Improper claim structure invalidates patent

Pfizer v. Ranbaxy (Fed. Cir. 2006).

In an infringement suit arising from Ranbaxy’s ANDA filing, Pfizer’s two patents covering Lipitor (atorvastatin) were upheld as infringed and not-invalid by the district court Judge Farnan. On appeal, the CAFC affirmed on one patent (leaving Lipitor covered), but found the asserted claim of the second patent (the ‘995 patent) invalid under Section 112, paragraph 4 because of an improper dependent claim structure.

Section 112 of the Patent Act requires, inter alia, that a dependent claim incorporates by reference all the limitations of the parent claim and also specifies a further limitation. 

Pfizer’s asserted claim 6 depends from claim 2, which depends from claim 1.  The subject matter of the claims is as follows:

Claim 1: (1) atorvastatin acid; or (2) atorvastatin lactone; or (3) pharmaceutically acceptable salts thereof.

Claim 2: atorvastatin acid but not atorvastatin lactone or a pharmaceutically acceptable salts.

Claim 6: the hemicalcium salt of the compound of claim 2.

Thus, claim 6 fails to "specify a further limitation of the subject matter" of the parent claim (as requred by 112) because the subject matter of claim 6 is completely outside the scope of claim 2.

Consequently, based on violation of the statute, the court found that the claim was invalid.

Although the district court was reluctant to find the fourth paragraph of ยง 112 to be an invalidating provision, doing so does not exalt form over substance. Rather, it is consistent with the overall statutory scheme that requires applicants to satisfy certain requirements before obtaining a patent, some of which are more procedural or technical than others.

Reversed in part. Pfizer may be able to correct this defect through a reissue application, although there is some question as to whether the defect here falls within an identified category.

18 thoughts on “Improper claim structure invalidates patent

  1. 18

    The invalidity of claim 6 followed from the stipulation of the parties that is was dependent. My question is, can parties stipulate to claim construction, which is supposed to be a conclusion of law?

    I also have a problem with the court being so nonchalant about the ‘995 patent, which so far as I can tell has no terminal disclaimer, providing Pfizer with a total of 23 years of patent protection for a single composition of matter.
    Pfizer argued in prosecution of patent B that “no, our earlier patent A does not cover this compound”, thereby getting patent B (with its later expiration date) granted, and then went on to assert in this litigation, “please ignore what we said in prosecuting patent B — patent A really does cover this compound after all.” The court seems to be saying they can have it both ways, merely because the two applications were “unrelated”.
    Am I alone in thinking that something is very wrong with this picture?

  2. 17

    Thank you Elizabeth for that additional piece of information regarding the other, though less common, meaning of “depend.” It may indeed (along with Judge Markey’s oft-cited opinion in Arrowhead Indus. v. Ecolochem, where the patentee also performs a “dance macabre”)have contributed to the confusion. My view has always been that, because section 112 speaks in terms of “dependent” and “independent” claims, “dependent” is used in the context of its more common meaning: that it relies or is contingent upon the independent claim, rather than hanging down from the independent claim. I suppose Damocles could have worried that his sword might become “independent,” in which case he would have had more to worry about than an additional USPTO fee, even though he would have been reduced to a small entity!

  3. 16

    Thank you Elizabeth. I had the same point to make. In my first week as a trainee I learned that you can write that a claim is dEpendEnt ON another, or AppendAnt TO another. Choose which you prefer but don’t write dependAnt because that means something else (like Principal and principle). “Depend” is useful in the specification, because something like a skirt depends from a belt.

  4. 15

    “By the way, where did the patentese ‘depends from’ originate, anyway? Honestly, it has always bugged me. Things depend ‘on’ or ‘upon’ another, not ‘from’ another.”

    Actually, it’s perfectly proper for the correct definition of “depend”. In addition to “rely upon,” it also means “hang from” (e.g., The sword of Damocles depended from the ceiling).

  5. 13

    The analysis of the case is more complex than it needs to be.

    Chemical practitioners know that “but not” can be used to exclude certain members of a class or genus when it is desired to claim part or all of the balance of the class or genus. I’d note in passing that once a species is claimed (I take it for granted that “atorvastatin acid” is a species) there’s nothing specific that needs to be excluded — everything else in the world is already excluded from the scope of the claim (save for equivalents).

    However, although the original comment suggests otherwise, claim 2 doesn’t recite “but not,” but simply claims atorvastatin acid, period. The court pointed out that claim 6 would have been valid if claim 2 had be drafted in that manner or if claim 6 had been presented in independent form, but held that compliance with Sect. 112 para. 4 is a statutory requirement for validity just like any other.

    It appears undisputed that claim 6 complied with paras. 1 and 2 of Sect. 112, and that’s all most of us would worry about. The lesson for practitioners it that the trend towards the imposition of draconic rules upon claim drafting continues.

