During the week of January 7, the Patent Trial and Appeal Board (“the Board”) issued four decisions in TC 1600, one denying institution of inter partes review (“IPR”) and three final written decisions. Summaries of the decisions follow:
CSL Behring LLC, CSL Behring GMBH, and CSL Behring Recombinant Facility AG (“CSL”) v. Bioverativ Therapeutics Inc. (“Bioverativ”), IPR2018-01313 (Decision Denying Institution of IPR Entered January 9, 2019).
In its petition, CSL challenged claims 1-28 of U.S. Patent No. 9,623,091 (“the ’091 patent”), directed to methods of administering Factor IX using chimeric polypeptides comprising Factor IX and an FcRn binding partner to treat hemophilia B.
CSL challenged claims 1-28 of the ’091 patent as obvious over the combined teachings of: (1) Peters and Shapiro and/or (2) Metzner and/or U.S. Pat. Pub. No. 2008/0260755 (the ’755 publication) in view of Shapiro and Carlsson. The challenged claims are directed methods of treating hemophilia B in humans by intravenously administering doses of 50 IU/kg to about 100 IU/kg of a chimeric Factor IX peptide. Doses are administered about intervals of about 10 to about 14 days. In particular, the chimeric Factor IX peptide comprises Factor IX and an FcRn binding partner, such as Fc or albumin. Following administration of the chimeric Factor IX peptide, the human subject’s plasma Factor IX activity is maintained above 1 IU/dL. IPR2018-01313, Paper 10 at 3-4.
Peters was relied upon by the examiner in support of an obviousness rejection during prosecution; however, the examiner later admitted that Peters was not prior art following Bioverativ’s arguments and 1.132 declaration in support of the same. Id. at 10. CSL argued that the examiner’s removal of Peters as prior art was improper at least because (1) the relevant inquiry is whether the Peters disclosure was attributable only to inventors of the ’091 patent, and (2) the “averments that coauthors were carrying out experiments under Peters’ direction and control are insufficient” in view of the authorship section of Peters, “which indicates that other coauthors designed research, and ‘paragraphs 4-5 of Peters’s declaration, which suggest that he took credit for directing and controlling all the work.’” Id. at 10-11.
Analyzing CSL’s arguments regarding Peters, the Board turned to the Federal Circuit’s Katz decision which “determined that a reference by an inventor co-authored with non-inventors was not § 102(a) prior art on the basis that the co-authors contribution fails to rise to joint inventorship, and not on the basis that they made no contribution. Katz, 687 F.2d at 455–56 (‘the board . . . should have accepted that [the co-authors] were acting in the capacity indicated, that is, students working under the direction and supervision of appellant’).” Id. at 11 (quoting In re Katz, 687 F.2d 450, 455 (C.C.P.A. 1982)). The Board ultimately disagreed with CSL and determined that the record supports Bioverativ’s position that Peters was responsible for the content of the Peters reference, resulting in the reference not being prior art to the ’091 patent.
CSL further contended that the ’091 patent cannot properly claim priority to any of the provisional applications in its priority chain because they “do not satisfy the written description and enablement requirements for the ’091 patent claims.” IPR2018-01313, Paper 10 at 13. CSL’s arguments regarding written description and enablement largely focus on the Factor IX and FcRn binding partner disclosures in the provisional applications as including a large number of proteins having a wide variety of structural variations. CSL pointed to data for only one molecule of this class in the provisional applications and alleged that the limited data was insufficient to meet both the written description and enablement standards for the claimed invention. However, as Bioverativ pointed out in its response, CSL’s allegations are “based on an incomplete reading of the Priority Applications” and, thus, ignore disclosure relevant to the adequacy of the written description, such as the PCT applications, U.S. patents, and other references incorporated therein. Id. at 15. The Board agreed with Bioverativ and determined that CSL failed to meet their burden to show that the ’091 patent was not entitled to its priority date, which eliminated Peters as 102(b) prior art, and, accordingly, found CSL’s obviousness grounds based on Peters to be unpersuasive.
With respect to CSL’s allegation that claims 1-28 are unpatentable for being obvious over Metzner and/or the ’755 publication in view of Shapiro and Carlsson, the Board disagreed. According to the Board, CSL failed to establish (1) that once weekly doses taught by Shapiro maintained a subject’s Factor IX activity above 1 IU/dL, (2) that “inconsistencies in Metzner and the ’755 publication as to half-life data in animal models, and the application of that half-life data to dosing and dosing intervals in human patients” are not immaterial, and (3) why “one skilled in the art would not have found the half-life data for FIX-deficient mice to undercut an expectation of half-life extension in human patients suffering from hemophilia B.” IPR2018-01313, Paper 10 at 25-26.
As such, the Board concluded that CSL failed to demonstrate by a preponderance of the evidence that claims 1-28 the ’091 patent would have been obvious.
Nuevolution A/S (“Nuevolution”) v. Chemgene Holdings APS (“Chemgene”)
In all three of the IPRs summarized below, the Board instituted trial only on certain grounds. Nevertheless, after the Supreme Court’s decision in SAS Institute, Inc. v. Iancu, 138 S. Ct. 1348 (2018), the Board modified the institution decision in each of the following IPRs to include the remaining grounds in the trial proceedings.
As an initial matter, the Board construed one claim term, found in all independent claims of the challenged patents, consistently across all three IPR proceedings. All three decisions turn on the Board’s interpretation of this claim term, which is, “one or more reaction wells . . . each of said reaction wells.” According to the Board, “‘one or more reaction wells . . . each of said reaction wells’ means that the synthesis of any particular bi-functional molecule . . . is conducted within the same reaction well because ‘said’ indicates that the reaction wells . . . are the same.” See, e.g., IPR2017-01598, Paper 47 (Final Written Decision) at 21. The Board reiterated this point when explaining that “a particular bi-functional molecule is synthesized by a linker becoming attached to a respective molecule fragment and oligonucleotide identifier is, for example, a tube or a nanocompartment in that tube, it must be the same reaction well — the same physical containment and localized space.” Id.
