Claim Construction in a Post-Teva World – Observations from the First Year of Federal Circuit Review

April 2016

By: Sheri L. Gordon and Larry S. Nixon

In a U.S. patent proceeding, be it before a district court, the Patent Trial and Appeal Board, or another federal agency authorized to decide patent-related issues, development of claim construction positions is one of the most critical and strategically important tasks to undertake. The construction adopted by the tribunal can be a good indicator of the direction the case will likely take, and may be dispositive on patentability/validity issues, as well as infringement.

For many years, parties have been able to develop claim construction positions based on intrinsic and extrinsic evidence, knowing that if a term’s construction landed before the Federal Circuit on appeal, all aspects of that construction would likely be reviewed de novo, including any fact-based matters. [1] This standard of review changed as of January 20, 2015, with the Supreme Court’s decision in Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 135 S. Ct. 831 (2015). [2]

In Teva, the de novo standard was held improper for certain factual findings made in connection with claim construction on appeal. Now, if the record shows that the lower tribunal made subsidiary factual findings to resolve a dispute over extrinsic evidence and those factual findings underlie the construction at issue on appeal, those findings must be given deference and accepted by the appellate court unless clearly erroneous. [3] If the lower tribunal’s construction was (or is found to have been) based solely on intrinsic evidence (i.e., the claims, specification and prosecution history), then the de novo standard still applies. [4] Irrespective of the existence or nonexistence of subsidiary findings of fact, the ultimate conclusion of a claim’s scope continues to be reviewed de novo. [5]

A review of Federal Circuit decisions involving claim construction that issued during the first year Teva deference has been in force reveals that deference is rarely applied. Most often, the construction at issue is found to have been determined from intrinsic evidence alone, or resort to extrinsic evidence is deemed unnecessary - even if the lower tribunal relied on extrinsic evidence. Although the majority of Federal Circuit decisions involving claim construction reflect a purely de novo review of the construction, some takeaways in the decisions can be instructive to development of a claim construction strategy in a post-Teva world.

1. When will Teva deference most likely come into play?

Teva deference is likely where the record shows the construction depended on an understanding of the state of the art (background science) or the meaning a term had to those skilled in the art at the relevant time (e.g., a word has a common understanding in the particular technology field), and the court made factual findings based on extrinsic evidence for such issues (e.g., expert testimony) that is in dispute. These are two areas specifically cited by the Supreme Court in discussing “evidentiary underpinnings” of claim construction. [6] The Supreme Court recognized, for example, that extrinsic evidence such as expert testimony may be indispensable to the understanding of what a technical term means to those skilled in the art and necessary to the construction of the term in context. [7] The Federal Circuit has found instruction in Teva that “‘how the art underst[ands] [a] term ... [is] plainly a question of fact.’” [8] Where there is a dispute over these subsidiary facts, the tribunal will need to make factual findings about the extrinsic evidence, which “are the ‘evidentiary underpinnings’ of claim construction….” [9]

Whether the Federal Circuit will consider Teva deference only where extrinsic evidence in dispute is directed to one of the above categories of “subsidiary facts” is unknown; these are traditionally recognized as the primary areas where consulting extrinsic evidence, if necessary, is proper. [10] Thus far, it seems that these are the two categories the Federal Circuit focused upon in decisions expressly applying deference. In Cephalon, Inc. v. Abraxis Bioscience, LLC, No. 2014-1411, -1442, 618 Fed. Appx. 663 (Fed. Cir. June 17, 2015), for example, the Federal Circuit observed that “microparticles” and “nanoparticles” are technical terms and that “how the relevant scientific community understands them is therefore a question of fact reviewable for clear error.” [11] The district court’s detailed findings of fact regarding extrinsic evidence of record on the issue were not found clearly erroneous on appeal. In Prolitec, Inc. v. Scentair Techs., Inc., 807 F.3d 1353 (Fed. Cir. 2015), the Federal Circuit gave Teva deference to the PTAB’s reliance on testimony from Prolitec’s expert about “the nature of adhesive bonding” to construe the term “mounted” in the limitation “a diffusion head mounted to the reservoir.” [12] The PTAB observed that use of an adhesive encompasses both permanent and non-permanent methods of bonding, and relied on the expert testimony to support this factual determination. [13] The specification used the term “mounted” according to its customary meaning in the context of adhesive bonding methods and did not limit the term to permanent bonding. [14] The Federal Circuit concluded that the “subsidiary factual determination regarding the nature of adhesive bonding was supported by substantial evidence” and found no clear error in the PTAB’s decision to reject limiting the term to “permanently joined”. [15]

