has at various times identify specific categories of fundamental principles, including natural phenomena, physical phenomena, principles of nature, abstract ideas, and mental processes.
As a practical matter, after Chakrabarty patent eligibility was not a major issue for biotechnology. So long as there was some degree of human intervention, biotechnology inventions were generally deemed patent eligible by the patent office and the courts. While living organisms and other biological organisms could not be patented as they exist in nature, genetic engineering of an organism was deemed sufficient to render the non-naturally occurring modified organism patentable. Similarly, isolation of a naturally occurring DNA sequence was deemed sufficient human intervention to render the isolated molecule patentable.
It was not until the Supreme Court granted certiorari in LabCorp v. Metabolite in 2005 that patent eligibility once again took on significance in the context of biotechnology. The patent at issue in lab were was based upon the discovery of a natural correlation between the level of total homocysteine (a naturally occurring metabolite) in the human body and existence of a vitamin B deficiency. Based on this discovery, the patent broadly claimed any method of detecting a vitamin B deficiency that involves assaying a patient's blood for total homocysteine content and correlating elevated total homocysteine with such a deficiency. Although the issue of patent eligibility was never directly addressed in the lower courts, the Supreme Court agreed to hear the case specifically to address the question of whether or not the patent violated the prohibition against the patenting of fundamental principles, i.e., natural phenomena, principles of nature and abstract ideas.
Upon further consideration, however, the Supreme Court had second thoughts about taking the case, and dismissed the LabCorp appeal without deciding it. However, three of the Justices signed on to a dissenting opinion penned by Justice Breyer stating that they considered the claimed process patent ineligible for embodying the natural phenomenon of a correlation between total homocysteine and vitamin B deficiency. Although this dissenting opinion by three of the nine justices has absolutely no binding legal authority, it has proven highly significant because at least established the viability of asserting the doctrine of patent eligibility to challenge patents based on fundamental biological discoveries, particularly patents broadly encompassing the use of natural correlations in a diagnostic method, which had become increasingly common with the emergence of genetic diagnostic testing and personalized medicine.
Patent eligibility after LabCorp
No doubt encouraged by Justice Breyer's dissent, accused infringers began raising the defense of patent ineligibility in patent infringement litigation. Three of these cases, Ariad v. Lilly, Classen v. Biogen, and Prometheus v. Mayo, have been extensively discussed in previous articles on my blog. Significantly, in all three of these cases the case for patent ineligibility was based upon an argument that the challenged claims effectively preempted all practical uses of a biological natural phenomenon, essentially the same rationale propounded by the dissenting justices in LabCorp.
These patent ineligibility-based challenges met with mixed success at the district court level. In Ariad, the district court rejected the challenge, based on a rationale that frankly made absolutely no sense to me. In any event, the issue was rendered moot because the court ruled the claims invalid on the alternate grounds that the claims violated the written description requirement, decision that was ultimately affirmed by the en banc Federal Circuit.
However, in Classen and Prometheus the district court found for the accused infringers and held the asserted claims patent ineligible for broadly preempting the practical exploitation of biological natural phenomena. The patent owners both appealed their cases to the Federal Circuit, but before the Federal Circuit heard the cases issued its en banc decision in Bilski I, which effectively changed the test for patent eligibility, as discussed in the next section.
Patent eligibility in the time between Bilski I and Bilski II
In Bilski I, the Federal Circuit essentially ignored Supreme Court precedent establishing that the test for patent eligibility is based on an inquiry into whether the patent claims a fundamental principle, and announced that the test for patent eligibility of any process claim was whether or not the claimed process included sufficient involvement of a machine or transformation. Bilski I was flawed in a number of regards. For one thing, the holding that no process could be patent eligible without substantial involvement of a machine or transformation was entirely inconsistent with Supreme Court precedent. In attempting to justify its holding, the Bilski I majority pointed to Supreme Court precedent suggesting that a process involving a machine or transformation must be patent eligible, and concluded that this meant that a patent eligible process must include a machine or transformation, a fundamental error in logic that effectively transformed a safe harbor for patentees into an absolute prerequisite for patent eligibility.
As I explained in an earlier article, the machine or transformation test is fundamentally inappropriate for assessing patent eligibility of some processes, especially those involving biotechnology. In crafting the test, the court seemed focused on the perceived problem of so-called business method patents and the like, such as the specific patent claims issue in Bilski I, and erred by declaring the test applicable to all process claims, without adequately considering the implications.