  6. 12

    Kevin & Jonnie:

    What is missing is that claim 6 was held to be invalid, not because it failed to distinguish the subject matter of claim 2 but, rather, because it failed to also include its limitations. As the court stated at pages 11 and 13 (see link above:

    “Claim 2 itself, however, only recites atorvastatin acid. Notably, it does not include the pharmaceutically acceptable salts of atorvastatin …. In other words, claim 6 does not narrow the scope of claim 2; instead, the two claims deal with non-overlapping subject matter. * * * Ranbaxy correctly argues that claim 6 fails to “specify a further limitation of the subject matter” of the claim to which it refers because it is completely outside the scope of claim 2.”

    Hope this helps.

  7. 11

    Kevin: I believe the logic of the claim construction looks something like this:

    1. A OR B OR C.

    (any of A, B, or C in scope)

    2. A but not B or C.

    (only A in scope left)

    6. A specific species of C as stated in claim 2 (which states C is not in its scope)

    C is not in the scope anymore, thus the claim is invalid. I’m just a simple country physicist though. In this example, A is the acid, B is the lactone, and C is a PA salt.

  8. 10

    Maybe I’m missing something, but doesn’t claim 6 recite a specific pharmaceutically-acceptable salt, so that the additional limitation is this particular species, rather than ALL PA-salts of the compound claimed in claim 2?

  9. 9

    What if Pfizer had not made the stipulation mentioned above and had argued that claim 6 is an independent claim? Wouldn’t Ranbaxy only be able to raise the fee deficiency problem only as part of an inequitable conduct defense? If so, Pfizer may have made a tactical mistake. I think it would be a tough job to convince a trier of fact that a company like Pfizer, whose very existence depends on its patents, misrepresented the dependent/independent status of claim 6 to the PTO with intent to deceive so that it could save the $100 or so fee for an extra independent claim.

  10. 8

    Even though Pfizer agreed that claim 6 was dependent, I’m not sure why the Federal Circuit didn’t address whether this characterization was, in fact, correct. As others have pointed out, this seems to me to be the key issue here: does mere reference to another claim (see MPEP 2173.05(f)) make that claim dependent (and thus subject to 112, 4)? By addressing this issue head on, the Court might have a) reached a different outcome (by finding claim 6 to be an independent claim), b) avoided casting uncertainty on many issued claims of this form. The Court certainly could have (and should have) answered this question.

  11. 7

    Here is a link to the decision, which Dennis often includes in his comment but which was omitted here:

    link to

    It indeed appears that claim 6 is not a dependent one. Why Pfizer would have stipulated to that and, even more puzzling, why the court agreed, is not apparent. The situation is not unlike assessing whether a claim is written in “means plus function” format pursuant to paragraph 6 of section 112: the court should not find that a claim is not written in that form and then invalidate it because it does not comply with that law that defines the form; it should, rather, and does, treat the statute as a means to determine in what form the claim is written, and then to construe it accordingly. The claim was not written in dependent form as defined in the statute, so it merely should have been construed as independent, regardless of what the patent owner thought.

    By the way, where did the patentese “depends from” originate, anyway? Honestly, it has always bugged me. Things depend “on” or “upon” another, not “from” another. What would we have thought of Al Jolson singing, “It all depends from you”? Maybe we understandably think of the dependent claim as conceptually derived from the independent. OK. It depends from how you look at it.

  12. 6

    The invalidation appears to be good to me. However, my experience has all been in the mechanical arts. If you write a dependent claim in the mechanical arts, it absolutely must limit the claims from which it depends, or you will get an antecedent rejection.

    Only on the basis of the claim as presented, it appears that Pfizer either needed to create a whole new independent claim to capture the matter of claim 6, or they needed to provide additional information in claim 2 to provide the proper antecedent for claim 6.

  13. 5

    A further description applicable to this decision:

    “..the kind of meticulous verbal analysis in
    which lawyers are too often tempted by their training to indulge.” (Lord Diplock, UK House of Lords, Catnic Components Ltd
    v Hill & Smith Ltd [1982] RPC 183, 243).

  14. 4


    (i) Is claim 6 an independent claim?

    (ii) Is non-payment of fee for this independent claim likely to affect validity / existence?

  15. 3

    With hindsight (so much easier) maybe Pfizer shouldn’t have agreed that the claim was dependent. If they’d been found not to have paid enough claim fees, would that have invalidated the patent?

    This decision appears to throw doubt on the validity of any claim which refers to an earlier claim but uses it as part of the definition of new subjectmatter rather than further restricting the earlier claim’s scope. And there are lots of claims like that (try a search for ‘product produced by the process of claim’, for example). So to avoid this, you have to recite the whole of the previous claim. What’s the point?

    There are words for decisions like these – I propose ‘pettifogging’.

  16. 2

    Pfizer stipulated at the District Court that claim 6 is a dependent claim. Had they argued that it’s an independent claim, they would be faced with the problem of a defeciency fees.

  17. 1

    Isn’t it possible to say that claim 6 is an independent claim? Does mere reference to another claim automatically make a claim dependent, such that it is subject to Sec. 112, para. 4? I know that it is permissible to recite “A method of using the apparatus of claim 1, comprising:”. (Disclaimer: I have not read this case completely; I have only read what is posted on Patently-O)

Comments are closed.