In Transgene v. Replimune Ltd., PGR2022-00014, Paper 38 (P.T.A.B. May 25, 2023), Transgene challenged claims 1-8 and 10-26 of Replimune's U.S. Patent No. 10,947,513 ("the '513 patent"). Replimune filed a statutory disclaimer of claims 1-8, 10-12, and 14-26 as part of its Patent Owner Preliminary Response.

Claim 13, the only remaining claim in the proceeding, depended on claims 1 and 11, each of which is reproduced below.

1. An oncolytic virus comprising:

(i) a heterologous GM-CSF-encoding gene; and

(ii) a heterologous CTLA-4 inhibitor encoding gene, wherein both heterologous genes are inserted into the genome of the virus.

11. The virus of claim 1, which is a herpes simplex virus (HSV).

13. The virus of claim 11, wherein the GM-CSF-encoding gene and the CTLA-4 inhibitor encoding gene are inserted into the ICP34.5 encoding locus, either by insertion, or partial or complete deletion, in a back to back orientation in relation to each other, each under separate regulatory control.

Anticipation

The Board first addressed the merits of the challenge to claim 11, despite the statutory disclaimer, and was not persuaded that claim 11 was anticipated. Silvestre, the asserted reference, did "not disclose a single embodiment of an oncolytic HSV that includes both a heterologous GM-CSF encoding gene and a heterologous CTLA-4 inhibitor encoding gene." Id. at *19. Instead, Transgene attempted to "randomly" select from "broader lists of possible oncolytic virus classes, immune checkpoint targets, and therapeutic genes," but the Board found this tactic involved too much "picking and choosing" and undermined Transgene's anticipation ground. And although the Board ultimately concluded that Transgene was too late in raising the argument that one skilled in the art could "immediately envisage" the HSV of claim 11, the Board nevertheless commented that Transgene's argument was not persuasive because it required reading too much into statements about preferred embodiments in the specification and was still based on too much "picking and choosing" from different portions of the reference. Id. at *21-23.

Since challenged claim 13 depended from claim 11, the Board considered its anticipation analysis sufficient for claim 13. Id.at *19.

Obviousness

The Board also found Transgene did not carry its burden with respect to proving claim 13 obvious because Transgene's motivation to combine arguments were unconvincing.

First, the Board addressed Transgene's expert's testimony that the proposed modification to Silvestre was "the simplest and most common way to" insert two genes into a locus. At institution, the Board credited this testimony, given that it was unrebutted. However, during a deposition after institution, Transgene's expert admitted that this position was an "overstatement." Id. at *32. Thereafter, in its Reply, Transgene shifted its argument that the proposed modification was generally "simple" and "commonly used" but the Board found the revised, new theory lacking any support and contradicted by credible testimony from Replimune's expert identifying multiple other "common" insertion strategies at the time of the invention. Id.

Second, the Board rejected Transgene's "known-technique" argument that the proposed modification was a known method to insert two genes into a locus and "among a 'finite number of identified, predictable solutions' within a POSITA's technical grasp." In doing so, the Board analyzed two recent Federal Circuit decisions: (i) Intel Corp. v. PACT XPP Schweiz AG, 61 F.4th1373, 1380 (Fed. Cir. 2023) ("Intel"), which the Board cited as permitting petitioner to use "'universal' motivations known in a particular field to improve technology" to sufficiently show motivation to combine "even absent any hint of suggestion [of the universal motivations] in the references themselves,'" and (ii) Xerox Corp. v. Bytemark, Inc., IPR2022-00624, Paper 12 (PTAB Feb. 10, 2023) ("Xerox"), a Director-issued opinion which the Board cited for giving little weight to expert testimony that "is an exact restatement of the [p]etition's arguments without any additional supporting evidence or reasoning." The Board found the facts of this case similar to those in Xerox Corp. v. Bytemark, Inc., IPR2022-00624, Paper 12 (P.T.A.B. Feb. 10, 2023), where the evidence did not support a "known-technique" rationale. Id. at *36-39. "In particular, as with the petitioner's expert declaration in Xerox, we find that Dr. Bell's declaration in this case merely offers conclusory assertions without underlying factual support as to what techniques were known in the art about insertion of multiple heterologous genes into the same locus. Indeed, Dr. Bell fails to cite any evidence in his declaration . . . to support his opinion that back-to-back insertion is the 'simplest and most common' and 'most straightforward' strategy a POSITA would use to insert two genes into a given locus." Id. at *39. Transgene, in the Board's view, provided only "uncorroborated, conclusory, and contradicted" testimony in support of its proposed modification Id. at *28-29.