Despite its flaws, Bilski I was an en banc decision of the Federal Circuit, and thus controlling law on the subject absent Supreme Court intervention. Even though Classen and Prometheus were both decided at the district court level prior to Bilski I, based on arguments that focused on alleged preemption of biological natural phenomena, by the time the appeals made it to the Federal Circuit the law had shifted, and as a consequence the parties had to change their arguments to focus upon the involvement (or lack thereof) of a machine or transformation in the claimed processes.
The Federal Circuit decided Classen first, and upheld the district court determination, concluding that the claim, which was originally held invalid or in compassing a natural phenomenon, was invalid for failing to satisfy the machine or transformation test. The decision was only a few sentences in length, and provided absolutely no explanation as to the basis for the court’s decision.
The Federal Circuit then issued a lengthy decision in Prometheus, this time reversing the district court and finding the challenged patent claims to be patent eligible. Specifically, the Federal Circuit found that the steps of administering a drug and determining the level of drug metabolites in a patient's body were both sufficiently transformative to satisfy the machine or transformation test.
The Bilski I machine or transformation test was also the basis relied upon by the district court in invalidating the method claims in the ACLU/Myriad gene patent case, as discussed in previous posts to this blog.
Bilksi II
In Bilski II, the Supreme Court identified the flawed logic in Bilski I, and rejected the Federal Circuit’s machine or transformation test, holding that while the presence of a machine or transformation in a process claim can be relevant in assessing the patent eligibility, there is no general requirement that a patent eligible process include a machine or transformation. The Court reiterated that the test for patent eligibility hinges upon whether or not the claim encompasses a fundamental principle, and held that because the claims at issue were directed towards an abstract idea, one of the fundamental principles expressly identified in earlier Supreme Court decisions, they were patent ineligible.
In essence, Bilski II has voided the Bilski I decision and turned back the clock to where we were prior to Bilski I. The patent eligibility of biotechnological inventions is still very much in play, and there will be much for the courts and patent office to sort out over the next years, but the focus will again be upon figuring out what it means for a patent claim to cover a fundamental principle, not whether it involves a machine or transformation.
The Supreme Court has vacated the Federal Circuit decisions in Classen and Prometheus, and sent those cases back to the Federal Circuit to decide based on Bilski II. Since Bilski II is essentially the law prior to Bilski I, this will require the Federal Circuit to consider the arguments that were actually made a district court level which focused upon preemption of biological natural phenomena, the correct test for patent eligibility. To a large extent, the Federal Circuit will be working on a blank slate. Although the Supreme Court has made clear that the determination of patent eligibility should be based on an inquiry into whether the patent claims encompass a biological natural phenomenon, the Court has provided very little specific guidance as to how this test is to be applied in practice.
An alleged failure to comply with the machine or transformation test was the basis for the invalidation of Myriad’s method claims in the ACLU gene patent challenge. These were the broadest and most relevant of the claims challenged in the case. In my view, Myriad’s product claims directed to polynucleotides are either highly susceptible to invalidation based on prior art (this applies to the fragment claims) or clearly not infringed by BRCA testing (the full-length coding sequence claims). I also think it is highly likely that the Federal Circuit will reverse the lower court’s ruling these product claims are patent ineligible.
However, I thought that the Federal Circuit would probably uphold the finding of patent ineligibility for some of Myriad’s method claims broadly claiming the detection of mutations in the BRCA genes. The processes recited in these claims did not seem to require the use of any machine or transformation, and the district court was probably correct in holding that the claims failed the machine or transformation test.
I think it is pretty apparent that if the district court had decided the case prior to Bilski I, or subsequent to Bilski II, it would have found the method claims patent ineligible for preempting natural phenomena. It just so happened that the district court decided the case during the brief tenure of the machine or transformation test, and thus through no fault of its own conduct the analysis based on the incorrect test. It remains to be seen how the claims will whether a patent ineligibility challenge based on the correct test as set forth in Bilski II.
Bilski II will also force the PTO to alter its examination practices. After Bilski I, the PTO began applying the machine or transformation test as a basis for rejecting patent claims, including claims relating to biotechnology methods, particularly with respect to inventions relating to diagnostics and personalized medicine. Since Bilski I was an en banc decision of the Federal Circuit, the PTO was required to implement it, but now that the Supreme Court has rejected Bilski I rejections based solely on the lack of a machine or transformation are improper. The PTO will need to either forgo rejections based upon a patent ineligibility, or if it chooses to continue on this route, to ground them in a finding of preemption of a fundamental principle, such as a biological natural phenomenon.
