Author: Jon Siderits, Associate Member, University of Cincinnati Law Review
On January 20, 2015, the Supreme Court announced its decision in Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., and did away with the Federal Circuit’s longstanding application of a purely de novo standard when reviewing a district court’s construction of a patent claim.[1] Specifically, the Supreme Court held that the Federal Circuit must review the district court’s resolution of any subsidiary factual matters under a clear error standard of review, and only then may it resolve the ultimate legal question of claim construction de novo.[2] Marking a significant shift in the role of the district court judge in claim construction, and realigning the appellate court’s role with the Federal Rules of Civil Procedure, the Teva decision ensures that the lower court’s construction will play a more dispositive, rather than disposable, part in patent infringement cases.
Claim Construction and the Infringement Analysis
Any patent infringement claim can be broken down into two major parts: construction of the asserted patent claims and comparison of the accused product to the claims as construed.[3] While the second part is a question of fact to be resolved by a fact finder, the Supreme Court decided in Markman v. Westview Instruments that the first part is a question of law to be resolved by a judge.[4] The court in Markman reasoned that judges are better suited than juries to construe written documents, and that the Seventh Amendment’s right to a jury does not include the step of claim construction.[5] Notably, labeling an issue as a question of law means that it is subject to appellate review under a de novo standard, and so the appellate court is not required to give any deference to the lower court’s resolution of the issue.[6] Still, the Markman Court recognized that the practice of claim construction involves certain “evidentiary underpinnings” that are factual in nature—for example, a party to the infringement suit may present an expert witness to support its proposed claim construction, and the judge may have to make “credibility judgments” about the witness and his testimony.[7]
In many cases, a patent claim can be construed simply by referring to the plain meaning of its included terms, or by looking to the language of the patent specification or the prosecution history. Such evidence is called “intrinsic” evidence, and is preferred because it suggests the meaning intended by the patentee. Moreover, this type of evidence is subject to a purely legal analysis. However, if the meaning of a certain word cannot be determined solely based on intrinsic evidence, a judge may need to look to “extrinsic” evidence such as expert testimony and specialized dictionaries to construe a claim. This type of evidence is commonly required in patent cases involving highly technical subject matter, where terms of art and complex jargon far beyond a judge’s general knowledge frequently run rampant. These terms, if left undefined by the patentee, can therefore give rise to a factual dispute. Nevertheless, the Federal Circuit has consistently disregarded district courts’ determinations on these subsidiary factual questions by reviewing those issues de novo.[8] This amounts to a violation of Rule 52(a)(6) of the Federal Rules of Civil Procedure, which states that an appellate court “must not . . . set aside” a district court’s “[f]indings of fact” unless they are “clearly erroneous.”[9]
The Supreme Court Corrects the Federal Circuit in Teva
In Teva, the main dispute concerned the meaning of the phrase “molecular weight” as it appears in the asserted patent claim.[10] The defendant had argued at the district court that the asserted patent claim was invalid for being indefinite, because it described a particular ingredient as having a molecular weight in the range of 5 to 9 kilodaltons, but never specified which of three potential methods of calculation was used to determine that weight.[11] The plaintiff countered with expert testimony in support of its proposed claim construction, which indicated that a skilled artisan would understand which particular method of calculation is to be used, taken in context.[12] Therefore, a question of fact arose as the two parties presented their conflicting expert testimonies on the legal question of claim construction.[13] Ultimately, the district court sided with the plaintiff, adopting its proposed claim construction with respect to the molecular weight limitation, and subsequently found that the patent claim was not invalid for being indefinite.[14] Without its invalidity defense, the defendant was unable to establish non-infringement, and the court found for the plaintiff.[15]
On appeal, the Federal Circuit reviewed the district court’s claim construction and concluded that the term “molecular weight” was indefinite and therefore invalidated the patent, completely reversing the district court’s decision.[16] In making this determination, the Federal Circuit reviewed all aspects of the district court’s construction de novo.[17]
After granting certiorari, the Supreme Court vacated and remanded the Federal Circuit’s decision.[18] The Court found that Federal Rule of Civil Procedure 52(a)(b) sets out a “clear command” and “does not make exceptions . . . or exclude certain categories of factual findings” from the court of appeals’ obligation to review findings of fact under the clear error standard, and therefore it applies to both subsidiary and ultimate facts.[19] Moreover, the Court found that the clear error standard is of particular importance in patent cases, due to the district court judge’s level of involvement.[20] In particular, a district court judge who has presided over the entire proceeding has a significantly greater opportunity to gain the necessary familiarity with the “specific scientific problems and principles” at issue than does an appeals court judge who merely reads the transcripts.[21]
