Category Archives: Chemistry

Enanta v. Pfizer: Alcohol analogy best for explaining 2≠1?

Enanta Pharma Inc. v. Pfizer Inc., No. 2025-1427 (June 23, 2026)

“[Enanta] did not invent what they did not disclose.”

Boom!  Enanta’s provisional patent application did not disclose -NCH(O)-C1-alkyl. It only recited -NCH(O)-C2-C12-alky. Therefore, it did not support Enanta’s non-provisional patent disclosure of -NCH(O)-C1-C12-alkyl. In between filing dates, Pfizer disclosed Nirmatrelvir, a.k.a. PF-073213321 in Paxlovid®, comprising a -NCH(O)-CF3 substituent, which is a -NCH(O)-C1-alkyl. One carbon made a big difference.

The court followed this conclusion with the following analogy:

“The issue in this case is akin to asking whether a disclosure of ethanol, a two-carbon alcohol regularly consumed by people, would provide adequate written description support for methanol, a one-carbon alcohol that is highly toxic to people.”

Fair point. Chemistry is often understood as an unpredictable art. Simply adding one atom to a molecule can significantly change its activity.  Magic methyls are indeed real for various properties of various chemical compounds. Indeed, here the methanol/ethanol analogy was decidedly persuasive. Logically in this case, if ethanol (CH3CH2OH), which is a -C2-alkyl-OH, cannot support methanol (CH3OH), which is a -C1-alkyl-OH, then -NCH(O)-C2-C12-alkyl cannot support -NCH(O)-C1-C12-alkyl.

However, even though this methyl/ethyl alcohol analogy was persuasive, was it the best analogy? Perhaps not. It was definitely good enough. Such an analogy may be persuasive only in respect to which property you are observing. In this analogy, that property is toxicity. What about other properties? Is toxicity itself even the best property to focus on?  

I believe there are at least two reasons why the alcohol analogy may not be the best analogy.

  • Ethanol is toxic, albeit to a lesser degree than methanol.
  • Both ethanol and methanol are highly flammable.

Firstly, Ethanol is indeed toxic. Just ask someone who is drunk and praying to the porcelain god. Moreover, heavy long-term alcohol use is a common cause of cirrhosis.2 Granted, compared to ethanol, methanol is drastically more toxic. DO NOT DRINK METHANOL. You will go blind.3 Literally. My discomfort with the alcohol analogy is that it appears to imply that ethanol is as safe to drink as distilled water and has no toxicity at all, which is not accurate. Most skilled artisan parents would likely not give their toddlers shots of 100 proof vodka. Rum perhaps, if they are pirates.

Secondly, both ethanol and methanol are highly flammable. Let’s extend the analogy to even simpler molecules such as ethane (two carbons) and methane (one carbon). Both are also highly flammable. Just ask Blue Origin about their New Glenn Rocket.4 Granted, methanol and ethanol indeed have different burn rates depending on various factors such as concentration. However, though the lens of flammability, I think that a provisional application claiming a flammable composition and describing ethanol would likely provide adequate written description support for methanol.

A good “magic methyl” analogy might have been better than the alcohol analogy. However, while a magic methyl might be technically better, it might not be as persuasive the alcohol analogy. I fret that it may be dismissed as pedantic. Kudos to the alcohol analogy. It is simple to understand and impactful. Dare I say elegant? Most people, skilled in the art or not, understand the axiom of whisky good, wood alcohol bad.

  1. “Pfizer unveils its oral SARS-CoV-2 inhibitor” B. Halford, C&EN Global Enterprise 2021 99 (13), 7-7. DOI: 10.1021/cen-09913-scicon3. C&EN, 2021, 99 (13), p 7 April 12, 2021 ↩︎
  2. https://www.mayoclinic.org/diseases-conditions/cirrhosis/symptoms-causes/syc-20351487 ↩︎
  3. P.F. Suit, at al., “Methanol intoxication: clinical features and differential diagnosis” Cleve Clin J Med 1990; 57:464-471 https://www.ccjm.org/content/ccjom/57/5/464.full.pdf ↩︎
  4. https://apnews.com/article/blue-origin-rocket-explosion ↩︎

2 ≠ 1: Enanta v. Pfizer

The U.S. Court of Appeals for the Federal Circuit (CAFC) issued another decision where they held an applicant to a narrow interpretation of one their application’s definitions.  

