As discussed in a previous blog post, since Mayo v. Prometheus, critics of medical treatment patents have advocated that such patents should be banned from patenting. While such arguments seemed futile based on the consistent position taken by the U.S. Court of Appeals for the Federal Circuit (CAFC) that treating a disease or other condition is not merely a “mental process” or “natural law,” the CAFC’s recent decision in INO Therapeutics, LLC v. Praxair Distribution Inc. may have changed that. Indeed, if this decision is followed by other panels of the CAFC, it could have a profound effect on the patenting of personalized medicine in the United States.

The Old Rule: Medical Treatment Informed by Diagnostic Testing Could be Patented

For the First Time, a Medical Treatment Patent Is Ruled Invalid Under Mayo/MyriadPrior to INO, the CAFC consistently ruled that methods of medical treatment are patent-eligible subject matter, regardless of whether they are combined with a diagnostic step. In Classen Immunotherapies Inc. v. Biogen IDEC – the seminal case in this line – the CAFC considered the patent eligibility of a claimed method of immunization of a subject that involved first immunizing a number of other subjects according to two immunization schedules and determining which has the least risk of side effects, then immunizing the subject according to the schedule with the lowest risk.

The unanimous panel found the act of immunizing two groups of subjects and comparing the results to determine the best immunization schedule, without applying that schedule to a subject, not eligible for patenting. However, with the extra step of actually immunizing a subject based on the schedule, the majority of the panel held that the claimed method was patent eligible. Notably, while the majority held that a method of medical treatment is eligible for patenting, even if it involves mental steps, a dissenting judge thought the claims were “a monopoly over the scientific method itself.” The CAFC followed this analysis in several later decisions, including in Natural Alternatives International, Inc. v. Creative Compounds, LLC; Endo Pharmaceuticals Inc. v. Mallinckrodt, LLC; and Vanda Pharmaceuticals Inc. v. West-Ward Pharmaceuticals Int’l Ltd.

In contrast, the CAFC has consistently found patent claims that include only diagnostic steps, without any form of medical treatment, ineligible for patenting. Such cases include Myriad (by the Federal Circuit, before it went to the Supreme Court); PerkinElmer Inc v. Intema Ltd.; Univ. of Utah Research Foundation v. Ambry Genetics Corp.; Cleveland Clinic Foundation v. True Health Diagnostics, LLC; Roche Molecular Systems, Inc. v Cepheid; and Ariosa Diagnostics, Inc. v. Sequenom, Inc.

The INO Rule: Withholding Treatment Dooms the Patent

The invention claimed in INO was directed to a method of treating hypoxia in newborns with nitric oxide while reducing the risk of pulmonary edema (a deadly side effect). The inventors discovered that infants with impaired function of the left ventricle are at elevated risk for pulmonary edema if treated with nitric oxide. The asserted claims of the patents in suit featured a method of diagnosing a hypoxic infant for left ventricular dysfunction (LVD) by echocardiography and administering an appropriate concentration of nitric oxide only in the absence of left ventricular dysfunction.

As in Classen and its progeny, the claims involved a medical intervention based on the result of the diagnostic step. Unlike Classen, the trial court in Mallinckrodt held that the use of previously known treatment methods based on the diagnosis does not add “significantly more” to the understanding of the natural law. You can read more about the trial court decision here.

Two of the three judges on the CAFC panel agreed with the trial court. The majority felt that invention in INO was different from those in Classen, Endo, Natural Alternatives, and Vanda in one significant way: in INO, the diagnostic step determined to which patients nitric oxide should be administered and to which patients it should be withheld, whereas in the other decisions, treatment was administered in some form, although the form differed depending on the results of the test.

The majority cited numerous, separate reasons that the distinction caused the invention to fall under the Mayo/Myriad ban. First, the majority explained that “[t]he effect of [nitric oxide] gas on a newborn with LVD is a matter of human physiology.” The majority then stressed that the use of nitric oxide was not new and, therefore, “the patented method does not propose a new way of treating LVD patients.” Although the patentee argued that the claim in its entirety was new, the majority dismissed this “whole claim” approach, opining that “…the focus of the invention is screening for a particular adverse condition that, once identified, requires iNO treatment be withheld.”

The majority contrasted this case with Vanda and Endo, because in the latter cases, the diagnostic test determined the dosage at which a specific treatment is given. Because treatment is administered regardless of the test results, the majority believes these are properly interpreted to be “methods of treatment.” The INO invention, on the other hand, involves either administration of a specific drug at a specific dosage or non-administration of that drug.

