Bad Science Makes Bad Patent Law—No Science Makes It Worse (Part II)

In Part I, I explained some general criteria for laws of nature, considering the prototypes of Newton's laws and Einstein's E=mc2. Now I'll turn to whether there are laws of nature in biology. Biological generalizations, such as those identified by the courts (and similarly found in many rejected patent applications), do not meet the various criteria used to identify laws of nature. Most significantly, they are not universal—they apply only to specific characteristics of human biology. And while in some cases they allow for predictions, they are not reductionist: They do not explain more specific phenomena in more general terms. In many cases, the generalizations are not objective and independent of human concerns but entirely dependent on them—for example the alleged law relating specific genes to elite athletic performance. Mother Nature does not concern herself with Usain Bolt's sprinting prowess. The reason that biological generalizations are not universal is that they are the result of evolution—they are contingent outcomes, they could have been otherwise. Indeed it is possible that they could not have arisen at all. This thesis is known as the Evolutionary Contingency Thesis.

Let's work backward to demonstrate. All of us learned as school children that cheetahs are the fastest land animals, capable of running between 60 and 75 mph (though that number is disputed). Using the Supreme Court's definition, the fact that cheetahs on average have a top speed of "about" 70 mph (remember "about" in Mayo?) "exists in principle apart from any human action." But obviously the average top speed of cheetahs is not a law of nature: It is the result of evolutionary forces such as mutations, adaptation, and environmental conditions. Cheetahs could have evolved to be slower or faster, or not evolved as a species at all. Further, the cheetah's speed is obviously dependent on its genetic makeup. That some unidentified genes are the cause of the cheetah's speed is not a law of nature. What goes for cheetahs goes for all animals—nature (and DNA) do not differentiate. I've discussed at length why the claim in Mayo was not a law of nature, so I'll not repeat that here. Let's consider instead the so-called law of nature in Genetic Veterinary Sciences v. Canine EIC Genetics, the relationship between a mutation at position 767 of the DNM1 gene and whether a dog (especially a Labrador Retriever) is susceptible to "exercise induced collapse" (EIC), the loss of control of its legs after strenuous exercise. The relationship between genetic mutations and biological traits is a product of evolutionary development. First, it could have turned out that it was a mutation at some other position on the same chromosome, or on some other chromosome that was responsible for EIC, since there were three other candidate locations on the same chromosome that were eventually ruled out. Thus, it was an essentially random event that this particular mutation was related to EIC. Second, Labradors themselves were bred by humans: By more careful selection, we could have bred them so that none of them had this particular genetic defect and did not suffer from EIC. One could argue that the "natural relationship" here only arose as a direct result of human action, the specific, intentional breeding of dogs that led to the Lab. In other words, that a particular mutation of a particular gene makes a Lab "susceptible" to EIC was more a result of human activity, dog breeding—not even a naturally occurring result, let alone a necessary one.

And what's true for dogs (and ducks) is true for humans. That a given gene or allele causes a specific disease (e.g., CFTR gene causes cystic fibrosis (CF)) is a result of random changes (mutations) that occurred at some point in the historical past. About 85-90% of CF patients have a mutation of just three DNA nucleotides at a specific location. Those mutations did not have to occur at all, but once they did, they continued to be propagated. And it is conceivable that these mutations may be eliminated in the future, given developments in gene therapy for CF. This temporal aspect further demonstrates that the "natural relationship" between genes and diseases is not a law of nature. Consider Genetic Technologies v. Merial. The alleged law of nature in this case was "linkage disequilibrium between the non-coding and coding regions—i.e., the tendency of these regions to be linked." The court held that this relationship is a law of nature, since it"is indisputably a universal, inherent feature of human DNA." Let us assume that it is an "inherent feature of human DNA." That does not make it a law of nature, any more so than the mutation of a specific gene being correlated with a specific disease, such as EIC, is a law of nature. More generally, the 'inherent' nature of the relationship is itself disputable. Prior to the inventors' discovery, it was generally assumed that non-coding loci were "junk DNA" and thus there was no tendency of specific coding and non-coding loci to be associated with each other. We could have evolved such that there was no linkage at all—again the linkage was a result of the evolutionary processes. Indeed Francis Crick, who discovered the structure of DNA, held essentially this view of the randomly evolved and thus arbitrary nature of the human genome. Crick's "frozen accident" hypothesis argues that the genetic code—the relationship between specific codons and the amino acids they code for—is an arbitrary relationship that resulted from all life having a common genetic ancestor. (While variants of the genetic code have been discovered, nonetheless a significant portion of the code is common across all life forms). Let's look at the alleged law in Merial case with respect to the criteria of laws of nature set forth above. The claim in that case recited:

  1. A method for detection of at least one coding region allele of a multi-allelic genetic locus comprising: a) amplifying genomic DNA with a primer pair that spans a non-coding region sequence, said primer pair defining a DNA sequence which is in genetic linkage with said genetic locus and contains a sufficient number of non-coding region sequence nucleotides to produce an amplified DNA sequence characteristic of said allele; and b) analyzing the amplified DNA sequence to detect the allele.

