Roche Molecular Sys., Inc. v. Cepheid

905 F.3d 1363

ROCHE MOLECULAR SYSTEMS, INC., Plaintiff-Appellant

v.CEPHEID, Defendant-Appellee

2017-1690

United States Court of Appeals, Federal Circuit.

Decided: October 9, 2018

Stephen S. Rabinowitz, Hughes Hubbard & Reed LLP, New York, NY, argued for plaintiff-appellant. Also represented by James W. Dabney, Mitchell Epner, Patrice Polyxene Jean, David E. Lansky, LYNN M. RUSSO.

Erik R. Puknys, Finnegan, Henderson, Farabow, Garrett & Dunner, LLP, Palo Alto, CA, argued for defendant-appellee. Also represented by Michael Paul Barker ; Grant L. Kim, Wesley Ellsworth Overson, Morrison & Foerster LLP, San Francisco, CA.

Before O’Malley, Reyna, and Hughes, Circuit Judges.

Concurring opinion filed by Circuit Judge O’Malley.

Reyna, Circuit Judge.

Appellant Roche Molecular Systems, Inc. ("Roche") owns U.S. Patent No. 5,643,723 ("the ’723 patent"), titled "Detection of a Genetic Locus Encoding Resistance to Rifampin in Microbacterial Cultures and in Clinical Specimens." The United States District Court for the Northern District of California found that the asserted claims of the ’723 patent are directed to patent-ineligible subject matter and are therefore invalid under 35 U.S.C. § 101. Roche appeals from a grant of summary judgment of invalidity. We affirm .

I. THE ’723 PATENT

The ’723 patent is directed to methods for detecting the pathogenic bacterium Mycobacterium tuberculosis ("M. tuberculosis" or "MTB"). ’723 patent col. 2 ll. 50–54. MTB infection is a major cause of tuberculosis. Id . col. 1 ll. 13–30. In 1994, before the priority date of the ’723 patent, the general method of MTB detection in a tuberculosis patient was known as sputum examination by the acid-fast bacilli smear. For this test, a biological sample taken from a patient is subjected to cell culture in a process that can take three to eight weeks. Id . col. 2 ll. 9–11. This test has limitations: it can identify the presence of bacterial cells in a biological sample, but cannot identify the cells as MTB. There is a need to know whether the MTB from a patient is resistant to antibiotics. The standard of care for MTB treatment at the time involved a regimen of antibiotics, with rifampin being a first-line anti-tuberculosis drug. Id. col. 1 ll. 31–33. Tuberculosis outbreaks, however, still resulted because of delays in diagnosis and reporting of rifampin-resistant tuberculosis due to the inability to rapidly identify MTB strains that are resistant to rifampin and put a patient on an appropriate alternative therapy. Id. col. 1 ll. 61–65.

Prior to the ’723 patent, scientists in the field had been working on diagnostic tests for faster detection of MTB, particularly rifampin-resistant MTB strains. Id . col. 2 ll. 18–46. It was speculated that "[g]enotypic detection of multi-drug resistant MTB [strains] directly from clinical specimens is theoretically the fastest and most direct step toward determining effective therapy for patients." Id . col. 2 ll. 39–42. It was known in the art that rifampin has a unique site of action on a particular gene that encodes the ß subunit of bacterial RNA polymerase ("the rpoB gene"). Id . col. 1 ll. 31–42. The rpoB gene is present in MTB and other bacterial species, and its deoxyribonucleic acid ("DNA") sequences were known to be highly conserved, with little variation from one bacterial species to another. In 1994, single site mutations in the rpoB gene that confer rifampin resistance in some bacteria, such as Escherichia coli ("E. coli"), were well characterized, making rpoB a prime candidate for studying rifampin resistance in MTB. Id . col. 1 ll. 42–52.

The inventors of the ’723 patent —scientists from Roche and the Mayo Foundation for Medical Education and Research ("Mayo")—sequenced the rpoB gene from various bacteria species, including MTB, obtained from a commercial vendor.¹ Id . col. 8 ll. 1–3 and col. 8 l. 15–col. 9 l. 20. After comparing rpoB DNA sequences across different species, the inventors discovered that the rpoB gene in MTB contains eleven "position-specific ‘signature nucleotides’ " (i.e., naturally occurring single nucleotide mutations) that are only present in MTB but not in other bacteria. Id . col. 2 l. 60–col. 3 l. 2. In other words, these naturally occurring signature nucleotides are like finger-prints of MTB: if an investigator detects one of the eleven signature nucleotides from a biological sample, she knows the sample contains MTB, and vice versa. These signature nucleotides, therefore, could be used to identify MTB using genetic testing, which is both faster and more accurate than the traditional MTB detection methods. Id . col. 2 ll. 9–31.