2. Is Teva deference triggered just because the court receives extrinsic evidence?

No. The mere fact that the lower tribunal hears or receives extrinsic evidence on claim construction does not automatically trigger Teva deference. Only fact findings made to resolve a factual dispute over extrinsic evidence that underlie the specific construction at issue may qualify for deference. [16] In Vasudevan Software, Inc. v. MicroStrategy, Inc., 782 F.3d 671, 676-679 (Fed. Cir. 2015), for instance, expert testimony and other extrinsic evidence was of record below for construction of “disparate databases.” There was no indication that the district court relied on extrinsic evidence in construing the term. The Federal Circuit concluded that the district court “essentially limited its review to intrinsic evidence” even though there was a stipulated construction in other litigation. [17] The Federal Circuit noted that while the existence of the stipulated construction is an undisputed fact, the import of the stipulation goes to a matter of law subject to de novo review. [18] Regarding expert testimony of record, the Federal Circuit observed that the expert actually admitted the term at issue does not have a “consistent use” and that the meaning “depends on the context.” [19] Other extrinsic evidence - marketing materials alleged to inform the plain and ordinary meaning of the term - was found to be of “scant import” as “[l]anguage used in marketing materials directed to potential customers can mean something quite different from the language used in a patent directed to persons skilled in the art.” [20]

3. Can the appellate court undertake a purely de novo review despite the existence of underlying factual findings that would trigger Teva deference?

Yes. Federal Circuit decisions reflect de novo review in this scenario when, in the Court’s opinion, the proper construction can be reached from intrinsic evidence alone. In Biosig Instruments, Inc. v. Nautilus, Inc., 783 F.3d 1374 (Fed. Cir. 2015), for instance, the Federal Circuit undertook a de novo review based on its finding that the construction on appeal was reached primarily from consideration of intrinsic evidence. [21] While extrinsic evidence was considered and relied upon, the Federal Circuit found that it merely “underscored” the intrinsic evidence. [22] Review of extrinsic evidence of record was therefore unnecessary. [23] Importantly, the Federal Circuit determined that “a skilled artisan would understand the inherent parameters of the invention as provided in the intrinsic evidence.” [24] In Cambrian Sci. Corp. v. Cox Commc’ns, Inc., No. 2014-1686, 617 Fed. Appx. 989 (Fed. Cir. June 29, 2015), the Federal Circuit declined to reach the issue of whether the district court improperly relied on expert testimony to construe a term. The district court did not identify specific testimony by the expert but, rather, made a generalized statement as to what it confirmed regarding the construction. The testimony, though, does not appear to have been necessary to construction as intrinsic evidence was found dispositive – the Federal Circuit explained that intrinsic evidence “fully determines the proper construction” and thus there was no need to address alleged reliance by the district court on expert testimony to construe the term at issue. [25] In Eidos Display, LLC v. AU Optronics Corp., 779 F.3d 1360 (Fed. Cir. 2015), the Federal Circuit determined that extrinsic evidence considered by the district court to construe “a contact hole for source wiring and gate wiring connection terminals” was “ultimately immaterial to the outcome because the intrinsic record is clear.” [26]

4. Is there a high frequency of purely de novo review of claim construction on appeal since Teva?

Yes. While the Federal Circuit has applied Teva deference in some cases, more often claim construction review has been purely de novo. In such cases, the Federal Circuit generally found some rationale to support a conclusion that the construction at issue was based solely on intrinsic evidence. [27] A purely de novo review has also been found appropriate when extrinsic evidence of record was deemed cumulative or unnecessary, as the proper construction could be determined from intrinsic evidence alone. [28]