In its petition, Nuevolution challenged claims 1, 2, 4, 5, 7-9, 16, 18-23, 27-30, 32, 33, 35, 36, 38-41, 42, 43, 45, and 46 of U.S. Patent No. 8,168,381 (“the ’381 patent”), which is directed to synthesis of encoded molecule fragments using mix and split techniques as well as template directed synthesis.
Independent claim 1 is representative and recites:
1. A method for synthesizing an encoded molecule comprising the steps of:
a) Adding a linker molecule L to one or more reaction wells;
b) Adding a molecule fragment to each of said reaction wells;
c) Adding an oligonucleotide identifier to each of said reaction wells;
d) Subjecting said wells to conditions sufficient to allow said molecule fragments and said oligonucleotide identifiers to become attached to said linker molecule, or conditions sufficient for said molecule fragments to bind to other molecule fragments and sufficient for said oligonucleotide identifiers to bind to other oligonucleotide identifiers, so as to form bi-functional molecules consisting of an encoded molecule and an oligonucleotide;
e) Combining the contents of said one or more reaction wells, to produce an admixture of said bi-functional molecules;
f) Optionally, distributing the combined product to one or more new reaction wells;
g) Optionally, repeating steps b) to f) one or more times; and
h) Contacting the resulting bifunctional molecule(s) of step e) or g) with one or more templates each capable of hybridizing to at least one of the oligonucleotide identifiers added in step c);
the linker molecule L contains at least one reactive group capable of reacting with a reactive group in the molecule fragment and at least one reactive group capable of reacting with a reactive group in the oligonucleotide;
the molecule fragments each contain at least one reactive group capable of reacting with a reactive group in the linker molecule L or a reactive group in another molecule fragment, and the reactive groups of each molecule fragment may be the same or different;
the oligonucleotide identifiers each contain at least one reactive group capable of reacting with a reactive group in the linker L or a reactive group in another oligonucleotide identifier, and the reactive groups of each oligonucleotide identifier may be the same or different;
the region of the oligonucleotide identifier added to each well in step c) which hybridizes to said template identifies the molecule fragment added to the same well in step b);
the steps a) to d) may be performed in any order;
the steps b) to d) in step g) may also be performed in any order;
the number of wells in steps a) and f) may be the same or different; and
the oligonucleotide template optionally is associated with a reactive group.
IPR2017-01598, Paper 47 (Final Written Decision) at 11-12.
The claims were challenged on eleven grounds of anticipation and/or obviousness in view of one or more of Gouliaev ’627, Gouliaev ’994, Pedersen, Franch ’929, Liu, and Freskgård. The Board instituted trial on three of the eleven grounds and its Final Written Decision discusses the same in detail. While the remaining eight grounds were added following the Supreme Court’s decision in SAS, the Board’s decision turned on the institution grounds.
First, the Board addressed Nuevolution’s asserted anticipation ground based on Freskgård and pointed out that “a key dispute here is whether the asserted prior art discloses that particular bi-functional (carrier) molecules are synthesized according to the claimed steps in the same reaction well.” Id. at 32. While Nuevolution argued that Freskgård disclosed that the claimed steps take place in the same reaction well (in particular—at Figure 12 which shows a 96-well microtiter plate where the disclosed process occurs in each well), Chemgene argued that Freskgård disclosed the use of different reaction wells for preparation of bi-functional carriers. The Board agreed with Nuevolution’s additional arguments regarding the claimed linker molecule, finding that “‘linker molecule L’ as claimed encompasses, at minimum, Freskgård’s nascent bi-functional complexes, such as shown in Freskgård’s Figure 13, to which particular molecule fragments and codon oligos are reacted and attached in individual wells of a microtiter plate.” Id. at 44. Unfortunately for Nuevolution, the Board ultimately agreed with Chemgene’s arguments that even if Freskgård discloses all features of the claimed method, Nuevolution relied upon individual steps from protocols of unrelated embodiments within Freskgård in violation of Net MoneyIN.
Second, in advancing its position that the claims are obvious in view of Freskgård, Nuevolution argued that each and every element of the claims is disclosed by Freskgård and it would be advantageous to combine the protocols of each of Freskgård’s embodiments (Modes 1 and 2) because Mode 2 provides versatile reactions, and Mode 1 provides high local concentrations of reactants. Id. at 50. The Board agreed because, while picking and choosing from embodiments may not be proper for an anticipation analysis, it may be for an obviousness rejection where, as here, there are “several evidence-backed reasons that we [the Board] find explain why the skilled person would have predictably and advantageously combined Mode 2 and Mode 1 techniques to arrive at the subject matter of the challenged claims.” Id. at 53. Chemgene’s arguments to the contrary did not persuade the Board otherwise at least because the “argument that there is no sufficient motivation or reasonable expectation of success in combining those techniques … is inconsistent with at least Freskgård’s express teachings and substantial expert testimony in this case.” Id. Accordingly, while the Board concluded that Nuevolution failed to demonstrate, by a preponderance of the evidence, that claims 1, 4, 5, 7–9, 16, 19–22, 27–30, 32, 33, 35, 36, and 38–41 of the ’381 patent are anticipated by Freskgård, the Board determined that Nuevolution had successfully demonstrated that these claims are obvious in view of Freskgård.