Based on these eleven MTB-specific signature nucleotides, the Roche inventors devised a diagnostic test that could (1) identify whether or not a biological sample contains MTB, and (2) if MTB is present, predict whether that MTB is a strain that is resistant to rifampin treatment. The diagnostic test of the ’723 patent involves subjecting DNA extracted from a biological sample taken from a patient (e.g., a tissue or fluid sample) to amplification by polymerase chain reaction ("PCR") using a short, single-stranded nucleotide sequence (a "primer") that can hybridize (i.e., bind) to at least one of the eleven position-specific signature nucleotides in the MTB rpoB gene.

PCR is a method of amplifying DNA exponentially. See Roche , 2017 WL 6311568, at *2. In PCR, a pair of primers effectively "flanks," or marks the start and finish of, the DNA segment—e.g., the rpoB gene or a portion of it—to be copied. Strands of DNA are then replicated between the primer pair by a DNA polymerase. This process is repeated until a sufficient number of copies of the desired DNA segment are generated. These copies, known as "amplification product," make it possible to detect whether a specific type of DNA is present. Id . It is undisputed that by the time of the invention in 1994, PCR had become a well-understood, routine, and conventional technique. Id .

After PCR is performed, the presence of DNA amplification product in sufficient copies from the reaction indicates that MTB is present in the biological sample. The absence of DNA amplification product (i.e., below the detection limit using standard assays) indicates that MTB is absent from the biological sample. The amplified rpoB DNA segment from the PCR can, in turn, be tested for the presence of known genetic mutations associated with rifampin resistance. Thus, the ’723 patent represents an improvement over the traditional sputum examination method for detecting MTB, as its genetics-based diagnostic method is faster and more accurate.

The ’723 patent provides two types of claims: (1) composition-of-matter claims for the primers used in the PCR, which could hybridize to the rpoB gene of MTB at a site that includes at least one of the eleven signature nucleotides ("the primer claims"); and (2) process claims for methods for detecting MTB that include amplifying target sequences by PCR and detecting amplification products, which, if present, indicate the presence of MTB ("the method claims").

Claims 1–13 are the method claims. Claim 1, the sole independent method claim, recites:

1. A method for detecting Mycobacterium tuberculosis in a biological sample suspected of containing M. tuberculosis comprising:

(a) subjecting DNA from the biological sample to polymerase chain reaction [PCR] using a plurality of primers under reaction conditions sufficient to simplify a portion of a M. tuberculosis rpoB [gene] to produce an amplification product, wherein the plurality of primers comprises at least one primer that hybridizes under hybridizing conditions to the amplified portion of the [gene] at a site comprising at least one position-specific M. tuberculosis signature nucleotide selected, with reference to FIG. 3 (SEQ ID NO: 1), from the group consisting

a G at nucleotide position 2312,

a T at nucleotide position 2313,

an A at nucleotide position 2373,

a G at nucleotide position 2374,

an A at nucleotide position 2378,

a G at nucleotide position 2408,

a T at nucleotide position 2409,

an A at nucleotide position 2426,

a G at nucleotide position 2441,

an A at nucleotide position 2456, and

a T at nucleotide position 2465; and

(b) detecting the presence or absence of an amplification product, wherein the presence of an amplification product is indicative of the presence of M. tuberculosis in the biological sample and wherein the absence of the amplification product is indicative of the absence of M. tuberculosis in the biological sample.

’723 patent col. 25 l. 57–col. 27 l. 6.² Dependent claims 2–13 add various limitations to claim 1 concerning PCR, PCR analysis, and primer preparation details.

Claims 17–20 are the primer claims. Independent claim 17 is representative and recites:

17. A primer having 14–50 nucleotides that hybridizes under hybridizing conditions to an M. tuberculosis rpoB [gene] at a site comprising at least one position-specific M. tuberculosis signature nucleotide selected, with reference to FIG. 3 (SEQ ID NO: 1), from the group consisting of [the same 11 nucleotides at the positions disclosed in claim 1].

Id . col. 28 ll. 14–31. Dependent claims 18–20 each add further limitations.³ Id. col. 28 ll. 32–46. Dependent claim 20, for example, discloses full DNA sequences of certain primers. Id . col. 28 ll. 44–46.

II. DISTRICT COURT PROCEEDING

Appellee Cepheid makes and sells "Xpert® MTB/RIF Assay," an assay that can detect MTB in a biological sample and can identify rifampin-resistant MTB. Roche brought a patent infringement case against Cepheid asserting that Cepheid’s product infringed the ’723 patent. Roche , 2017 WL 6311568, at *1. Cepheid filed a motion for summary judgment, arguing that all of the asserted claims claim patent-ineligible subject matter under 35 U.S.C. § 101. Id . at *9.

The district court granted Cepheid’s motion. Id . at *19. The district court found that the primer claims of the ’723 patent, "which have genetic sequences identical to...