Prosecution History Estoppel: Defining a Foreseeable Alternative

Duramed Parmaceuticals (Teva Women’s Health) v. Paddock Labs. (Fed. Cir. 2011)

By Dennis Crouch

Note: This case supports the already accepted notion that the presumptive bar to the doctrine of equivalents created by a narrowing amendment during patent prosecution is very difficult to overcome.

Teva is the largest worldwide manufacturer of generic drugs. In this case, however, the tables are turned. In 2008, Teva purchased Barr and its subsidiary Duramed. In 2009, the company sued Paddock – alleging that the defendant’s application with the FDA to distribute a generic estrogen-derived hormone replacement constituted patent infringement under 35 U.S.C. § 271 (e)(2)(A).

(2) It shall be an act of infringement to submit – (A) an application under section 505(j) of the Federal Food, Drug, and Cosmetic Act … for a drug claimed in a patent or the use of which is claimed in a patent.

The asserted patent – No. 5,908,638 – covers Duramed’s Cenestin ERT product and appears to be set to expire in 2015.

Narrowing Amendment: During prosecution of the patent, Duramed amended narrowed its claim 1 in response to an office action rejection. The amendment took a limitation from a dependent claim and added it to claim 1 so that the independent claim now required that the pharmaceutical composition be “coated with a moisture barrier coating comprising ethylcellulose.”

In its abbreviated new drug application (ANDA), Paddock indicated its plans to use a barrier made of polyvinyl alcohol rather than ethylcellulose. So, when it sued, Duramed alleged infringement under the doctrine of equivalents (DOE).

Festo and Prosecution History Estoppel: The district court rejected the DOE claim – holding that Duramed was barred from claiming equivalents over the ethylcellulose coating based upon the doctrine of prosecution history estoppels. (Citing Festo).

On appeal, the Federal Circuit affirmed – holding that the district court had properly ruled the patent not infringed.

A patent is literally infringed only when the accused action or device meets all of the elements of the invention as defined by the patent claims. Long ago, courts added some leeway to the infringement analysis through the Doctrine of Equivalents. Through application of the DOE, infringement can be found even when the accused device is outside of the literal scope of a patent claim – so long as there is “equivalence between the elements of the accused product or process and the claimed elements of the patented invention.” Warner-Jenkinson Co. v. Hilton Davis Chem. Co., 520 U.S. 17 (1997).

DOE and PHE: The courts have limited the DOE in a number of ways. One of the most important limitations is that of prosecution history estoppel (PHE). In Festo and other cases, the courts held that it would be unfair to allow a patentee to use the DOE to recapture patent scope that had been voluntarily surrendered in order to acquire the patent.

The process for determining whether prosecution history estoppel applies first asks whether estoppels presumptively applies and then asks whether the patentee can overcome the presumption. Thus:

  • PHE is presumed when a patent claim is narrowed for reasons relating to patentability;
  • However, the presumption can be overcome by showing that the patentee did not intend to surrender the particular equivalent.

In Festo, the Supreme Court offered two specific arguments that could be used to overcome the presumption: (1) that the alleged equivalent would have been unforeseeable at the time the amendment was made; and (2) that the reason for the amendment only tangentially related to the equivalent in question. In addition, the Supreme Court noted that other explanations may also work to overcome the presumption.

Here, the Federal Circuit agreed that the presumption of PHE applied because Duramed had narrowed its claims in response to a prior art rejection. The court also agreed that Duramed had failed to overcome the presumption. In particular, the court held that the polyvinal alcohol alternative should have been foreseeable because it was already known to be used in other pharmaceutical compositions.

What is Foreseeable?: Courts will find an alternative “foreseeable” if it is known in the prior art pertinent to the claim scope before the amendment. This definition means that a foreseeable alternative is somewhat more broadly defined than is an obvious alternative. Here, Duramed argued that it was not foreseeable at the time of the amendment to use the polyvinal alcohol alternative in combination with the estrogen drug. On appeal, the Federal Circuit held that foreseeability for PHE analysis does not require such precision – rather, it was sufficient for the court that the alternative was known for use with other pharmaceutical compounds.

