Guest Blog: Greg Vetter, Oral Argument in Teva v. Sandoz: Patent Law Exceptionalism and Judicial Process

Law or Fact?

The sometimes vacuous distinction between law and fact reminds us that nuance influences legal process. These two labels seem distinct when juxtaposed with an example: a patent claim is invalid when drawn to an inoperable invention, such as a perpetual motion machine. This proposition in law needs a factual finding to fulfill its purpose: that a specific “perpetual motion” machine is indeed inoperable as a matter of science.

An issued patent on an inoperable perpetual motion machine would be an embarrassment for the patent system. In the view of some, during the last fifteen years, the patent system’s judicial process for claim construction invites similar derision.

What is that process and why is it that way? And how does it relate to the Supreme Court’s forthcoming opinion in Teva v. Sandoz? The Court heard oral argument in the case on Wednesday, October 15, 2014. In what follows below, I discuss the implications for patent law and its most important court: The United States Court of Appeals for the Federal Circuit.

Concern with United States technological competiveness was among the many factors that influenced Congress to create the Federal Circuit. Since its inception in 1982, the Federal Circuit has been perhaps the primary institutional focus for patent law. Its efforts to unify and regularize patent law have waxed and waned in effectiveness all during a time when the patent system has grown in importance during the last three decades. Innovation and technological competiveness in a global setting remain top priorities for Congress with the patent system, as evidenced most recently with the changes to the patent system implemented by the America Invents Act (AIA), passed in 2011 and fully effective in the Spring of 2013.

The AIA is not the focus of the issues in Teva, but the AIA alters the institutional landscape for the patent system, likely raising the importance of the United States Patent and Trademark Office (PTO) in relation to the courts and the Federal Circuit. Relatedly, Teva may alter the allocation of judicial authority and power between the district courts and the Federal Circuit. How this might impact outcomes for competiveness and technological process are background issues in the case, while in the forefront Teva examines issues similar to those from law school hypotheticals about what is a fact versus what is an issue of law.

Claim Construction Under Markman/Cybor

The patent at issue in Teva is typical in that it has many pages of scientific front matter, called the “specification,” and a much smaller number of pages containing claims at the end of the instrument. The claim is the focus for patent law’s exclusive rights. Typically, lawyers draft patent claims. A typical patent will have several dozen claims (each claim is, per PTO rules, a single sentence). In litigation, the district court interprets the words in the claim, in light of the specification and in light of the exchanges between the patent applicant and the PTO. The record of those exchanges is called the “prosecution history” – because the applicant is prosecuting her application before the PTO. Collectively, patent law calls these three things “intrinsic evidence” to help determine the meaning of the claim terms: (1) the words of the claim; (2) the specification; and (3) the prosecution history. Anything else used to determine claim meaning is of less import, and is called “extrinsic evidence.” A prime example of extrinsic evidence is testimony by an expert.

In the mid‑1990s the Federal Circuit, affirmed by the Supreme Court in Markman v. Westview Instruments (1996), held that claim construction was a matter to be undertaken by the judge. It was not the province of the jury. The Markman opinions changed the patent litigation system as district courts began holding Markman “hearings” to construe the meaning of claim terms dispositive in patent litigation. Markman hearings vary among the district courts as to their form and structure. Some include trial‑like ritual, such as oral testimony before the judge. Some are completely based on the paper submittals made by the parties, and these do not really resemble a “hearing.”

In 1998, in an en banc opinion, Cybor Corp. v. FAS Tech., the Federal Circuit further held that claim construction determinations by the district court would be reviewed de novo. No deference would be given to the district court judge’s claim construction. This is incongruent in the eyes of some, particularly when a Markman process strongly resembles a trial‑like hearing. In any event, after Cybor, the model of claim term meaning as a complete issue of law took hold. That model continues to this day, but Teva challenges it with a venerable source of law: the Federal Rules of Civil Procedure (FRCP).

The Issue of FRCP 52

Remorselessly, FRCP 52(a)(6) states: “[f]indings of fact, whether based on oral or other evidence, must not be set aside unless clearly erroneous . . . .”

The Federal Circuit in Cybor deemed claim interpretation to be “a purely legal question, [reviewed] de novo on appeal including any allegedly fact-based questions relating to claim construction.” FRCP 52 was not considered by the Federal Circuit in the Cybor opinion.

In Teva, the question presented to the Court exposes the potential conflict between FRCP 52 and Cybor:

“Whether a district court’s factual finding in support of its construction of a patent claim term may be reviewed de novo . . . or only for clear error, as Rule 52(a) requires.”