To be clear, the Court acknowledged that when a district court reviews only intrinsic evidence, then its claim construction is purely a question of law and the Federal Circuit will review that construction de novo; however, when a district court must turn to extrinsic evidence “to understand, for example, the background science or the meaning of a term . . . where those subsidiary facts are in dispute,” then its claim construction involves underlying questions of fact.[22] Thus, while the ultimate claim construction is a legal issue that must be reviewed by the Federal Circuit de novo, the Federal Circuit may only overturn the district court’s resolution of an underlying factual dispute by finding that the district court made a clear error with respect to those factual findings.[23]
The Impact of Teva on Future Cases
Though claim construction is only half of the infringement analysis, it can be (and frequently is) dispositive of the entire issue. After all, following the judge’s claim construction order, the remaining question (comparison) simply requires a fact finder to determine whether the accused product falls within the scope of, or is equivalent to, the construed claim.[24] Typically, each side has advocated for a claim construction that clearly benefits its case, so after the court construes the claim, it is often apparent which way a reasonable jury must decide. Settlement discussions and dispositive summary judgment motions are commonplace following the judge’s claim construction order. Therefore, patent infringement cases are often won or lost on claim construction—that is, until the case is appealed and the Federal Circuit substitutes its own claim construction for the district court’s. Interestingly, however, the Federal Circuit, as a rule, does not usually accept interlocutory appeals on the issue of claim construction.[25] Therefore, a party that wins on claim construction, presents its case at trial, and wins a favorable verdict is prone to having the outcome reversed if the Federal Circuit disagrees with the manner in which the district court construed a particular claim limitation. Of course, this is an entirely just outcome if the district court has erred as a matter of law. However, as the Supreme Court has now pointed out, this does not comport with the Federal Rules of Civil Procedure when the Federal Circuit’s reconstruction of the patent claim is the result of a mere disagreement with the district court’s not-clearly-erroneous determination of an underlying factual issue. Yet that had been the Federal Circuit’s common practice, until now. Critics of the Federal Circuit have bemoaned this problem for years, with some commentators now hailing the Supreme Court’s decision as ending the Federal Circuit’s “love affair” with de novo review.[26]
Thus, the Supreme Court’s decision to realign the Federal Circuit’s role in claim construction with the Federal Rules of Civil Procedure gives district courts the deference that they are owed. Giving deference to a district court’s resolution of subsidiary factual disputes involved in claim construction will serve a number of important purposes. For example, it will improve predictability for the parties at the district court level. With less likelihood that the Federal Circuit will adopt an entirely new claim construction, parties can feel more certain that the district court’s construction will control the final outcome of the case. This may, in turn, improve judicial expediency by encouraging a party that loses on claim construction to settle the case, rather than push through the entire trial in hopes of the Federal Circuit ordering an entirely new construction. Time, money, and judicial resources will be conserved.
Most importantly, applying the clear error standard of review to factual questions underlying the claim construction analysis will ensure that they are resolved by the party in the best position to do so—the district court—so that the claims are properly construed. Regardless of what the Federal Circuit may think about a particular underlying factual issue, the district court is the judicial entity that has presided over the entire procedure, observed the witnesses, and, as the Federal Rules of Civil Procedure recognize, is in the best position to evaluate witness credibility and other factual issues. While a district court may ultimately err as a matter of law on the issue of claim construction, its resolution of any disputed evidentiary underpinnings must still be accorded deference because it was better situated to resolve those issues than the Federal Circuit, which merely reviews the written record. Accepting the district court’s determination of any underling factual disputes, in the absence of a finding of clear error, will help to inform the Federal Circuit of the proper claim construction of the patent, and thus more likely result in a just outcome.
Teva Brings Swift Change
Of course, it is not yet certain how the Federal Circuit will implement the Supreme Court’s decision. Perhaps it will merely attempt to repackage its de novo rulings under the guise of the clear error standard, but hopefully it will give district courts the deference that they are owed. It will not take long to find out, as the Supreme Court has, in addition to Teva, already vacated and remanded three other patent infringement cases involving subsidiary questions of fact to the Federal Circuit for further consideration in light of Teva.[27] Indeed, it appears that the de novo love affair is truly coming to an end.
