In early 2015, the U.S. Supreme Court changed the standard of review for patent claim construction with its decision in Teva Pharmaceuticals USA v. Sandoz Inc., 135 S. Ct. 831 (2015) (Teva I). Previously, the U.S. Court of Appeals for the Federal Circuit reviewed all claim construction rulings de novo—irrespective of whether they involved any factual findings—without giving any deference to the lower court's factual findings. In Teva I, the Supreme Court held that factual findings during claim construction such as findings related to extrinsic evidence must be given deference on appeal and should be reviewed for "clear error" instead.
In two recent cases, including the remand from Teva I, however, the Federal Circuit came to opposite conclusions when reviewing lower court decisions under the new clear error standard.
The Federal Circuit on remand from the U.S. Supreme Court, held on June 18, 2015 that the district court's claim construction was incorrect and that the asserted patent claims are invalid as indefinite. Teva Pharm. USA v. Sandoz Inc., 789 F.3d 1335 (Fed. Cir. 2015) (Teva II). While the majority of the panel (led by Judge Kimberly Moore and joined by Judge Evan Wallach) did give deference, in accordance with the Supreme Court's decision, to the district court's particular fact-finding regarding the claim term, "molecular weight," the majority still found that the lower court had drawn incorrect conclusions of law from the factual bases. Furthermore, it appears that the invalidity holding was easier to justify under the U.S. Supreme Court's new "reasonable certainty" standard decided in Nautilus v. Biosig Instruments, 134 S. Ct. 2120 (2014).
The Teva cases involve Copaxone, a drug approved for treatment of multiple sclerosis. When Sandoz marketed a generic version, Teva sued for patent infringement. Sandoz said the patent is invalid because the claim term that required the active ingredient having "a molecular weight of 5 to 9 kilodaltons" was indefinite under 35 U.S.C. §112 because it did not state which of the three possible methods was used to determine the active ingredient. Teva I, 789 F.3d at 835–836. The active ingredient, copolymer-1, is a mixture of individual polymer molecules with different constituent ratios, and different molecular weights; it is often expressed as "average molecular weight." There are at least three ways to calculate average molecular weight: peak average molecular weight, number average molecular weight, and weight average molecular weight. Peak average is the molecular weight of the most abundant molecule in the sample; number average is the combined total mass of all of the molecules in the sample divided by the total number of molecules; and weight average is similar to number average except instead of dividing the sample by total number of molecules, one takes into account the weight fraction of each molecule.
The patent failed to disclose which method was used for determining the molecular weight of the active ingredient. While it did mention the use of size exclusion chromatography (SEC), that does not help as all three ways use SEC. Statements in related cases magnified the issue. In one case, the applicant said the average molecular weight is peak average, while in another related case they stated that it is weight average. Teva's expert testified that because specification only mentions SEC, one of skill in the art would have understood that average molecular weight could only mean peak average, since the other methods require additional calculations. The district court found Teva's expert credible and concluded that the claim is not indefinite. On appeal, the Federal Circuit reviewed the claim construction de novo and concluded that the claim is indefinite.
The Supreme Court vacated that decision and remanded the case back to the Federal Circuit to apply its new standard of review when reviewing the trial court's claim construction ruling involving subsidiary factual disputes. While the majority concluded that the ultimate task of construing patent claims is a legal question reviewed de novo, it also explained that to the extent the claim construction involves resolution of subsidiary factual disputes, review of such factual findings is for clear error.
The Supreme Court provided some guidance regarding what issues are considered to be legal determinations as opposed to subsidiary factual findings. It noted that interpretation of evidence "intrinsic to the patent," i.e., "the patent claims and specifications, along with the patent's prosecution history," is to be treated as a legal determination and reviewed de novo. And in cases where the district court consults extrinsic evidence such as expert testimony to understand the technical background involved in the case, findings on such issues are factual in nature and therefore are reviewed for clear error. The dissent argued that the need for uniformity in patent claim construction supports the de novo review standard and further that because patents are more analogous to statutes that do not involve subsidiary factual issues as opposed to contracts that do, patent claim construction does not involve findings of fact.