First, the claim does not describe a "universal" condition that is true under all conditions and independent of contingent facts. While some specific alleles may be linked with some specific non-coding regions, others may not be. Does this claim enable any specific prediction that can be tested and reproduced? Reproducibility and prediction are hallmarks of a scientifically valid law of nature. From a true law, a specific prediction of an outcome can be made, and experiments can be run to test the prediction and those experiments can be repeated with the same outcomes. The more confirmatory the outcomes, the more we believe the underlying statement is a law of nature. In fact, we continue to this day to experimentally validate Einstein's law. Even laws of nature that are probabilistic, such as those in quantum mechanics, enable precise equations for the probability distributions of outcome overall states (e.g., the exact probability of finding an electron in a region of space). But Genetic Technologies’ claim above does not enable predictions about the linkage between specific coding regions and non-coding regions, or even the probabilities of such linkages. Instead it defines a methodology for identifying the relationships. Thus, it does not have predictive power. Further, because it does not allow for specific predictions, it cannot be disproved in scientific terms. This last point leads to a more critical insight: In science, laws of nature and theories they are part of are never proven "true" or "correct." Rather, science can only disprove a theory and whatever underlying law or principle it attempts to explain. Even if we conduct a thousand or hundred thousand experiments that confirm a particular theory, the next experiment may well disprove it—and it takes only a single such instance to bring down a proposed theory. The history of science is the history of theories and laws of nature being disproved. All scientific knowledge is provisional: What we believe today to be the true and correct explanations and law may be falsified tomorrow. Albert Einstein himself stated that "If a single one of the conclusions drawn from [his theory of general relativity] proves wrong, it must be given up." Consider this: If a scientist believes she has discovered a new biological principle (let's be careful and not yet call it a law), she typically conducts a number of experiments and then publishes the results and explanation of the principle. The publication is subject to peer review—critical analysis by other scientists to identify errors or problems in the author's research methods, computations or other details. Once her findings are published, other scientists will conduct experiments as well to verify the findings. Over time, if repeated experiments and analysis fail to refute the principle, a consensus will emerge that a new law of nature has been discovered. The key thing here is that the determination of whether the newly discovered principle is a law of nature is an empirical process: It often takes years, sometimes decades, for a law to be accepted as such. For example, Einstein's theory of special relativity, published in 1905, assumed that the speed of light is constant, but this was not experimentally confirmed until 1964. Whether something is a law of nature is an empirical question, not a question of a priori logic. In fact, most proposed laws of nature end up being falsified by experimentation. Now contrast this to the typical patent eligibility case. A company research scientist discovers some "naturally occurring relationship" between various genetic markers (e.g. specific mutations on specific genes) and a particular disease. She files a patent application for a method for diagnosing the disease based on the presence of the markers. She makes no statements in the patent application to having discovered any law of nature. The patent application is reviewed by a patent examiner. The examiner, while not likely having the same level of expertise as a scientist who would peer review a research paper, is nevertheless technologically skilled and understands that the invention is not directed to a law of nature, but to a specific useful application of a particular biological relationship, and so grants the application. The company sues to prevent a competitor from offering a competitive diagnostic. The competitor moves to dismiss, arguing that the claims are ineligible because they recite a law of nature. The court grants the motion and invalidates the patent, without a shred of evidence offered.