Third, the Board agreed with Nuevolution regarding its allegation that the claims having features of template-directed synthesis methods are obvious in view of Freskgård and Pederson. The Board explained that, “[w]hen combined with bi-functional molecules synthesized according to Freskgård’s Mode 2 (split-and-mix) technique,” they find that “the skilled artisan would have had reason to combine the split-and-mix and template-directed techniques to obtain the known advantages of each of those techniques.” IPR2017-01598, Paper 47 at 63. Since “the skilled artisan would have had a reasonable expectation of success in carrying out the method claimed,” at least for the same reasons discussed with respect to the anticipation and obviousness grounds based on Freskgård alone, Nuevolution demonstrated by a preponderance of the evidence that claims 1, 2, 4, 5, 7–9, 16, 18–22, 27–30, 32, 33, 35, 36, 38–41, and 46 are unpatentable for being obvious over Freskgård and Pedersen. Id. at 63-64.
In sum, the Board found Nuevolution had demonstrated by a preponderance of the evidence that claims 1, 2, 4, 5, 7-9, 16, 18-23, 27-30, 32, 33, 35, 36, 38-41, 42, 43, 45, and 46 were unpatentable.
In its petition, Nuevolution challenged claims 1, 3, 5, 6, 10-15, 17, 23-26, 31, 34, 37, 44, and 45 of the ’381 patent. Representative claim 1 is provided above. The claims were challenged on seven grounds of anticipation and/or obviousness in view of one or more of Pedersen, Franch ’929, Franch ’427, and Freskgård. The Board instituted trial on five of the seven grounds and its Final Written Decision discusses the same in detail. While the remaining two grounds were added following the Supreme Court’s decision in SAS, the Board’s decision turned on the institution grounds.
Similar to its reasoning on anticipation and obviousness grounds based on Freskgård in IPR217-01598, the Board found Nuevolution had demonstrated by a preponderance of the evidence that claims 1, 3, 5, 6, 10-15, 17, 23-26, 31, 34, 37, 44, and 45 were unpatentable.
In its petition, Nuevolution challenged claim 1 of U.S. Patent No. 8,951,728 (“the ’728 patent”), which is directed to methods for synthesizing encoded molecules.
Independent claim 1 is representative and recites:
1. A method for synthesizing an encoded molecule comprising the steps of:
a) Adding a linker molecule L to one or more reaction wells;
b) Adding a molecule fragment to each of said reaction wells;
c) Adding an oligonucleotide identifier to each of said reaction wells;
d) Subjecting said wells to:
conditions sufficient to allow said molecule fragments and said oligonucleotide identifiers to become attached to said linker molecule, or
conditions sufficient for said molecule fragments to bind to other molecule fragments and sufficient for said oligonucleotide identifiers to bind to other oligonucleotide identifiers;
e) Combining the contents of said one or more reaction wells;
wherein at least one reactive group of the linker molecule L reacts with a reactive group in the molecule fragment, or with a reactive group in the oligonucleotide;
wherein at least one reactive group of the molecule fragments reacts with a reactive group in the linker molecule L, or with a reactive group in another molecule fragment,
wherein at least one reactive group of the oligonucleotide identifiers reacts with a reactive group in the linker L, or with a reactive group in another oligonucleotide identifier; and
wherein the region of the oligonucleotide identifier added to each well in step c), which hybridizes to said template identifies the molecule fragment added to the same well in step b).
IPR2017-01603, Paper 45 (Final Written Decision) at 10-11.
The claims were challenged on fourteen grounds of anticipation and/or obviousness in view of one or more of Gouliaev ’627, Gouliaev ’994, Pedersen, Franch ’929, Franch ’427, and Freskgård. Id. at 11-12. The Board instituted trial on four of these grounds and its Final Written Decision discusses the same in detail. While the remaining ten grounds were added following the Supreme Court’s decision in SAS, the Board’s decision turned on the institution grounds.
Similar to its rationale discussed above with respect to Nuevolution’s anticipation and obviousness grounds relying on Freskgård in IPR2017-01598, the Board found Nuevolution had demonstrated by a preponderance of the evidence that claim 1 is unpatentable.
During the week of January 14–19, 2019, the Patent Trial and Appeal Board (“the Board”) issued 5 IPR institution decisions, 1 IPR final written decision, 1 PGR institution decision, and 1 PGR final written decision in TC 1600.
Rimfrost AS v. Aker BioMarine Antarctic AS, IPR2018-01178 and IPR2018-01179: The Board granted institution of Rimfrost’s two petitions. IPR2018-01178 challenged as obvious claims 1–32 of U.S. Patent No. 9,375,453 B2 (“the ’453 patent”); IPR2018-01179 challenged as obvious claims 33–61 of the ’453 patent. Aker BioMarine declined to file a Preliminary Response in either action. The challenged claims were directed to methods of producing polar krill oil extracts by denaturing the lipases and phospholipases, adding a polar solvent to obtain specific oil compositions, and formulating the polar krill oil with a carrier for oral consumption.
Across both petitions, Petitioner Rimfrost asserted that certain claims would have been obvious over the combined teachings of Breivik II, Catchpole, Bottino II, and Sampalis I (“the primary references”).
Breivik II disclosed applying heat to denature krill and result in a denatured krill product for E. superba krill by teaching an extraction step at supercritical pressure, using CO2 containing 10% ethanol, and isolating the lipid fraction from the ethanol; formulating with a carrier for oral consumption was well-known, as exemplified by Sampalis I. IPR2018-01178, Paper 7 at 16–17. The specific percentage compositions recited in claim 1 were naturally present, as evidenced by Catchpole, Bottino II, Breivik II, and expert testimony. See id. at 17–19. From there, the relative proportions of the krill oil constituents would have been varied in predictable ways with a reasonable expectation of success. Id. at 20.