It would seem that the majority’s decision is based on the premise that (1) the claimed treatment with nitric oxide of infants without LVD cannot be considered in the patent eligibility analysis because nitric oxide treatment was already known and (2) the claimed withholding of nitric oxide treatment of infants with LVD is not really a method of treatment because no treatment is administered. In fact, the majority explained that a claim not to treat is akin to not disturbing the naturally occurring processes within the LVD patient’s body and “risks monopolizing the natural processes themselves.” Therefore, the claimed method is merely “instruction” to apply the “law of nature” of how nitric oxide affects human physiology.

Judge Newman – the most senior judge on the CAFC – dissented. Her dissenting opinion took a familiar direction, pointing out that the claimed invention was not a “law of nature” as “it was designed by and is administered by humans.” She further pointed out that all medical treatments rely on the response of human physiology to an artificial intervention. She finally pointed out that the Supreme Court has repeatedly cautioned against dissecting claims into component parts and considering only specific parts in the patent eligibility analysis. As noted by Judge Newman in her dissent, while the claimed method may in fact be invalid under the tests of novelty or obviousness, to declare it patent-ineligible is contrary to preceding decisions from the Supreme Court and the CAFC.

Where Does this Leave Personalized Medicine?

This decision leaves many key questions about patent eligibility unanswered and the future uncertain for personalized medicine patents. Prior to INO, the Vanda decision was regarded as a beacon in the foggy post-Mayo world when it came to patenting personalized medicine. It introduced an unaccustomed level of certainty to the question of subject matter eligibility since Mayo. In fact, published USPTO guidance relies on Vanda to teach that “it is not necessary for method of treatment claims that practically apply natural relationships to include nonroutine or unconventional steps to be considered patent eligible under 35 U.S.C. § 101.”

INO calls all of that into question for a number of reasons. First, the majority did not explain why the Mayo/Myriad ban applies to treatment regimens where a treatment is completely contraindicated by test results but not to regimens where the dosage is changed.

Second, the majority opined that Vanda claimed “an improved set of specific doses” but that INO did not. But in Vanda, the lower dosage in the claim was not more effective to treat the disorder; rather, it had less side effects. This appears to be eerily similar to the claims in INO, except that the INO claims effectively require a “zero dosage” for infants with LVD.

Third, the majority gives no explanation of the relevance of the fact that the use of nitric oxide without reference to LVD was already known. Indeed, the treatment steps in Vanda and Endo were also already known if severed from the diagnostic steps.

In summary, although the majority opinion repeatedly attests that there is no inconsistency between INO and the CAFC’s previous opinions, INO seems to be at odds with established precedent and appears to signal another period of uncertainty and inconsistency regarding the patentability of personalized medicine in the United States. Inventors of personalized medicine inventions would be wise to consider maintaining their innovations as a trade secret or explore markets outside of the United States to protect their inventions.

Detecting Disease Is Not a “Tangible and Useful Result” Eligible for PatentingThe federal appeals court with jurisdiction over questions of patent law has consistently held that methods of diagnosing a disease or other biological condition violate the Supreme Court’s ban on patenting “natural phenomena.” A recent decision reaffirmed this position (in the veterinary sphere), and for the first time in many years clearly articulated the test that a method of diagnosis fails: In the court’s opinion, mere diagnosis of a disease is not a “tangible and useful” result. However, due to the history of this test, it is of questionable value to patent stakeholders.

Diagnosis without Treatment Is Banned from Patenting

Readers of this blog are aware that the Supreme Court decided that a method of determining whether a patient receiving a drug is overdosed or under-dosed was banned from patenting in Mayo Collaborative Services v. Prometheus Laboratories, Inc., for the stated reason that the method was nothing more than a recognition of a “natural phenomenon” (i.e., that the artificial drug in question is effective at certain dosages and toxic at other dosages). The U.S. Court of Appeals for the Federal Circuit (CAFC) has interpreted Mayo to ban patents on methods of diagnosis, but to allow methods of medical treatment that depend on a diagnosis (i.e., personalized medicine). This trend, and the cases that define it, were blogged about in late 2018 in “Patenting Diagnostics and Biomarkers Six Years After Mayo.”

Since 2018, the majority of a triumvirate panel of the CAFC decided that a method of diagnosing previously un-diagnosable cases of myasthenia gravis by detecting a newly discovered biomarker fell under the Supreme Court’s Mayo ban (Athena Diagnostics, Inc v. Mayo Collaborative Services, LLC). The longest serving judge in the CAFC, Pauline Newman, offered a lengthy and thorough dissent. She emphasized that the patented method of diagnosis was new, used a wholly unrecognized approach, and produced a “tangible and useful result.” The “tangible and useful result” standard was taken from the earlier decision in Rapid Litigation Management Ltd., v. Cellzdirect, Inc., in which the CAFC upheld the patent-eligibility of a method of freezing cells. Because effectively preserving cells was considered a “tangible and useful result,” the court allowed the patent in that case to stand.

