The U.S. Court of Appeals for the Federal Circuit on June 23, 2026 affirmed a summary judgment that every claim of Enanta Pharmaceuticals' U.S. Patent 11,358,953 is invalid as anticipated, in Enanta Pharmaceuticals, Inc. v. Pfizer Inc., No. 2025-1427. The panel of Circuit Judges Lourie, Bryson, and Chen, in an opinion by Judge Lourie, traced the outcome to a one-character difference between Enanta's provisional application and its issued patent. The court framed the question narrowly: whether the provisional's disclosure of a two-to-twelve-carbon alkyl group provided written description support, under 35 U.S.C. Section 112, for a one-carbon group that the issued patent later claimed. The appeal arose from the U.S. District Court for the District of Massachusetts, where Judge Denise J. Casper granted Pfizer's motion for summary judgment on December 23, 2024.
According to the opinion, the '953 patent issued from a non-provisional application filed November 9, 2021, and states that it claims priority from Enanta's U.S. Provisional Patent Application 63/054,048 (the '048 provisional), which has a priority date of July 20, 2020. Both documents define the term \u201csubstituted\u201d and list chemical moieties that qualify as substituents. The court states that the definitions are substantially the same \u201cbut with one critical difference\u201d: the '048 provisional recites \u201c\u2014NHC(O)\u2014C2-C12-alkyl,\u201d whereas the '953 patent recites \u201c\u2014NHC(O)\u2014C1-C12-alkyl.\u201d The subscripts identify the number of carbon atoms in the alkyl group. As the court explains, \u201cC2-C12\u201d denotes alkyl groups containing a range of two to twelve carbon atoms, while \u201cC1-C12\u201d additionally includes a one-carbon alkyl group.
The timing of an intervening disclosure made that single carbon dispositive. The opinion recounts that on April 6, 2021, Pfizer made publicly available a presentation disclosing a protease inhibitor, nirmatrelvir, which was eventually incorporated into Pfizer's Paxlovid product. The court notes that nirmatrelvir has an \u201cA\u201d group substituted with a \u2014NHC(O)\u2014C1-alkyl group \u2014 the one-carbon species. Enanta did not dispute that this disclosure anticipates the '953 patent's claims if the patent is not afforded a priority date before April 2021. The litigation therefore turned entirely on priority: if the '953 patent could reach back to the July 20, 2020 provisional, Pfizer's April 2021 disclosure would not count as prior art; if it could not, the disclosure anticipated the claims.
\u201cWe respect applicants\u2019 statements in their specification that they invented what was specifically disclosed in the \u2019048 provisional, but similarly we conclude that they did not invent what they did not disclose.\u201d\u2014 U.S. Court of Appeals, Federal Circuit, opinion (No. 2025-1427)
Why the provisional did not describe the claimed species
Enanta argued that the \u201cC2\u201d in the provisional was a typographical error that should have read \u201cC1,\u201d and that the non-provisional application corrected it without adding new matter. The opinion records that, per Enanta, the company realized the asserted error on July 9, 2021, and filed the non-provisional application on July 19, 2021. The Federal Circuit declined to treat the dispute as one about correcting errors in an issued patent. It distinguished its line of cases under 35 U.S.C. Section 251 (reissue) and Novo Industries, L.P. v. Micro Molds Corp., observing that this case is not before the Patent Office and that, in its words, a change from C2 to C1 \u201chas not been shown here to be a correction of an obvious error.\u201d The court added that even if the district court's error-correction framework applied, the existence of an error in the provisional is \u201csubject to reasonable debate.\u201d
Instead, the court applied the written description standard for priority drawn from Lockwood v. American Airlines and New Railhead Manufacturing v. Vermeer: to gain the benefit of an earlier filing date, each application in the priority chain must comply with Section 112's written description requirement. The standard, the court reiterated, requires that the earlier application describe the later-claimed invention in sufficient detail that a skilled artisan can conclude the inventor was \u201cin possession\u201d of the claimed invention as of the filing date sought. Applying it, the panel concluded that the '048 provisional does not convey to a skilled artisan that the inventors possessed \u2014NHC(O)\u2014C1-alkyl at the time of filing. \u201cC2 is simply different from C1,\u201d the opinion states, and the provisional's explicit disclosure of a two-to-twelve-carbon range \u201cnotably does not include an alkyl group with one carbon atom.\u201d
The court also addressed Enanta's expert declaration, which it characterized as pointing to a purported inconsistency in the provisional's general definition of \u201calkyl\u201d \u2014 where a numeral \u201cC2-C12\u201d appeared alongside written words describing \u201cone to twelve\u201d carbon atoms. The panel found that argument unpersuasive because the expert identified a possible error in the general definition of \u201calkyl,\u201d not in the specific disclosure of \u2014NHC(O)\u2014C2-C12-alkyl within the definition of \u201csubstituted.\u201d The opinion notes that the provisional specification is \u201chighly detailed, providing dozens of specific chemical moieties presumably described with great care,\u201d and that an expert opinion about a typographical error located elsewhere in the provisional does not create a genuine dispute about whether the specific substituent at issue was disclosed. The court underscored the governing principle from Lockwood that \u201centitlement to a filing date extends only to that which is disclosed.\u201d
What the holding turns on, and what it does not
The opinion is explicit about the boundaries of its analysis. The court opened its discussion by stating \u201cwhat this case is not\u201d: it distinguished the genus-and-species written description disputes of Ariad Pharmaceuticals v. Eli Lilly and In re Ruschig, framing the present question instead as whether the provisional's disclosure of \u201cC2-C12\u201d alkyl provides adequate written description support for the patent's \u201cC1\u201d \u2014 that is, in the court's phrasing, \u201cwhether \u20182\u2019 provides adequate written description support for \u20181.\u2019\u201d To illustrate the point, the opinion offers a chemistry analogy: it likens the question to asking whether a disclosure of ethanol, a two-carbon alcohol, would provide written description support for methanol, a one-carbon alcohol, observing that a disclosure of one compound or integer \u201ccannot necessarily be a disclosure of another, even one close by structurally.\u201d
Procedurally, the panel reviewed the grant of summary judgment de novo under First Circuit law, applying Federal Circuit law to the substantive patent questions, and treated written description compliance as a question of fact. It noted that although the district court had focused on its power to correct a purported typographical error rather than on the written description standard, the appellate court could apply the correct standard itself on de novo review. The result, the court held, is that the '048 provisional \u201cprovided no written description support for the '953 patent, so the '953 patent cannot be afforded the '048 provisional's priority date,\u201d and the district court therefore properly granted summary judgment that the claims were anticipated by Pfizer's disclosure of nirmatrelvir.
For the record, the case caption identifies Barbara A. Fiacco of Foley Hoag LLP as arguing counsel for plaintiff-appellant Enanta, and David M. Krinsky of Williams & Connolly LLP as arguing counsel for defendant-appellee Pfizer. The district court matter is captioned Enanta Pharms., Inc. v. Pfizer, Inc., No. 1:22-cv-10967-DJC, with the underlying summary judgment decision reported at 2024 WL 5203036 (D. Mass. Dec. 23, 2024). The Federal Circuit affirmed that decision in full.
Comments
Loading comments…