In Enanta Pharmaceuticals. Inc. v. Pfizer Inc., No. 2025-1427 (June 23, 2026), the court held that Enanta’s U.S. Patent No. 11,358,953 (“the ’953 Patent”) could not claim the benefit of priority to U.S. Provisional Patent No. 63/054,048 because the patent’s recitation of “NCH(O)-C1-C12-alkyl” in its definition of “substituent” could not be supported by the provisional applications disclosure of only “-NCH(O)-C2-C12-alkyl” in its own definition. Without that priority, Pfizer’s disclosure of a species that falls within the Enanta’s generic claim is invalidating, novelty-defeating prior art.   

Timeline:

  • July 20, 2020: Enanta filed a provisional patent describing a substituent comprising a “C2-C12 alkyl.”
  • April 6, 2021: Pfizer disclosed PF-07321332 (nirmatrelvir), which has a “C1” for the substituent.
  • November 9, 2021: Enanta filed the non-provisional correcting C2 to C1.

Enanta argued that “C2” was merely a typographical error that a skilled artisan would recognize the error as such and clearly conclude that the inventor was in possession of the full range of “C2-C12.” The court disagreed saying that the “disclosure of one chemical compound, or integer in this case, cannot necessarily be a disclosure of another, even one close by structurally.” They backed it up with the analogy of “asking whether a disclosure of ethanol, a two-carbon alcohol regularly consumed by people, would provide adequate written description support for methanol, a one-carbon alcohol that is highly toxic to people.”

Fair point. Magic-methyls are indeed real for various properties of certain chemical compounds. However, I think there is more nuance to discuss. I also think there are other lenses that the court could have used to view the facts in this case.

I intend to follow up this post with further analysis of this case.

Check back soon!

Patent Lexicography: When “Branched Alkyl” ≠ “Branched Alkyl” – A Cautionary Tale

Before fancying yourself as your own lexicographer, read the cautionary tale told by the U.S. Court of Appeals for the Federal Circuit’s decision in Alnylam Pharmaceuticals, Inc. v. Moderna, Inc., No. 2023-2357. In this case, the court held Alnylam to their narrow definition of “branched alkyl” instead of its ordinary meaning. The narrow definition allowed Moderna’s COVID-19 vaccine SPIKEVAX® to avoid infringement of Moderna’s patents (U.S. Patent Nos. 11,246,933 and 11,382,979).

Estimated reading time: 2 minutes

Introduction: Conflict between a term’s ordinary meaning and an applicant’s own definition

Generally, words or phrases are given their ordinary and customary meaning in a patent claim. However, an inventor may act as their own lexicographer and define terms inconsistently from the term’s ordinary meaning. While drafting definitions of terms more narrowly than their ordinary meaning may be advantageous at times, such definitions may be disadvantageous if the broader ordinary and customary meaning is needed in litigation.

High Level Summary

In Alnylam’s case, Moderna did not infringe Alnylam’s patents because Alnylam defined its term “Branched Alkyl” too narrowly. The court held Alnylam to its patent definition of “Branched Alkyl” to mean an “alkyl in which one carbon atom in the group (1) is bound to at least three other carbon atoms, and (2) is not a ring atom of a cyclic group,” instead of Alnylam’s later suggested ordinary meaning of “a saturated hydrocarbon moiety that is not a straight chain.” The patent’s definition limited Alnylam’s claims to covering only lipids that comprised tertiary or quaternary carbons. Therefore, Alnylam’s claims could not cover Moderna’s cationic lipid SM-102, which only comprised primary or secondary carbons, as used in SPIKEVAX®.

The court treated the term “Branched Alkyl” as having a binding lexicographic definition because:
(1)  the term was located in a “Definitions” section,
(2)  the term was in quotation marks,
(3)  the definition included the phrase “refers to,” which confers definition intent,
(4)  other definitions used non-liming phrases such as “for example” and “including, but not limited to,” and
(5)  the definition included the phrase “unless otherwise specified,” which suggests that the rest of the sentence establishes a generally applicable rule or definition.