Diagnosis and Genotyping Not “Tangible and Useful”

The CAFC has again applied the “tangible and useful result” standard in the recent case of Genetic Veterinary Sciences v. LABOKLIN GmbH & Co. KG. In Genetic Veterinary Sciences the patent at issue claimed methods of diagnosing the canine disease known as hereditary nasal parakeratosis, a disease that causes unsightly crusting and fissures in the dog’s nose. The claims were directed to genotyping the SUV39H2 gene, which was discovered to cause the disease. Dependent claims included limitations for specific sequencing techniques, all of which had previously been known and used to sequence other genetic loci. A unanimous panel of the CAFC decided that the invention, as claimed, was not eligible for patenting because the invention was nothing more that the recognition of a natural phenomenon, notwithstanding claims to specific techniques such as PCR (or even the use of specific DNA primers). In comparing this invention to the cell cryopreservation technique in Cellz Direct, the court concluded that “Here, the Asserted Claims provide no tangible result save the observation and detection of a mutation in a dog’s DNA. While a positive and valuable contribution, these claims fall short of statutory patentable subject matter.”

A New Standard of Patent Eligibility?

The CAFC seems to be relying on a test for patent eligibility that rests on shaky ground. The “tangible and useful result” test is actually an old standard, first established by the CAFC in the State Street Bank case. Although this standard was never explicitly overturned, Associate Justice Stephen Breyer has written disparagingly of it in his concurring opinion in Bilksi v. Kappos and his dissent from the refusal of the petition for certiorari in Labcorp v. Metabolite Labs. The USPTO’s internal procedure forbids the use of this standard, as the USPTO interprets Supreme Court precedent as incompatible with it. The CAFC’s recent reliance on a standard that is apparently disfavored by the Supreme Court would seem to be setting the scene for yet another Supreme Court reversal of CAFC precedent. It could be risky for patent stakeholders to rely on this test to evaluate the patent-eligibility of inventions for two reasons. First, the CAFC is notorious for internal divisions, and a panel of different judges from those on the Veterinary Sciences panel might reject the “tangible and useful result” test. Second, this Supreme Court reverses the CAFC frequently, and it is foreseeable that the “tangible and useful result” test could be the subject of such a Supreme Court reversal. Apart from the potential demise of the test, it is anything but clear when an invention produces “tangible and useful” results based on the CAFC’s decisions. Why does a method freezing cells produce a tangible and useful result, but not a method of genotyping dogs? Until questions such as these are resolved, the “tangible and useful result” test remains of dubious value to the regulated public.

Eliminate Medical Treatment Patents? An Effort Approaches the Supreme CourtWill the Supreme Court’s banning of methods of medical diagnosis from patenting in Mayo v. Prometheus be extended to patents for medical treatments? Since Mayo some have argued that some methods of medical treatment should also be banned from patenting. Up until recently these arguments have seemed futile, as the U.S. Court of Appeals for the Federal Circuit has consistently held that treating a disease or other condition is not merely a “mental process” or “natural law.” Critics of medical treatment patents are now making their case to the Supreme Court.

Will the Supreme Court Ban Medical Treatment Patents in Hikma?

The U.S. Supreme Court is now considering whether to rule on this question in the Hikma Pharmaceuticals USA Inc. v. Vanda Pharmaceuticals Inc. case. The Hikma case involves “personalized medicine” combining diagnosis and treatment. The invention is a method for treating schizophrenia with the drug iloperidone that is tied to determining the patient’s tolerance for iloperidone with a new genetic test. The inventors discovered certain genotypes that indicate poor metabolism of iloperidone, that in turn increase the patient’s risk of cardiac complications of iloperidone treatment at doses of at least 12 mg/day (the risk at lower doses is not increased). The claims comprise determining whether the patient has the poor metabolizer gene, administering 12 mg/day or less iloperidone if the patient has the poor metabolizer gene, but otherwise administering 12-24 mg/day iloperidone. This is a classic application of personalized medicine, in which the side effects of treatment are minimized by varying the dosage based on the patient’s genotype.

After losing at the Federal Circuit, the infringer in Hikma urged the Supreme Court to invalidate the patent. The infringer’s argument is that the patent protects only a medical diagnostic method, and that the course of treatment based on the test results are mere “window dressing” to fool the court.