Importantly, though extrinsic evidence is heard or received by the lower tribunal, Teva deference is unlikely if it is not clear that the tribunal based its construction of the term at issue on factual findings made regarding such evidence. In Williamson v. Citrix Online, LLC, 792 F.3d 1339 (Fed. Cir. 2015), an expert declaration purporting to evidence how a term would be understood by one skilled in the art was of record but not discussed in the district court’s construction rulings on appeal. [29] The Federal Circuit found the constructions were based solely on the intrinsic record and review was de novo. [30] While the de novo review included consideration of the declaration, it was not found persuasive. [31]

5. What level of detail in the construction ruling below is necessary for Teva deference?

There is a greater likelihood that Teva deference will be applied if the lower tribunal provides detailed factual findings that include specific citation to and discussion of extrinsic evidence relied on to support such findings, and that tie factual findings to the construction adopted. In other words, it should be clear that the construction on appeal was based on factual findings, what those factual findings are, and how they were determined. [32], [33]

To the extent a party wants the lower tribunal to make and rely on factual findings about extrinsic evidence that may be given deference on appeal, it is important that the evidence be directed to background science and/or how the term would be understood by one skilled in the art, and that a solid foundation be made for the tribunal to consider and rely on the evidence. [34] The tribunal may still find that intrinsic evidence alone provides enough for a proper construction and thus extrinsic evidence is unnecessary. [35] Nevertheless, the odds that extrinsic evidence will play a crucial role in the construction of a given term for purposes of Teva deference on appeal are better if the tribunal is given a good reason to look to extrinsic evidence (e.g., necessity) and extrinsic evidence addresses an issue identified by the Supreme Court as one that is a factual component of claim construction. In the Southco appeal, for example, the Federal Circuit observed that the district court did not rely on extrinsic evidence in construing the claims, although such evidence was presented. [36] As mentioned above, the district court expressly found that intrinsic evidence supplied all that was needed for the proper constructions.


[1] Cybor Corp. v. FAS Techs., Inc., 138 F.3d 1448, 1456 (Fed. Cir. 1998) (en banc) (de novo review of “fact-based questions relating to claim construction”).

[2] The Federal Circuit has acknowledged that the standard set forth in Teva applies to decisions by the Patent Trial and Appeal Board. In re 55 Brake LLC, No. 2014-1554, 608 Fed. Appx. 942 (Fed. Cir. April 13, 2015) (inter partes reexamination); Prolitec, Inc. v. Scentair Techs., Inc., 807 F.3d 1353, 1358 (Fed. Cir. 2015) (inter partes review).

[3] Teva, 135 S. Ct. at 838, 841-43.

[4] Id. See also, Sociedad Espanola de Electromedicina y Calidad, S.A. v. Blue Ridge X-Ray Co., No. 2015-1102, 621 Fed. Appx. 644, 646-47 (Fed. Cir. July 31, 2015).

[5] Shire Dev., LLC v. Watson Pharm., Inc., 787 F.3d 1359, 1364 (Fed. Cir. 2015); Teva, 135 S. Ct. at 841-42; Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 789 F.3d 1335, 1340 (Fed. Cir. 2015) (on remand).

[6] Teva, 135 S. Ct. at 841.

[7] Id.

[8] Cephalon, Inc. v. Abraxis Biosciences, LLC, No. 2014-1411, -1442, 618 Fed. Appx. 663, 665 (Fed. Cir. June 17, 2015) (quoting Teva, 135 S. Ct. at 838 (internal quotations omitted)).

[9] Teva, 135 S. Ct. at 841. See also, Teva, 789 F.3d at 1342 (factual components of claim construction include background science and meaning a term has in the relevant technology field at the time of the invention); FenF, LLC v. SmartThingz, Inc., 601 Fed. Appx. 950, 952 (Fed. Cir. Feb. 6, 2015) (reliance solely on intrinsic evidence to construe “separators”, not any testimony by one skilled in the art about the meaning the term had in the relevant art at the time of the invention); Flexiteek Ams., Inc. v. Plasteak, Inc., 603 Fed. Appx. 994, 996-97 (Fed. Cir. Mar. 19, 2015) (“

[W]hen the district court looks beyond the intrinsic evidence and consults extrinsic evidence, for example to understand the relevant science, these subsidiary fact findings are reviewed for clear error”); In re Papst Licensing Digital Camera Patent Litigation, 778 F.3d 1255, 1261 (Fed. Cir. 2015) (justifying de novo review because the “district court relied only on the intrinsic record, not on any testimony about skilled artisans’ understanding of claim terms in the relevant field”).