For further reading, see these other links to blogs discussing Alnylam v. Moderna:

Home » Chemistry

Updated Reference for Protecting Groups

Greene’s Protective Groups in Organic Synthesis, P. G. M. Wuts, 6th edition, John Wiley & Sons, 2025.
ISBN: 978-1-394-23318-2

Sample description for adding to a patent application specification:
[0001] Nitrogen protecting groups, oxygen protecting groups and sulfur protecting groups are well known in the art and include, but are not limited to, those described in Greene’s Protective Groups in Organic Synthesis, P. G. M. Wuts, 6th edition, John Wiley & Sons, 2025, incorporated herein by reference.

Caution! Before fancying yourself as your own lexicographer, read the cautionary tale told by the U.S. Court of Appeals for the Federal Circuit’s decision in Alnylam Pharmaceuticals, Inc. v. Moderna, Inc., No. 2023-2357. In this case, the court held Alnylam to their narrow definition of “branched alkyl” instead of its ordinary meaning.
Look out for an more detailed post on the case shortly.

SCHRÖDINGER’S POLYMORPH: WHEN A CRYSTALLINE FORM CLAIM IS OBVIOUS AND INVALID

Salix Pharmaceuticals, Ltd. v. Norwich Pharmaceuticals Inc., No. 22-2153 (Fed. Cir. 2024)

“Schrödinger’s cat,” is a paradoxical thought experiment devised by physicist Erwin Schrödinger, while debating quantum superposition in 1935. Oversimplified, in the thought experiment, there is a cat in a box, and you have no way of knowing what physical state the cat is in, like a quantum wave function, until you open the box and observe it.

While studying polymorphic forms, like Schrödinger’s hypothetical [albeit a long stretch], you do not know what form a solid compound is until you measure it with x-ray crystallography. However, unlike Schrödinger’s hypothetical, you may have a reasonable expectation of succeeding to figure out what form your compound is depending on the prior art and knowledge of a person of ordinary skill.

Shifting from quantum to organic chemistry patent law, if there is a reasonable expectation of success in characterizing a known crystalline product for potential polymorphism using routine, conventional methods and skill, then a claim to that crystalline product would be invalid for obviousness. Such was the decision in Salix Pharmaceuticals, Ltd. v. Norwich Pharmaceuticals Inc., No. 22-2153 (Fed. Cir. 2024).

Salix claimed a specific crystalline form (form β) of rifaximin. The United States patent and Trademark Office (USPTO) granted Salix a patent on it.  However, when litigated, the United States Court of Appeals for the Federal Circuit (CAFC) affirmed the District Court of Delaware’s finding that Salix’s claims to rifaximin form β were invalid for obviousness.*

“The difference between the prior art and the claims is thus effectively nothing more than the performance of routine characterization to identify the polymorphic forms that result from the known Cannata processes.”

Salix Pharmaceuticals, Ltd. v. Norwich Pharmaceuticals Inc., No. 22-2153 (Fed. Cir. 2024), at 15.

In other words, you cannot claim ownership of a compound just by making the expected crystalline form using someone else’s process and taking an x-ray crystal structure of it and giving a Greek letter name.

Background

  • Salix markets Xifaxan® (rifaximin), and Norwich wanted to sell a generic version.
  • Xifaxan® is a formulation of rifaximin crystalline form β.
    • Polymorph β is a commonly produced polymorph.
    • Polymorph β the most stable form of rifaximin.
  • Disputed patent claims: Claim 4 of U.S. Patent No. 7,612,199 and Claim 36 of U.S. Patent No. 7,902,206.
    • Claim 4. Rifaximin in polymorphic form β, wherein the rifaximin has x-ray powder diffraction pattern peaks at about 5.4°; 9.0°; and 20.9°2θ and wherein the rifaximin has a water content of greater than 5%.
  • Prior art: Cannata et al. U.S. Patent No. 4,557,866.
    • Although Cannata does not discuss rifaximin’s crystal structure in detail, it does disclose several preparation protocols for rifaximin that include solvents used for crystallization.
    • Experts testified that Cannata’s preparation of crystalline Rifaximin would have yielded its β form. But they could not say it would every time.
  • Rifaximin was a known compound with a known, useful activity (antibiotic).
    • While rifaximin was a known antibiotic, Salix owns rights to patents covering new method of use (hepatic encephalopathy). Salix won on appeal for such method-of-use claims.
  • Experts testified that Cannata’s preparation of crystalline rifaximin would have easily made and taken an X-Ray diffraction pattern to determine that rifaximin was in its β form.
    • The experts could not convince the judge that the prior art process would produce rifaximin form β “every time.” If they could, then Salix’s claims would have been anticipated (not novel) as inherently present in the prior art.‡
    • Although the experts could not convince the judge of the certainty of form β (novelty), they at least convinced the judge that form β was reasonably expected (obviousness).