At this point you should see the problem: If the scientific community would require years of experimentation before acknowledging that something is a law of nature, how can a court on a motion to dismiss, with no evidence at all, decide exactly, merely because patent eligibility is said to be a question of law? This kind of judicial alchemy happens all the time. Courts turn mere patent claims and assertions of discovery into golden laws of nature without any consideration of how the scientific process actually works. For example, the so-called law of nature in Mayo was based on just two studies with a total of a 182 patients. The specific limitation that the court said was the law—"if the levels of 6-TG in the blood (of a patient who has taken a dose of a thiopurine drug) exceed about 400 pmol per 8x108 red blood cells, then the administered dose is likely to produce toxic side effects"—was based on the toxicology results of just six people! That’s hardly sufficient evidence of a law of nature, and would not survive serious peer review if so asserted. Scientists take note: Forget publication in Science, Nature, Cell or Physical Review. Get a patent and have it invalidated by a district court instead and achieve the pinnacle of scientific recognition. Thus, the issue of the empirical basis of laws of nature leads directly to the question of evidence. Even accepting the Supreme Court's definition of a law of nature there should be evidence in the record that supports the conclusion that law of nature is in fact being claimed. But this never seems to happen. District courts are often willing to announce a new "law of nature" where neither the inventor nor any other scientist would make such a claim and there is no evidence in the record to support such a finding. And the Federal Circuit upholds these decisions (Ambry, Ariosa, Exergen, Genetic Technologies, Hemopet) or overlooks the core problems (Rapid Litigation Management). Several years ago I chaired a conference on patent eligibility (what else?). I suggested to Rebecca Eisenberg, one of the panelists, that the Supreme Court should look to what philosophy of science has to say about laws of nature. "God help us," she said, if the Court should consider philosophy of science and have to read up on Karl Popper and Thomas Kuhn. The funny thing is this: The Court has relied on philosophy of science, and Popper in particular. In Daubert v. Merrill Dow Pharmaceuticals, the Court addressed the question of what are the standards for admitting expert scientific testimony in federal court under F.R.E 702: "If scientific, technical, or other specialized knowledge will assist the trier of fact to understand the evidence or to determine a fact in issue, a witness qualified as an expert by knowledge, skill, experience, training, or education, may testify thereto in the form of an opinion or otherwise." What counts as "scientific knowledge"? The Court's analysis is worth quoting at length:

Ordinarily, a key question to be answered in determining whether a theory or technique is scientific knowledge that will assist the trier of fact will be whether it can be (and has been) tested. "Scientific methodology today is based on generating hypotheses and testing them to see if they can be falsified; indeed, this methodology is what distinguishes science from other fields of human inquiry." Green, at 645. See also C. Hempel, Philosophy of Natural Science 49 (1966) ("[T]he statements constituting a scientific explanation must be capable of empirical test"); K. Popper, Conjectures and Refutations: The Growth of Scientific Knowledge 37 (5th ed. 1989) ("[T]he criterion of the scientific status of a theory is its falsifiability, or refutability, or testability"). … Another pertinent consideration is whether the theory or technique has been subjected to peer review and publication. Publication (which is but one element of peer review) is not a sine qua non of admissibility; it does not necessarily correlate with reliability, see S. Jasanoff, The Fifth Branch: Science Advisors as Policymakers 61-76 (1990), and in some instances well-grounded but innovative theories will not have been published, see Horrobin, The Philosophical Basis of Peer Review and the Suppression of Innovation, 263 J. Am. Med. Assn. 1438 (1990). Some propositions, moreover, are too particular, too new, or of too limited interest to be published. But submission to the scrutiny of the scientific community is a component of "good science," in part because it increases the likelihood that substantive flaws in methodology will be detected. See J. Ziman, Reliable Knowledge: An Exploration of the Grounds for Belief in Science 130-133 (1978); Relman and Angell, How Good Is Peer Review?, 321 New Eng. J. Med. 827 (1989). The fact of publication (or lack thereof) in a peer reviewed journal thus will be a relevant, though not dispositive, consideration in assessing the scientific validity of a particular technique or methodology on which an opinion is premised.