[10] See e.g., Phillips v. AWH Corp., 415 F.3d 1303, 1314 (Fed. Cir. 2005) (en banc) (approving consideration of extrinsic evidence to understand the “meaning of technical terms, and the state of the art”).

[11] Cephalon, 618 Fed. Appx. at 665.

[12] Prolitec, Inc. v. Scentair Techs., Inc., 807 F.3d at 1358-59.

[13] Id.

[14] Id.

[15] Id. at 1359.

[16] See e.g., Cardsoft, LLC v. Verifone, Inc., 807 F.3d 1346, 1350 (Fed. Cir. 2015) (“

[I]t is not enough that the district court may have heard extrinsic evidence during a claim construction proceeding–rather, the district court must have actually made a factual finding in order to trigger Teva’s deferential review”); Shire Dev., 787 F.3d at 1368. See also, Teva, 789 F.3d at 1342 (“Teva cannot transform legal analysis about the meaning or significance of the intrinsic evidence into a factual question simply by having an expert testify on it”).

[17] Vasudevan Software, 782 F.3d at 676.

[18] Id.

[19] Id. at 678.

[20] Id.

[21] Biosig, 783 F.3d at 1382-84.

[22] Id.

[23] Id.

[24] Id. at 1384.

[25] Cambrian Sci. Corp., 617 Fed. Appx. at 993.

[26] Eidos Display, LLC, 779 F.3d at 1365.

[27] See e.g., Pacing Techs., LLC v. Garmin Int’l, Inc., 778 F.3d 1021, 1023 (Fed. Cir. 2015) (“only evidence at issue on appeal and presented to the district court…was intrinsic”); Lexington Luminance LLC v. Inc., 601 Fed. Appx. 963, 970 (Fed. Cir. Feb. 9, 2015) (“intrinsic record fully determines the proper constructions and the district court’s constructions were not based on expert testimony”); FenF, LLC, 601 Fed. Appx. at 952 (claim construction reached solely on intrinsic evidence, not any testimony from one skilled in the art).

[28] See e.g., takeaway point 3 supra and cases cited therein.

[29] Williamson, 792 F.3d at 1351.

[30] Id. at 1346.

[31] Id. at 1351.

[32] See e.g., Lighting Ballast Control LLC v. Philips Elecs. N. Am. Corp., 790 F.3d 1329, 1336-39 (Fed. Cir. 2015) (term “voltage source” had an understood meaning to those skilled in the relevant art, the district court made findings of fact based on extrinsic evidence to construe the term “voltage source means,” the findings were determined on appeal to be supported by the record – specifically, by expert testimony explaining what the term would mean to one skilled in the art that was clearly identified and discussed in the district court decision, and Teva deference was applied); Eon Corp. IP Holdings LLC v. AT&T Mobility LLC, 785 F.3d 616, 624 (Fed. Cir. 2015) (construing means-plus-function terms at issue, the district court made “explicit factual findings, based on expert testimony, that each of the … terms at issue recited complicated, customized computer software” without sufficient structure and ultimately found the terms indefinite, and Teva deference was applied on appeal); Cephalon, 618 Fed. Appx. 663 (district court made detailed findings of fact regarding extrinsic evidence presented on the construction of “microparticles” and “nanoparticles,” and Teva deference was applied on appeal).

[33] Compare the foregoing with Williamson (district court’s construction of “distributed learning control module” under 112, para. 6 did not discuss an expert declaration alleged to evidence how the term was understood by those skilled in the relevant art, and review was de novo).

[34] See e.g., Southco, Inc. v. Fivetech Tech., Inc., 2012 WL 246253 at *4 n.7 (E.D. Pa. Jan. 25, 2012).

[35] Id.

[36] Southco, Inc. v. Fivetech Tech., Inc., No. 2014-1390, 611 Fed. Appx. 681, 685-86 (Fed. Cir. April 10, 2015) (discussing construction of “attached”).