Third, with respect to whether one of ordinary skill in the art would have had a reasonable expectation of success in implementing the proposed modification and arriving at the invention of claim 13, the Board agreed with Replimune that Transgene's arguments were unpersuasive. In asserting that the claims lacked enablement, Transgene admitted to the "difficulty in creating functional oncolytic virus" and the "high level of unpredictability in the art." The Board found these admissions to clash with Transgene's obviousness positions because, according to the Board, "[c]laim 13, by incorporating all the features of claims 1 and 11, expressly requires an 'oncolytic virus.'" Id. at *44. As such, the reasonable expectation of success inquiry must examine whether one skilled in the art would have achieved the claimed functional oncolytic virus as a whole and, therefore, Transgene's unpredictability and difficulty arguments undermined making such a finding. Id. at *43.

Unexpected Results Supporting Nonobviousness

The Board also addressed Replimune's objective evidence of nonobviousness, in particular the unexpected results shown by Example 13 of the '513 patent. Id. at *45. Example 13 compared the anti-tumor effects of an oncolytic HSV1 expressing an anti-CTLA-4 antibody on Virus 27 and Virus 16. Id. at *45. The Board concluded that expert testimony supported Replimune's position that "these results would not have been predicted by a POSITA given the complex interplay of the anti-tumor and anti-viral immune responses induced by the oncolytic viruses," and that Transgene's unsupported attorney argument was insufficient to rebut Replimune's evidence of unexpected results. Id. at *48-49.

Take-aways

"[A] reference must clearly and unequivocally disclose the claimed compound or direct those skilled in the art to the compound without any need for picking, choosing, and combining various disclosures[.]" In re Arkley, 455 F.2d 586, 587 (CCPA 1972). Practitioners are cautioned, however, that if the set of possible combinations is small enough that one of ordinary skill in the art could "immediately envisage" the claimed arrangement, a prior art disclosure could be found to anticipate a claimed invention. See Kennametal, Inc. v. Ingersoll Cutting Tool Co., 780 F.3d 1376 (Fed. Cir. 2015).

The petitioner in this case tried to rely on Kennametal, but did not raise that argument until its reply, which the Board considered was too late. The Board also reminded practitioners that no new evidence may be presented at oral argument. Id. at *40, FN 23.

Another lesson for practitioners is to watch out for arguments relying on common sense as the basis for establishing obviousness, and consider whether such arguments can be attacked as conclusory and unsupported. "[R]eferences to 'common sense' -whether to supply a motivation to combine or a missing limitation-cannot be used as a wholesale substitute for reasoned analysis and evidentiary support[.]" Arendi S.A.R.L. v. Apple Inc., 832 F.3d 1355, 1362 (Fed. Cir. 2016).

This case is a good reminder of the importance of fully developing all theories for unpatentability in the petition. In particular, Transgene shifted from originally arguing that "back-to-back insertion was '[t]he simplest and most common method,' and contend[ed] in its Reply that motivation to use that technique comes from that method merely being 'simple' and 'commonly used.' . This is a different theory than what was originally presented in the Petition." Id. at *32.

Finally, this case is also a good example of successfully using unexpected results to show nonobviousness. When making such arguments, it could be helpful to include an example embodiment from the specification showing unexpected results and an associated claim drawn to that embodiment. This could help show how the unexpected results have a nexus and coextensiveness with the claimed invention.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.