It boils down to whether the Court will view the activities underlying claim construction as “factual” in the sense of FRCP 52. If that characterization is applied, the approach of Cybor will fall.

The Patent Claim Term at Issue in Teva

The Court will determine the fate of Cybor in a case about a patent for Teva’s blockbuster drug Copaxone. The litigation commenced when Sandoz sought to market a generic competitor drug. At the district court level, and at the Federal Circuit, the case involved multiple patents, but only one Teva patent related to Copaxone remains unexpired at the time of oral argument.

As is typical in patent infringement litigation, the defendant, here Sandoz, raised various arguments that the patent’s claims are invalid. Because there are multiple legal tests to satisfy to obtain a patent originally, these same legal tests are the basis for invalidity arguments.

One way to allege that a patent claim is invalid is to argue that it is indefinite. Indefiniteness is related to claim construction: if there is no claim construction that will, with reasonable certainty, inform an artisan/technologist in the field of the invention about the scope of the claim, then it is indefinite and thus invalid.

Virtually all aspects of patent infringement litigation depend on the claim construction. The scope of coverage of a patent claim will influence both the infringement question and the various arguments that a claim is invalid. For example, oftentimes infringement defendants will argue that a patent claim is obvious. Obviousness is an issue, unlike claim construction, that the Federal Circuit has explicitly stated is a mixed question of law and fact. The ultimate determination of obviousness is a question of law, but based on underlying factual findings. For the question of infringement, the district court judge construes the claims using canons of claim construction emanating from the Federal Circuit. The comparison of the interpreted claim to the accused infringing product or method, however, can be an issue that goes to the jury. But, under Markman, the claim construction stays with the judge. However, judges can make factual findings.

In Teva, the district court found that the claim term “molecular weight” was not indefinite. This determination was cast in doubt because there are apparently at least three possible ways to interpret that term. The patent claim covers a complex molecule and molecular weight is a way to describe it both qualitatively and quantitatively. The three possible ways used differing concepts to represent an average molecular weight. The district court interpreted the claim to be understood to use “peak average molecular weight” as the measurement standard. This determination was made in part by reference to information in the specification. Therefore, under the district court’s logic, the claim was not indefinite. The claim was also infringed by Sandoz’s planned generic drug, and thus Teva was in a position to block the generic competitor from the market until the patent expired.

At the Federal Circuit, however, Sandoz prevailed on its indefiniteness challenge. The Federal Circuit viewed the information in the specification as less definitive, and perhaps in conflict with information in the prosecution history. Finding this supposed incongruence in the intrinsic evidence, it did not view the claim as definite. Because indefiniteness invalidates the claim, it no longer mattered that Sandoz had been adjudged below to have infringed it.

In its case before the Court, then, Teva hopes to benefit by convincing the Court that the FRCP 52 standard applies. The more claim construction is a mixed issue with subsidiary questions of fact, the more chance for remand in which the Federal Circuit would give deference to the district court’s work to understand what that claim meant when it recited the term “molecular weight.” The deference would come from the clear error standard applicable to judge‑made factual findings. For example, when a judge hears direct testimony by an expert witness to understand some scientific or technological concept related to the invention, this would be in the “sweet spot” for characterization as a fact‑based finding.

Pros and Cons of the Markman/Cybor Regime

One criticism of the Cybor approach was that final certainty for claim term meaning only occurred near the end of the litigation, with the Federal Circuit appeal. During the fifteen years under Cybor, numerous studies of claim construction reversals showed the problem peaking at about fifty percent under some studies, but with the problem seeming to diminish to about half that rate more recently. For those concerned with the claim construction reversal rate, greater deference to the district court on claim construction is viewed as a possible way to further reduce the reversal rate.

Claim construction reversals are especially problematic for the patent litigation process. A new claim construction undoes a fundamental determination that can upend the rest of the case, sometimes resulting in a new trial. Patent litigation is already the “sport of kings” with tremendous costs. Thus, many stakeholders in the patent system are concerned by anything that contributes to these costs. Claim construction reversals also create tension between the Federal Circuit and the district courts. The energy and resources invested into the claim construction process are perceived to be much greater at the district court level. And they are sometimes dressed up in trial‑like ritual of a Markman hearing, furthering a feeling that these are factual inquiries. All these process considerations sweeten the invitation that FRCP 52 should apply to aspects of claim construction.

