On 16 July 2025, the Court of Appeal dismissed AstraZeneca’s appeal and upheld the first instance decision, finding that AstraZeneca’s compound patent for dapagliflozin, an SGLT2 inhibitor used to treat diabetes, was invalid for lack of plausibility and lack of technical contribution. 

Background

The Defendant, AstraZeneca (“AZ”) was the proprietor of European Patent (UK) No. 1 506 211 entitled “C-aryl glucoside SGLT2 inhibitors and method” (“The Patent”) which expired on 14 May 2023, and UK supplementary Protection Certificates Nos. SPC/GB13/021 and SPC/GB14/050 (expiry 13 and 14 May 2028). The Patent claims relate to a compound called dapaglifloxin, marketed by the Defendant under the trade mark Forxiga pursuant to marketing authorisation granted 14 November 2012, which is used to treat diabetes.

The Claimants, generics manufacturers Viatris, Teva and Glenmark, contended that the Patent was invalid, and therefore that the SPC’s were invalid. No challenge was claimed to the priority date of 20 May 2002. The Defendant acquired the Patent from Bristol-Myers Squibb (“BMS”) in 2014.

The Appeal

AstraZeneca appealed on eight grounds. The Court of Appeal categorised these into: (1) grounds 1-3 concerning the Judge’s interpretation of the patent; (2) grounds 4-6 concerning the law on plausibility; and (3) grounds 7-8 concerning arbitrary selection.

Judge’s interpretation of the patent

AstraZeneca argued that the judge had wrongly assessed the disclosure of the patent when he found that it did not disclose a verbal statement of an experimental result confirming the efficacy of dapagliflozin.

The Court of Appeal upheld the judge’s interpretation, noting that the patent merely referenced an “inhibitor” being tested in an assay without any experimental results or reasoning to support the assertion that dapagliflozin was the inhibitor being tested. Lord Justice Arnold concluded that if the patent was intended to record a test that had been performed on dapagliflozin, then it should have said so.

AstraZeneca further contended that the judge wrongly assessed the disclosure of the patent when he found that it did not disclose enough to make it plausible that dapagliflozin would effectively treat diabetes.

The Court of Appeal again agreed with the judge, finding that even if the skilled team had understood the disclosure of the patent to mean that dapagliflozin had been tested, the patent contained no information about: (i) the level of SGLT2 inhibitor potency obtained from dapagliflozin; (ii) its EC50 value; or (iii) whether the (undisclosed) EC50 value of dapagliflozin was sufficient to confer utility.

The Court held that the Judge had correctly followed Sandoz Ltd v Bristol-Myers Squibb Holdings Ireland UnLimited Company [2023] when concluding that the patent did not make it plausible that dapagliflozin was an SGLT2 inhibitor with sufficient efficacy to treat diabetes, as the specification contained only bare and unsupported assertions, lacking experimental data or theoretical reasoning to support these assertions. This was therefore insufficient for the skilled person to conclude that dapagliflozin had been tested in the assay or that it had the necessary inhibitory activity to be effective.

Plausibility

AstraZeneca advanced several grounds in respect of the law of plausibility. First, it asserted that the test set by the Supreme Court in Warner-Lambert set too high a standard for plausibility and must be revisited in light of G2/21. This was rejected outright by the Court of Appeal as the decision in Warner-Lambert is binding on the court with respect to the sufficiency of medical use claims.

Second, AstraZeneca argued that claim 2 was to a product per se and, as such, the first instance judge should have assessed plausibility in the context of inventive step rather than insufficiency. AstraZeneca maintained that product claims should be assessed using the ab initio implausibility standard in accordance with the Enlarged Board of Appeal’s decision in G2/21.

Ab initio implausibility permits a technical effect to be claimed when it is not implausible in light of the technical content of the application read with the skilled person or team’s common general knowledge (that is, when there is no legitimate reason to doubt that the purported technical effect can be achieved).

AstraZeneca accepted that the Court of Appeal’s decision in Sandoz v BMS was a prima facie binding authority against this argument. However, it invited the court to depart from Sandoz v BMS using the limited exception recognised in Actavis v Merck.

Lord Justice Arnold acknowledged the importance of adopting a harmonised approach to the interpretation of the EPC among contracting states. However, he concluded that there was no settled view on how the test in the Enlarged Board of Appeals Decision in G2/21 was to be interpreted and applied. In his view, the case law of the courts in other EPC contracting states were “some way from unanimity” and did not provide justification for departing from Sandoz v BMS. Therefore, there was no reason to depart from previous UK case law.

The Court of Appeal also held that even if AZ had prevailed in establishing that the correct standard to be applied was the ab initio implausibility standard, the disclosure of the patent did not even meet this standard. Based on the available evidence concerning the efficacy of structural analogues, the skilled team would have had a legitimate reason to doubt that dapagliflozin was an SGLT2 inhibitor.

Arbitrary selection

AZ further argued that the judge was wrong in his approach to the issue of arbitrary selection. Specifically, it maintained that where prior art merely asserts that a genus has an advantageous property without rendering it plausible, then a subsequent patent may make a technical contribution by making the same assertion in respect of a selected compound and rendering it plausible.

The Court held it was correct to hold that the patent did not make a technical contribution over the prior art, which was very similar to the patent in suit and disclosed a class of compounds encompassing dapagliflozin. The Court confirmed a selection from the prior-disclosed genus is only inventive if the selection makes a technical contribution because the selected compound has some useful property that means the selection is a technical advance.  Mere plausibility is not enough for this purpose. It concluded that the Judge’s decision was rightly reached, despite the fact that dapagliflozin had proved to be successful, and emphasised the lack of basis in the patent in suit to distinguish dapagliflozin over the class disclosed in the prior art.

AstraZeneca’s request to appeal the decision was refused by the Supreme Court on 31 July 2025.