In a landmark 7-2 decision issued on January 20, 2015, the Supreme Court held that factual determinations underlying a trial court's construction of a patent claim are entitled to deference on appeal, and may only be overturned if they are clearly erroneous.  Teva Pharm. USA, Inc. V. Sandoz, Inc., 574 U. S. ____ (2015).

The Court's ruling in Teva upset the Federal Circuit's long-standing rule that claim construction is a pure issue of law, reviewed entirely de novo on appeal.  While a trial court's factual determinations underlying claim construction must now be reviewed on appeal for clear error, the Court confirmed that the ultimate question of claim interpretation remains an issue of law reviewed de novo.

Teva: A Matter of Indefiniteness

Teva's patent covered a method of manufacturing its multiple sclerosis drug Copaxone.  When Teva sued Sandoz-who marketed a generic version of Teva's drug-for infringement, Sandoz responded that Teva's patent claims were invalid for indefiniteness. 

Sandoz argued that Teva's claims were indefinite because they referred to "molecular weight" without specifying to which of three possible weight-calculating methods the claims were referring.  Teva countered that the claims were indeed definite and were referring to one particular weight-calculating method.  Sandoz pointed in response to a figure in Teva's patent and argued that the figure was plainly inconsistent with the weight-calculating method that Teva was advocating. 

The District Court heard conflicting testimony from the parties' experts regarding the arguably inconsistent figure in Teva's patent.  Ultimately, the District Court credited Teva's expert, who testified that there was a reasonable explanation for the slight discrepancy in the figure, and held that Teva's claim was sufficiently definite.1

The Federal Circuit reversed on appeal.  The three-judge panel reviewed all of the evidence-including the patent itself, its prosecution history, and testimony of the parties' experts-and concluded that the claims were indefinite and invalid.2  Teva petitioned the Supreme Court for review, and the Court granted Teva's petition to answer the question "whether a district court's factual finding in support of its construction of a patent claim term may be reviewed de novo."3

Clear Error, As a Matter of Fact

After introducing the indefiniteness proceedings in the District Court, the Court opened its substantive discussion by quoting Federal Rule of Civil Procedure 52(a)(6).  That rule "states that a court of appeals 'must not . . . set aside' a district court's '[f]indings of fact' unless they are 'clearly erroneous.'"4  The Court emphasized that the rule "sets forth a 'clear command' that "does not make exceptions" and "applies to both subsidiary and ultimate facts."5 The Court further noted that "appellate courts must constantly have in mind that their function is not to decide factual issues de novo."6

The Court next turned to its decision in Markman.7 The Court accepted Markman's  "conclu[sion] that it was proper to treat the ultimate question of the proper construction of the patent as a question of law in the way that we treat document construction as a question of law."8  The Court analogized claim construction to the construction of "other written instruments, such as deeds, contract, or tariffs."9 It dismissed the suggestion that claim construction is more analogous to interpretation of statutes.10

Critically, the Court acknowledged that, while it is ultimately a question for the judge and not the jury, the construction of a written instrument may sometimes "give rise to a factual dispute," such as when the writing uses "technical words or phrases not commonly understood."11  In those circumstances, the Court said, "the determination of the matter of fact will precede the function of construction."12  "And . . . Rule [52(a)(6)] requires appellate courts to review all such subsidiary factual findings under the 'clearly erroneous' standard."13

While emphasizing that facts underlying construction of a patent claim are reviewed on appeal for clear error, the Court stressed that the "ultimate interpretation is a legal conclusion" and "[t]he appellate court can still review the district court's ultimate construction of the claim de novo."14

Obviousness Lookalike

Having acknowledged that that claim construction is ultimately a legal issue-albeit one with "evidentiary underpinnings" that sometimes require "subsidiary factfinding"-the Court analogized claim construction to the issue of obviousness, also ultimately a question of law.15  Obviousness,  the Court said, likewise "involves subsidiary factfinding subject to Rule 52(a)'s clear error review."16  Interestingly, the Court analogized claim construction not only to obviousness but also "to a judge (sitting without a jury) deciding whether a defendant gave a confession voluntarily."17 