It is commonly taught in law school that courts fluidly characterize issues as either law or fact in order to impact the balance of importance in the judicial hierarchy. There are limits to this gamesmanship, but it is not surprising that the Federal Circuit has continued to favor treating claim construction as entirely an issue of law. Doing so gives it greater influence over patent law generally, and over each case on appeal specifically. The Federal Circuit recently revisited its Cybor opinion, but, with some dissents, Cybor stood after the Federal Circuit’s en banc review to reconsider it in early 2014.

During the last several years, the Court has been very active in reviewing the Federal Circuit’s patent law jurisprudence. In the view of some, the reversals arising from this review signal a preference by the Court to have Federal Circuit jurisprudence not be too specialized or exceptional for patent law. As to its coverage of areas of technology, patent law is unitary. We have mostly the same patent law doctrine for mechanical devices as we have for chemicals as we have for other areas of technology. There are exceptions to this, but a mostly unitary patent system as to the technology coverage perhaps suggests keeping patent law procedural concepts in line with general procedural devices in law. This thinking also sweetens the invitation to apply FRCP 52.

Court’s Questioning of the Federal Circuit Rule at Oral Argument

On the other hand, however, there may be valid reasons to consider the Cybor view as correct. Some of these arose during the oral argument at the Court. For example, there is the natural inclination to analogize the claims of a patent to other legal instruments, such as public law in the form of a statue, or private law instruments such as a deed for land or a contract. In addition, one could reasonably view the process of understanding the technology (to help interpret the claims) as something different from the types of facts contemplated by FRCP 52.

One concern raised by some Justices was whether a particular claim term in a particular patent might be construed differently by two different district court judges. With fact‑issue deference by the Federal Circuit, this facial conflict might be harder to resolve, leading to a particularly stark form of disuniformity. There are procedural and other reasons to consider this scenario as unlikely, but it showed some concern by the Justices for uniformity in patent law and patent litigation outcomes.

A larger reason one might consider Cybor to be correct relates to institutional purpose, and relates to the desire for uniformity in patent law. The Federal Circuit was created as a specialized appellate court with a specific mandate to add uniformity and certainty to patent law. Despite the recent reversals by the Court of the Federal Circuit’s work, in earlier opinions it has acknowledged the Federal Circuit’s special role. The mandate by Congress to centralize patent infringement appeals to a single court is a strong policy influence that the Court might very well use to justify an “exception” to FRCP 52’s application to claim construction. Such an “exception” would not be explicitly stated because doing so would be unworkable as a matter of jurisprudence. The “exception” would be in the form of the Court joining the Federal Circuit in the view that none of the activities underlying claim construction are the type of activities contemplated by FRCP 52 in the context of patent law.

Potential Unintended Consequences for Subsidiary Fact Finding within Claim Construction?

In Markman, the Court famously characterized claim construction as a “mongrel practice.” The metaphysical lens of the issues in Teva demonstrates the difficulties: the regime under Cybor has its problems of judicial diseconomy and potentially driving unnecessary costs into patent litigation, but will fact versus law distinctions further complicate claim construction in a way that outweighs the benefits of appellate deference? A new regime that embraces FRCP 52 might solve some of these problems, but what new problems might it create? To conclude, I discuss three possibilities.

First, district court forum shopping might be amplified. A mixed law/fact approach to claim construction enables district courts to further specialize in patent infringement litigation. An important trend in patent litigation over the last decade has been the emergence of district courts garnering disproportionate numbers of infringement cases. Deference on fact finding might further empower district court judges, inviting greater specialization with the attendant forum preferences by litigants. The perceived problems with district court forum shopping have been sufficiently significant to attract the attention of Congress as it considers additional post‑AIA revisions to patent law.

The second potential consequence is related to the first. While perhaps costs will be reduced for unnecessary retrials after claim construction reversals, it seems that the use of experts in patent litigation will increase. Thus, will cost savings from less rework due to reversals be swamped by increased use of experts or other types of fact‑finding in front of the district court judge?

Finally, there is the possibility of “satellite litigation” to cleave the law/fact divide. This might be a cost that reduces itself as the new approach comes to equilibrium. The insight is that during the early years in a mixed approach to claim construction the litigants continually argue about whether specific activities subsidiary to claim construction are law or are fact.

It remains to be seen whether the Court will change claim construction to a mixed question of fact and law such that FRCP 52 must apply. If it does not, claim interpretation will remain the mongrel practice we have come to know under the Markman/Cybor approach. If it does, it may be a new species of mongrel practice with some unexpected characteristics and impacts on the patent system.