In IPR2018-01179, the Board agreed that Breivik II taught or suggested the claimed method steps for producing polar krill oil from E. superba krill by short-term heating to provide a denatured krill product and extracting polar krill oil from that product using a polar solvent; “Catchpole disclosed the components of polar krill oil and the amounts thereof”; Bottino II disclosed “the triglyceride content of polar krill oil”; and Sampalis I disclosed polar krill oil formulated for oral consumption. IPR2018-01179, Paper 7 at 20–22. While the Board disagreed that Breivik II expressly disclosed the astaxanthin esters percentages, the Board found that Breivik II as a whole, combined with Rimfrost’s expert’s testimony, nonetheless suggested the claim limitation. See id. at 21–22. The Board’s analysis in IPR2018-01179 tracked that of its institution decision in IPR2018-01178. Thus, split across the two actions, the Board agreed that Petitioner Rimfrost had demonstrated a reasonable likelihood that the contested claims were obvious by the combined references, and ordered institution on all claims (claims 1–61) of the ’453 patent. See IPR2018-01178 at 22–23; IPR2018-01179 at 22–23.
Novartis Pharm. Corp. v. Plexxikon Inc., IPR2018-01287: The Board denied institution of Novartis’s petition. Novartis challenged claims 1, 2, 4–6, 9, and 11–12 of U.S. Patent No. 9,469,640 B2 (“the ’640 patent”), a compound and method of treatment patent titled “Compounds and Methods for Kinase Modulation, and Indications Therefor,” as anticipated by U.S. Patent No. 7,994,185 B2 (“the ’185 patent”). As part of its petition, Novartis also challenged the ’640 patent’s priority date. IPR2018-01287, Paper 17 at 10. [We note that institution of PGR2017-00069 on related U.S. Patent No. 9,844,539 B2 was denied the same day—see following case summary.]
The ’640 patent describes a genus of compounds with a given chemical formula (claim 1), specific substitutions of that compound (claims 2–8), a pharmaceutical composition containing the chemical compound (claims 9–10), and methods of treating certain cancers using the chemical compound (claims 11–12). See id. at 2–4.
Novartis specifically argued: (1) A certain priority application, “P2,” disclosed an “enormous genera of chemical compounds” covering “an effectively incalculable number of compounds”—it did not provide sufficient disclosure of the claimed subgenera (species), id. at 10; (2) P2 provided no “blaze marks” to direct a POSA to focus on the bond or –N(H)C(O)– options for a particular substituent (L1) on the claimed compounds, particularly when the claims covered “trillions” of compounds (indeed, only three examples in P2 were within the claim scope, and zero used L1 as a bond), id.; and (3) without data showing activity on kinase proteins where L1 is a bond, there was no enablement for a compound where L1 is a bond, as a kinase inhibitor, or for any other pharmacological purpose, id. at 10–11. After discussing the prosecution history in depth, see id. at 5–10, the Board gave short shrift to Novartis’s arguments and found that these arguments were substantially the same as those raised during prosecution, see id. at 10–13. Given that determination, the Board declined to institute review, noting that Petitioner Novartis failed to “point [the Board] to additional facts or evidence that would warrant [its] reconsideration of the arguments on the basis of Examiner error.” Id. at 13–15.
Novartis Pharmaceuticals Corp. v. Plexxikon Inc., PGR2018-00069: The Board denied institution of Novartis’s PGR petition. Petitioner Novartis challenged claims 1–2, 4–9, 11–12, and 14–19 of U.S. Patent No. 9,844,539 B2 (“the ’539 patent”) as unpatentable as anticipated (Ground 3), premised on the ’539 patent being eligible for PGR based on either lack of written description (Ground 1) or enablement (Ground 2). [We note that, on the same day, the Board denied institution in IPR2018-01287 on related U.S. Patent No. 9,469,640—see previous case summary.]
Similar to the ’640 patent contested in IPR2018-01287, the ’539 patent related to compounds capable of modulating kinases. The ’539 patent generally disclosed two sets of compounds, “formula Ib” and “formula Ic,” with various proposed substitutions to their structures. PGR2018-00069, Paper 16 at 3–4. Because PGR is only available if the patent had an “effective filing date” on or after March 16, 2013, Petitioner Novartis contended that the challenged claims were not adequately described or enabled in any parent application prior to March 16, 2013. Id. at 8–9. As explained below, the Board rejected Novartis’s approach.
Ground 1, Lack of Written Description: Similar to IPR2018-01287, Novartis contended that the specification of a certain priority application (“P2”) lacked any “blazemarks” providing adequate description of the subgenus of compounds recited in the challenged claims, in particular the specific combination of claimed L1 and R1 substituents. Id. at 10. Rejecting Novartis’s approach, the Board explained that “‘blaze marks’ do not need to identify what species among disclosed species are preferred; it is only necessary that the species or sub-species is identified.” Id. at 14. Furthermore, for written description, the Board considers “the entire claimed subject matter in view of the disclosure.” Id. at 15. And finally, a “‘broad and complete generic disclosure, coupled with extensive examples fully supportive of the limited genus now claimed’ can provide written descriptive support for a later claimed subgenus that excludes some species from the original genus.” Id. at 17 (citation omitted). The Board agreed with Patent Owner Plexxikon that formula Ib taught a genus of which one option—N(H)C(O)—was a preferred substituent for L1, and formula Ic described L1 as a bond, and thus a POSA would have been guided to focus on those options. Id. at 15. Regarding R1, the Board rejected that a POSA would not have “be[en] able to clearly discern the sub-genus now claimed from the generalized formula Ib or Ic disclosed in P2,” which described 23 options. Id. at 16–17. Additionally, the Board agreed with Plexxikon that the specification provided numerous examples showing the “breadth of options” for R1, with three compounds falling within the claims. Id. at 17 (citation omitted). Accordingly, the Board found adequate written description in P2 and determined that the ’639 patent was not PGR-eligible on that basis.
