Supreme Court to Consider De Novo Review of Claim Construction

By Dennis Crouch

Teva Pharmaceuticals USA, Inc. v. Sandoz, Inc. (Supreme Court 2014)

The Supreme Court has granted Teva’s petition for writ of certiorari on the issue of de novo review of district court claim construction. Rather than mounting a facial challenge to all de novo review, the petitioner here argues that, at minimum, a district court’s explicit factual conclusions should be given deference. The case will be heard next term.

The question presented is as follows:

Rule 52(a) of the Federal Rules of Civil Procedure provides that in matters tried to a district court, the court’s “[f]indings of fact . . . must not be set aside unless clearly erroneous.”

The question presented is as follows: Whether a district court’s factual finding in support of its construction of a patent claim term may be reviewed de novo, as the Federal Circuit requires (and as the panel explicitly did in this case), or only for clear error, as Rule 52(a) requires.

In its responsive brief, Sandoz re-characterized the question as follows:

Whether Federal Rule of Civil Procedure 52(a) requires an appellate court reviewing a district court’s construction of patent claims to defer to conclusions that are based on a litigation expert’s declaration that expressly contradicts the patent record.

The case involves Teva’s multi-billion-dollar drug Copaxone that is arguably covered by a set of nine different patents. The appellate decision by the Federal Circuit rejected the lower court’s claim construction and, as a consequence, found several of the patents invalid. The result is that the remaining patents expire in May 2014 rather than September 2015. Teva’s daily US sales of the drug are about $8 million.

I’ll speculate here that the result will be a unanimous rejection of the Federal Circuit’s no deference policy.

24 thoughts on “Supreme Court to Consider De Novo Review of Claim Construction

  1. 7

    The “Federal Circuit’s” no deference policy? I thought the original Markman 517 US 370 (1996) was the one that had the judge riffling through dictionaries and expert testimony, and all of this was supposed to be considered findings of law. I’m not so sure they will reverse their own holding. SUre, the CAFC has been groaning for years abouyt the sheer weight of the caseload: If it’s de novo review, it’s malpractice for the loser NOT to appeal. Yet, is the possible nitpicking over what is fact and what is law going to swallow up even more time? I think it’s too close to call and admonish all that the best way to gamble is to be the casino.

  2. 6

    “In Markman I, we held that, because claim construction is purely a matter of law, this court reviews the district court’s claim construction de novo on appeal. See Markman I, 52 F.3d at 979, 981, 34 USPQ2d at 1329, 1331. In reaching this conclusion, we recognized that:

    [t]hrough this process of construing claims by, among other things, using certain extrinsic evidence that the court finds helpful and rejecting other evidence as unhelpful, and resolving disputes en route to pronouncing the meaning of claim language as a matter of law based on the patent documents themselves, the court is not crediting certain evidence over other evidence or making factual evidentiary findings.” – Cybor

    Sometimes all you can say about this court is, hey, at least they aren’t sitting elsewhere.

  3. 4

    Odd that the Federal Circuit should hold this. In other contexts, it accepts a two-tier review. For example, in evaluating likelihood of confusion in appeals from the TTAB. Here is what they said in one case:

    “We evaluate for correctness the Board’s ultimate legal conclusion as to the likelihood of confusion between the application and opposer’s marks, but we review the underlying factual findings of the Board for clear error. See Kenner Parker Toys v. Rose Art Indus., 963 F.2d 350, 352, 22 U.S.P.Q.2d 1453, 1455 (Fed.Cir.1992); Weiss Assocs. v. HRL Assocs., 902 F.2d 1546, 1547-48, 14 U.S.P.Q.2d 1840, 1841 (Fed.Cir.1990).”

    (Some Circuits treat the whole question of likelihood of confusion as one of fact, so on appeal from regional circuits the Fed. Cir. follows their law.)

    Why should claim construction be different?

    1. 4.1


      It is not. What you are seeing is a conflation of the final question. In truth, the same two-step process still applies in patent law. However, the conflation is being perpetrated by attempting to mash the two separate questions into one question. A question of law will always be a question of law, even of that question reflects a mongrel makeup involving some factual matters.

      No matter how much (or how little) factual matters are at stake, those factual matters necessarily feed into a question of law, and it is that question of law that will (must?) remain open to De Novo review. See post 1.