As we spelled out in Schwarz, when the language of both original and issued claims begins with the words “[a] pharmaceutical composition,” that language defines the field of the invention for purposes of determining foreseeability. Accordingly, PVA MBCs need only to have been known in the field of pharmaceutical compositions as of the time of Duramed’s narrowing amendment. . . . [F]oreseeability does not require such precise evidence of suitability. And even if the [prior art] disclosure indicates that PVA is less than ideal in some pharmaceutical uses as an MBC, it is still disclosed to be useful as such, and that renders it foreseeable for purposes of prosecution history estoppel. Foreseeability does not require flawless perfection to create an estoppel.

Because Duramed was unable to overcome the presumption, prosecution history estoppels applies to bar Duramed from asserting that the accused Polyvinal alcohol barrier is equivalent to the claimed ethylcellulose barrier.

28 thoughts on “Prosecution History Estoppel: Defining a Foreseeable Alternative

  1. It would seem the application itself teaches that the one MBC is obvious.

    Supposing that the application itself does actually teach that, and further supposing that the teaching in the application is factually correct, I don’t have any problem denying the applicant a claim to the obvious.

    That’s just the sort of statement we should want to see in the application. It all comes back to Max’s wacky Euro-style suggestion that the applicant should know what he has invented before he seeks patent protection for what he has invented.

  2. What are you talking about, IANAE? Are you suggesting that “support” is not the same thing as “in the original spec.”

    No, I’m saying the opposite. I’m saying that if you want to claim it, you should put it in the spec. If your complaint is that you are unable to literally claim a foreseeable equivalent because you don’t have 112 support for it, your complaint is actually that you should have foreseen the foreseeable and drafted a better spec.

  3. More to this point, if the application teaches that any MBC would work and that is actually claimed, how does the claim become patentable if limited to just the one MBC? It would seem the application itself teaches that the one MBC is obvious.

  4. What are you talking about, IANAE? Are you suggesting that “support” is not the same thing as “in the original spec.”

    Btw, I agree with Max’s summary of the issue.

    Now, may I ask, if one discloses but one MBC, how can one claim the genus MBC? This, after all, is not a mechanical invention, or is it? What was all the argument about that it was not known that all known MBCs would work for this application? How can one claim the genus when it is not known that all MBCs would work in this application?

    Something doesn’t smell right.

  5. I always think drafters have to choose, at the outset, between implying that their species (here EC) is special, and other members of the genus (like PVA) are not so good, and implying that their species is not different from any other member of the class. That choice has to be conscious and deliberate.

    For it matters, at least when it turns out that at least one member of the class is old.

    If you drafted that your species was better than other class members all well and good. But don’t then complain when a competitor uses one of those other class members.

    If you wrote that other class members will also serve, then you should have wider enforceable scope, but woe betide your validity, when some class members turn out to be old.

    Of course, if you don’t even mention other class members in your app, it beats me why any court should even give you the time of day, when you say that PVA infringes your EC species claim.

  6. So, if something is foreseeable, it has to be literally claimed. But one needs support for that, so it has to be in the original spec.

    Fixed that for you.

    I guess my 10:55 post on the subject was too cryptic.

  7. Squeeze, indeed. MBC as a genus was obvious. Particular species, such as the claimed ethylcellulose, not. That, apparently, was the only MBC disclosed. By limiting the claim to it, all other known MBC were excluded. This seems right.

    So, if something is foreseeable, it has to be literally claimed. But one needs support for that, so it cannot be claimed.

    If it was not foreseeable, it could be an equivalent, but it strikes one that this cannot be right in principle, just as Malcolm suggested in the first post above.

    So, while we have the figment of a flexible rule of prosecution history estoppel, in reality, it is not all that flexible.

  8. How does one literally claim PVA when one discloses only ethylcellulose?

    That’s a very good question, but it’s far too late to ask it when the time comes to amend the claims in prosecution.

    But, again, it’s not the rest of the world’s problem if for whatever reason the applicant fails to literally claim it.

  9. IANAE: “If so, you’ve surrendered it by not literally claiming it.”

    How does one literally claim PVA when one discloses only ethylcellulose?

    Do we have a Catch 22 here?

  10. IANAE yes, indeed. BTW, in Germany (another logical country) they call it the Formstein Defence.

    Glad to know you appreciate English squeeze arguments. In Germany, with its separate tracks for infringement and validity, such squeeze arguments are harder to run.