Ground 2, Lack of Enablement: Novartis raised three main arguments: (1) P2 provided no guidance on “how to make the trillions of claimed compounds wherein L1 is a bond,” particularly with only a prophetic example and purported difficulties in using the referenced Suzuki reaction, see id. at 18; (2) P2 provided no indication that the scheme would work for the entire claimed genera where L1 was a bond, id. at 18–19; and (3) the solubility and kinase interactions would not support the large genus of R3 substituents, id. at 19. The Board again sided with Patent Owner Plexxikon, explaining that despite possible difficulties with the Suzuki reaction, it was a well-known reaction scheme in the pharmaceutical industry and P2 provided an example of how to use it to make compounds of the claimed genus where L1 is a bond. Id. at 21. Notably, the Board found that even if some of the compounds within the claim scope were “difficult or impossible to make,” the claims would not necessarily fail to be enabled. Id. at 21. In the Board’s view, P2 disclosed a generic class of compounds with a common core structure, common utility (modulating protein kinases), and representative assay results—the number of inoperative embodiments within the claim scope would be relevant only if it forced a POSA to experiment unduly to practice the invention, and the Board was not persuaded that was the case. Id. at 21–22. Accordingly, the Board found adequate enablement in P2 and thus determined that the ’639 patent was not PGR-eligible on that basis.
Thus, the Board found that Petitioner Novartis failed to meet its burden to show lack of written description or enablement, and thus failed to establish that the ’539 patent was eligible for post-grant review. Id. at 22.
Natera, Inc. v. Illumina, Inc., IPR2018-01317: The Board denied institution of Natera’s petition. IPR2018-01317 challenged as obvious certain claims U.S. Patent No. 9,493,831 B2 (“the ’831 patent”), which claims a method to detect fetal aneuploidy by selectively enriching non-random polynucleotide sequences, generating libraries of these selectively enriched sequences, and using the selectively enriched sequences for detection of fetal aneuploidy (an abnormal number of chromosomes or parts of a chromosome, as seen in, for example, Down Syndrome). Petitioner Natera asserted obviousness of some claims over Fluidigm (Ground 1); others over Fluidigm and Sequenom (Ground 2); and one claim over either Fluidigm and Harismendy (Ground 3) or Fluidigm, Sequenom, and Harismendy (Ground 4). IPR2018-01317, Paper 15 at 5–6.
The Board agreed with Natera that Fluidigm taught to analyze tagged target nucleotide sequences by DNA sequencing, id. at 10, but found that it only taught this for the second amplification step, not the third as required by the claims, see id. at 9–10. Furthermore, the Board determined that Natera did not “sufficiently address why an ordinary artisan would have used sequencing to analyze the products of the third amplification step, which is itself a detection step.” Id. at 10. Accordingly, the Board found that Natera had not shown a reasonable likelihood of prevailing on claim 1, and therefore its arguments as to dependent claims 2–13 also failed. Id. at 10–11. For independent claim 14 and dependent claims 15–24, the Board found Natera failed to support why a POSA would have used an indexing sequence from the prior art in the claimed method. See id. at 12–13. The Board swiftly denied Natera’s remaining grounds. See id. at 13–14.
A note on the Board’s reasoning: the Board explained that the obviousness inquiry should focus not on “whether each element existed in the prior art, but whether the prior art made obvious the invention as a whole. It is not proper to dissect a claim and reconstruct it in piecemeal fashion by picking and choosing from among different methods using the challenged patent as a blueprint.” Id. at 11–12. Furthermore, where a prior art reference teaches that something “can” or “should” be done, the burden remains on the Petitioner to explain why a POSA would have modified the disclosed method. Id. at 12–13.
CSL Behring LLC v. Bioverativ Therapeutics Inc., IPR2018-01345: The Board denied institution of CSL’s petition. IPR2018-01345 challenged claims 1–17, 20, 22, 24, and 28 of U.S. Patent 9,623,091 B2 (“the ’091 patent”), a method for treating hemophilia B. The ’091 patent is “directed to methods of administering Factor IX using chimeric polypeptides comprising Factor IX and an FcRn binding partner . . . to treat hemophilia B” in humans. IPR2018-01345, Paper 10 at 3.
Petitioner CSL asserted one ground of unpatentability: that claims 1–17, 20, 22, 24, and 28 were obvious over Peters 2007 and U.S. Patent No. 7,404,956 (“the ’956 patent”) in view of Shapiro and Carlsson. Id. at 4. CSL contended that Peters 2007 disclosed “the existence of a recombinant fusion protein rFIXFc, its intended purpose, animal testing demonstrating efficacy and an extended half-life, and a prediction based on the testing as to half-life in humans,” see id. at 8; the ’956 patent disclosed rFIXFc as a chimeric protein involving Fc and FIX in the first chain and a second chain with unmodified Fc with a purpose to increase serum half-life, see id; and Shapiro and Carlsson taught FIX prophylaxis regimens, see id. at 9. Given this knowledge, CSL asserted that a POSA would have been motivated to administer the extended half-life rFIXFc molecule from Peters 2007 and the ’956 patent in similar dose amounts as used for prophylaxis with rFIX, but with longer intervals of about 10–14 days. Id. at 9.