While the Federal Circuit did not find any clear error in the district court's factual conclusion that one of skill in the art would understand that average molecular weight would be peak average using SEC without further calculations, it nevertheless concluded that the patent claim is indefinite as the extrinsic evidence (i.e., expert testimony) did not resolve the ambiguity of the term "molecular weight." The Federal Circuit then reasoned that the extrinsic evidence did not change the fact that there are at least three different ways of calculating "average molecular weight" that could result in three different values. Considering only the intrinsic evidence including the claim language, specification, and prosecution history, the Federal Circuit, citing the Nautilus case, found that a skilled artisan "would still not be reasonably certain in light of the entire record as to which type of average [weight] was intended" in the patent claims. Teva II, 789 F.3d at 1345.
Shortly after deciding Teva II on remand, the Federal Circuit came to almost the opposite conclusion in Lighting Ballast v. Philips Electronics, 790 F.3d 1329 (Fed. Cir. 2015), also on remand. In Lighting Ballast, the district court first concluded that the claim term "voltage source means" was a means-plus function element and is indefinite because the specification failed to disclose sufficient structure. Later, while deciding a motion for reconsideration, the district court reversed itself stating that it did not give enough deference to the expert testimony regarding how a skilled artisan would understand the specification. On appeal, a panel of the Federal Circuit disagreed and reversed. Sitting en banc, the Federal Circuit reaffirmed the de novo review standard for all claim construction issues, which the U.S. Supreme Court vacated and sent for remand in light of its Teva I decision.
On remand, the Federal Circuit reviewed the district court's factual determinations under the clear error standard and affirmed the lower court's determination that the specification provided sufficient structure. While the outcome in Lighting Ballast is opposite to that of Teva II, these cases are distinct enough in at least one important aspect. In Lighting Ballast, extrinsic evidence was introduced to explain what the intrinsic evidence (i.e., specification) meant, whereas in Teva II, extrinsic evidence was introduced to fill in the gaps within the intrinsic evidence that failed to describe which method was used for measuring the claimed "molecular weight."
It appears that the new Teva I standard would only apply in cases where the district court had relied on extrinsic evidence to make a factual finding during claim construction. For example, in CardSoft LLC v. VeriFone, Inc., 769 F.3d 1114 (Fed. Cir. 2014), the Federal Circuit, on remand, used the de novo standard as opposed to Teva I's deferential clear error standard by reasoning that the district court did not make any factual findings based on extrinsic evidence while construing the claims. Even though extrinsic evidence was admitted at trial, the Federal Circuit concluded that the district court did not rely on the extrinsic evidence and that mere submission of extrinsic evidence is not enough to mandate deference to a district court's claim construction.
While it may take years to appreciate the impact of Teva I on claim construction, the case raises a few questions regarding patent practice going forward. For example, does the change in the review standard incentivize patent applicants going forward to provide more detailed disclosure at least for those patent terms that are likely to be construed? One advantage of such detailed disclosure would be to avoid expert testimony at trial; even if one were to be used, such testimony would likely be used to explain what the specification actually meant (similar to Lighting Ballast) as opposed to supplement what the specification might be missing (similar to Teva I). Another question is whether trial court judges would be more willing to allow expert testimony during claim construction hearings now that their factual findings can only be overturned for clear error. Yet another question is whether the perceived higher bar of the deferential Teva I standard reduces the number of appeals to the Federal Circuit.
Finally, the new claim construction standard might increase a likelihood of forum shopping, which the Federal Circuit was created to avoid. For example, under the previous de novo review for both legal and factual aspects of claim construction, plaintiffs do not have any incentive to forum shop—at least as far as claim construction issues are concerned—because appeals arising from any district court are reviewed under the de novo standard without deference. With the more deferential Teva I standard, plaintiffs might have an incentive to select a district court empirically more amenable to rely on extrinsic evidence for claim construction such that the claim construction has a lower likelihood of reversal at the Federal Circuit.
In any case, it appears likely that the new patent claim construction review standard will affect future patent practice, the full impact of which will only be known with the passage of time
The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.