The Court readily acknowledged that "in some instances, a factual finding may be close to dispositive of the ultimate legal question of the proper meaning of the term in the context of the patent," but the Court was not bothered:  "Simply because a factual finding may be nearly dispositive does not render the subsidiary question a legal one."18

A Supreme How-To

The Court endeavored to explain how to implement this fact-law division in practice.  "[W]hen the district court reviews only evidence intrinsic to the patent [(claims, specification, prosecution history)], the judge's determination will amount solely to a determination of law, and the Court of Appeals will review that construction de novo."19  On the other hand, when the district court consults "extrinsic evidence in order to understand, for example, the background science or the meaning of a term in the relevant art during the relevant time period," the court "will need to make subsidiary factual findings about that extrinsic evidence" which "must be reviewed for clear error on appeal."20

As an example, the Court considered a hypothetical situation where a district court resolves a dispute between experts "and makes a factual finding that, in general, a term of art had a particular meaning to a person of ordinary skill in the art at the time of invention."21  The corresponding legal analysis, the Court said, is "whether a skilled artisan would ascribe that same meaning to that term in the context of the specific patent claim under review."22  "That is because experts may be examined to explain terms of art, and the state of the art, at any given time, but they cannot be used to prove the proper or legal construction of any instrument of writing."23 

The Dissent

Justice Thomas, joined by Justice Alito, dissented.  In the dissent's view, the interpretation of patent claims is purely a conclusion of law that does not involve "findings of fact" within the meaning of Fed. R. Civ. P. 52(a)(6).24

Patent claims, the dissent opined, are more like statutes and land patents (which govern public rights and duties) than contracts and deeds (which effect private intent).25  The dissent observed that interpretation of statutes and contracts both, at times, require analysis of extrinsic evidence.  But the dissent draws a distinction between "subsidiary evidentiary findings underlying statutory construction"-which it characterizes as "analytically legal"-and "subsidiary findings of fact" made in construing contracts, where the "purpose . . . of the court is to ascertain the intention of the parties" and this "search for intent becomes factual in nature."26

"Real intentions," the dissent says, "have an existence outside the written instrument . . . that the instrument merely records,"27 while, in contrast, "the 'fact' of how a skilled artisan would understand a given term or phrase [in a patent claim] at a particular point in history is a legal fiction; it has no existence independent of the claim construction process."28  A "patent holder's actual intentions," the dissent points out, "have effect only to the extent that they are expressed in the public record."29

The dissent went on to criticize the majority's analogy between claim construction and obviousness: "[T]his analogy is even further off the mark because obviousness turns on historical facts about the circumstances of the invention, rather than on the construction of a written instrument."30

The dissent voiced several concerns with the majority's holding.  First, the dissent expressed concern that the majority's holding "could be read to cast doubt" on the existing practices of interpreting "terms in statutes [that] have technical meanings . . . as involving only conclusions of law," "as well as on [the] holding in Markman that claim construction is exclusively for the court."31

Second, the dissent pointed out that "patents are authoritative government dispositions," so "when a judge construes a patent, he is , in a very real sense, 'say[ing] what the law is.'"32  Consequently, the dissent said, "the majority's rule will distort the appellate court's construction of the law by requiring it to defer to subsidiary determinations that are dispositive as to meaning."33 

Third, the dissent stressed the "need for uniformity in claim construction."34  "If the boundaries of the patent right could shift from case to case," the dissent cautioned, "then the result would be a zone of uncertainty which enterprise and experimentation may enter only at the risk of infringement."35

Finally, the dissent said this: "The majority's rule provides litigants who prevail in district court a significant opportunity and incentive to take advantage of this [fact vs. law] uncertainty by arguing on appeal that the district court's claim construction involved subsidiary findings of fact.  At best, today's holding will spawn costly-and, if the majority is correct about the [low] frequency with which these evidentiary determinations make a difference, meritless-collateral litigation over the line between law and fact."36

Take-Aways

Teva has wide-ranging implications both in district court and before the Federal Circuit. 

With a new emphasis on factfinding, district courts may push claim construction further into the discovery process.  Claim construction (Markman) hearings might become even more like mini-trials, involving reams of evidence and extensive expert testimony.  District courts may also feel compelled to make individual factual findings, clearly indicated in claim construction orders and distinct from their conclusions of law.  On appeal to the Federal Circuit, we may see both a lower incidence of appeals involving issues of claim construction, and fewer reversals of district court decisions on such issues.  This moves the approach closer to that currently followed in the United Kingdom and away from that of the European Patent Office.