The Board rejected CSL’s arguments. First, the prior art supported more frequent dosing, such as 2–3 times per week, to maintain the recited FIX activity level—which CSL’s expert did not directly rebut. See id. at 16–17. Second, neither Shapiro nor Carlsson taught that a specific IU/dL value was necessary for prophylaxis, and indeed the consensus after Shapiro was that at least twice-weekly dosing was required. See id. at 17–18. Third, while Peters 2007 adequately taught that rFIXFc has good in vivo clotting activity, Peters 2007 did not address how the limited duration of clotting effect is consistent with maintaining plasma FIX activity levels above 1 IU/dL over a 10–14 day interval—and indeed, later publications contradicted Peters 2007. See id. at 18–19. Fourth, the Board was suspicious of the link between animal data and human administration, thus weakening CSL’s reasonable expectation of success arguments. See id. at 19 (“the animal data from Peters 2007, and associated reasoning as to its predictive value, falls short in the absence of an explanation how the data relating to the ability of the FIX-deficient dogs to clot is not contrary to a dosing interval of 10–14 days”). Finally, the Board agreed that CSL failed to establish a reasonable expectation of success that routine or predictable optimization of existing regimens would have resulted in the claimed dose and dosing schedule. See id. at 19–21. For example, Peters 2007’s prediction of a more than two-fold longer half-life for rFIXFc than for rFIX, see id. at 7, 19–20, contradicted Peters 2007’s statement that this was “thus consistent with a once weekly dosing regimen,” id. at 20 (citation omitted). The ’956 patent’s disclosure to administer “at any interval” was, according to the Board, “insufficient to invite routine optimization.” Id. (citation omitted). CSL’s claims to routine optimization were further undermined by the Examiner’s statement that “’optimization is not routine or predictable to arrive at what is claimed in the absence of human data’ where there is ‘evidence that dosing studies in animal models are often different from what occurs in humans.’” Id. at 21 (citation omitted).
The Board thus found that Petitioner CSL did not establish a reasonable likelihood of prevailing on claim 1. Given that nothing in the dependent claims remedied the issues identified for claim 1, the Board denied institution. See id. at 22.
Elysium Health Inc. v. Trustees of Dartmouth College, IPR2017-01795:
In this Final Written Decision, the Board found challenged claims 1 and 3–5 unpatentable. Petitioner Elysium Health challenged claims 1–5 of U.S. Patent No. 8,383,086 B2 (“the ’086 patent”) as anticipated. Although originally instituted only on claims 1 and 3–5, post-SAS, the Board instituted review on all five claims. IPR2017-01795, Paper 39 at 2–3. Elysium challenged patentability on two § 102 grounds: Goldberger (Ground 1) and Goldberger & Tanner (Ground 2). Id. at 5.
The ’086 patent related to the production of nicotinamide riboside (“NR”) and compositions containing NR, as well as compositions containing an effective amount of NR to treat NR-deficiency related disorders. Id. at 4.
- Claim Construction: The Board provided claims constructions of three terms. First, “pharmaceutical composition” was construed to mean “a composition, including a food composition, which contains NR as an active agent in an amount effective for the treatment or prevention of a disease or condition associated with the nicotinamide riboside kinase pathway of NAD+ biosynthesis.” See id. at 10–11. Second, the Board construed the limitation that the NR “is isolated” to mean “that the nicotinamide riboside is separated or substantially free from at least some of the other components associated with the source of the molecule such that it constitutes at least 25% (w/w) of the composition.” See id. at 12. Third, the Board reiterated its earlier, lengthy (61-word) construction of the term “carrier.” See id. at 14–15.
- Prior Art References: Goldberger determined that skim milk was preventative for canine blacktongue, see id. at 15–16, with subsequent research showing that NR, a precursor to NAD+, was one of the milk components, and that NR increased biosynthesis of NAD+, see id. at 16. Goldberger & Tanner disclosed a clinical study where humans were administered buttermilk to treat and prevent NAD+ deficiency, pellagra, with subsequent research showing that milk contained NR, a NAD+ precursor, and that NR increased biosynthesis of NAD+. See id. at 31–32.
- 102 Anticipation of Claims 1–5 (Ground 1, Goldberger): The Board found claims 1 and 3–5 anticipated by Goldberger. The Board found that Goldberger taught using skim milk to treat an NAD+ deficiency, the record showed that the NR in milk was bioavailable, and thus it met the limitation of claim 1. See id. at 17. It was also, clearly, administered orally. See id. at 24–25. The Board rejected Dartmouth’s argument that the admixture needed to be “purposefully mixed” or created via “some affirmative purposeful act.” Id. at 23–24. Goldberger also anticipated claim 3’s limitation of being “food,” see id. at 27; claim 4’s limitation that the pharmaceutical composition comprised “one or more of tryptophan, nicotinic acid, or nicotinamide” since it was well-known that NAD+ precursors in milk included nicotinamide and tryptophan, see id. at 28–29; and claim 5’s limitation that the pharmaceutical composition “increases NAD+ biosynthesis upon oral administration” given that oral NR administration boosted NAD+ production and prevented blacktongue, see id. at 29–31. Claim 2 was the only claim not anticipated by Goldberger because the skim milk was not isolated from a natural or synthetic source with at least 25% (w/w) NR. See id. at 26–27.
- 102 Anticipation of Claims 1–5 (Ground 2, Goldberger & Tanner): The Board also found claims 1 and 3–5 anticipated by Goldberger & Tanner. The Board’s analysis of Goldberger & Tanner largely tracked that of Goldberger, above. See generally id. at 32–42. The Board thus found claims 1 and 3–5 anticipated by Goldberger & Tanner, but not claim 2, since there was “no persuasive evidence in the record to show that NR constituted at least 25% (w/w) of the buttermilk used by Goldberger [&] Tanner.” See id. at 37–38.
Nippon Suisan Kaisha Ltd. v. Pronova Biopharma Norge AS, PGR2017-00033:
In this Final Written Decision, the Board found challenged claims 1–21 and 26 of U.S. Patent No. 9,447,360 B2 (“the ’360 patent”) unpatentable; granted-in-part and denied-in-part Patent Owner’s motion to amend and cancel claims 22–25; and denied-in-part and dismissed-in-part Patent Owner’s motion to exclude evidence. Having granted cancellation of claims 22–25, the Board considered only claims 1–21 and 26. See PGR2017-00033, Paper 37 at 2, 7–8, 26–27. Thus, at issue were seven grounds of unpatentability relating to § 112(b) (Ground 1); § 102(a)(2) (Grounds 3, 5, and 6); and § 103(a) (Grounds 7–9). Id. at 7–8.
