ModernaTX v Pfizer (EWCA, [2025] EWCA Civ 1032, 2025)

Dhenne Avocats · 11 October 2026

Our analysis

Summary

The Court of Appeal dismissed Pfizer and BioNTech’s appeal and upheld the validity of claim 3 of Moderna’s European Patent (UK) No. 3 590 949 (EP949), which claims an mRNA in which 100% of uracil containing nucleotides are replaced with N1-methyl-pseudouridine (m1Ψ) nucleotides (paras 15 and 121). The University of Pennsylvania’s international application WO 2007/024708 (UPenn) contained no individualised disclosure of m1Ψ combined with 100% replacement: the combination required selection from two lists with no pointer towards it (paras 65 and 72). The judge’s assessment of obviousness, resting in particular on the absence of any expectation of success, disclosed no error of law or principle (para 120). The judgment is relevant to any novelty attack built on lists in the prior art.

Facts and procedure

After a 19 day trial, Meade J held EP949 valid and infringed by the Comirnaty vaccine, and Moderna’s EP565 invalid ([2024] EWHC 1695 (Pat)) (paras 1 and 3). Pfizer and BioNTech appealed only against the finding that EP949 was valid, contending that it lacked novelty over, or was obvious in light of, UPenn; the 1 October 2010 priority date was not challenged and Moderna relied only on claim 3 (para 2). UPenn, whose inventors are Karikó and Weissman, defines « pseudouridine » in [0056] as including m1Ψ and four other modifications, lists in [0074] 33 replacement percentages from 0.1% to 100%, and proposes in prophetic Example 31 to test a long list of modifications including m1Ψ (paras 17, 24, 28 and 36).

The applicable law

Prior art anticipates only if it discloses subject matter which, if performed, would necessarily infringe (Paroxetine); this requires clear and unambiguous disclosure, and it is not enough that it would be obvious to modify the prior art so as to fall within the claim (paras 39 and 40). For an item selected from a list or class, the test is whether there is an individualised description, with no distinction of principle between lists and classes (paras 43, 44 and 47); selection from multiple lists is generally novel (T 12/81) (para 53). Novelty is a binary question, whereas obviousness, a multifactorial evaluation, can only be disturbed for an error of law or principle (paras 58 and 94).

Question

Did claim 3 of EP949 lack novelty over UPenn, or was it obvious in light of it?

Decision

On novelty, [0056] of UPenn is a definition containing no technical teaching about the listed items; the list is open ended and m1Ψ is not said to be preferred (para 63). [0074] in substance covers everything from 0.1% to 100% with no preference for 100% (para 64). Crucially, the appellants’ case was one of selection from two lists with no pointer whatsoever to the combination; the need to choose a percentage, and the obviousness of choosing 100%, do not amount to clear and unambiguous disclosure (para 65). Example 31, a prophetic and woolly proposal with a long and open list, contains no individualised disclosure of m1Ψ or of 100% replacement (paras 69 to 72). On obviousness, the court endorsed the judge’s identification of the skilled person by reference to a practical interest in mRNA (paras 110 and 111) and held that he was entitled to give less weight to the appellants’ expert, a basic scientist, since an expert must be able to speak to the skilled person’s common general knowledge (paras 105 and 112). It upheld the reading of [0056], the finding of hindsight in the expert evidence and the absence of any expectation of success, the experiment not being routine for the skilled person (paras 114, 116 and 119). The appeal was dismissed (para 121).

Key points for practice

• Disclosure of one item in one list and another item in a second list does not anticipate their combination absent a pointer to it, even if the choice was obvious (para 65).
• An extended definition of a term in a prior application is not, in itself, a technical teaching about the items it lists (para 63).
• The court may assess how closely an expert’s experience matches that of the skilled person and weigh the evidence accordingly (paras 105 and 112).
• Practical point: a novelty attack combining items from several lists must show a specific pointer to the combination; otherwise the debate moves to obviousness and the expectation of success (paras 57, 65 and 119).

Provisions applied

European Patent Convention

Art. 54; Art. 56

National law

Patents Act 1977, section 3

Case law cited

SmithKline Beecham plc’s (Paroxetine Methanesulfonate) Patent [2005] UKHL 59; Dr Reddy’s Laboratories v Eli Lilly [2009] EWCA Civ 1362; EPO, T 296/87 Hoechst/Enantiomers; EPO, T 12/81 Bayer/Diastereomers; EPO, T 783/09; Actavis Group v ICOS Corp [2019] UKSC 15; Pozzoli v BDMO [2007] EWCA Civ 588; Technip France SA’s Patent [2004] EWCA Civ 381; Mölnlycke v Procter & Gamble (No 5) [1994] RPC 49; Illumina Cambridge v Latvia MGI Tech [2021] EWHC 57 (Pat)

Prepared by Dhenne Avocats from the text of the decision (National Archives, Find Case Law), consulted on 11 October 2026. Only the official text is authoritative.