  11. Max, this would be the Gillette case of Gillette Defense fame?

    Yeah, good luck with that in the US. You need to pick a side (too broad to be valid or too narrow to be infringed) before you get to the Markman hearing, and then try your luck with that argument at trial.

    You see, logic was invented in Europe, and has yet to make the trip across the pond. Maybe when it does, it’ll bring Gillette along for the ride.

  12. Now, if the patentee genuinely believed at the time of the amendment that PVC would not work as an MBC for estrogen and could prove it, I think that the use of PVC for the application would in fact not be foreseeable.

    Not foreseen, certainly. But that’s not the same thing.

    Obviously it was not foreseen by the applicant, because the applicant didn’t disclose or claim it. That’s a trivial result in every case where the question even arises.

    The important question here is whether a person of ordinary skill (not the particular applicant) would have expected that this known coating would work about the same as that known coating intended for the same purpose. If so, you’ve surrendered it by not literally claiming it.

  13. Pete, I was hoping to goad somebody into confirming (or denying)the notion “That which was already obvious in the USA at the date of the claim can’t infringe, if it first appears after the date of the claim”. Too early to say, yet, whether I failed.

  14. What a long way we have travelled, since 1913.

    Yes, we have finally arrived in the States, where the subject of this thread is focused on, from the UK, where MaxDrei is pulling his comparative case law.

    MaxDrei’s arms must really be tired.

  15. Pure coincidence. The English case Convatec v Smith just issued includes a reference to the famous Gillette case from 1913, quoting its:

    “It is impossible for an ordinary member of the public to keep watch on all the numerous patents which are taken out (this from 1913) and to ascertain the validity and scope of their claims. But he (in 1913) is entitled to feel secure if he knows that that which he is doing differs from that which has been done of old only in non-patentable variations such as the substitution of mechanical equivalents or changes of material, shape or size”

    What a long way we have travelled, since 1913.

  16. Sorry readers. I should have read the Decision before offering a comment on it. On page 1, it recites: “Original claim 1 recited a conjugated estrogen pharmaceutical composition coated with a moisture barrier coating” and the Examiner indeed found it obvious. So there was a squeeze on the Applicant. Select ethylcellulose out of the set of moisture barrier materials, or be told again that any wider claim embraces subject matter that is obvious.

    I suppose that, sooner or later, it would have been possible for Paddock to have run that squeeze too, except for that good old Presumption of Validity.

  17. Afterthought: was it by any chance obvious, merely to coat with a moisture barrier? So the recitation of cellulose as the specific barrier was necessary, just to get to issue? Is that the point?

  18. Haven’t read the Decision yet. This comment is made purely to tug the thread.

    With hindsight, a better set of dependent claims would have got the patent owner home. All it needed was to write two dependent claims, the first to the moisture barrier, the second to the particular barrier material.

    So, courts have to decide how much slack to give to those not competent to write a set of dependent claims of ordinary quality. In declining to use the DOE to rescue such situations for the patent owner, is not the court serving the public policy objective of encouraging those writing dependent claims to reach a basic minimum standard of quality, to the ultimate benefit of all inventors?

    Prosecuting in Europe cases drafted in the USA, I am often struck by the abysmal quality of the dependent claims. In Europe, the dependent claims are likely going to be of decisive importance to the case, years down the line. I gather that, in the USA, they are seen as being in all probability of no practical importance (other than to inflate the scope of the independent claim from which they depend, on a claim differentiation theory).

    Above I wrote “with hindsight”, but no hindsight was needed here, to write sensible dependent claims. Just the competence of an average patent agent of ordinary intelligence and with at least a few months of work experience.

    Readers?

  19. Malcolm,

    A fine point, I don’t think the patentee said that he didn’t think PVA would not work as an MBC for an estrogen drug. I think he said that it was not known that PVA would work. This is slightly different.

    Now, if the patentee genuinely believed at the time of the amendment that PVC would not work as an MBC for estrogen and could prove it, I think that the use of PVC for the application would in fact not be foreseeable. How else would you prove lack of foreseeability?

    But just because it was not foreseeable, it does not follow that it was an equivalent. It just means that it is not excluded from being an equivalent.

    But why does it follow that if someone else discovers that PVC works, where the applicant did not foresee that it could work, that it should not be an equivalent?