    2. 4.2

      Mr. Benschar, what “facts” are decided by a District Court in conducting claim construction?

      1. 4.2.1

        Well, how about a different part of the same Teva v. Sandoz opinion in the CAFC (not challenged in the Supreme Court) in which the CAFC stated:

        “The parties dispute whether the district court’s consideration of the percentages in conjunction with its consideration of the “approximately 6:2:5:1” limitation constitutes a “derivative” claim
        construction or a part of its infringement analysis. The former is a question of law; the latter is a question of fact. We hold that whether the amino acid percentages in the accused products meet
        the “approximately 6:2:5:1” limitation is a part of the district court’s infringement analysis. Thus, we review the district court’s conclusions for clear error.”


          Mr. Benshar, experts in claim construction explain background technology and give their opinion on what one of ordinary skill in the art would understand a claim term to mean. But the latter question really is a legal question, the ultimate question, to be decided by a court. The expert’s most beneficial role is to explain the background art to the court. But this should not involve questions of fact.

          A term is construed on a record that speaks for itself, like a contract. If a term is indefinite, the claim should be declared invalid. If the term has two reasonable meanings, the law says chose the narrower.

          I just do not see where the court is deciding an issue of fact that should bind the appellate court in any way.


            Ned, the 1962 English case of Allied Ironfounders v Barking Brassware is a nice simple one, on fact vs law.

            The claim was to a mixer valve for two “fluid” streams but the “field” of the Invention was mixer taps for bathtubs (so the “fluid” was, in reality, water. Prior art D1 had exactly the claimed structure but happened to be a gas mixer (for an oxy-acetylene welder, as I recall). The court held that D1 did not take away novelty, because of the fact (established by the evidence adduced by the patent owner) that, in the field, fluid = liquid.

            On the infringement issue, the claim required “co-axial” fluid feeds. Now, the accused infringement mixer tap also had nested feeds, hot water sleeved by cold, but without the two fluid flows sharing the same single longitudinal central axis. The court nevertheless found infringement because of the “fact” (again, established by the evidence of the patent owner) that, in the field, “co-axial” is wide enough to include arrangements that are nested, whether or not they share the same axis.

            One might suppose there is plenty of room for an appeal court to find “No infringement, and no novelty anyway” except that to do that it has to interfere with the lower court’s finding of fact.

            In the Barking case, the Defendant was so sure of its position that it omitted to adduce any evidence of its own, as to the meaning in the field of “fluid” and “co-axial”.

            Valid, infringed, and no chance to reverse on appeal, all because it seemed unnecessary to file evidence as to the meaning of “fluid” and “co-axial”. Is this what one might call a Triple Whammy?


              In the Barking case, the Defendant was so sure of its position that it omitted to adduce any evidence of its own, as to the meaning in the field of “fluid” and “co-axial”.

              You did say the defendant was Barking, right?


              Perhaps only a nit, but not sure how much help an English case will be, MaxDrei, in comparison (for example) to the US case of In re Schreiber, 128 F.3d 1473.

              Perhaps not a nit, then, is the distinction I made – see post 1.


              Yes IANAE, it is indeed a source of pleasure that the name of the litigant is “Barking”. But now that I have found my copy of the case report I discover that Barking was the winning plaintiff.

              I suspect this was a case in which David won because Goliath was excessively complacent.

              anon, I offer only some facts of the case, to stimulate debate, in any English language jurisdiction. I am not being disrespectful to the sovereignity of the United States of America no matter how much that “fact” might “seem” clear to you or be “apparent” to you. As to the Schreiber case (is vs does) I suppose AI omitted to prove to the court that the gas welding equipment was inherently capable of mixing fluid streams.


              Thanks MaxDrei, I appreciate your nod of respect to US sovereignty.

              But the case I presented – do not the facts speak for themselves? Is “what is” a fact or a legal issue? (and yes, I am reminded of a certain US President, who might testify to an answer of “That depends on what the meaning of “is” is).


              Thanks Max.

              With respect to the “fluid issue,” I think the testimony regarding the background art, while findings of fact, is not an issue of fact that has to be resolved. The legal conclusion of what the claim term means based upon the record and the background art is not a finding of fact but a conclusion of law based upon the facts that are otherwise undisputed.


            Ned in reply to your below, I much enjoyed your clever pun, putting the close quotes after “issue” rather than after “fluid”. Very good.