Patent eligibility post-Bilski II
I predict that patent eligibility will be a relevant issue for biotechnology in the foreseeable future. Now that the Supreme Court has squashed the red herring of the machine or transformation test, the focus is again upon whether a claim preempts a fundamental principle. For claims like those at issue in Bilski, including claims to so-called business methods, financial methods, and the like, the category of fundamental principle most implicated has been "abstract idea."
Biotechnology inventions do not implicate the claiming of abstract ideas, so much as the patenting of a natural phenomenon. Bilski II provided almost no guidance with respect to what it means for a patent claim an abstract idea, other than pointing out that methods of hedging risk are well known in the economics literature, and concluding that a claim directed to a method of hedging risk improperly encompasses an abstract idea. The Supreme Court has provided even less guidance with respect to what it means to claim a biological natural phenomenon.
Moving forward, the courts, and the PTO if it chooses to base rejections of biotechnology patent claims based on lack of patent eligibility, will need to grapple with two fundamental questions. First, what exactly is a biological natural phenomenon? And second, what does it mean for a patent to impermissibly claim a natural phenomenon?
The first question was addressed in Prometheus. The inventors had discovered that for certain drugs there was a relationship between the level of specific drug metabolites in a patient's body and the optimal dosage of the drug. The patent claimed a method that involved determining the level of the specific drug metabolites in a patient's body and using that information to adjust the dosage of the drug, thereby tailoring the dosage to the physiology of an individual patient, an example of a personalized medicine. The district court held that the relationship between the level of drug metabolites in a person's body and optimal dosage was a patent ineligible "natural phenomenon."
On appeal to the Federal Circuit, I filed an amicus brief arguing that the District Court had erred in this regard, and that the interaction of a man-made, non-naturally occurring molecule (such as a drug metabolite) with the human body is not a natural phenomenon. In my brief, I pointed out that characterizing the interaction of a non-naturally occurring molecule with the human body as a natural phenomenon simply because the interaction is governed by natural principles of chemistry and physiology is akin to characterizing the interaction of an airplane with the atmosphere as a natural phenomenon simply because the interaction of airplane with the air is governed by fundamental scientific principles. I also pointed out that if the interaction of a drug metabolite with the human body is treated as a natural phenomenon, then logically this implied that in general the interaction of drugs with the human body are natural phenomena, which would generally cast doubt upon the validity of many drug patents.
On appeal, the Federal Circuit avoided the issue by simply relying on the machine or transformation test, but now in the wake of Bilski II the Supreme Court has remanded the case to the Federal Circuit, and on remand I think the court needs to address head-on the question of whether or not the interaction of a drug metabolite with the human body is a natural phenomenon.
In contrast, in LabCorp the claim was based on a correlation between a naturally occurring metabolite (homocysteine) and a physiological condition (vitamin B deficiency). This looks much more like a natural phenomenon to me than the correlation between a non-naturally occurring drug metabolite and optimal drug dosage. All of the examples of natural phenomena that have been provided by the Supreme Court involve phenomena that are present absent any human intervention, such as the law of gravity or E=mC2. In Chakrabarty to Supreme Court made clear that for biological inventions human intervention is the key to determining patent eligibility. By drawing a line between biological phenomena that occurs absent human intervention and phenomena that occurs as a result of human intervention, one could have a principled basis for finding the LabCorp claim patent ineligible while upholding the eligibility of the Prometheus claims, and drug patents in general.
This approach could also be a way of dealing with gene patents. Myriad's method claims are directed towards identification of naturally occurring mutations in the BRCA gene. Like a correlation in LabCorp, the correlation between naturally occurring genetic mutation and susceptibility to cancer, or even simply the presence or absence of a national mutation, could plausibly be characterized as a natural phenomena, and therefore a claim preempting the phenomena would be patent ineligible. On the other hand, a patent claim directed towards a correlation between a genetic mutation and a person's optimal medical treatment might be fine because it is not implicate a natural phenomenon. For example, genetic testing is used to identify patients that will benefit from treatment with the cancer drug Herceptin. Since Herceptin is a non-naturally occurring drug, the correlation between genotype and response to Herceptin treatment should not be considered a natural phenomenon, and thus a patent claim directed towards this correlation would not be susceptible to the same sort of challenge that a patent claim directed more broadly simply to the mutation itself.
