imageRisky business

Patent infringement cases involve an immense amount of risk. Why?

Answers by people who know — especially plaintiffs-side attorneys who work on a contingent-fee basis — will cluster around three facts of life in patent disputes:

  1. high stakes,
  2. huge costs, and
  3. the propensity of the Federal Circuit to undo trial court rulings.

This post deals with the hazards awaiting you behind door number three.


In Teva Pharm. USA, Inc. v. Sandoz, Inc., 135 S. Ct. 831 (Fed. Cir. 2015), the Supreme Court tried to curb the Federal Circuit’s dread of deferring to district courts.

The Federal Circuit had erred in Teva, the high court ruled, because it “reviewed de novo all aspects of the District Court’s claim construction, including the District Court’s determination of subsidiary facts.” Teva, 135 S. Ct. at 836. The justices held 7-2 that the Federal Circuit must accept a trial judge’s “determination of subsidiary facts” about the meaning of patent language unless the judge “clearly” made a mistake.

Resistance to discipline

The court of appeals has given Teva a narrow reading. On remand of that very case, for example, a 2-1 panel of the Federal Circuit brushed aside the idea that it needed to defer to the district court’s finding about what the phrase “molecular weight” meant “in the context of patents-in-suit.” Teva Pharm. USA, Inc. v. Sandoz, Inc., 789 F.3d 1335, 1342 (Fed. Cir. 2015). “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.” Id. That the trial court credited the expert’s testimony to make “key factual determinations” (in the view of the dissenting judge) did not matter. Id. at 1346 (Mayer, J., dissenting).

Indeed, in all the cases involving a Teva issue, the Federal Circuit has rejected arguments that it had to defer to trial courts’ factual determinations in claim constructions under Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996).

Latest effort

The court’s most recent reading of Teva likewise gives the case a grudging scope.

In CardSoft, Inc. v. VeriFone Holdings, Inc., No. 14-1135, slip op. at 6 (Fed. Cir. Dec. 2, 2015), the court recited the holding in Teva and then quickly added:

But as we have repeatedly held after Teva, it 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. See, e.g., Shire Dev., LLC v. Watson Pharm., Inc., 787 F.3d 1359, 1364 (Fed. Cir. 2015) (citing Teva, 135 S. Ct. at 840–42); Teva Pharm. USA, Inc. v. Sandoz, 789 F.3d 1335, 1342 (Fed. Cir. 2015) . . . ; see also Teva, 135 S. Ct. at 840 (“[S]ubsidiary factfinding is unlikely to loom large in the universe of litigated claim construction.”). And even then, we may nevertheless review the district court’s constructions de novo if the intrinsic record fully determines the proper scope of the disputed claim terms. See, e.g., Shire, 787 F.3d at 1364 (citing Teva, 135 S. Ct. at 840–42); Microsoft Corp. v. Proxyconn, Inc., 789 F.3d 1292, 1297 (Fed. Cir. 2015) (reviewing claim construction de novo, and declining to consider “findings on [extrinsic] evidence because the intrinsic record [was] clear”); Eidos Display, LLC v. AU Optronics Corp., 779 F.3d 1360, 1365 (Fed. Cir. 2015) (“To the extent the district court considered extrinsic evidence in its claim construction order or summary judgment order, that evidence is ultimately immaterial to the outcome because the intrinsic record is clear.”).

“In this case,” the panel concluded, “we review the district court’s construction de novo, as the district court did not make any factual findings based on extrinsic evidence that underlie its constructions of the disputed claim term.” CardSoft, slip op. at 6.

An explicit “finding” test?

The panel’s statement that “the district court must actually have made a factual finding in order to trigger Teva‘s deferential review” precedes citation to two cases — Shire and Teva on remand — that nowhere mention any need for an “actual[] . . . factual finding”.

Does “actual” mean “explicit”? I think not.

The district court in CardSoft referred to extrinsic evidence in its Markman opinion on claim construction, but in the course of discussing the crucial language — “virtual machine” — the court spoke only of intrinsic proof (claim language, the specification, and prosecution history). CardSoft, Inc. v. VeriFone Holdings, Inc., No. 2:08-cv-98, 2011 WL 4454940, at *3-*8 (E.D. Tex. Sept. 29, 2011). The trial court therefore did not find — explicitly or implicitly — that extrinsic evidence supported its construction of “virtual machine”.

An implicit factual finding should suffice under Teva. Although we will all agree that the parties should request, and trial courts should make, explicit findings that support claim constructions on the basis of extrinsic evidence.