            Not sure I understand your writing, that testimony is a “finding” of fact. But never mind. I agree of course with your comment (despite all this discussion being about an “issue” that’s “fluid”) that the finding of the ultimate meaning of the claim, based on the found facts, is a conclusion in law.


              Max, one wonders at time how we moderns would have faired at Versailles post the Sun King and prior to the gaieties of Paris circa 1792.


              Speaking of the Reign of Terror, I am rereading A Tale of Two Cities. I highly recommend it if you haven’t had the pleasure.

              I have also read a number of first-hand accounts of the revolution by survivors. Simply heartrending. Did you know that whole families were butchered? Pregnant women were spared until they gave birth. Immediately, they were taken to the guillotine. Can you imagine?

              At other times mobs to people apart while still alive, carrying body parts about on display. The worst of the Roman Empire could not have been so violent.

  4. 3

    Dennis, is it your view that claim construction based on intrinsic written record involves an issue of fact?

    We have a contract. Is its interpretation an issue of fact? If the contract is ambiguous on a point, I would agree that parol testimony as to intent would make that testimony an issue of fact.

    But, where in claim construction is there room to turn a construction based on the record into a issue of fact? If there are two reasonable construction, the rule is to adopt the narrower.

    1. 3.1

      “Is its interpretation an issue of fact? If the contract is ambiguous on a point, I would agree that parol testimony as to intent would make that testimony an issue of fact.

      But, where in claim construction is there room to turn a construction based on the record into a issue of fact?”

      You seem to answer your own question. Interpretation of the claims of a patent is a lot closer to parol than straight contract interpretation. Contract language is legal language and there is no legal/factual dichotomy. The judge may sit as a judge and is imbued to be his own expert. Patent language by definition is industry language as it is viewed from one of ordinary skill in the art. The judge isn’t and cannot sit as an industry expert, so from where would he draw his expertise? That invites expert testimony which requires a factual finding.

      It’s only in the case of a particular, special definition within the patent that you might be able to get away with it only being a question of law, but even then there might be a fact question to put the defined term into a broader context.

      1. 3.1.1

        The judge uses, or should use expert testimony among other sources to put herself into the place of the skilled artisan. She should take the time to understand the technology and the status of the technology at the time of filing. This, to me, is at a minimum a quasi-factual finding to which the Fed. Cir. owes deference. Reasonable minds may differ as to the views of the hypothetical skilled artisan, but the district court judge has more intimate contact with the subject matter and is better placed to make this determination.

        The Fed. Cir.’s review standard, in my view, has caused district court judges to become complacent about really learning the technology and fully understanding the complexities of the case, because they are rightfully concerned their efforts are for naught. In the end, we get poor constructions, and confusing, contradictory rulings out of the Fed. Cir.

      2. 3.1.2

        RandomGuy, if there’s going to be a debate, the factual debate, on the meaning of particular terms the people of ordinary skill in the art that presents a genuine issue of fact that has to be resolved, I do not believe this is a role of the court to resolve it, but the role of a jury.

        Think about it. If there is a constitutional right to a trial by jury for patent validity, how can the most critical facts involved in determining whether the patent is valid are not, it’s claim construction, be decided by a court and not by a jury. I don’t think this can be possible.

        I would hope someone on this appeal to the Supreme Court will races issue. There cannot be genuine issues of fact involved in claim construction lest there be a right to trial by jury in those disputed facts.

        However if the court does decide that there are factual disputes in the district court must be given some deference, this will open the issue up that I just raise on whether these factual disputes have to be tried to a jury in the validity case. Remember, Beacon Theaters is also involved to the extent that the court cannot decide factual issues that are critical to a decision that bears a right to a trial by jury.

  5. 2

    I think the Court is a little perturbed that an appellate court would revisit factual findings of a d/ct judge without deference, particularly, when the judge’s fact finding was made after listening to the experts of either side. The appellate court panel is in no position to judge the credibility of each witness, and therefore, the factual finding must stand absent clear error.

    1. 2.1

      But J, is that the question presented?

      Whether or not the Court is perturbed a little or even a lot, the question presented is not whether a fact stands absent clear error, but whether a legal conclusion is to be treated at that same level of fact.

  6. 1

    Appears to be an easy question, as there is a clear difference between a fact and a legal conclusion based on fact.

Comments are closed.