“Personalized medicine” is of increasing importance in the medical field today. Personalized medicine, broadly speaking, refers to the design of medical treatments for the individual characteristics of each patient. As of March 2013, there were approximately 119 drugs approved by the U.S. Food and Drug Administration (FDA) with pharmacogenomics markers in their labels. Scientific research directed to finding “biomarkers,” or indicators of whether a patient is susceptible to, or has, a particular disease, or whether a patient will be responsive to a particular drug, has increased exponentially. This research has resulted in the formation of early-stage companies to commercialize diagnostics tests. Laboratories and early stage companies have, predictably, filed for patent protection for these diagnostics-type inventions.
Typical diagnostics-method claim types include “a method to diagnose Disease X in a patient, comprising detecting Biomarker Y and diagnosing the patient with disease X if Y is present,” and, “a method to determine whether patient with Disease X will be responsive to Drug Z, comprising detecting Biomarker Y and predicting efficacy with Z if Y is present.”
In 2012, the Supreme Court decided the case of Mayo Collaborative Services v. Prometheus Laboratories, Inc. 132 S. Ct. 1289 (2012) (“Prometheus”). In Prometheus, the Supreme Court invalidated a claim that was directed to a method of optimizing thiopurine dosage by determining the level of the thiopurine metabolite in the blood and adjusting the dosage accordingly. The way the Court appears to have reasoned the case is that the relationship between thiopurine metabolite and the efficacy of the drug is a “law of nature”, and thus not patentable. Further, the step of adjusting the dosage of the drug in accordance with the metabolite, while an active method step and thus potentially patentable subject matter, did not confer patentability to the claim, because the step of adjusting the dosage in accordance with the metabolite was a “conventional” step. Therefore, it was clear that correlations between biomarkers and a disease and/or efficacy for a particular drug were an unpatentable “law of nature,” and “conventional” methods steps based upon the correlation did not confer patentability upon such a claim.
This case has brought a significant chill to the fledgling diagnostics industry. By limiting the patent protection that can be obtained for diagnostics-type inventions, this decision has made investments in early-stage diagnostics start-up companies significantly less attractive.
The Prometheus decision is part of a trend in the 2000s by the courts, including the Supreme Court, to reduce the scope of 35 U.S.C. § 101, regarding patentable subject matter. 35 U.S.C. § 101 determines what kinds of discoveries and inventions can be patented. Up until the mid-2000s, 35 U.S.C. § 101 was interpreted quite broadly to include many types of subject matter; however, “laws of nature”, “natural phenomenon” (products of nature), and “abstract ideas” (performable by a mental process) were never patentable subject matter. But diagnostics method claims were, until Prometheus, considered patentable subject matter; diagnostics claims were routinely granted by the USPTO, and challenges to their status as patentable subject matter were not typically raised during litigation.
Two other cases, one older, but cited with approval in Prometheus, and two more recently decided cases, help give further insight into the meaning of Prometheus. One such case is Diamond v. Diehr, 450 U.S. 175 (1981). The claim at issue was directed to testing the temperature of rubber in a mold and automatically opening the mold at a certain temperature where the rubber was known to set. The Court in Prometheus held that the relationship of the temperature to the setting point of the rubber was a natural law, but that the step of opening the mold transformed the process into an application of the natural principle. In Perkin Elmer v. Intema 2012 U.S. App. LEXIS 23845 (Fed. Cir. Nov. 20, 2012), the claim was directed to a method of diagnosing Down’s syndrome by comparing the measured levels of a first ultrasound screening marker to the marker in the second semester and comparing the levels of the markers, thereby diagnosing Down’s syndrome. The Federal Circuit determined that the correlation between the marker and Down’s syndrome was a “natural law”, the comparison and diagnosing step was a “mental step,” and the measurement of the marker was a “routine and conventional activity.” This claim was not patentable. In Classen v. Biogen Idec, 659 F.3d 1057 (Fed. Cir. 2011), (decided before Prometheus, but still relevant) the claim was directed to methods of immunizing a patient based on detection of markers in a screening step. This claim was held patentable. Although the screening step was a natural principle (law) the active immunization step according to the screening step was a nonconventional specific application of the principle.