The ’360 patent related to a process for “purifying crude organic oils, most preferably fish and other marine oils.” See generally id. at 3–5. The disclosed process comprises an aqueous fluid processing step to remove undesired hydrophilic components from the crude marine oil, followed by a stripping processing step (e.g., distillation) to remove undesired lipophilic components. The ’360 patent explains that in commercial production, processed volumes are high thus causing scaling build-up, and the crude oil id deacidified by alkali refining prior to stripping—this removes the free fatty acids (“FFAs”) present in the crude oil and also requires external working fluid be added for the stripping process. The invention purportedly reduced the risk of scaling by performing partial alkali refining (i.e., using less than the amount of base needed to neutralize all the free fatty acids to soaps) before stripping.
- 112(b) Indefiniteness of Claims 1–21 (Ground 1): The Board agreed that claims 1–21 were indefinite under either the Nautilus standard or the Packard standard. Id. at 12–15. Claims 1, 4, 6, and 11, in particular, recited both “broad” limitations and “narrow” limitations for the same components, making “the scope of coverage for these claims unclear,” see id. at 12–13, and the specification and prosecution history left certain terms “amenable to two or more plausible claim constructions,” see id. at 14 (citation omitted). The Board rejected Patent Owner Pronova’s rebuttal as to claims 4, 6, and 11, which would have required a “backward ‘fixing’ of the claim” by adopting the broadest reasonable construction of certain terms and thus “rewriting the claims to take out the narrower limitations, which we will not do,” noting that this approach conflated claim construction and indefiniteness. Id. at 14–15.
- 102(a) Anticipation of Claims 1–7 and 11–26 (Ground 3, Doisaki); 1, 2, 4–7, 11–15, 22, and 26 (Ground 5, Hata); and 1, 2, 4–7, 12, 13, 22, and 27 (Ground 6, Yamanouchi): The Board agreed that claims 1–7 and 11–26 were anticipated by Doisaki (Ground 3) and accordingly, declined to address whether they were also anticipated under Grounds 5 or 6. Id. at 26–27. Doisaki disclosed a process to reduce environmental pollutants in marine and feedstock oils where the oil would be subjected to distillation to remove undesired hydrophilic components, such as FFAs, from the crude marine oil composition. Id. at 22. Although Doisaki was “silen[t]” on certain conditions of the wash step, the Board found that claim 1 did not require specific parameters such as duration, and Pronova failed to show how any specific parameters would have affected the separation of undesired hydrophilic components. Id. at 22–23. The Board likewise agreed that Doisaki’s Example 4 warm water wash read on claim 26. See id. at 24–25. Given that Nippon’s dependent claim arguments were unrebutted by Pronova, the Board agreed that independent claims 1 and 26 and dependent claims 2–7 and 11–21 were unpatentable under 35 U.S.C. § 102(a)(2) as anticipated by Doisaki. Id. at 26–27.
- 103(a) Obviousness of Claims 1–26 (Grounds 7, 9) and Claims 1–25 (Ground 8): Given the Board’s anticipation decision and the similarity between Nippon’s anticipation and obviousness rationales, the Board focused on the obviousness of claims 8–10. (The Board declined to reach whether the challenged claims were obvious under Ground 8 or 9.) For claim 8, the main question was whether a POSA “would have had a reason to contact the crude marine oil composition with an aqueous fluid comprising a sub-equimolar amount of base relative to the FFA content of a marine oil composition,” see id. at 32, and for claim 10, “whether a POSA would have been ‘led away’ from using NaOH as a base to partially neutralize the FFA content of an oil composition,” id. The Board found that the ’360 patent itself acknowledged using alkali refining to remove FFAs in crude oil, and Nippon’s secondary references taught neutralizing FFAs, including with NaOH use as part of degumming. Id. at 35. Because traditional degumming required “extremely small amounts of base” and would remove “almost no free fatty acids,” the Board agreed with Nippon’s expert that the amount would have to be sub-equimolar, id. (citation omitted), and thus was nothing more than routine optimization and a desire to improve upon what was generally known, see id. at 35–36. Rejecting Pronova’s arguments regarding the scope of prior art, the Board pointed out that “the test for obviousness is not that the claimed invention was ‘expressly suggested in any one or all of the references,’ but ‘what the combined teachings of the references would have suggested to those of ordinary skill in the art.’” Id. at 37 (citation omitted). In rejecting Pronova’s rebuttal evidence, the Board would not read in an unclaimed purpose for the method such as quality, see id. at 40, declined to limit the motivation for the method to one claimed rationale (“In determining whether the subject matter of a patent claim is obvious, neither the particular motivation nor the avowed purpose of the patentee controls.”), id. at 40 (quoting KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 419 (2007)), and rejected Pronova’s evidence of teaching away, see id. at 41–43 (further recognizing that “a given course of action often has simultaneous advantages and disadvantages, and this does not necessarily obviate motivation to combine,” (quoting Medichem, S.A. v. Rolabo, S.L., 437 F.3d 1157, 1165 (Fed. Cir. 2006))).
- Motion to Exclude Evidence: The Board issued routine rulings on certain motions to exclude evidence. See generally id. at 53–57. Of note, the Board declined to exclude two references submitted in reply to Patent Owner’s argument to show the state of the art, reasoning that the proper challenge for “belatedly presented evidence” should have been an authorized motion to strike or authorization for further briefing to address newly-raised evidence. See id. at 55.
During the week of January 21, 2019, the Patent Trial and Appeal Board (“the Board”) issued two decisions in TC 1600. In those, the Board denied institution of inter partes and post-grant reviews.
