UK Patents Court reinforces the UK threshold for plausibility

Accord Healthcare Ltd v Novartis AG & Novartis Pharmaceuticals UK Ltd [2026] EWHC 2127 (Pat) (Meade J) – 13 August 2026.

Meade J rejected all of Accord’s attacks on validity with the result that Novartis’ patent and SPC protecting its blockbuster combination product Entresto, were valid.  The Judge held that the claimed valsartan/sacubitril combination was not obvious, that the patent made the claimed therapeutic benefit plausible, and that the invention disclosed a genuine technical contribution and was not an arbitrary selection or collocation.  This Judgment provides further guidance on the UK’s threshold for plausibility where experimental data have been generated and relied upon, but the data are not disclosed in the specification.

Background

The dispute concerned Novartis’ European Patent no. EP (UK) 1 422 217 and corresponding SPC protecting Entresto, which used a combination of two compounds (valsartan an angiotensin receptor blocker (“ARB”), and sacubitril a neutral endopeptidase (“NEP”) inhibitor) for the treatment of heart failure and hypertension.   The two active ingredients in Entresto were in the form of a co-crystal complex.   Accord sought to clear the way for the launch of its own generic product (a simple admixture where the two actives were separate salts, mixed together) and brought revocation proceedings before expiry of the SPC in 2028.

Although Accord denied infringement, the Judge noted that by the time the matter reached trial, that did not have any significant impact on the action, which was essentially all about validity.  The Judge noted that the Dutch Patents Court had held the patent to be valid but that the way the arguments were made meant that neither side relied on the decision in this case.

Findings

A notable preliminary issue concerned the identity of the skilled team.  Meade J rejected the notion that patent’s references to heart failure could be set aside for the purposes of assessing obviousness.  Instead, he held that the patent had to be assessed as the skilled reader would understand it, meaning that obviousness fell to be considered from the perspective of the skilled team interested in both hypertension and heart failure.

Meade J also reiterated the distinction between the role of expert evidence and the role of the UK Patents Court in construing a patent specification.  While expert evidence was admissible to assist the court in understanding the underlying science and the significance of experimental results, the meaning of the language used in the patent was ultimately a matter for the judge.  Although a substantial amount of the cross-examination had been directed to the meaning of the application itself, the interpretation of the patent’s language was a judicial task not the experts’.

On plausibility (in the context of sufficiency), Meade J followed the Supreme Court in Warner Lambert [1]and the Court of Appeal in Dapagliflozin[2] and held that the patent met the relatively modest plausibility threshold.  A key issue was whether the patent merely asserted that the valsartan/sacubitril combination would provide a therapeutic benefit, or whether it conveyed that such a benefit had been observed.  Meade J held that, read as a whole, the patent qualitatively disclosed that the inventors had carried out animal studies and obtained positive results for the combination before filing.  Although the patent did not disclose the underlying numerical data, it communicated that the claimed therapeutic effect rested on completed experimental work rather than a research hypothesis.  Meade J also rejected the suggestion that any disclosed effect could be disregarded as “uselessly low”, noting (with reference to Apixaban[3]) that the relevant question was whether the patent disclosed a real technical effect supported by experimental work, not whether the reader was provided with the underlying measurements.  This distinguished the case from authorities such as Apixaban and Dapagliflozin, where the patent disclosure was held not to provide an adequate foundation for the asserted technical contribution.

Meade J rejected the contention that the patent was a mere collocation of known features. A collocation arises where known components are simply placed side-by-side, with each performing its conventional function and without any interaction giving rise to a technical contribution.  In contrast, the Judge found that the patent empirically demonstrated that valsartan and sacubitril interacted positively rather than adversely.  Whilst the patent did not quantitatively demonstrate synergy in the sense that the combined effect exceeded the sum of the individual effects, Meade J held that this was unnecessary.  It was sufficient that the patent dispelled the uncertainty surrounding the interaction between the two compounds and demonstrated that the combination “worked”.  This disclosed interaction constituted a genuine technical contribution and meant that the invention could not properly be characterised as a mere collocation.

Meade J likewise rejected the suggestion that the invention represented an arbitrary selection from known classes of compounds.  In his view, the patent contributed more than the identification of particular members of the ARB and NEP inhibitor classes; it disclosed that the specific valsartan/sacubitril combination achieved a beneficial therapeutic effect.  That experimentally supported technical contribution was sufficient to justify the selection and prevent it from being characterised as arbitrary.

Meade J’s findings on plausibility, collocation, and arbitrary selection all turned on what the patent disclosed about the inventors’ experimental work in the application as filed as a whole.  Although he described the issue as a close call, he concluded that the skilled person would understand the application as communicating that the inventors had carried out experimental work  and had the results of that work in hand when drafting the application.  Meade J held that the patent qualitatively disclosed that the inventors had tested the valsartan/sacubitril combination in animal models and obtained positive results.  Such disclosure was sufficient because it conveyed to the skilled person that the inventors had tested the claimed combination and observed positive results.  Warner Lambert did not require the disclosure of numerical results.

Meade J stated that his view shifted both during trial and when preparing the judgment.   Ultimately, he found in favour of Novartis but did observe that this risk could have been avoided by Novartis through explicit disclosure of experimental details and data in the application.

Meade J also rejected the SPC challenge. A central issue was the identification of the relevant “product” for the purposes of Article 3(a) of the SPC Regulation. He held that the product was the combination of valsartan and sacubitril, rather than the specific sacubitril/valsartan co-complex crystalline form used in Entresto.  Meade J found that the evidence overwhelmingly showed that valsartan and sacubitril were the pharmacologically active ingredients and that the co-complex itself had no independent pharmacological significance.  Having identified the product in this way, he concluded that the product was specifically disclosed and identifiable from the patent, with the result that the SPC was valid.

Conclusion

This judgment provides reassurance that patents based on actual experimental findings may satisfy UK plausibility and technical-contribution requirements even where the patent does not include the full underlying dataset, provided that it credibly communicates that the claimed therapeutic effect was supported by such data.  While it is helpful confirmation of the Supreme Court’s statement that plausibility does not necessarily require the disclosure of detailed quantitative data, Meade J’s reasoning suggests that this was a close case on its facts. Patentees should consider carefully which available experimental data to disclose to strengthen their specifications from such validity attacks.

[1] Warner-Lambert Company LLC v Generics (UK) Ltd [2018] UKSC 56

[2] Generics (UK) Ltd v AstraZeneca AB [2025] EWCA Civ 903

[3] Sandoz Ltd v Bristol-Myers Squibb Holdings Ireland Unlimited Co [2023] EWCA Civ 472