Similar arguments have been made to the Federal Circuit in several recent cases without success.

The Federal Circuit’s Defense of Medical Treatment Patents

Classen Immunotherapies Inc. v. Biogen IDEC was the seminal case in this line, although arguably not the strongest. A Federal Circuit panel considered the patent eligibility of a method of immunization of a subject that involves first immunizing a number of other subjects according to two immunization schedules and determining which has the least risk of side effects, then immunizing the subject according to the schedule with the lowest risk. The panel was unanimous that the act of immunizing two groups of subjects and comparing the results to determine the best immunization schedule, without applying that schedule to a subject, was not eligible for patenting. However, with the extra step of actually immunizing a subject based on the schedule, the majority of the panel held the method was patent eligible. While the majority held that a method of medical treatment is eligible for patenting, even if it involves mental steps, a dissenting judge thought the claims were “a monopoly over the scientific method itself.”

In Natural Alternatives International, Inc. v. Creative Compounds, LLC, the patent claimed methods of increasing a human subject’s anaerobic capacity by the administration of one of three amino acids (β-alanine, an ester of β-alanine, or an amide of β-alanine). The infringer argued that the method was merely a recognition of the natural principle that the listed compounds increase human anaerobic capacity, and therefore was a patent-ineligible natural law. Although the trial court agreed and declared the patents invalid, the majority of a three-judge panel for the Federal Circuit reversed the trial court.  The Federal Circuit panel decided that the method of increasing anaerobic capacity was more than a mere “natural phenomenon,” but rather a specific and useful application of a natural law: All drugs function according to natural principles, and a ban on methods of medical treatment that rely on natural principles would be tantamount to a ban on all methods of medical treatment.

In Endo Pharmaceuticals Inc. v. Mallinckrodt, LLC, the patent claimed a method of treating pain in renal patients by administering a controlled-release form of oxycodone, measuring the patient’s creatinine clearance rate, and then administering another lower dose of oxycodone until the AUC of oxycodone drops below 21 ng hr/mL. The inventors observed that a patient’s clearance rate of oxycodone was directly correlated with the creatinine clearance rate (a proxy for kidney function), and developed methods of preventing excessive levels of oxycodone by customizing the dosage regiment based on the patient’s ability to clear creatinine. This is another classic application of personalized medicine, in which the treatment is varied based on the metabolic idiosyncrasies of the individual patient. Although the trial court felt this was merely a natural law that could not be patented, the Federal Circuit panel again reversed. The Federal Circuit viewed the patent as not claiming a natural law, but as claiming a method of safely treating pain that applies a natural law.

The Federal Circuit’s Invalidation of Diagnosis Patents

There are several counter-examples in which attempts to patent methods of medical diagnosis alone have been denied by the Federal Circuit. These include Myriad (by the Federal Circuit, before it went to the Supreme Court), PerkinElmer Inc v. Intema Ltd., Univ. of Utah Research Foundation v. Ambry Genetics Corp., Cleveland Clinic Foundation v. True Health Diagnostics, LLC, Roche Molecular Systems, Inc. v Cepheid, and Ariosa Diagnostics, Inc. v. Sequenom, Inc. The facts vary in each of these cases, but in every case the end result of the method is diagnosis, not the treatment of a patient.

Will the Supreme Court Eliminate Patents for Medical Treatment Tied to Diagnosis?

This Supreme Court is hostile to patents, and if the Court agrees to review this decision, there is a real possibility that personalized medicine patents are in jeopardy. The Supreme Court has a nearly unbroken streak of deciding every patent it considers to be at least partially ineligible for patenting since the Chakrabarty case in 1980; of the four times the Court has considered that question since, it has at least partially invalidated every patent it considered except one (in J.E M. Ag Supply, Inc. v. Pioneer Hi-Bred International, Inc.). Both sides have arguments that are at least plausible at first blush.  Opponents to medical treatment patenting proffer that a method of diagnosis is of no use without a treatment; therefore adding a treatment step to a banned method of diagnosis actually adds nothing.  Proponents of the status quo argue that if a method of treating a medical condition may be patented in the absence of a diagnostic step, it makes no sense to ban the same method of treatment in combination with a diagnostic step.

One indication that the Supreme Court might take the case is that it requested the U.S. Department of Justice to provide its opinion on the case. Typically, the Supreme Court requests an opinion from the DOJ if it considers a given case to be of significance. The DOJ has not responded yet, but it will be interesting to see what position it takes. All in all, this case and its possible consideration by the nation’s highest court are additional illustrations of the instability and constant uncertainty surrounding current U.S. patent law.