    I think the reason is pretty close to what I said in my earlier post. That discovery is the invention of another is not the invention of the patentee.

  20. The majority, IIRC, held that just because an alternative was patented, did not remove it from being an equivalent.

    The majority was correct to reject that particular per se rule. As any defendant in a patent suit will tell you, there is a big difference between a patent claim and a valid patent claim.

  21. It appears that the use of PV alcohol for use as a coating with the claimed drug was not expected by the applicant to work — or at least that was the argument for lack of forseeability. If that in fact was true, the alternative coating was in fact not foreseeable. But, nevertheless, it was still not an equivalent for an entirely different reason: the DOE cannot or at least should not extend to the independent inventions of others.

    This point was argued recently in a dissent. The majority, IIRC, held that just because an alternative was patented, did not remove it from being an equivalent. However, I think the Supreme Court case law on this point is abundantly clear that independent inventions of other is grounds for removing that alternative from the scope of equivalents.

    The result here appears to be right, but for the wrong reasons. The error may have been forced by the prior case, which indicates that that case was itself wrongly decided.

  22. I asked myself the same thing. Assuming a pharmaceutical composition with compound X was known or obvious and you have to limit your claim; Let’s say your specification discloses a coating Y in combination with X but not a coating in general. The EPO would never allow “composition with compound X and a coating”, would the USPTO? If no, would it help to state in the response to the US Examiner that you limit the claim to “composition with compound X and coating Y” but only because you have no basis in the specification for anything broader? After all, the answer to the Examiner is part of the file history, isn’t it … Would this still be a “voluntary surrender” of patent scope? Well, it’s kind of a joke because people would undoubtedly start presenting two sets of claims to the Examiner: The claim they would like to have and the one they can have without adding subject matter that was not disclosed in the specification…

  23. The question I have is whether or not a different result would arise if the patentee had amended the dependent claim to include therein all of the limitations of its parent claim(s)to recast it as an independent claim? Since by definition the dependent already includes all of the limitations of its parent claim(s), then making it independent means that the scope of the claim was not narrowed to avoid prior art, and therefore, DoE should not apply?

  24. I haven’t read the case, but does it suggest how the applicant could have amended the claim to include a known alternative in the claim’s scope if that alternative was not described in the specification?

    Perhaps weasel-words? What words would have worked in this case?

  25. “Duramed argued that it was not foreseeable at the time of the amendment to use the polyvinal alcohol alternative in combination with the estrogen drug”

    Lulz. Considering “the use of the of the PVA alternative in combination with the estrogen drug” is not the equivalent in question then what possible relevance could the foreseeability of that have to do with anything?

    The equivalent in question is the use of PVA as a coating as opposed to the explicitly claimed ethylcellulose coating. And clearly it was a foreseeable equivalent which they neglected to claim (and indeed, it appears they neglected to invent at all as of the time of filing).

    I swear, how do lawlyers get up in the morning and make such frivolous claims? Rule 11 should smacked down upon them more often as you mentioned MM.

  26. “Long ago, courts added some leeway to the infringement analysis through the Doctrine of Equivalents.”

    Long ago, courts added a windfall to the infringement analysis through the Doctrine of Equivalents for application writing impared patentees.

    “Clearly it’s to put an end to the plague of doctrine of equivalents arguments in patent cases.”

    I sure hope so. The ones I’m seeing happening in “licensing negotiations” irl are totally absurd.

  27. The court held that the polyvinal alcohol alternative should have been foreseeable because it was already known to be used in other pharmaceutical compositions

    Duramed argued that it was not foreseeable at the time of the amendment to use the polyvinal alcohol alternative [a known alternative coating] in combination with the estrogen drug.

    In other words, even though everyone know about polyvinyl alcohol drug coatings when the patentee’s claim was amended to exclude polyvinyl alcohol coating, what’s most important is that the patentee didn’t think that polyvinyl alcohol would work as a coating with an estrogen drug. Therefore, equity requires that patentee owns the rights to make and use polyvinyl coatings on an estrogen drug.

    Everyone got that?

    Clearly it’s to put an end to the plague of doctrine of equivalents arguments in patent cases. And whatever happened to Rule 11?

    What happened to Paddock Labs here is disgusting.

Comments are closed.