Benson Hill Biosystems, Inc. v. The Broad Institute Inc., Presidents & Fellows of Harvard College & Massachusetts Institute of Technology, No. PGR2018-00072 (Decision Denying Institution of Post-Grant Review Entered January 22, 2019). The Board refused to institute a post-grant review sought by Benson Hill Biosystems, Inc. (“Benson Hill”) against U.S. Patent No. 9,790,490 (“the ’490 patent”) owned by The Broad Institute, Inc., President and Fellows of Harvard College, and Massachusetts Institute of Technology (“Patent Owner”). Benson Hill raised seven grounds for challenge: two based on lack of written description, two based on lack of enablement, one based on indefiniteness, one based on lack of utility, and one based on obviousness. IPR2018-00072, Paper 11 at 5. Benson Hill relied on the last two as alternative grounds if the Board disagreed with its claim construction position. Id. The § 112 grounds also depended on a claim construction issue, and the Board’s decision centered on that aspect of the case.
The claims pertained to non-naturally occurring Clustered Regularly Interspaced Short Palindromic Repeats (“CRISPR”) systems for targeting nucleic acid sequences. The systems comprised a Cpf1 effector protein, a novel RNA-endonuclease, and an engineered guide polynucleotide, which together form a complex designed to hybridize to a target nucleic acid sequence. The systems also lack a trans-activating crRNA (“tracr”) sequence. Id. at 3–4.
Benson Hill sought a construction under which the Cpf1 effector protein performs the specific function of cleaving a target sequence. Id. at 9. But the Board rejected this construction. Neither the plain language of the claims nor the term “effector protein” required this functionality, and the written description described systems with functions other than cleaving, including preferred embodiments with no cleavage activity whatsoever. Id. at 11–14. The Board also rejected Benson Hill’s prosecution history estoppel arguments, reasoning that Patent Owner’s arguments to overcome an anticipation rejection relied on the claimed systems’ hybridization function (not any cleaving function), and Patent Owner’s amendments required the systems only to lack a tracr sequence, not have cleaving activity. Id. at 14–16.
The Board therefore dismissed Benson Hill’s § 112 claims, which were premised on this “cleavage” limitation. Id. at 17–18. Benson Hill’s two alternative grounds—obviousness and lack of utility—also failed. The obviousness ground required the Board to construe the systems as having no functional activity at all, id. at 19, but the Board refused to limit the claims this way because they required hybridization activity and the written description contemplated other additional functionality (including functions not involving cleavage). Id. The lack of utility ground failed for the same reasons as the § 112 claims. The Board thus denied the petition.
Thermo Fisher Scientific Inc. v. Regents of the University of California, No. IPR2018-01347 (Decision Denying Institution Entered January 22, 2019). Thermo Fisher Scientific, Inc. (“Thermo”) filed a petition for inter partes review (“IPR”) directed to five claims of U.S. Patent No. 9,085,799 (“the ’799 patent”), owned by The Regents of the University of California (“UC”). Thermo raised one ground of unpatentability, anticipation under § 102(b) by U.S. Patent Application No. 2004/0142344 (“Bazan”), the first published application in the ’799 patent’s family. IPR2018-01347, Paper 10 at 5. Thermo’s petition relied on a priority challenge: it asserted the ’799 patent was not entitled to the August 26, 2002 filing date of the family’s provisional application or the August 26, 2003 filing date of the first non-provisional application in the chain, and was thus entitled to a priority date no earlier than September 12, 2007, the filing date of the second non-provisional application. Id. at 18. Bazan, the asserted anticipatory reference, published before this date, on July 22, 2004. Id.
The claims at issue covered methods of analyzing polynucleotides in a sample using light-harvesting multichromophore systems based on fluorescence resonance energy transfer (“FRET”). Id. at 3. The samples are contacted with systems comprising a “signaling chromophore” and a “water-soluble conjugated polymer” that can transfer energy to the signaling chromophore to provide a “greater than 4 fold increase in fluorescence emission” from the signaling chromophore over what can be achieved by “direct excitation of the signaling chromophore” absent the polymer. Id. at 5. The claims then require applying a light source to the sample and detecting whether light is emitted from the signaling chromophore. Id.
After construing “greater than 4 fold increase in fluorescence emission” in accordance with its plain and ordinary meaning and “in the absence of the polymer” broadly to encompass “without polymer present” and “lack of energy transfer (via FRET),” id. at 15–18, the Board turned to the priority issue. UC first claimed that Thermo raised impermissible grounds under § 311 (which allows challenges under only § 102 and § 103) because the anticipation claim was a disguised attack on written description under § 112. Id. at 21. But the Board explained that it may assess written description issues bearing on priority claims and the applicability of prior art for allowable challenges under § 311. Id. at 22–23.
The Board then determined the claims were entitled to a priority date prior to Bazan’s publication date because Thermo had failed to establish that the provisional application and first non-provisional application lacked adequate written support for the claims. Id. at 33. Thermo had argued the “greater than 4 fold increase in fluorescence emission” limitation represented a broad genus of emission increases, but the provisional and first non-provisional application provided only one species within that genus. Id. at 28. The Board disagreed, finding instead that four covered species were disclosed. Id. at 29. Thermo had asserted three examples were outside the claims, including one with two signaling chromophores, id. at 29–30, but according to the Board, the claims were broad enough to cover systems with multiple signaling chromophores. Id. at 30. Thermo neglected to address a second covered example, id. at 31, and discounted a third because there was no mention of an “absence of the polymer,” a required limitation in the claims, id. at 32. But the Board reasoned that this example involved no polymer because it used the same term as the preceding example, “direct C* excitation,” which referred to excitation in the absence of a polymer. Id. Thermo thus failed to establish that the four disclosed species were inadequate support for the claims, particularly given the predictable nature of the art. Id. at 33. Bazan therefore was not § 102(b) prior art, and the challenge failed. Id.