[1] Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc., 574 U.S. at ___ (2015) (slip op. at 1-2).
[2] Id.
[3] Markman v. Westview Instruments, Inc., 517 U.S. 370, 384 (1996).
[4] Id.
[5] Id. at 384-390.
[6] See, e.g., Teva, 574 U.S. at ___ (slip op. at 1).
[7] Markman, 517 U.S. at 389-390.
[8] See, e.g., Shire Dev., LLC v. Watson Pharm., Inc., 746 F.3d 1326 (Fed. Cir. 2014) cert. granted, judgment vacated sub nom. 2015 WL 303221 (U.S. Jan. 26, 2015); Lighting Ballast Control LLC v. Philips Electronics N. Am. Corp., 744 F.3d 1272 (Fed. Cir. 2014) cert. granted, judgment vacated sub nom. 2015 WL 303220 (U.S. Jan. 26, 2015); Butamax(TM) Advanced Biofuels LLC v. Gevo, Inc., 746 F.3d 1302 (Fed. Cir. 2014) cert. granted, judgment vacated sub nom. Gevo, Inc. v. Butamax Advanced Biofuels LLC, 2015 WL 303219 (U.S. Jan. 26, 2015).
[9] See, e.g., Gene Quinn, Supremes End Federal Circuit Love Affair with De Novo Review, IP Watchdog (Jan. 20, 2015), http://www.ipwatchdog.com/2015/01/20/supremes-end-federal-circuit-love-affair-with-de-novo-review/id=53873/.
[10] Teva, 574 U.S. at ___ (slip op. at 2).
[11] Id. Under 35 U.S.C. § 112, each patent claim must particularly point out and distinctly claim the subject matter regarded as the invention. If it fails to do so, the claim is said to be indefinite and invalid.
[12] Id. at ___ (slip op. at 3).
[13] Credibility determinations are factual in nature, and are therefore typically left to the district court judge (in a bench trial) or jury (in a jury trial).
[14] Teva, 574 U.S. at ___ (slip op. at 3).
[15] Teva Pharm. USA, Inc. v. Sandoz, Inc., 876 F. Supp. 2d 295, 419 (S.D.N.Y. 2012) aff’d in part, rev’d in part, 723 F.3d 1363 (Fed. Cir. 2013), vacated, 135 S. Ct. 831 (2015).
[16] Teva, 574 U.S. at ___ (slip op. at 3).
[17] Id.
[18] Id. at ___ (slip op. at 4, 16).
[19] Id. at ___ (slip op. at 4) (quoting Anderson v. Bessemer City, 470 U.S. 564, 574 (1985) and Pullman-Standard v. Swint, 456 U.S. 273, 287 (1982)).
[20] Id. at ___ (slip op. at 7-8).
[21] Id.
[22] Id. at ___ (slip op. at 12).
[23] Id.
[24] See, e.g., Wolverine World Wide, Inc. v. Nike, Inc., 38 F.3d 1192, 1199 (Fed. Cir. 1994).
[25] See, e.g., Regents of Univ. of Cal. v. Dakocytomation Cal., Inc., 517 F.3d 1364, 1371 (Fed. Cir. 2008) (Stating that “we have not generally certified motions for interlocutory appeal of claim construction”).
[26] See, e.g., Quinn, supra note 9. Moreover, this is certainly not the first time that the Supreme Court has had to correct the Federal Circuit’s misapplication of de novo review. See, e.g., Highmark Inc. v. Allcare Health Mgmt. Sys., Inc., 134 S. Ct. 1744, 1748 (2014) (holding that all aspects of a district court’s exceptional-case determination for awarding attorney’s fees should be reviewed under the abuse of discretion—not de novo—standard).
[27] See Shire Dev., LLC v. Watson Pharm., 2015 WL 303221, at *1 (U.S. Jan. 26, 2015); Lighting Ballast Control LLC v. Universal Lighting Technologies, Inc., 2015 WL 303220, at *1 (U.S. Jan. 26, 2015); Gevo, Inc. v. Butamax Advanced Biofuels LLC, 2015 WL 303219, at *1 (U.S. Jan. 26, 2015).