The Court is quite clear that scientific evidence must be evaluated against the norms of scientific research. Consider a toxic tort case in which the plaintiff made the assertion that exposure to the defendant's chemicals caused a particular gene mutation, which in turn caused particular disease, and the relationship between the gene mutation and the disease was a law of nature. The mere assertion of this relationship would fail. The plaintiff would have to have an expert testify that it is a law of nature that the mutation caused the disease. That evidence would be evaluated by the court under the Daubert factors, including whether the relationship has been empirically validated, and the research subject to peer review. Now imagine a patent case where the patentee has a patent claim to diagnosing the same disease based on the presence of the same mutation. Here, it is the defendant who asserts that the relationship is a "law of nature," and thus the patent is invalid. The proposition before the court is the same in both cases, and yet most courts would accept this assertion uncritically, relying at most on whatever statements the patentee made in the patent about the discovery of the relationship. (I have never seen a patent in which the inventor claimed to have discovered a law of nature). For example, in Oxford Immunotec v. Qiagen, on a motion to dismiss, the court held that it was a law of nature that specific T-cells generate interferon in response to exposure to tuberculosis bacterium, even though there was no expert evidence in support of that holding, only the argument of the defendant. One would hope that if the requirements are so rigorous in a liability case, equal rigor would be applied in a patent case. In short, I believe that the district courts should demand scientific evidence from a patent defendant that a claimed natural relationship is recognized by the scientific community as a law of nature before invalidating a patent. Several former district court judges with whom I have spoken agree with this proposition. One response to my analysis is to acknowledge that what the courts have identified are not really laws of nature, but are instead natural phenomena. Consequently, the claims are ineligible and the end result is the same. This argument fails. Natural phenomena and laws of nature are very different kinds of things. Lightning, earthquakes, the aurora borealis, Mammatus clouds are natural phenomena, but they are not laws of nature. And what counts as an "inventive concept" beyond merely applying these different types of subject matter will necessarily be different in kind. Thus, it is necessary to articulate clear differences in these categories so that the U.S. Patent and Trademark Office and the courts can properly identify the subject matter of a claim and determine if it is indeed directed to an ineligible concept and whether there is an inventive concept beyond the exception. A related argument is that it is the role of the courts to terms for legal purposes, and the courts' approach to defining judicial exceptions is no different. But unlike purely legal constructs like probable cause or wilful infringement or inequitable conduct, concepts like laws of nature, natural phenomena, natural products, mathematical algorithms, abstract ideas have significance and meaning entirely outside of the law. As I noted in Part I, the term law of nature dates back to 1250, hundreds of years before the patent law even existed. But the courts are so accustomed to defining the meaning of legal terms—something wholly and properly in their purview—that they do not realize that they cannot reach into the real world and define at will extra-legal concepts in the same way. The Supreme Court cannot define laws of nature etc. in a scientifically incorrect way, any more than they can define π to be 3, or a gene to be simply "a discrete unit of heredity." Just as it is necessary, in my view, to understand the nature of science and innovation to properly promote the useful arts, it is equally necessary to define the terms of the patent law to achieve that goal as well. Fixing the Problem The Supreme Court in Mayo acknowledged that "Courts and judges are not institutionally well suited to making the kinds of judgments needed to distinguish among different laws of nature." Indeed. And it is precisely because the courts cannot make such distinctions, that the Supreme Court needs to correct the problem it created by adopting a more scientifically coherent approach to laws of nature. It's been argued that it's too soon for the Court to take up another patent eligibility case, having only recently decided Alice. But it's been just over four years since the Mayo decision. The Supreme Court "corrected" Parker v. Flook (1978) only three years later in Diamond v. Diehr (1981). And fixing this problem is necessary before more patents (and patent applications) are improperly invalidated for important inventions in diagnostics and treatments. The Court had that opportunity in Ariosa but it denied Sequenom's cert. petition. Now the Court has the opportunity again. Genetic Technologies has filed for certiorari. The Court should take up the case for the reasons I've articulated in these posts. More specifically, the Court can address two issues. First, the Court can articulate a more complete and "patently" useful definition of a law of nature. In the past, the Court has expressed a particular distaste for bright line rules in the patent law, preferring instead flexible standards. Consider the Court's rejection of the "machine-or-transformation" test in Bilski, and the rejection of the "teaching-suggestion-motivation" test in KSR. However, the Court's current definition is such a bright-line rule, by making any natural relationship a de jure law of nature. A revised definition need not be perfect, only more in concert with current scientific theory and practice. If the Court is concerned that providing a revised definition would be too difficult and still "rule" like, then there is an easily achievable target. The Court can simply clarify the methodology for evaluating alleged laws of nature, by following their own example in Daubert. By focusing on this issue, the Court can give the district courts a flexible methodology that is consistent with how the courts handle all other types of scientific evidence. As Judge Hughes noted in his dissent in SCA Hygiene Prods. Aktiebolag v. First Quality Baby Prods., LLC., "the Supreme Court has made abundantly clear that there must be a particular justification in the statute before this court may announce special rules for patent cases that depart from the rules for other areas of civil litigation. See, e.g., Teva Pharm. USA, Inc. v. Sandoz, Inc., (2015); Octane Fitness, LLC v. ICON Health & Fitness, Inc., (2014); eBay Inc. v. MercExchange, LLC, (2006)." In that case, arguments that claims recite "laws of nature" should be subject to the same evidentiary requirements as in other areas of civil litigation. This would be a simple and clear outcome and would make the judicial approach to the patent law consistent with other domains.

*The perspectives expressed in the Bilski Blog, as well as in various sources cited therein from time to time, are those of the respective authors and do not necessarily represent the views of Fenwick & West LLP or its clients.