Non-Prior Art Evidence in IPR - Yeda v. Mylan

January 28, 2019

Stuart V.C. Duncan Smith, Kevin R. Mosier

(as published by Bloomberg Law)

Section 311(b) of the America Invents Act (“AIA”) provides that a petitioner for inter partes review (“IPR”) may request cancellation of one or more claims of a patent “only on the basis of prior art consisting of patents or printed publications.” In light of that language, one might assume that all of the petitioner’s evidence during an IPR trial must be prior art. That assumption would be incorrect. The Federal Circuit’s October 2018 decision in Yeda Research & Development v. Mylan, 906 F.3d 1031, held that in some situations a petitioner may use non-prior art evidence to establish the state of the prior art.

While on one hand, Yeda resolves a relatively simple question regarding the evidence available to the petitioner in an IPR trial, the decision also illustrates some of the bigger issues facing IPRs more generally. Due to court decisions and new regulations, IPR is evolving in some respects that make it increasingly similar to district court litigation. Yeda continues this trend and shows that these changes, both big and small, can impact IPR strategy. Now in its seventh year, IPR practice is not yet fixed, and practitioners should take heed.

Petitioner’s Non-Prior Art Evidence

In Yeda, the petitioner challenged patent claims concerning treatment of multiple sclerosis as obvious in light of the prior art. The Patent Trial and Appeal Board (“the PTAB”) concluded that the petitioner’s prior art evidence disclosed all limitations of the challenged claims except for a limitation regarding the frequency that the patient had to receive injections. The challenged patent called for six injections over two weeks, while the prior art disclosed seven injections over two weeks.


To support its argument that modifying the prior art’s seven injections to use the claimed six injections was obvious, the petitioner cited a reference that described research regarding ways to reduce the number of such injections. This reference was not prior art, since it was published just three weeks after the threshold date to determine whether a reference is prior art, but it described research that occurred two years before that date.

The patent owner objected to the non-prior art reference under Section 311(b) and sought to have it excluded. The PTAB acknowledged that the reference was not prior art, but nonetheless overruled the objection and considered the non-prior art reference as probative of the motivation to modify the treatment disclosed in the prior art to involve fewer injections. The PTAB ultimately found that the patented claims were invalid as obvious.

The Federal Circuit’s Decision

On appeal, the patent owner challenged the PTAB’s conclusion on the basis that its consideration of the non-prior art reference violated Section 311(b). In particular, the petitioner argued that IPRs are unlike district courts, which can consider the full range of prior art under Section 102, including both prior art patents and publications and information “otherwise available to the public.” By contrast, petitions for IPR may be requested “only on the basis of prior art consisting of patents or printed publications” under Section 311(a).

The Federal Circuit disagreed, and it held that the statutory scheme allows the PTAB to “rely on evidence other than just the prior art.” Section 311(b) merely defines what type of prior art may form the basis of a ground for canceling a patent claim (i.e., “patents or printed publications”), the Court explained, but does not prevent the PTAB from considering other relevant evidence. Notably, the Court reasoned that Section 312(a)(3) suggests that a petition could properly identify both prior art and other “supporting evidence.”

The Federal Circuit explained that the non-prior art reference could be relied upon in “proper supporting roles.” These proper roles include “indicating the level of ordinary skill in the art, what certain terms would mean to one with ordinary skill in the art, and how one with ordinary skill in the art would have understood a prior art disclosure.” Here, the Federal Circuit held that the PTAB did not err when it considered the non-prior art reference, reasoning that it is relevant evidence of the motivations of those skilled in the art to reduce the injection frequency.

Non-Prior Art Evidence of a Likelihood of Success?

Practitioners looking to Yeda for clear guidance should note that while the Federal Circuit endorsed the PTAB’s use of non-prior evidence to support its finding regarding the motivation to modify the prior art, the Federal Circuit did not endorse the PTAB’s use of the same non-prior art reference to support its finding regarding the expectation of success of that change. Rather, the Court reasoned that if the PTAB erred by considering the non-prior art reference on the issue of the expectation of success, the error was harmless, because the PTAB also relied on other evidence.

The Federal Circuit did not explain why it took different approaches to the motivation to modify the prior art (where the PTAB properly considered the non-prior art) and the expectation of success with that modification (where the PTAB may have improperly considered the non-prior art). The parties’ briefing did not highlight any difference. As a consequence, we will not know whether the Court will treat those concepts differently until another case provides guidance.

A Lesson in Statutory Construction?

Yeda is notable for the Court’s approach to interpreting Section 311(b). The Court emphasized that its interpretation of Section 311(b) as not barring non-prior art evidence of the motivation to modify the prior art is consistent with Section 312(a)(3), which permits the petition to include other “supporting evidence.” To the patent owner and others that would limit IPR, the decision could be viewed as choosing an interpretation that favored a holistic view of the statute over the literal language of Section 311(b).