    CAFC Opinion Discussion

    The CAFC decided that “the district court found a reasonable expectation of success in characterizing the crystalline product of Cannata for potential polymorphism using routine, conventional methods and skill [Salix Pharm., Ltd. v. Norwich Pharm., Inc., Civil Action No. 20-cv-430-RGA, (D. Del. Apr. 29, 2021), at *6–7].†  We see no clear error in that conclusion. Indeed, Salix has done no more than combine known elements of the prior art to verify readily accessible information concerning a compound already in the hands of those of ordinary skill in the art, and such routine efforts do not justify removing this polymorph from the public domain. See KSR Int’l Co. v. Teleflex Inc., 550 U.S. 398, 427 (2007); see also Pfizer, 480 F.3d at 1367−68.”

    The District Court supported its factual determinations of obviousness with expert testimony that:

    (1)  a skilled artisan would have had good reason to characterize the crystalline rifaximin obtained by following the Cannata protocols,

    (2)  such characterization was routine and could have been performed “in one day,” and

    (3)  doing so would have led the skilled artisan to have “detected rifaximin β.” 

    Salix argued that a person of ordinary skill in the art would not have “expect[ed] to succeed” because, as of the critical date, the polymorphic nature of rifaximin had not yet been reported and the identity of the β form remained undisclosed. They further argue that a skilled artisan would not have been able to predict which polymorphic form Cannata made. This argument was flawed. The CAFC states that Salix’s is incorrect in framing of the issue, as it suggests that no unknown entity could ever be obvious, as one cannot reasonably expect what was hitherto unknown.

    * In the same appeal, The CAFC did affirm the District Court’s finding that Salix’s other claims to methods of treating hepatic encephalopathy were valid and nonobvious.

    † Whether or not there would have been a reasonable expectation of success is a question of fact. “Whether or not there would have been a reasonable expectation of success is a question of fact, (IXI IP, LLC v. Samsung Elecs. Co., 903 F.3d 1257, 1262 (Fed. Cir. 2018)), which we review for clear error, Hospira, 946 F.3d at 1328. We review the ultimate conclusion of obviousness de novo.” Did the Justices themselves believe that the claims were obvious? Unknown. The CAFC perhaps equivocated or sidestepped the question. The justices did not say that they themselves believe that the polymorph claims were obvious. Rather, they merely said that they found no clear error in the District Court’s finding of obviousness. 

    ‡ The CAFC did not opine on whether or not the polymorph claims were invalid for being inherently anticipated (§102). The court said that they did not need to go that far since they already determined that the claims were invalid for obviousness (§103). The district court said that the polymorph claims were not anticipated as being inherent because Norwich did not prove with clear and convincing evidence that the prior art method would produce rifaximin form β “every time.” The inherency issue may deserve its own post. Question for discussion: Does the “every time” requirement go too high for “clear and “convincing evidence?” Perhaps requiring a specific result Every time may go up to the realm of “beyond a reasonable doubt,” which is not the level of burden of proof for patent law. Recall last year’s discussion of the Amgen case where there were arguments about how many examples enable the “full scope” of a claim. Stay tuned.

    Additional Information

    OrganicPatents.com page on Obviousness of Stereoisomers &Purified Compounds

    Obviousness of Stereoisomers & Purified Compounds

    See also PatentDocs’s post: https://www.jdsupra.com/legalnews/salix-pharmaceuticals-ltd-v-norwich-3362384/.


    May 8, 2024