The elephant in the room may be the question of what is left in claim construction as an issue of law?-how much of claim construction will factual determinations swallow?  While the Supreme Court was not deterred by the potentially determinative nature of underlying factual issues, this possibility is particularly acute in the construction of patent claims.  If a court must construe a patent claim by asking, "What was the meaning of this term, in context, to a person of ordinary skill in the art at the time of the invention?"-and the court purports to answer this question as a matter of fact-what is left for review as a matter of law?

One potential answer to these questions, which seems to find support in Teva, lies in existing Federal Circuit case law.  The Federal Circuit has consistently held that claim construction starts with the language of the claims and other evidence intrinsic to the patent.37  Courts should consult extrinsic evidence only when needed, for example, if the intrinsic record is ambiguous.38  Thus, as Tevaacknowledged, "a district court's construction of a patent claim, like a district court's interpretation of a written instrument, often requires the judge only to examine and to construe the document's words."39  In other words, consideration of extrinsic evidence in claim construction should be the exception rather than the rule.

But when claim construction excludes extrinsic evidence and draws only on the intrinsic record, what are the "findings of fact"?  As a matter of fact, there are none:  "[W]hen the district court reviews only evidence intrinsic to the patent . . . the judge's determination will amount solely to a determination of law, and the Court of Appeals will review that construction de novo."40 Teva thus represents an opportunity to renew the importance of-and perhaps clarify-principles governing admissibility of extrinsic evidence in claim construction.  This approach could mitigate the risk that parties will "take advantage . . . by arguing on appeal that the district court's claim construction involved subsidiary findings of fact."41  It could also reduce the risk that the majority's approach will "spawn costly . . . meritless . . . collateral litigation over the line between law and fact."42

The European Perspective - Diverging Decisions

Justice Thomas in dissent highlighted the importance of uniformity in claim construction.  Given that many patents are litigated on both sides of the Atlantic, it's interesting to compare the approach laid down in Teva with the way two other important venues for patent disputes-the European Patent Office and the English courts-deal with mixed questions of fact and law.

In the EPO and the English courts there are no juries, and there is no equivalent to the Markmanhearing.  All of the factual and legal issues are dealt with by the judge or EPO member, who are normally patent specialists, and claim construction is dealt with at the same time as the decision on the merits.

The EPO deals only with validity and does not hear infringement cases, which are left to the individual European national courts.  The right to appeal is automatic and the appeal board will review all of the issues, including claim construction, de novo; there is no formal deference to the factual findings of the first instance tribunal (the Opposition Division).  EPO cases, however, involve a much less detailed factual examination than a US or UK patent case: witness evidence and expert reports are typically relatively short, witnesses are rarely heard, there is no cross-examination, and hearings usually last no more than a day.  There is therefore less reason to defer to the Opposition Division's assessment of the facts, since the Opposition Division has not been immersed in the details over a lengthy trial.  The parties have a second bite at the cherry and can re-run all of their original arguments on appeal.

The English courts deal with all aspects of infringement and validity, including claim construction, at a single hearing.  The factual examination is detailed, with multiple rounds of written evidence and cross-examination in front of the judge at a trial lasting a week or more.  The English courts have therefore developed a similar approach for all issues, including claim construction, to that laid down in Teva: findings of fact will only be interfered with if plainly wrong, because the trial judge is in the best position to assess the nuances of the factual evidence.  Where the judge makes a factual determination and then applies an established legal standard, the appellate court can review that decision.  What is now the UK's Supreme Court, however, held in Biogen v Medeva ([1997] RPC 1) that, "When the application of a legal standard such as obviousness involves no question of principle but is simply a matter of degree, an appellate court should be very cautious in differing from the judge's evaluation."  In light of this approach, where there is a mixed question of fact and law and the legal principles are well established, the first instance judge can deliver a judgment where the factual findings decide the ultimate issue and the decision is effectively "appeal-proof."  In the wake of Teva a similar trend may emerge in claim construction decisions in the US, with district courts increasingly framing their decisions as fact-based and therefore more difficult to overturn.