The patent claim construction process requires district court judges to reach legal conclusions about the meaning of claim terms. Courts have long recognized these legal conclusions may be informed by subsidiary findings of fact. In view of these “evidentiary underpinnings” and the mandate of the Federal Rules of Civil Procedure that “findings of fact . . . must not be set aside unless clearly erroneous,” the Supreme Court in Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc. clarified the appropriate standard of review that applies to findings of fact made during claim construction.1 After the Supreme Court’s ruling, the Federal Circuit has made a point of separately evaluating findings of fact made at the district court level for clear error, while at the same time emphasizing that these findings of fact may not be dispositive in the ultimate construction of a claim, which is a legal conclusion reviewed de novo. In addition to the controversy that may arise from situations in which the legal conclusion does not align with the findings of fact, the issue of what constitutes a finding of fact subject to review for clear error appears ripe for debate.

According to Teva, when a district court “reviews only evidence intrinsic to the patent (the patent claims and specifications, along with the patent’s 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.”2 When a district court “looks beyond” the intrinsic record such that the claim construction is informed by extrinsic evidence, however, the Teva approach diverges with respect to the applicable standard of review.3 Subsidiary findings of fact that “underlie a district court’s claim construction” are reviewed for clear error, while the ultimate legal conclusion remains subject to de novo review.4

So long as it is clear what constitutes a finding of fact, the Teva decision appears to adopt a workable framework for appellate review of claim construction decisions. In some cases, however, it may be difficult to determine whether a factual finding “underlie[s] a district court’s claim construction.”5Similarly, it may not be clear when a district court has reviewed only evidence intrinsic to the patent, especially in light of the ubiquitous nature of expert testimony in patent cases. Teva explained that extrinsic evidence may be necessary to “understand . . . the background science or the meaning of a term in the relevant art during the relevant time period,” which are issues common to many patent infringement cases.6 Although the court provided the example of “resolv[ing] a dispute between experts and mak[ing] a factual finding that . . . a certain term of art had a particular meaning to a person of ordinary skill in the art at the time of the invention,” other, perhaps less clear-cut, examples of subsidiary findings of fact can be envisioned.7

The Supreme Court outlined the subsidiary findings of fact subject to review under the clearly erroneous standard in Teva. Thus, on remand, the findings of fact at issue were the same as those used in the Supreme Court’s example above, (i.e., “how a skilled artisan” would understand certain aspects of the technology at issue, and a determination of which expert’s testimony should be credited over another).8 Despite determining that the district court’s findings of fact were not clearly erroneous, the Federal Circuit found the claim at issue invalid for indefiniteness after determining that, despite the factual findings, “in light of the specification and prosecution history, the patentee has failed to inform with reasonable certainty those skilled in the art about the scope of the invention."9

Not all cases will involve subsidiary findings of fact matching the fact pattern ofTeva. In Lighting Ballast Control LLC v. Phillips Electronics North America Corp., the Federal Circuit initially reversed the district court’s construction of the term “voltage source means.”10 On remand from the Supreme Court, however, the Federal Circuit affirmed the district court’s construction after finding its factual findings supported by the record.11 Although the district court does not appear to have resolved a conflict in expert testimony like in Teva, the district court did rely on testimony from the inventor as well as expert testimony indicating that the term “voltage source means” “suggest[ed] a sufficient structure, or class of structures . . . support[ing] a conclusion that the limitations convey a defined structure to one of ordinary skill in the art.”12 The court found that the district court’s factual findings were “based on extrinsic evidence” and were “supported by the record.”13 In Cephalon, Inc. v. Abraxis Bioscience, LLC, the Federal Circuit recognized that the district court made a number of findings of fact in construing the terms “nanoparticles” and “microparticles.”14 For example, the district court found that the constructions were supported by “the ‘widely accepted definition’ of the terms nanoparticles and microparticles,” and “that other Acusphere patents in the same field, many credited to the inventors” of the patent at issue, defined the terms using the same ranges.15 The district court also relied on a textbook to confirm this understanding of the claim terms.16 In rejecting Acusphere’s argument that the district court’s construction constituted clear error, the court emphasized that “how the art underst[ands] [a] term . . . [is] plainly a question of fact,” and “[t]echnical words ‘may give rise to a factual dispute’” that “must be reviewed for clear error.”17

These cases illustrate variations in the fact finding performed at the district court level, moving from resolving conflicts in expert testimony, to relying on expert testimony, to considering extrinsic evidence during claim construction. Shire Development, LLC v. Watson Pharmaceuticals, Inc., also on remand from the Supreme Court, provides some indication as to when the Federal Circuit may determine that the district court only reviewed intrinsic evidence.18 Absent a certain level of fact finding, the Federal Circuit may conclude that the de novostandard of review applies.19 In Shire, the court notes at the start that “[b]ecause this case does not involve factual findings to which we owe deference underTeva, we again reverse the district court’s constructions of the disputed claim terms.”20 Shire argued that “because the district court ‘heard’ testimony from various expert witnesses during a Markman hearing and at trial,” the court should “defer to the district court’s constructions of the appealed terms.”21 The court dismissed this argument, pointing out that “there was no indication that the district court made any factual findings that underlie its constructions.”22 The court reasoned that the deferential standard was not triggered merely because the district court heard or received extrinsic evidence.23 Rather, only “factual findings that underlie a district court’s claim construction” should be reviewed for clear error.24

As the above cases illustrate, whether a subsidiary factual finding will be dispositive or instead outweighed by the legal analysis of claim construction will have to be determined on a case-by-case basis. Until the case law on the issue is more developed, the same can be said for the point at which a district court’s consideration of extrinsic evidence “underlies” its claim construction. Given thatShire has petitioned for en banc review of the decision, more guidance may be yet to come.25 In the meantime, these decisions emphasize the impact subsidiary findings of fact relating to extrinsic evidence can have on appellate review of claim constructions at the Federal Circuit, as well as the importance of establishing a record that clearly indicates the factual findings underlying the district court’s claim construction.