Reviewing the relevant cases discussed above, and also reviewing a selection of recently issued diagnostics-type method claims at the USPTO (in recently issued patents), some guidelines can be obtained regarding how to obtain a valid diagnostics claim in the wake of Prometheus.
- Detect a novel molecule (e.g., new single nucleotide polymorphism, or SNP). Detecting a novel (not previously disclosed) molecule is not a well-understood, conventional step, and thus this step confers patentability to the method. See, e.g., U.S. Patent No. 8,431,345 (April, 2013), which is directed to a method for determining risk of glaucoma progression comprising detecting a G nucleotide instead of an A nucleotide at position 31 of SEQ ID NO:X. A novel molecule is also potentially a candidate for a “kit” claim.
- Use a novel detection reagent (e.g., a new antibody). Use of a novel (not previously disclosed) detection reagent, such as an antibody, is not a well-understood conventional step, and thus confers patentability to the method. See, e.g., U.S. Patent No. 8,540,995 (September 2013), which is directed to detecting influenza A by contacting the specimen with monoclonal antibody IC9 with Accession No. PTA-8759 and detecting the binding of said antibody. One additional potential strategy is to add a detectable label to a known reagent, thus conferring novelty to the reagent. However, whether mere addition of a label to a reagent provides “enough” to confer patentability to the claim is questionable. A novel detection reagent may also potentially be the subject of a “kit” claim.
- Add a treatment or other active step based on the result of the diagnostic step. This should work where the correlation between the biomarker and the disease and/or the drug efficacy was not previously known. The treatment step may even be quite generic (directed to a class of drugs, rather than to a specific drug.) See, e.g., U.S. Patent No. 8,492,096 (July 2013), which is directed to a method for treating cancer in a patient by measuring for presence of reduced level of TGFBR1, followed by colonoscopy and treating said cancer with chemotherapy and/or radiation if cancer is found. Another strategy is to tie the result to treatment with a particular drug. This is useful for the entity which owns rights to the drug. See, e.g., U.S. Patent No. 8,445,535, which is directed to treating Alzheimer’s and other diseases by measuring for the presence of an ApoE4 allele, and treating the ApoE4(-) patient with medium chain triglycerides. A special note for claims of this kind is that there is likely no single entity that will perform all of the steps of the method; the testing laboratory will perform the screening, and the physician will perform the treatment step. These entities are likely not under each other’s control. Prior to 2012, the law of “induced infringement” required that one entity must perform all steps of the method, thus, a claim of this type would be very difficult to infringe. However, changes in the law in 2012 relaxed this standard, so that now, it is not required that one entity perform all of the steps. Instead, induced infringement can be found where one party, e.g., the testing laboratory, knowingly induces others (e.g., doctors) to commit the acts necessary to infringe the patent.
- Test for a novel combination of markers. In this case, testing for a novel “combination”, or markers that have not been tested together before, can confer patentability to a diagnostics method. The step of measuring for the new combination is not a well-understood, conventional step and therefore this step may confer patentability to the method. See, e.g., U.S. Patent No. 8,465,923 (June, 2013), which is directed to a method to assess response to an antibiotic treatment, comprising measuring the levels of a combination of six markers both before and after the antibiotic treatment and predicting a response to the antibiotic based on the levels. Another potentially available claim is that to a “kit” containing a non-obvious combination of probes, antibodies, or other reagents.
Thus, over a year after the Prometheus decision, some insight has been gained on how to obtain meaningful scope of coverage to a “diagnostics” claim. Although some types of diagnostics-type claims can no longer be obtained, the addition of certain factors may render a “diagnostics” type claim patentable. These factors include detection of a new marker or SNP, use of a new detection reagent, detection of a combination of biomarkers, and addition of an active treatment-type step based on the outcome of the screening to the method. This list is certainly not an exhaustive list of potential factors which can confer patentability to a diagnostics-type claim. No doubt further factors which will confer patentability to diagnostics claims exist and will come to light in the future. These modified diagnostics-type claims will hopefully continue to allow early stage companies involved in developing and commercializing diagnostics tests to obtain meaningful patent coverage, allowing further development of the field of personalized medicine.