The benefit of this approach is that would allow some principled means for drawing a line between patent claims broadly encompassing the practical exploitation of a naturally occurring biological phenomenon, which a sizable number of people find objectionable, and patent claims directed towards drugs, personalized medicine, and specific diagnostic applications, which most would agree warrant patent protection. Many people do not like Myriad’s claims broadly directed towards identification of naturally occurring mutations in the BRCA genes, nor did they like claims of the type at issue in LabCorp, which broadly encompass the observation of naturally occurring biological correlations, but would not object to patent protection denied for more targeted protection of methods of treatment or diagnosis, or the combination of the two embodied in personalized medicine
Furthermore, the Supreme Court has failed to articulate what exactly it means to impermissibly patent such a natural phenomena. The admonition could be interpreted as a simple statement of the obvious, i.e., that it is impossible to patent a natural phenomena per se. It is well-established that only products or processes can be claimed in a patent, so inherently it would be impossible to patent a natural phenomenon per se. For example, it would be impossible to simply claim “gravity,” since gravity is neither a product or process. If the prohibition against patenting natural phenomena were interpreted this narrowly, the Supreme Court dictum regarding the patenting of natural phenomena would be entirely gratuitous, and provide absolutely no additional limitation on patenting beyond the statutory limitation of patent protection to products and processes.
However, with all the talk in Supreme Court decisions about the patent ineligibility of natural phenomena, we should assume that the Supreme Court envisions some actual limitation on the scope of patentable subject matter, implying that the prohibition must extend beyond the mere patenting of natural phenomena per se. Clearly Justice Breyer and the other dissenting justices in LabCorp saw its reach in much broader terms. The claim at issue in LabCorp was not directed towards the physiological correlation per se, but rather to a method that involves assaying the blood for total homocysteine level and using that information to diagnose the vitamin B deficiency, thus requiring substantial human and all. Nonetheless, the court concluded that this process amounted to an impermissible attempt to claim the natural phenomenon itself.
On the other hand, it cannot be the case that an invention is patent ineligible simply because it involves a practical application of a natural phenomenon. Indeed, it is difficult to imagine a technology that is not based to some extent on the practical exploitation of natural phenomena, a point made in my Prometheus amicus brief.
If there is indeed a prohibition against the patenting of natural phenomena, and the Supreme Court has repeatedly stated that there is, logically it must go further than simply barring the patent eligibility of natural phenomena per se, but it cannot go so far as to declare all inventions involving the use of a natural phenomena patent ineligible. The line must be drawn somewhere, and there is language from the Supreme Court suggesting that the line is crossed when a patent claim effectively preempts all practical uses of a natural phenomenon. It seems to me this is the most sensible place to draw the line, and this is the criterion that was applied by the district court in Prometheus.
In that case, the district court held that the process claims effectively encompass all practical uses of the natural phenomena at issue, i.e., the correlation between drug metabolite and optimal dosage. In the ACLU gene patent decision, the district court also concluded that the claims at issue effectively encompassed all practical uses of the claimed genes.
If this is correct, a determination of patent eligibility based on the patenting of a natural phenomenon will require two showings: first, a true natural phenomenon will need to be identified, and second, if a natural phenomenon is identified, that must be determination of whether the claim effectively preempts all practical uses of the phenomenon. The impact of the patent eligibility doctrine on biotechnology will depend in large part on how "preemption" is defined. For example, arguably the claim at issue in LabCorp covers any practical method of exploiting the discovery of a correlation between total homocysteine levels and vitamin B deficiency. A court might conclude that this claim is so broad it entirely preempts the practical use of the correlation, thereby preempting a natural phenomenon and thus rendering the claim patent ineligible. However, one could alternatively find that since the claim has no impact on the correlation as it exists and functions naturally in the human body, the phenomenon is not entirely preempted by the claim.
Similarly, some of the Myriad gene patent method claims broadly cover identification of mutations in the BRCA genes. One could consider this preemptive, to the extent it covers any diagnostic use of the information regarding the mutations. On the other hand, the claim does not cover the mutations as they exist in nature, nor the effect of changes in the BRCA sequence on human physiology, so arguably the claim does not preempt the natural phenomenon. Depending on how a court conducts its preemption analysis, patent in eligibility based on preemption of a natural phenomenon could have little impact on these patent claims, or if interpret more restrictively, could constitute a new in substantial obstacle to broad patent claims such as those at issue in Prometheus, Classen in the ACLU patent lawsuit.
Although the natural phenomenon category is probably the most relevant for biological inventions, mental processes has also been identified as a category of patent ineligible fundamental principle, and some biotechnology process patents might be challenged for claiming a thought process. For example, it has been argued that the LabCorp claim, the Prometheus claims, and the Myriad claims directed to methods of detecting mutations all could be infringed by a doctor simply recognizing and thinking about the claimed correlation. Some of the Myriad claims appear to be broadly directed at simply identifying or observing a mutation in a BRCA gene, which could be challenged as improperly covering the mental process of simply noting the existence of the mutation.
Subscribe to:
Posts (Atom)
Posts