In several recent cases, the courts have disfavored similar holistic interpretations that may be contrary to a specific section’s literal text. For example, in SAS Institute v. Iancu, 138 S. Ct. 1348 (2018), the Supreme Court rejected the PTAB’s interpretation of the statute (which the Federal Circuit endorsed) as permitting partial institutions. Subsequently, the Federal Circuit sitting en banc in Click-to-Call Technologies v. Ingenio, 899 F.3d 1321 (Fed. Cir. 2018), rejected the PTO’s interpretation of the IPR time-bar.

In both SAS Institute and Click-to-Call, the courts relied on the literal text of the specific section at issue over what other sections could suggest. Those that disagree with the Court’s decision in Yeda might argue that the three-judge panel’s somewhat sparse reasoning and express reliance on Section 312(a)(3) ignored that SASInstitute and Click-to-Call counsel in favor of a more literal approach to interpreting the statute governing IPR.

Aligning IPR with District Court Litigation?

Yeda’s interpretation of Section 311(b) could also be understood as continuing a trend of aligning IPR’s procedures in some respects with those in district court litigation. Non-prior art evidence of the knowledge of a person of ordinary skill in the art (“POSA”) is not unheard of. Indeed, in the written description and enablement context, the Federal Circuit’s predecessor explained that non-prior art evidence could be relevant to show that “characteristics of prior art products were known.” In re Koller, 613 F.2d 819, 824 (C.C.P.A. 1980).

Precedent on the use of non-prior art evidence to support an obviousness finding is less common, but the petitioner identified at least one case. In Syntex (U.S.A.) v. Apotex, 407 F.3d 1371 (Fed. Cir. 2005), the Federal Circuit held that the district court erred by refusing to consider a non-prior art reference that described what was “well known” in the industry. There, similar to Yeda, the reference published just five days after the threshold date to be prior art. The Federal Circuit thought it “incredulous” that the disclosure could go from unknown to “well known” in those five days.

But by adopting that same rule for IPRs, the Federal Circuit rejected the patent owner’s argument that, with respect to the use of non-prior art, Congress intentionally limited IPR relative to district court litigation. Drawing parallels to reexamination, the patent owner asserted that Congress similarly calibrated IPR to provide a narrower but more efficient inquiry: petitioners can choose IPR rather that district court litigation but must accept IPR’s limits. Although other evidence might explain the prior art references in IPR, the patent owner argued, it should not supplement the prior art references.

By permitting at least some use of non-prior art evidence of the state of the art, Yeda might be understood to embrace a recent trend of aligning IPR’s practices with those of district court litigation. In SAS Institute, for example, the Supreme Court reasoned that the petitioner in IPR, “[m]uch as in the civil litigation system it mimics,” is the “master of its complaint” and entitled to judgment on all claims it challenges. In DuPont v. Synvina, 904 F.3d 996 (Fed. Cir. 2018), the Federal Circuit held that IPR follows the rule in district courts regarding burden shifting after a prima facie case of obviousness.

And recently, the PTO amended the rules to require the PTAB to use the traditional Phillips claim construction standard that district courts have long applied. See, 37 C.F.R. 42.100(b) (effective Nov. 13, 2018). The regulations explaining the rule change expound on the merits of “[m]inimizing differences” between IPR and district court litigation promoting “greater uniformity” between the proceedings. 83 Fed. Reg. 51,340, 51,342 (Oct. 11, 2018).

Of course, even a trend of aligning IPR in some respects with district court litigation will not eliminate all differences. The statute sets the evidentiary standard in IPR lower than in District Court litigation (i.e., preponderance of the evidence in IPR versus clear and convincing evidence in district court litigation). 35 U.S.C. §316(e). Further, some Federal Circuit decisions recognize other differences between the proceedings. See,e.g., Arista v. Cisco, 908 F.3d 792 (Fed. Cir. 2018) (holding that assignor estoppel does not apply in IPR, unlike in District Court).

Not everyone will agree whether the evolving character of IPR will promote Congress’s goal of providing a “quick and effective alternative[] to litigation.” H.R. Rep. No. 112-98, at 48 (2011). Critics of IPR might argue that it has morphed into a proceeding that Congress did not intend, and that changes such as using the Phillips claim construction standard are needed to protect patent owners. Supporters of IPR might argue that IPR was created to address a problem of bad patents, and care should be taken not to prevent IPRs from fulfilling that mission.

Looking Forward

Practitioners should keep in mind that any change to IPR can impact the strategies that petitioners and patent owners should employ. Although not all changes are as far reaching as those prescribed by SAS Institute or the PTO’s new claim construction rules, even the more particular tweaks to IPR’s scope like the burden shifting in Dupont and the use of non-prior art evidence in Yeda have an impact on strategy.

Yeda provides practical lessons and raises important questions. For now, at least, non-prior art evidence of a POSA’s motivation to modify the prior art is permitted in IPR. But the decision conspicuously left open the question of whether non-prior art evidence of a POSA’s expectation of success will be permitted. The decision raised questions about the correct way to construe the statute in light of statutory mandates that, if read narrowly and literally, might favor a different result. And the decision informs the discussion about IPR’s role as an alternative to district court litigation.

Thus, as IPR enters its seventh year, its impact on the concerns and goals that animated Congress to create IPR remains in flux, and important questions remain unanswered.